-------
60--
A(lat)
A(amp)
—Control Air
— Diesel Exhaust
Q Area of S.E.M. for latency and amplitude
A p<0050
B p < 0010 N-128 responses per animal
C p < 0 006 14 animals per group
C(lal)
Aldiff)
- -40
- -20
-0
. ^20
20
40
60
80
100
120 140
TIME (MILLISECONDS) from Stimulus Artifact
Figure 4. Group composite waveforms of day 35 post concep-
tion (pc) double plus somatosensory evoked poten-
tials from animals exposed to either control air
or diesel exhaust from day 22 or 23 pc. Inter-
stimulus interval between top and bottom wave-
forms is 300 msec. Double pulse delivered at
a rate of 0.5 Hz.
721
-------
VISUAL EVOKED POTENTIALS
The VHP waveforms are visually similar to the SEP wave-
forms already presented. Tables 3 and 4 are summaries
of the VEP latency and amplitude data, respectively,
obtained on days 35, 42 and 49 pc. On day 35 pc, the
latencies of all VEP peaks occurring before 300 msec
were greater than controls, but the differences reached
statistical significance for only the P2 peak of the
response to the first stimulus.
DISCUSSION
Significant differences between control and diesel ex-
haust exposed neonates were detected for developing SEPs
and VEPs. The electrophysiological differences can be
attributed to differences between the nervous systems of
the two groups. The exhaust exposure may affect both
the central and peripheral nervous systems (CNS and
PNS). For example, the slower latencies to the various
peaks may be an indication of decreased myelination in
both the CNS and PNS. CNS effects can be hypothesized
by regarding the recoverability differences. The ability
to follow rapid stimulation is limited by synaptic
phenomenon. Because all of the synapses involved with
the SEP are located in the CNS, the recovery disparity
between the control and exposed groups can be attributed
to a lesion(s) in the CNS.
SEP alterations were found in the diesel exhaust exposed
group on day 35 pc and not on other days of testing.
This may be related to the extremely rapid rate of neural
growth and development which occurs in the rat around
day 35 pc (approximately day 14 postparturition) (4).
Among the processes occurring prior to and during this
period are cell migration, axon myelination, cell
differentiation, synaptogenesis, and elimination and
stabilization of neuronal connections (5). In particular,
the rate of increase in rat brain weight peaks around
day 10 postparturition, and the in brain cholesterol
around day 16 postparturition (4). The rapid rate of
722
-------
ro
£ 71
0
i-
u_
c
O)
-^
h-
in m
cv
rtJ •
o
C -1-
Ol +J
+J CL
O QJ
Q- O
-o§ g
O) U
O L/l
> o
UJ 0_
( —^
— ,
^
CO "O
'> « £
4— •*
O C\J
00
fO LT>
O) ro
O_
c/1 g
13
u en -r-
h- o-o
D. C
X O
ii i (_>
Q < >-
o1* co CM cn
• > • •
i-H rH i-H CM
+ 1+1 +1+1
LO ^- LO ^
LO i— « r^. «e3-
LO LO ro ro
ro ro ro ro
CO • LO •
• O • LO
LO i-H LO rH
+ 1+1 +1+1
O LO LO r-«
•— i r-.. LO P--
ro «=f CM t-H
CM CM CM CM
CM
r»- LD • 01
^- LO i — i ro
+ 1+1 +1+1
r-* r-- LO ro
LO 01 en co
<_> rH *d-
LO O ' •
* • LO CM
LD LO ro r*^
CM t-H r-H r^.
^- LO LO LD
ro ^j- t^. co
^ LO *3- ^~
+ 1+1 +1+1
«* r- co co
CM o o ro
O i — 1 i-H rH
t-H ^H t-H t-H
«d- LO ro I-H
ro ro ro ro
+ 1+1 +1+1
CM CT> CM rH
^3- en r^ 01
r*- r*. r-* r-»
tn i/i
QJ QJ
en 1/1
4— 13 •+— 13
O Q_ O O-
-w -o "o "o
to QJ CO)
i-H i_ CM t-
Q_ Q-
r— i—
Or- O r-
Sw QJ S- QJ
C QJ C QJ
O Q 0 O
ro
i— 1 i-H
LO *3"
+ 1+1
CO LO
- ro LO
en en
»-H i— t
r — r^.
^- «^-
+ 1+1
O CM
co cn
ro ro
rH t-H
cn co
rH O
+ 1+ 1
<^ rx.
rH CO
ID ^J-
CM LO
rH 0
ro LO
CO LD
ro ro
y^ vo
rH .— 1
+ 1+1
CO 01
^- ^3-
CM OJ
CM i-H
ro CM
+ 1+ 1
^J- 1^
LO LO
t— 1 .—1
LO
QJ
LO
•
0 D_
+J "O
rH J_
(C
Q_
,_
O •—
S- QJ
•*-> i/i
C QJ
O -r-
C_J C3
CM
CM rH
I""-. LO
+ 1+1
o ro
ro «— i
CO O
rH CM
O CO
LO LD
+ 1+ 1
CO <3-
CO rH
t— 1 CM
•— 1 rH
CM rH
CM CM
+ 1+ 1
ro r*-
oi cn
LD LO
LO O
t— 1 i-H
O rH
LD LD
CM rH
+ 1+ 1
CM Oi
CM cn
ro CM
ro «d-
CM ^~
+ 1+1
rH O
CM LO
CM rH
LO
CD
LO
' —
O D-
T3 -0
C 0)
CM S-
(TJ
D-
(
o •—
i- QJ
C QJ
o •<—
o o
r- co
LD LO
+ 1+1
LO CM
i— 1 ^~
rH O
rH t-H
CO LO
ro rj-
+ 1+1
r-. ro
CM ro
cn cn
r- co
O CM
+ 1+ 1
LO ^t
r- o
<3- LO
LO r^
o o
LO ro
ro «d-
co ro
LO O
O rH
+ 1+1
O LD
^- ^>
CM CM
CM rH
CM ro
+ 1+1
CM r-t
ro *d-
t-H rH
r-. co
+ 1+1
cn cr»
^- LO
O CM
rH rH
O CO
ro CM
+ 1+1
""t °^
^- f*^
cn o*>
CO LO
o ro
+ 1+1
co «a-
ro cri
LD LO
cn ro
0 rH
^t LO
ro ^r
*a- <^-
tOLn
O rH
+ 1+1
^^
ro 10
ro ro
en CD
ro CM
+ 1+1
t-H t — 1
LO •— 1
CM ro
L/l
OJ
LO
1 —
O Q_
-Q -rj
C QJ
CM S-
fU
Q.
r_
O •—
S- QJ
C QJ
O ••-
O f~~\
LD
LO O
O O
00
V V
Q. Q_
II II
< 0
723
-------
O rO LO LT)
r-H r-H i — 1 r-H
ro +1+1 +1+1
^ 21
QJ . ' . . .
-^ ro o ro CM
rO ill
1—
^ C <£> r-. «a- in
i — O ....
rO T— r-H ro CM ^ir
!Jj "Q. ro + + +|+|
C <1J °~ *f^ Oro
O) o ....
4-j cz in cn ^o ^-
O O r-H r-H r-H r-H
Q- O
~o to ....
O^ CO ^" \O *t~
UJ «d~ • • co •
. -T-, r-H 1 CO
^T5 ' '
13 ro
1/1
•i — OJ
> ^j- in r-i u-> r-v.
CM CM 1*-. ^~
O LO CM + |+ | + 1+ 1
ro °-
oo •* -1 ^. ^
-^ to in r-H o ur>
rO >} ii<
C rO r-H IT) CO
0 ° g +1+1 +1+1
*r~ £.2 CO IO CO LT)
ro ^ co r— i ^j- CM
> O_ CM «a- cxj ro
i i i i
M— -t-J
O fO
C£ O CO CO CM
C/l ....
CU O> r-H CM r-< CM
~§ "fd ^ + 1+ 1 + 1+ 1
-P ^ ro r-. CO
c 4- rs ^- 3
O O Q_ O Q_
O.
^- co •*-* "O "O "O
(1) CO CD C QJ
QJ a: r-i i. CM s_
>~~ ra 03
£* c o- a.
ro CO
a. c c QJ ecu
X O O •!- O —
m
o - ro
LO I—I
CM ro
+ 1 +
CO 1^
CM r--.
r-H i— 1
1 1
r^ r-H
CM ^n
+ 1+
^j- cn
r-H CM
^
O r-l
^T CM
+ i7
CT> CO
CO O
CM <*&
1 1
<3~ vo
OJ ID
un ro
+ 1+ 1
«3- 0
^* r^
i-H r— i
1 1
OJ UD
OO ^1-
+ 1 +
ro a*
00.-H
r-H
<* 03
IO ID
+ 1+1
CT* r-»
co r-.
CM CM
i i
CM '^•
ro ro
+ 1+1 +1+1
* *o
ro ^J-
O O
1 1
IT) CO
O O
+ 1+
r^ ro
t-H CNJ
V}
01
VI
M- "3
o a.
+> 13
in Ul
C OJ
O •!-
O Q
r— CM
O O
1 1
r*«. cn
CM CM
+ 1+
CO CO
CM IT)
cn
QJ
4- 3
o a.
-o -o
C O
CM
O ro in in
r-H .—
+ 1 +
r^-r-
ro c
cn ^j
r*» r*.
+ +
r-H r-H
1 +1+1
VO.O
ro CM
i i
CM
r—t cn
1 +1+1
in CM CT* ro
CO CO
CM CM tn ro
CM CO VO r-H
cn c
r-H CN.
+ 1 +
r-< C
CT» r-
1 1
VO C\
CM u:
vo ro
CM r-H
+ l+l
C^ O
c,n 0">
*o in
i i
r- ro
10 ro
CM CM r-H
+ 1+
o «-
UD CC
O ir
r-H
CM r*
r-H CC
+ 1+
m <=j
r-H C
1
r-H C
i — 1 i—
+ 1 +
+ 1+1
00 **
r-l CTl
CO OJ
«s- co
o o
^H r-H
1 +1+1
CO O
ro 10
l-H 1
1
oj r^
CO ro
1 +1+1
O UD CM «d-
CM f-
O cn
CO CO
a
cu
CO CO
4- =
o d.
-*-> T:
t/i a
<*- r:
o Q.
13 T)
C OJ
r-H i- CM S-
rO ro
CL
O r—
s- a
•M t/>
C 01
O 'r-
) o a
cn
m
o
o
Cl~
n
724
-------
neural growth is reflected in rapidly changing evoked
potential waveforms recorded around day 35 pc. A toxic
effect which alters or retards development would produce
a larger absolute difference between control and exposed
waveforms around day 35 pc than on the other days tested.
Specifying the particular process(es) affected by diesel
exhaust without further study of development under exposure
conditions is difficult because each developmental stage
is highly dependent on the successful completion of pre-
vious events.
It should also be realized that disruption to any of the
growth processes may result in permanent sequale. Male
siblings of the animals used in this experiment were
monitored, as adults, for spontaneous locomotor activity
(SLA). The results of the SLA experiment indicated that
the adults, exposed to diesel exhaust as neonates, were
significantly less active than control animals (2).
Other behavioral experiments involving diesel exhaust
exposed animals have been reported in a companion paper(2).
The electrophysiological data reported here suggests
that the long-term behavioral alterations produced by
neonatal diesel exhaust exposure are related to the
failure to develop a normally functioning nervous system.
The long-term nature of the behavioral aberrations
demonstrates that: 1) developmental alterations are not
countered during subsequent nonexposure periods; 2)
neonatal rats are more sensitive than adults to diesel
exhaust exposure; and 3) the effects in neonates are
qualitatively different from the transient depression of
activity produced in exposed adult rats (1, 2).
Most investigations of diesel exhaust toxicity have
employed adult animal exposure. The investigations of
possible developmental effects have used classical
teratological endpoints and have not investigated the
functional competence of the pups (6, 7). With the
projected increase in the number of diesel-powered light
duty passenger vehicles, a growing number of developing
humans will be exposed to diluted diesel exhaust. In
order to help assure the protection of these sensitive
individuals, the effects of diesel exhaust on develop-
mental processes should be investigated further.
725
-------
ACKNOWLEDGEMENTS
The authors are grateful to Thomas Wessendarp and Julius
Williams for technical assistance and Juanita Jefferies
for clerical assistance.
REFERENCES
1. Laurie, R.D., J. P. Lewskowski, G.P. Cooper and L.
Hastings. Effects of Diesel Exhaust on Behavior of
the Rat. Air Pollution Control Association, Annual
Meeting, Houston, Texas, June 25-29, 1978.
2. Laurie, R.D., W.K. Boyes and T.K. Wessendarp.
Behavioral Alterations Due to Diesel Exhaust
Exposure. U.S.E.P.A. Sympoisum. Health Effects of
Diesel Exhaust Engine Emissions. Cincinnati, 0.,
Dec. 3-5, 1979.
3. Hinners, R.G., J.K. Burkart, M. Malanchuk and W.D.
Wagner. Animal Exposure Facility for Diesel Exhaust
Studies, U.S.E.P.A. Symposium. Health Effects of
Diesel Exhaust Engine Emissions. Cincinnati, 0.,
Dec. 3-5, 1979.
4. Dobbing, John. Undernutrition and the Developing
Brain. In Himwich, William A. (ed.). Developmental
Neurobiology. Charles C. Thomas, Springfield, 111.,
1970.
5. Jacobson, M. Developmental Neurobiology (second ed.).
Plennm Press, New York and London, 1978.
6. Werchowski, K.M., Chaffee, V.W., and G.B. Briggs.
Teratologic Effects of Long-Term Exposure to Diesel
Exhaust Emissions (Rats). EPA Pub. No. EPA-600/180-
010. January 1980.
7. Werchowski, K.M., Henne, S.P. and G.B. Briggs.
Teratologic Effects of Long-Term Exposure to Diesel
Exhaust Emissions (Rabbits). EPA Pub. No. EPA-600/ 1-
80-11. January 1980.
726
-------
General Discussion
C. RUDD: You indicated on a slide that the rat's brain
and the cholesterol levels are very imporatant on day 35.
Did you read slides of any rat brains and do cholesterol
levels in diesel exposed animals vary from controls?
D. LAURIE: No we didn't.
C. RUDD: In your report, were these changes on day 35
in all experiments which you did?
R. LAURIE: It only happened on day 35.
C. RUDD: How many experiments did you do?
R. LAURIE: This was the first experiment.
727
-------
EFFECTS OF SIX-MONTH EXPOSURE OF RATS TO
PARTICULATE CARBON AND NITROGEN DIOXIDE
Hastings, L., Vinegar, A., Finelli, V.N,
Leng, J., and Cooper, G.P.
Department of Environmental Health
University of Cincinnati
Cincinnati, OH 45267
Laurie, R.D., Pepelko, W., and Orthoefer, J
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio 41> 2 6 8
ABSTRACT
To evaluate the toxic effects of specific compo-
nents of diesel exhaust, rats were exposed to
carbon participate matter, carbon particulate
matter with adsorbed N02> or N02 alone. Exposure
was 8 hours per day, 7 days a week, for 6 months.
Animals were removed from the study every two
months and evaluated for changes in pulmonary
function, pulmonary alveolar macrophages,
alveolar wall permeability and lung pathology.
Other rats were examined for alterations in
behavior (running wheel activity) and reproductive
function. Although the data is still being
analyzed, no obvious overt toxic effects were
found. In one phase of the activity study a
significant depression of activity occurred but
was not replicated in the second phase. Before
any definitive statement can be made, analysis of
the data must be completed.
*Supported by EPA Contracts 68-03-0492 and
68-03-2325
728
-------
INTRODUCTION
The data to be reported here result from a
cooperative effort between the Department of
Environmental Health at the University of
Cincinnati and the E.P.A. This collaboration
originally began during the TAME (Toxicological
Assessment of Mobile Emissions) studies with a
specific interest in sulfuric acid emissions
resulting from the use of catalytic converts.
With the increased use of diesel engines, atten-
tion has now been shifted to the health effects
of diesel exhaust. Several studies already
conducted by the EPA (1) have shown that exposure
to diesel exhaust at various stages of develop-
ment in the rodent result in definite neuro-
behavioral effects. Other studies have found a
wide variety of toxic effects resulting from
exposure to diesel exhaust (2).
Since the E.P.A. is conducting a very extensive
research program on complete diesel exhaust, the
studies described here were concerned only with
selected components within the exhaust. Specifi-
cally, the study was designed to investigate what
toxic effects would result from exposure to
carbon particulate matter alone or carbon parti-
culate matter in conjunction with gas N02- The
hypothesis tested was that the gas adsorbed on
respirable carbon particulate matter would
reside in the lungs longer than the gaseous
portion of the atmosphere and thus could poten-
tially have a much more toxic effect. Parameters
examined included measures of general growth,
spontaneous locomotor activity, pulmonary
function, pulmonary alveolar macrophages,
lysozyme, permeability of the alveolar wall and
lung pathology.
METHODS
Overa 11 Experimental Design: The study consisted
of two separate groups of rats which were run
concurrently—one group at the University of
Cincinnati, the other at the E.P.A. The 2 groups
were initially drawn from the same sample pool.
Limitations in facilities resulted in the
necessity of the study being conducted at both
institutions. A control group and a group
exposed to N02 were run at the EPA, while a
control group, a charcoal particulate only group,
729
-------
and a charcoal participate with adsorbed
group were studied at U.C. Finally, the
behavioral studies were divided into two separate
phases - one involving adults, the other neonates
born in the exposure chambers. The other para-
meters were collected on animals exposed from the
beginning of the study.
Treatment Groups: From an initial pool of 350
fifty g male CFE rats from Charles River, five
groups of 70 each were randomly chosen for each
exposure condition. For each exposure condition,
10 rats were placed in 35 cm Wahmann running
wheels, with the remaining 60 being placed in
hanging wire cages. Every two month 20 rats were
selected from each exposure condition; 10 were
evaluated for pulmonary function and 10 used in
assessing biochemical indices. In addition 12
females (225 g, same supplier) were included in
each treatment condition for reproduction studies.
Exposure Conditions: Animals at both facilities
were exposed for 8 hours per day, seven days a
week. Total duration of the exposure was 6
months.
U . C . - Animals were housed in three 21m^ chambers.
One treatment group received cocoanut dust pre-
treated with N02, as well as free N02 gas at 5.5
mg/m^ and 2% ppm., respectively. A second treat-
ment group received 5 mg/m3 of plain charcoal
powder only. A third chamber contained a clean
air control treatment group. Mass median aero-
dynamic diameter particle size of the dust in
these chambers was between 3-4 ym. Free N02
gas was delivered from cylinders containing 0.5%
N02 gas in air. Charcoal/N02 pretreatment was
performed by mixing 100% N02 gas with charcoal
powder to first yellow in a separatory funnel.
Approximate concentration by weight was 10% N02
and 90% N02 + charcoal. Charcoal dust was
generated by agitating a charcoal bed with air
currents, producing a fluidized bed. The charcoal
dust was then passed through a vertical elutriator
where charcoal powders + free N02 gas were diluted
with clean air flowing at approximately 2% m^/min.
Charcoal dust was sampled on Polyvic filters
for eight hours and weighed. A Sinclair-Phoenix
Smoke Detector was also used to obtain immediate
730
-------
approximations of dust concentration. Particle
size was determined with an Anderson Impactor.
N02 gas concentrations were determined manually
with Saltzman Reagent (N-IOSH P&CAM 108).
E . P.A . - Animals were housed in eight chambers,
each having a volume of 3m3. The exposure
facilities have been described indetail else-
where in this symposiurn.(3) The control group
was exposed to filtered air only, while the
experimental group received 2^ ppm N02- The
gas was obtained from the same supplier as that
used by U.C. and generated in the same manner.
Parameters Examined
Behavior: Weanling rats were placed in
Wahmann running wheels two weeks prior to
commencement of exposure in order to obtain
baseline data. They remained in the wheels
continously for the duration of the study.
Number of wheel revolutions was recorded daily.
Food and water intake and body weight were
recorded weekly. In the second phase of the
study, weanlings from the reproduction study
were placed in the wheels and maintained on the
same schedule. However since most of the rats'
activity occurs during nightime, and exposure
would be greatest during such periods of exercise ,
the light-dark cycle was reversed. Thus, although
exposure conditions remained the same, the
effective concentration was increased.
Pulmonary Function
Every two months, 10 rats from each treat-
ment were sacrificed and measurements made of
the following pulmonary parameters: lung weight,
carbon monoxide diffusion (D|_CO), vital capacity,
alveolar volume and residual volume by gas
dilution and water displacement. Methods used
were modified from those of O'Neil et al. (4).
The lungs were subsequently saved for patho-
logical examination.
Biochemi stry
As in tP
rats from eac
two months and the pulmonary alveolar macrophage
examined. Albumin - 131I (80 y Ci/Kg body
As in the pulmonary function studies, 10
rats from each treatment were sacrificed every
731
-------
weight) was injected into the caudal vein of the
experimental animals 6 hours before sacrifice.
After the animals were sacrificed, the lungs
were lavaged and aliquots of the lavage were used
for macrophage counting, 131j activity counting,
and protein and lysozome determinations.
Reproduction
After one months of exposure, the 12 females
in each treatment were mated with 12 males from
the same treatment (males were subsequently used
in studies of pulmonary function). Number of
successful pregnancies as well as litter size and
birth weights were recorded. On day 3, litters
were culled to no more than eight pups. At
weaning, 3 males per litter were saved; the rest
were sacrificed. At 40 days of age, 10 male off-
spring per treatment were placed in the running
wheels.
RESULTS
The general health stuatus of the experimental
animals was unaffected by exposure to any of the
treatment conditions. Food and water consumption
was the same across all conditions as was also
body weight.
Analysis of the activity data for the first group
of rats placed in the running wheels showed that
while there were no overall treatment differences,
there was a significant interaction of Treatment
X Time. (F = l.63 ,df = 30,405; p<.02) Inspection of
Fig. I. reveals that while the control group and
the Carbon + N0£ group started
showed the greatest separation
of highest activity (week 7j.
study the groups were once again nearly equal
There were no significant differences found any-
where between the control group and the N02 group
(Fig. 2).
out equal , they
during the period
By the end of the
In the second
again started
study, although the three groups
out equal (see Fig. 3), the
difference increased rapidly between the control
and the exposed animals. However, analysis of
the data revealed that the differences (Treatment
alone) only approached significance (F=2.92,
732
-------
§
UJ
70
50
50
5 3°
ix 20
10
• •
+ m •—•
'.ARBO'l
5 7
11 13 15
TIME (WEEKS)
Figure 1. Running wheel activity of rats
exposed to carbon particulate matter alone,
particulate matter + NOps or filtered air.
Exposure began on week 3 and continued for the
duration of the study.
733
-------
CONTROL • 0
70
g 6°
o
3 50
CD
LLJ
!! 40
LU
UJ
1 3°
ix 20
10
N0
2
1 I 1
1 i 1
135 7 9 11 13 15
TIME (WEEKS)
Figure 2. Running wheel activity of rats
exposed to N02 or filtered air. Exposure
began on week 3 and continued for the
duration of the study.
734
-------
70
•** 60
c/j
•ZL
O
50
.0
or filtered air.
735
-------
df=2,26; p<.07). Unlike the first run where the
groups had nearly equal activity levels at the
end, the groups in this run still showed marked
separation. It should be pointed out though,
that the carbon only group showed the greatest
reduction and not the carbon + N02 group as in
the first run. There were no significant differ-
ences between the N02 and control group (Fig. 4).
In the reproduction phase of the study there were
no significant differences found on any of the
parameters measured (see Table 1). Rate of
growth was also comparable across all three
groups .
Tables 2 and 3 contain the results of the pulmo-
nary function evaluation studies. These data are
only preliminary in nature and do not include the
results from those animals exposed for the full
six months. Although there are significant
differences in several of the parameters it is
premature at this time to interpret them until
results are available from the last study and
from the pathology reports.
In previous studies involving sulfuric acid expo-
sures (5), a drastic effect of inhaled Al2(S04)3
was observed on the pulmonary alveolar macrophage
which resulted in 3 to 5 fold increases in leak-
age of lysozyme into the extracellular fluid.
Moreover, aluminum sulfate caused a pronounced
increase in the leakage of intravenously injected
radi ol abel 1 ed albumin into the alveolar space.
These two parameters were highly correlated,
therefore, it was hypothesized that damage to
macrophages with consequent spillage of lysosomal
enzymes in the alveolar space lead to an inflam-
matory process resulting in an increase in
alveolar wall permeability.
In this study rats exposed to carbon showed only
a slight but significant increase in extracel-
lular lysozyme; however, during the first 4
months of exposure, the alveolar wall permeabiliy
to radio labelled albumin remained unchanged.
No effects were seen in the rats exposed to
Finally, studies of lung pathology have thus far
been done on a small sample from the two-month
exposure group. In general, very few effects were
seen-especi al ly with the Carbon alone or
736
-------
CO
o
/O
60
50
40
30
20
10
•_ •
NC'2
5 7 9
TIME (WEEKS)
11
13
Figure 4. Running wheel activity of rats born
in exposure chambers and exposed from birth on
to either N02 or filtered air.
737
-------
C
O
C ••-
•r- -P
13
-O -P
Ol 10
4-> CU
£
01
l/> .C
1 i Ij
03
1. -P
3
,
(O 'r-
O. to
^J
0)
> TJ
•r- CO
-P to
O O
3 Q.
-o x
O 01
s_
O.-O
0) C
o; to
•
01
.0
(O
1—
•p
-C O)
cn
•r* >~~*
CU O»
-C Q
| * C
S- CO
•r- >4-
J3 — •
IX
•P
-C
cn
•r- cn
2 —
J^;
•P r-
S- Q
a —
IX
s-
01
•P Ol
•P N
i — 10
IX
1 —
M- I/)
l/l O)
to •!-
O) U
Lf> VO
r*. t/5 oo
• • •
VO ID ^O
CM Lf> CO
• • •
«i- ro co
i— r— i —
CM CM CM
^-. ^-^ ^^
cn r~» cn
csj
o
•z.
+
O C C
s- o o
-P J3 JD
C S- S-
O lO (O
U U O
738
-------
01
c
0
03
S-
QJ
-(-> CO
4-> 01
03 3
E^^
r—
03
CU >•
res
3 CO
o ^
•r- CU
j_ 5
03
Q.
CO
O S-
-Q O
S- 4-
tO
0 S-
O 03
+J
T3
T3 O)
m c
QJ >-
CO Ol
O 4->
O-i—
X -r- •
CU 4- C
0
CO i_ •!—
4J O 4->
03 03
S- Ol -r-
C >
4— O "i—
Or- -O
03
i- CMS-
O) CD 03
4-> Z T3
Ol f"
E: CM O3
05 O 4->
S- 'Z. CO
O3
0. 4- +1
C S- C
O O) 03
v- 4-> O>
4-* 4-^ E
(J 03
CEO)
3 S-
4- Ol 03
4-J
>5 03 T3
i_ r— CD
03 ^ 4->
C (_) 03
O 'I- O
E 4-* "r—
3 03 C
D- Q.-I-
CM
Ol
r—
-Q
03
1—
c 3:
o E
•i— E
O l/»—
O 131
4- C
M- -^
5<§
0
^
v^.
, —
03 O)
~^ E *^~^
T3 13 •
5
f— 4-J
03 *i—
4-> 0 r-
••- 03 e
> Q-- •
03
CJ
^^
03 CO
13 E •
T3 3 O r—
•r- i— CMS
co o :c ---
01 ;=•
o;
4->
O^^I ^ — ^
c en en
^ .,— V .
—1 O)
s
4->
^"^ ^^
"O CD' — .
O -r- CD
CQ cu —
-=•
.^
i — i — C31CO COLO LOCO COID
1 — CO CXtr^ CTt^f OCO CMCM
t— O •— O i— O CMO CMO
'DO *sl" O ^-i — CM^ rOi —
i — f— r— r— r—
r*^ ^j~ t»o o r**^ ^o en o ^o ^^
• • •• •• •• ••
CMi— COCM OOr- ,— CM r\Jr-
r— f— . p— n— , — .
r^. co .0 ^D O 4->
C S_ S- 2: C CM
O 03 (O O O
0 0 0 + C_> -ZL
LO
O
V
f~^-
+J
03
! —
O
s_
4->
C
O
o
Ol
4->
03
S^
0.
O
Q.
Q.
03
E
O
s_
4-
4J
C
CU
s-
Ol
4-
4-
•^j
^
40
C
03
(J
4-
•r*
c
en
>r~
OO
4:
739
-------
«*
cu
c
o
to
cu ""
+J ^
^ to
=• >
cu
to ,/;
^ "cu
_<-> CU
4J 5
^~ LO
to p—
Q_
J_
d Q
S **~
S_ j_
O "U
-u cu
s-
-o cu
CU -!->
LO i —
O •!-
0.4-
X
CU S- C
0 0
4-> CU 4->
to c to
S- 0 -r-
1 — >
4- (0 •!-
O "O
CM
to O "O
£_ ^ 5^.
CU to
+J T3
CU « C
ro CD^-S
s- z: to
to
Q- + +1
C S- £Z
o cu to
•r- -t-> CL)
4J 4-> E
c E cu
rs i-
4- CU fO
>> tO "O
S- ,— CU
fO 3 -t-*
c o to
§•1- O
-(-> -1-
i— S- T3
3 (O C
Q. Q.T-
.
CO
cu
JO
rO
1—
1
O)
C T"
0 E
'o»—
CD 33 1
0 4- C
4- •,-
S CD
O
r—
E
r—
to CU
•a u cu i—
•r- ^ Z E
to o — '
cu >
a:
s_
fO ^
i — E -^-^
S-2^
i— >^
k
^— +•*
rO *i~~ ^~ -«•
2 «1
:1r— to C
<0
O
, —
to CU
^ E ^ — •
-a ?i o >—
•^ r- CME
to O 3C — '
CO >
cc
4J
01 g~ ^ — .
c a> en
ZJ -i
— 1 LD
+J
>^ f—
"O CO^ — -
O T- en
CQ CU
3
O i—
00 LO
CM O
LD CO
. •
i — CD
LD UO
f^ ,
^~
O CM
^Q ,
|
CO CM
VO LD
CM CO
— O
cn o
r~^ o
en CM
i— O
LO ^*
LD CO
LD
O
i.
+J
C
O
o
CM •* CM CM
O LD r^. LO
CO O CM O
^
•3- LD CO LD
• • • •
i— O O O
r^ co LO LD
r^ CM LD CM
f— p_
CTi *d" OO i —
LO CM LO CM
1— ) —
LD LD CM CO
LD LO r— r~~
O i— r— r-
r— 0 r— 0
CTl OO r— i—
«3- LO O LO
i— CM i— .—
CM O CM O
^e
•* •!*• 1 O
t-~ CO i— CO
LO LD
C C
O O CM
JO JD 0
S- S_ Z
(0 fO
0 O +
LO ID
r— CM
CO 0
CO LO
• •
i— 0
LO r~-
(V^ ,
i —
CM «d-
LD i—
i —
oo co
r^ en
LD CO
r— O
LO LO
LO LD
CM ^1-
CM O
^J- CO
co ^s-
LD
"^
i_
+J
C
0
o
•K
r-x oo
LD CM
CM O
CM LO
• •
r- O
LO LD
I-- i—
, —
LD *J"
LO r-
f—
*
LO CM
CO CM
i— O
en i—
r-- i —
01 i—
r- 0
*
cr> en
CO CM
^J-
CM
o
~z.
o
-t->
c
o
o
a.
£
Q.
O-
res
+J
C.
CU
cu
4-
4-
•»—
•o
>1
•p
c
c
CD
L/1
740
-------
alone groups. In the N02 + Carbon group a few of
the animals showed dilation of the peripheral air-
ways, indicating a weakening of the pulmonary
parenchyma or a loss of some alveolar septal
strength. Again these results are very preli-
minary and whether this trend continues after
longer exposure remains to be seen.
DISCUSSION
Since the data from this study are still being
analyzed, any conclusions that are drawn at this
time must still be considered tentative. With
this in mind, it can be stated that no obvious,
overt detrimental health effects were observed
resulting from exposure to carbon particulate
matter, N02, or a combination of the two. Growth
curves and food and water consumption was compa-
rable across the groups. There were no differ-
ences in any of the reproduction parameters,
either. A few scattered measurements involving
pulmonary function were significant, but they
appear to be due to random statistical variation.
Pathology also failed to detect any significant
toxic effects.
Behavior, as measured by running wheel activity,
was significantly depressed in the first run but
not the second. This was unexpected since 1.)
those in the second run had been exposed from
birth and 2 ) exposure occurred, on a daily basis,
during their most active period. While the
results of the second run were not statistically
different they followed the same trend as the
first run and of previous studies (1), i.e.,
depression of activity. From the available data
it does not appear that any dramatic affects can
be attributed to any of the treatment conditions
or that there was any significant difference
between carbon particulate matter alone or
particulate matter and
However it would be premature to state that
carbon particulate matter alone or in conjunc-
tion with N02 are without any toxic effects.
This would be tantamount to proving the Null
Hypothesis, which is clearly impossible. It is
very easy to develop an experimental design with
features which would make it difficult to obtain
741
-------
any significant differences. Unfortunately,
many of these factors often occur in toxico-
logical research--factors such as limitations
in sample size, very low exposures, short
duration of exposure, etc. Before any definitive
statements can be made, greater attention must
be given to these condiserations.
REFERENCES
1. Laurie, R.D., J.P. Lewkowsky, G. P. Cooper
and L. Hastings. Effects of Diesel Exhaust
on Behavior of the Rat. Air Pollution
Control Association, Annual Meeting, Houston,
Texas. June 25-29, 1978.
2. Santodonato, J., Basu, Dr., and Howard, P.
Health Effects Associated With Diesel Exhaust
Emissions: Literature Review and Evaluation.
EPA Publication No. EPA-600/1-78-063 ,
November, 1978.
3. Hinners, R.G., O.K. Burkart, M. Malanchuk and
W.D. Wagner. Animal Exposure Facility for
Diesel Exhaust Studies. U.S.E.P.A. Symposium.
Health Effects of Diesel Engine Emissions.
Cincinnati, Ohio, December 3-5, 1979.
4. O'Neil, J.J., J. Takezawa and T.D. Crapo.
Pulmonary diffusing capacity. Single breath
measurements compared to morphometric analysis
in rats exposed to NOp and Op. The Physio-
logist, 20(4):69, (1977).
5. Finelli, V.N., Lee, S.D. Danner, R.M.,
McMillan, L. and Cooper, G.P. Inhalation of
Sulphate Particulates II: Pulmonary Bio-
chemical Effects. Presented at Society of
Toxicology, 17th Annual Meeting, San Francis-
co, California, 1978, Abstract No. 63, pp. 55.
General Discussion
S. DUTTA: What were the gross appearances of the lung?
Do you see blackness of the lung and a lot of carbon par-
ticles?
L. HASTINGS: There is a definite grayish appearance to
the lungs. They are not as dark as animals exposed to
diesel exhaust but they are darkened.
742
-------
S. DUTTA: What about lymph nodu-les in the lung or lymph
nodes? We have just seen four weeks, as well as two weeks,
exposure at 6,000 micrograms per cubic meter. That would
be about six milligrams per cubic meter.- That exposure to
diesel exhaust causes a dense color change as well as clear
blackening of the lymph nodes close to the trachea. I was
wondering what kind of things you observed in the lymph
nodes as well as in the lung? If they are clear, then it
would appear to me that carbon particles are an entirely
different species so far as diesel particles are concerned.
L. HASTINGS: I didn't do the gross pathology on these.
Although I didn't see it, I am sure the lymph nodes would
have contained particles; the macrophages were loaded with
particles very similar to the diesel exposed animals and
there were what looked like carbon particles adhered to the
alveolar walls. They were very similar to the diesel ex-
posed lungs.
SPEAKER: Does the answer lie in your particle size? I
didn't hear you say particle size. What was the particle
size?
L. HASTINGS: Three to four microns - not as small as
the smallest diesel particles.
J. ORTHOEFER: Actually, they are quite a bit larger
than the diesel particles.
L. HASTINGS: The lymph nodes were black.
RIGGER: In your further data collection activities, how
much is your sample size going to increase in terms of the
number of animals to each of the four specific categories?
L. HASTINGS: I think we will have to look at the data
first and see what type of variability we are experiencing
in order to determine what future size the samples should
be.
RIGGER: I think the fact may be that you haven't found
anything, any trends, or anything significant to date. It
may be a function of numbers purely. The probability of
erroneously rejecting some known, or accepting some known
hypothesis is very high given the sample sizes you have
right now.
SPEAKER: Did you have a chance to determine the met-
hemaglobin concentration in the blood or nitrate since you
have used nitrogen dioxide as your gas component?
L. HASTINGS: No.
743
-------
ATMOSPHERIC COMPONENT CONCENTRATIONS IN THE ANIMAL
EXPOSURE CHAMBERS, TAPE II STUDY
M. Malanchuk, N. P. Barkley, and G. L. Contner
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
ABSTRACT
The atmospheric data from the various analyses - daily
monitoring of major pollutants and intermittent measurements
of other components in the animal exposure chambers - was used
to calculate whole study averages for the first year of
operation of the diesel engine. The concentration values, for
exhaust emissions diluted by clean air (16 to 18X), served as
reference data in the evaluation of animal health effects and
as a guide for controlling the engine system operation.
The atmospheres in the animal exposure chambers were analyzed
for both gaseous and particulate components in each of twelve
such chambers supplied with atmospheres containing diluted
engine exhaust emissions and of four reference chambers sup-
plied with clean air (C.A.). Daily monitoring prevailed for
the components of major concentration - carbon dioxide (CC>2),
carbon monoxide (CO), total hydrocarbons (THC), nitric oxide
(NO), nitrogen dioxide (N0£) and sulfur dioxide (802). The
atmospheres also were checked for levels of ammonia gas (NI^)
and ammonia products resulting from chamber animal contri-
butions interacting with the other atmospheric components.
Analytical procedures and instrumentation are identified in the
report by R. Hinners'1'.
RESULTS AND DISCUSSION
The chambers were monitored for temperature and relative
humidity as factors involved in the overall animal health
effects, whether directly or as a result of interactions in the
atmosphere.
The concentration values were used as reference data relating
to animal health effects and as a guide for controlling the
engine system operation within limits defined for the study.
744
-------
The engine conditions coupled with the level of dilution air
volume were used to maintain a suspended particulate con-
centration of about 6 mg/m-' in the animal exposure chambers.
The weekly averages, as shown in the figure, calculated as a 52-
week average of 6.39 +_ 0.78 mg/m .
Average values of the component concentrations for (1) the
exposure chambers supplied with diluted exhaust emissions and
(2) the chambers supplied with clean air, were used to
establish composite weekly averages for each of the two
atmosphere types.
A study average of each of several major components for the
first 52 weeks showed:
Table I. - Component concentrations
Exhaust Emissions (E.E.) Reference (C.A.)
Chambers Chambers
C02, % 0.29 + 0.03 0.05 + 0.01
CO, ppm 19.72 + 2.13 2.07 + 0.53
THC, ppm as C 7.84+0.99 3.49+0.38
NO, ppm 11.23 + 1.53 0.08 +_ 0.03
N02, ppm 2.65 +_ 0.55 0.05+0.03
S02, ppm 2.06 + 0.43 0.05*+_ 0.01
Particulate, mg/m3 6.39+0.78 - - -
* Based on clean air values obtained at beginning of the day.
Initial samples of the particulate mass in the C.A. atmosphere
chambers were barely measurable ( 0.1 mg/m3) and such measure-
ments were discontinued.
From the above table, it is evident that there was a con-
siderable contribution, percentagewise, of hydrocarbons (HC)
to the exhaust emissions chambers, 7.84 ppm, from the back-
ground clean air, 3.49 ppm. The difference, 4.35 ppm,
represents the actual input from the engine exhaust emissions.
The sulfur dioxide (S02) value for the reference chambers was
determined from measurements made on the clean air supply
before the sampling operation was locked into the cycling
program at the beginning of the day's run. During the cycling
program, the exhaust emissions chambers (n=12) with their
relatively high pollutant concentrations were sampled se-
quentially before the group of reference chambers (n=4). The
S02 monitoring instrument was very slow in responding to the
large drop-off of concentration from the last exhaust emissions
chamber to the nearly zero concentration of the first and
745
-------
CM
1)
Q.
co
o
in
co
CD
I
Q.
CO
O
O
LU
ce
CO
O
0.
X
LLl
o
ro
CO
fc
•*
Ul
UJ
CO
o
a
x
0) c
s- •-
3 «>
en «:
•i- CO
^ 3
.2
r
CO
a.
"5
CO
i
0)
o
o
o
co
a>
746
-------
succeeding reference chambers so that falsely high levels of
S02, averaging about 0.54 pptn, were recorded for those cham-
bers.
The particulate mass value was determined from the samples
collected on teflon-coated filters. The weights that were
measured for similar samples on glass fiber filters were
invariably higher by about 1 mg/m^; they are judged to be less
accurate because of the considerable sensitivity of that filter
material to humidity effects.
The particulate matter in filter samples examined by a scanning
electron microscope revealed an apparently basic unit of 0.05-
0.10 urn size, with agglomerates reaching almost 1.0 urn in
diameter.
Sulfate concentrations, as determined by ion chromatography
applied to the aqueous extracts of the filter samples, averaged
1.53 mg/m^ during the first quarter (3 months) of the study. It
represented 23.0 percent of the total-particulate concen-
tration measured over the same interval. A more thorough study
of the ion components in the particulate was done during a
period immediately preceding the start of the presently re-
ported study; the procedure of that type analysis is described
in a report by N. P. Barkley'2).
Ammonia gas (NH3) concentrations were measured in the various
exposure chambers primarily to minimize if not eliminate
episodes of high levels of the gas. Since the concentration
varied with the number and kind of animal in the chamber and
with the cleaning schedule for the chamber, there resulted
quite a large variation in NH3 concentration among the cham-
bers. Although the average value for the entire study was only
0.62 ppm inclusive of all the chambers, there did occur certain
daily episodes in a chamber that measured as much as 11.3 ppm
NH3.
The DNPH method (gas chromatography of the dinitrophenyl-
hydrazine derivatives of the aldehydes) for 5 to 6 individual
aldehydes was applied to the exposure chamber atmosphere
samplings. However, when considerable difficulty was en-
countered in obtaining reliable or reproducible results, that
method was set aside in favor of several other aldehyde
procedures - for total, aliphatic aldehydes'-^ for for-
maldehyde^), and for acrolein^). The study averages of those
components by the latter group of procedures showed 0.18 ppm
aliphatic aldehydes, 0.11 ppm formaldehyde and a little less
than 0.03 ppm acrolein.
747
-------
Hydrocarbon concentration measurements by means of the heated
(350°F) sampling line were made as daily checks on (the same)
two exposure chambers throughout the study. They averaged 0.75
ppm higher than the ambient temperature values (average of 7.60
ppm) monitored in all twelve of the exhaust emissions chambers
over the corresponding period of time.
REFERENCES
1. Hinners, R.G., J.K. Burkart, M. Malanchuk and W.D.
Wagner. "Animal Exposure Facility for Diesel Exhaust
Studies," Generation of Aerosols, edited by Klaus
Willeke, Ann Arbor Science Publishers, Inc., Ann Arbor,
Mich. pp 525-540 (1980).
2. Barkley, N.P., G.L. Contner and M. Malanchuk. "Simul-
taneous Analysis of Anions and Cations in Diesel
Exhaust Using Ion Chromatography," Ion Chromatographic
Analysis of Environmental Pollutants, Vol. 2, edited by
J.D. Mulik and E. Sawicki, Ann Arbor Science Pub-
lishers, Inc., Ann Arbor, Mich. pp 115-128 (1979).
3. Hauser, T.R., and R. Cummins. "Increasing Sensitivity
of 3-Methyl-2-Benzothiazolone Hydrazone Test for Anal-
ysis of Aliphatic Aldehydes in Air," Anal. Chem. 36:679
(1964).
4. Altshuller, A.P., D.L. Miller, and S.F. Sleva.
"Determination of Formaldehyde in Gas Mixtures by the
Chromotrophic Acid Method," Anal. Chcm. 33:621 (1961).
5. Intersociety Committee, Methods for Ambient Air Samp-
ling and Analysis, "Tentative Method of Analysis of
Acrolein Content of Atmosphere," 435-05-01-70T. H.-
L.S. 7:179 (1970).
748
-------
PULMONARY FUNCTION CHANGES IN CHINESE HAMSTERS
EXPOSED SIX MONTHS TO DIESEL EXHAUST
Allen Vinegar and Arch I. Carson
Department of Environmental Health
University of Cincinnati
Cincinnati, Ohio 45267
William E. Pepelko
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
ABSTRACT
Chinese hamsters were exposed for eight hours per day to
automotive diesel exhaust emissions which were diluted with
air (18 to 1) and had a particulate level of 6.4 mg/m^.
Pulmonary function measurements were made after six months
exposure. Body weight (BW), lung weight (LW), vital capacity
(VC), residual volume by water displacement (RVW) and by gas
dilution (RVp), alveolar volume (V/\), and carbon monoxide
transfer factor (D^CO) were measured. LW showed a sig-
nificant increase in the diesel exposed animals (P .01)
while VC, RVW, and DLCO showed decreases (P .01). Static
deflation volume-pressure curves showed depressed deflation
volumes for diesel exposed animals when volumes were cor-
rected for body weight and even greater depressed volumes
when volumes were corrected for lung weight. However, when
volumes were expressed as percent vital capacity, the diesel
exposed animals had higher lung volumes at 0 and 5 cm H20.
Results of the pathological examination of the lung tissue
will be necessary for final analysis of our findings.
However, preliminary interpretation indicates possible em-
physematic changes which are compatible with the observed
decrease in D| CO.
749
-------
INTRODUCTION
The effects of exposure to diesel exhaust upon the lungs
cannot be accurately predicted from the results of previous
studies using either individual pollutants or whole emissions
from gasoline engines. A portion of the gaseous pollutants,
for example, are adsorbed onto the surface of the particulate
matter (Stokinger, 1973), altering both the depth of pene-
tration (Task Group on Lung Dynamics, 1966) and residence
time in the lungs (Creasia et al, 1971). There is also little
information concerning the composition of sulfates present in
diesel exhaust. Certain sulfates, such as zinc ammonium
sulfate, have been found to be considerably more irritating
than sulfuric acid, the primary form in catalyzed automobile
exhaust (Amdur, 1970). Finally, diesel exhaust contains a
wide range of both aliphatic and aromatic compounds, the
concentration of which, and even the chemical structure, are
not well defined (Karasek et al, 1974).
The present study was designed to evaluate the effects of 6-
months exposure to a concentration of diesel exhaust con-
taining 6 mg/m^ particulate upon pulmonary function in the
Chinese hamster.
METHODS
General Procedure. Adult male Chinese hamsters were exposed
to diesel exhausf eight hours per day seven days per week at
a dilution ratio of approximately 1:18. Particulate con-
centration averaged 6.4 mg/m^. The animals were exposed in
stainless steel wire cages 11 inches square. Nine or ten
animals were housed per cage. Food and water were provided ad
libitum. Total length of exposure was 6 months. Following
completion of exposure, the animals were removed, weighed,
and body temperature measured via rectal probe. The hamsters
were then anesthetized with pentobarbital sodium admin-
istered intraperitoneally. The trachea was exposed and
cannulated just below the glottis.
Vital Capacity. The animal was placed in the supine position
on an insulated platform and the tracheal cannula was
attached to the breathing port of the D[_CO apparatus (see
Figure 1). The animal was hyperventilated, with the respira-
tory valve in the "B" position, for a period sufficient to
produce apnea lasting 10-15 seconds. With the valve in
position "A" fresh air was injected into the lungs of the
animal from a calibrated syringe until the pressure in the
airway plateaued at +25 cm hbO. The volume reached was
defined as total lung capacity (TLC). Air was then withdrawn
into the syringe until the pressure in the airway plateaued
750
-------
DLCO
APPARATUS SCHEMATIC
VITAL CAPACITY
SYRINGE
-10cmH2O
VACUUM
+25 cmH20
TEST GAS
Figure 1. D|_CO Apparatus Schematic.
751
-------
at -10 cm H20. The volume reached was defined as residual
volume (RV). The corrected difference between the two
syringe readings was defined as vital capacity (VC). The
animal immediately returned to breathing fresh air.
Alveolar Volume. The animal was again hyperventilated until
apnea was produced. The valve was returned to the "C"
position allowing the lungs to be deflated to RV. The valve
was turned to the "D" position allowing the lungs to be
inflated quickly to TLC with air containing a known quantity
of the inert gas neon. The valve was turned to the "E"
position and the gas in the lungs was pumped in and out ten
times to facilitate even distribution of the neon within the
lungs. The valve was turned to the "F" position allowing
passive expiration of about one-third of the vital capacity
into a preset ground glass syringe, effectively eliminating
the anatomical and mechanical deadspace air. The remaining
alveolar air could then be sampled into a gas-tight chro-
matography syringe for analysis. The animal returned to
breathing fresh air.
D| CO. The animal was hyperventilated until apnea was
produced. The valve was turned to the "C" position allowing
deflation of the lungs to RV. The valve was then turned to
the "D" position allowing rapid inflation of the lungs to TLC
with air containing known amounts of neon and carbon monox-
ide. Approximately 8-10 seconds after inflation, the valve
was turned to the "F" position and the alveolar gas sampled as
described above. During all of the above procedures, the
body temperature was monitored and an attempt made to keep it
as close to the preanesthesia level as possible by use of a
variable intensity heat lamp mounted above the platform. The
procedures and calculations for determination of D|_CO, al-
veolar volume and vital capacity as described above are
modified from those reported by O'Neil et al (1977).
Static Deflation Volumes-Pressure Curves. The lungs were
exposed by opening the chest cavity and retracting the ribs.
The lungs were connected in parallel, by way of the trachea!
cannula, to a syringe and a pressure transducer. Pressure
was monitored on an oscilloscope and volume read directly
from the syringe. The lungs were given a known volume history
by inflating twice from 0 to 30 cm h^O, maintaining for five
seconds, and then deflating in increments of five cm h^O, and
maintaining at each subsequent pressure for 15 seconds.
Deflation was taken to -10 cm 1^0. Volumes were recorded at
each pressure decrement.
Final Lung Treatment. The lungs were then removed and
attached to a pressure reservoir of -10 cm HzO. Lung weight
752
-------
and displacement volume were determined. This volume was
defined as residual volume (RVW) as compared with the above-
mentioned determination by gas dilution (RVp).
Finally, the lungs were instilled with 10% buffered formalin
and maintained at 30 cm H20 while submerged in a bath of the
same fixative. After 36 to 48 hours they were stored in
individual jars of buffered formalin to await preparation for
pathological examination.
RESULTS
Our preliminary results are pre-sented in Table 1 and Figure
2. The apparent difference in RV, as determined by water
displacement, may not be real since correction was not made
for tissue density. The diesel exposed animals had lungs
that were mainly coal black in color. The added density of
the deposited material made the lungs denser, but without
knowing the weight and density of deposited material a total
correction for lung density cannot be made.
The static deflation volumes, Figure 2 are plotted graph-
ically against pressure in four different ways: A) percent
vital capacity, B) absolute volume of air removed from lung,
C) lung volume corrected for body weight, and D) lung volume
corrected for lung weight. Where differences between diesel
and control animals exist at least at the P=.05 level this is
indicated with an asterisk.
DISCUSSION
Pathological investigation of the lung tissue has not yet
been performed. However, the elevated lung volumes at 0 and
5 cm H20 for the volume-pressure curve where volume is
expressed as %VC indicate a possible emphysematic condition
in the diesel exposed animals. This is consistent with the
significantly reduced D[_CO found in these same animals.
Exposure at the same concentration for longer periods of time
and exposure at higher concentrations may be necessary to
more carefully evaluate the potential effects of exposure to
diesel exhaust on lung function.
REFERENCES
1. Amdur, M.O. (1970). The impact of air pollutants on
physiologic response of the respiratory tract. Proc. Am.
Philosophical Soc., 14:3-8.
2. Creasia, D.A., Poggenburg, J.L. and Nettesheim, P. Jr.
(1976). Elution of benzo(a)pyrene from carbon particles
in the respiratory tract of mice. J. Toxicol. Environ.
753
-------
00
S_
O)
CL>
o
i — $_
=3 QJ
O_ -1-1
c E
O f\3
Q-DC
— i
CJJ
Q) 00
S- O)
=! C
00 -i—
O ^:
Q-CJ)
"^
3 ^
•~~ _J
CJ ^~
>•
X —
3 CD
CO \
^ — 1
C-> ^"
^> ^ — ^
un
CSJ
<^-
co
to
CSJ
o
0
to en
•
CO
ro
O en
O
O un o
0 „•
CSJ r- V
o_
r^ to un
r- -3- 0
CSJ O O „•
0 0 r- V
D_
X
UJ
to O
en .—
un csj o
O
cu
oo
cu
o
csj r^ i—
to r— O
d ^ ° V
i— O_
o
CJJ
CD CD
1 >—i s: i-~ csj
i uj -—• csj csj en
3 csj
o
S £ o v.
OO i— o_
I— O
>- ic a:
O 1—1 CD ^
co LJJ --- o
3 CJ
uu
oo
en LU
un ro
ro
CO
754
-------
% VC ML
100,
50
f 1 * 1-2 1
4 * '
A -8
* .6
*
A .4
, 1 •
. • •
. A # *
.A * *
A *
¥•
-10 -5 0 5 10 15 20 25 30 ' 1 £
CM H O
2
ML/G BW
.03
.02
.01-
-10 -50 5 10 15 20 25 30
CM H O
2
ML/G LW
6
4-^-
4
3
* - ^ * * t
-10 -5 0 5 10 15 20 25 30 -10 -5 0 5 10 15 20 25 30
CM H O CM H O
2 2
Figure 2. Static Deflation Pressure Volume Curves Controls
are indicated by circles and diesel exposed are
indicated by triangles. Asterisks indicate
significance at least at the P = .Ob level.
755
-------
Health, 1:967-975.
3. Karasek, F.W., Smythe, R.J., and Laub, R.J. (1974). A
chromatographic-mass spectropholometric study of organic
compounds adsorbed on participate matter from diesel
exhaust. J. Chromat, 101:125-136.
4. O'Neil, J.J., Takezaqa, and Crapo, J.D. (1977). Pul-
monary diffusing capacity: Single breath measurements
compared to morphometric analysis in rats exposed to NO?
and 0?. The Physiologist, 20:69.
5. Stokinger, H.E. (1975). Toxicology of diesel emissions.
Proceedings of the Symposium on Use of Diesel-Powered
Equipment in Underground Mining. 1C 8666. Bureau of
Mines. U.S. Dept. Interior, Washington, D.C., 366 pp.
6. Task Group on Lung Dynamics. (1966). Deposition and
retention models for internal dosimetry of the human
respiratory tract. Health Phys., 12:173-207.
756
-------
PULMONARY FUNCTION EVALUATION OF CATS AFTER ONE YEAR
OF EXPOSURE TO DIESEL EXHAUST
William E. Pepelko and Joan Mattox
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
William J. Moorman and John C. Clark
National Institute for Occupational Safety and Health
4676 Columbia Parkway
Cincinnati, Ohio 45226
ABSTRACT
Adult male, inbred, disease-free cats of uniform age and size
were exposed eight hours per day, seven days per week to a
1:18 dilution of diesel exhaust emissions. After one year of
exposure, the animals were removed from the chambers for
measurement of lung volumes, forced expiratory flow rates,
dynamic compliance and resistance, diffusing capacity, and
nitrogen washout. No important changes in pulmonary function
were detected with the exception of a decrease in closing
volume (P < .05). The inability to detect decrements in
pulmonary function may have been due to insufficient concen-
tration of exhaust, insufficient exposure length, or to the
use of a species resistant to diesel exhaust.
To test these possibilities, the cats are being exposed for
an additional year, and another species, hamsters, are being
exposed for future testing at exhaust dilutions of 1:18 and
INTRODUCTION
The use of diesel engines has increased greatly in the past 30
years. Much of this increase has been due to a changeover to
use of diesel engines in loconotives, heavy trucks, con-
757
-------
struction equipment, and farm tractors. With increasing fuel
costs, the percentage of diesel engine equipped automobiles
is also increasing and is expected to continue to increase in
the foreseeable future.
Diesel exhaust contains a wide variety of pollutants, many of
which are known to cause lung damage (4). Moreover, many of
these pollutants are adsorbed on the very fine carbon
particulate allowing for effective penetration into the deep
lung (8). For these reasons, it was considered important to
study the effects of chronic inhalation of diesel exhaust
emissions upon lung function.
The present experiment was designed to evaluate the effects
of chronic diesel exhaust inhalation upon pulmonary function
and pathology in cats.
METHODS
Young adult male cats, born and maintained in a disease free
environment and inbred for several generations were purchased
from Liberty Labs, Liberty Corners, New Jersey. The cats
were uniform in size and varied in age by less than two weeks.
The cats were exposed in chambers 1.38 meters square with an
interior volume of 2.83 cu m. Each chamber contained a wire
mesh floor and two half shelves of similar material at 2/3 and
1 1/3 meters above the floor. This allowed the cats about 4.7
m2 (50 sq. ft.) of horizontal space. Eight or nine cats were
housed per chamber. They were not caged but allowed to roam
free. Food and water was provided ad libitum. The cats were
exposed eight hours per day, seven days per week for one year
to diesel exhaust diluted to produce a particulate con-
centration of 6 mg/m^. Details of the exposure conditions
have been presented elsewhere in the proceedings by Hinners
et al.
Prior to pulmonary function testing, the animals were fasted
for one day. The testing followed 18-20 hours of no diesel
exposure. The animals to be tested were anesthetized with
Ketaset Plus (Ketamine 100 mg, Promazine 7.5 mg per ml) at a
dose of 42 mg/kg. Following the induction of anesthesia, an
esophageal balloon was placed in the lower third of the
esophagus and an 18-22F endotracheal tube was inserted into
the trachea with the aid of a laryngoscope. The cuff of the
endotracheal tube was inflated and excessive length of the
distal end trimmed even with the end of the mouth. The animal
was then placed into the chamber, ventral side up, for
compliance and resistance testing. For all other tests, the
animal was in the prone position, dorsal side up.
758
-------
Pulmonary mechanics were obtained from simultaneous volume,
flow, and transpulmonary pressure tracings displayed on a
twelve-channel photographic recorder (Electronics for Med-
icine, DR-12). Airflow through the pneumotachograph was
measured with a differential transducer and electrically
integrated to produce a volume trace. Dynamic pulmonary
compliance (CLpyN) was calculated from simultaneous volume
and transpulmonary pressure tracings at points of zero flow
(5). Average flow resistance (RL/\ve_ Flow) was calculated
from the change in transpulmonary pressure (at equal volumes)
divided by the sum of inspiratory and expiratory flow.
All mechanics were measured while the animal was breathing
spontaneously through the pneumotachograph only. The animal
was inflated for ten seconds initially and periodically
throughout the testing to expand atelectatic areas.
The pulmonary function tests requiring breathing maneuvers
lung volumes, maximum expiratory flow-volume curve (FEF),
diffusing capacity (CL^lSo), nitrogen washout (AN?), and
closing volume (CV) were performed using a variable pressure
plethysmographic chamber. The basic method employed was
similar to that used in an external tank respirator; however,
a hydraulic control system enabled the operator to bring
about inspiration, expiration, breath holding, and breathing
rate within the anatomical and physiological limits of the
animal. Both flow and volume were controlled secondarily by
changes in the pressure surrounding the animal. Inspiratory
and expiratory airflow could be controlled from very low
rates to the maximum within each subject. Likewise, volume
could be controlled for both maximum inspiration and ex-
piration. Inspiratory capacity (1C) was obtained by rapid
depressurization to -70 cml^O from testing tidal position.
Prior to the flow-volume testing it was determined that
plethysmograph pressures of +70 cmh^O would be used to
produce maximal expirations. Inspection of flow-volume
curves at increasing driving pressures showed that flow
limitation characteristics had been reached at volumes above
50% Total Lung Capacity (TLC) when the plethysmograph pres-
sure was greater than 70 cmHjO. Therefore, FEF at 50% and 40%
of TLC are values take at an effort-independent zone of the
flow-volume curve. The curves were highly reproducible in
each animal and demonstrated a low coefficient of variation
(2-2.5%) in the effortindependent zone.
To ensure that sufficient intrathoracic driving pressure was
developed, esophageal pressure was recorded during forced
expirations. A trans-chest-wall pressure gradient was ob-
served; however, intrapleural pressures of 30-35 cmHpO were
achieved which are efficient to produce flow maxima. A
759
-------
volume error, as a result of thoracic gas compression, was
calculated to be approximately 3% at 50% TLC with the
intrapleural pressure of 30-35 cmH20. This error was
considered to be irrelevant because the results were compared
in animals tested at the same driving pressures.
Breathing manipulations could be performed in anesthetized
animals because of the apnea produced on inflation as a
result of the inflation reflex documented by Hering and
Breuer (3). The inspiratory inhibition had been demonstrated
by recording action potentials from the phrenic nerve.
1C and forced vital capacity (FVC) were recorded during a
maximum inspiration, followed by a maximum expiration. Flow
and volume tracings were recorded, which provided the es-
sential data points for calculating forced expiratory flows
and volumes (FEVg.5 I.Q) ar|d peak expiratory flow (PF). This
procedure of maximum inspiration followed by maximum ex-
piration was performed initially and thereafter for all test
maneuvers, insuring equal volume and flow histories.
The methods of Brashear et al (1) and Mitchell et al (6), were
combined to obtain values for DL£l8rj and TLC. The calcula-
tions for DLCO were performed according to the method
described by Wagner et al (9), for C^O. Gas analyses were
done using a respiratory mass spectrometer (Perkin-Elmer
MGA1100).
Distribution was studied using the single-breath nitrogen
washout and closing volume according to the methods described
by Buist and Ross (2).
RESULTS
The pulmonary function results for exposed and control cats
are summarized in Tables 1 and 2. No significant differences
were found in mechanical properties, diffusing capacity,
uniformity of distribution, or ventilatory performance.
Closing volume (volume of Phase 4) was found to differ from
the control values; however, this change cannot be in-
terpreted as impairment. The exposed cats demonstrated a
lower closing volume than the controls indicating improved
function of small airways. This effect is believed to be a
Type 2 statistical error or some unexplainable adaptation
phenomena. The latter explanation is unlikely based on our
experience. It is probably due to high animal to animal
variation seen in some of the pulmonary function tests.
A preliminary study was conducted by the EPA in which several
animal species were exposed for 20 hours per day for up to two
months to a similar concentration of exhaust as used in the
760
-------
Table 1
Forced Expiratory Flow Rates in Cats
osed to Diesel Exhaust
N
21
21
21
21
21
21
21
21
21
21
21
21
19
21
21
19
21
EXPOSED
MEAN
278.7
158.2
69.2
86.1
415.3
20.26
348.4
84.3
97.6
1016.5
728.0
490.4
196.5
2.101
1.400
.569
486.3
S.D.
44.81
35.61
24.58
36.99
56.02
6.936
43.46
8.44
1.96
185.10
195.62
186.81
107.35
.5434
.4806
.3131
252.64
N
21
21
21
21
21
21
21
21
21
21
21
20
20
21
20
20
21
CONTROL
MEAN
301.9
265.4
67.0
104.1
449.5
22.72
368.9
81 .6
97.3
1041.8
661.1
481.4
222.2
2.072
1.314
.605
557.2
S.D.
49.56
42.22
19.05
37.67
74.49
5.896
42.14
6.39
1.74
174.17
160.43
199.49
156.82
.4249
.5310
.4294
248.05
LUNG VOLUMES
1C (ML)
FRC (ML)
ERV (ML)
RV (ML)
TLC (ML)
RV/TLC (%)
DYNAMIC LUNG VOLUMES
FVC (ML)
FEV 5/FVC (%)
FEV1/FVC (%)
PEFR (ML/5)
FEF 50% (ML/5)
FEF 25% (ML/5)
FEF 10% (ML/5)
FEF 50%/FVC
(FVC/SEC)
FEF 25%/FVC
(FVC/SEC)
FEF 10%/FVC
(FVC/SEC)
FEF 40% TLC (ML/5)
FEF 40% TLC/TLC
(TLC/SEC) 21 1.185 .6355 21 1.317 .6368
761
-------
Table 2
Dynamic Compliance and Resistance, Diffusion Closing
Volume and Nitrogen Washout in Cats Exposed to Diesel Exhaust
Exposed
N Mean S.D.
Control
Mean
MECHANICS
(CMH20/L/5)
(ML/CMH20)
DIFFUSION
S.D.
21 10.675 4.5785 21 10.323 4.4305
21 23.536 7.2045 21 23.698 9.2548
DLCO (ML STPD/
MIN/MM HG) 21 1.179 .3143 21 1.217 .3023
DLCO/VA
(1/MIN/MM HG) 21 .00345 .00125 21 .00335 .00084
CLOSING VOLUME AND UASHOUT UITH N?
21 25.6* 13.44 21 36.8
21 7.91 3.276 21 10.46
CV
CV/VC (*)
(CV+RV)/TLC(%) 21 26.83 * 6.993 21 20.05
%N2/25%VC
21
.32
.206 21
.29
16.00
4.6
7.192
.302
* Significantly different from Controls P<.05
** Significantly different from Controls P<.01
762
-------
present study (8). This initial study was designed to
provide preliminary data on tolerance levels, toxic effects,
and target organs. Some of the findings included detection
of the presence of black granular particles in the alveolar
macrophages, black pigment in the bronchial and carinal lymph
nodes, increased pulmonary flow resistance, increased lung
weight/body weight ratios, and sinus bradycardia in guinea
pigs.
In the present study, in which animals were exposed for a
longer period of time (one year), but only eight hours per
day, convincing evidence is present that inhalation of diesel
exhaust under these conditions does not result in functional
alterations in the lungs of cats. This is in contrast to the
preliminary study in which some positive responses were
noted. We believe cats in the exposed groups have function-
ally adapted to this exposure. Increasing the concentrations
of diesel exhaust or increasing the duration of this study
may produce significant chronic pulmonary disease; however,
the present results cannot provide a guide to those para-
meters which may be early indicators of functional al-
terations.
Minimal or no responses would be expected to result from the
individual gaseous pollutants at the concentrations present.
The particulate level might be expected to produce impaired
ventilation as a result of the site of deposition and
potential tissue response. The effect of the combination of
gaseous pollutants, interaction with, or adsorption to the
particulate, however, cannot be predicted.
The results reported here are from an ongoing study in which
cats are being exposed for a planned two years to diesel
exhaust emissions. Snortly after the start of the second
year of exposure, the dilution of exhaust was decreased to
produce a particulate concentration of 12 mg/m^. Pulmonary
function, pathology, and biochemical parameter will be as-
sessed following completion of exposure. In addition, lung
pathology is being evaluated from several "cats that died
during the year either from unrelated causes or during
testing.
763
-------
REFERENCES
1. Brashear, R.E., Ross, J.C., and Daly, W.J. (1966).
Pulmonary diffusion and capillary blood volume in dogs
at rest and with exercise. J. Appl. Physiol., .21:526-
520.
2. Buist, A.S., and Ross, B.B. (1973). Quantitative analy-
sis of the alveolar plateau in the diagnosis of early
airway obstruction. Am. Rev. Resp. Pis., 108:1078-1087.
3. Hering, E. and Breuer, J. (1968). Die Selbsteuerrung der
athmung durch den nervus vagus. Sitzber. Acad. Wiss.
Wien., 57:672-677.
4. Karasek, F., Smythe, R.J. and Laub, R.J. (1974). A gas
chromatographic, mass spectrophotometric study of organ-
ic compounds adsorbed on particulate matter from diesel
exhaust. J. Chromatograp., 101:125-136.
5. Mead, J. and Whittenberger, J.L. (1953). Physical
properties of human lungs measured during spontaneous
respiration. J. Appl. Physiol., 5:779-796.
6. Mitchell, N.M. and Tenzetti, A.D. (1968). Application of
the single breath method of total lung capacity meas-
urement to the calculation of carbon monoxide diffusing
capacity. Am. Rev. Resp. Pis., 97:581-584.
7. Task Group on Lung Dynamics: Deposition and Retention
Models for Internal Dosimetry of the Human Respiratory
Tract. Health Phys., 12:173-207, 1966.
8. U.S. Environmental Protection Agency. (1978). Health
Effects Associated with Diesel Exhaust Emissions. Lit-
erature and Evaluation. EPA-600/1-78-063.
9. Wagner, P.O., Mazzone, R.W. and West, J.B. (1971).
Diffusing capacity and anatomic dead space for carbon
monoxide [C180]. J. Appl. Physiol., 31:817-852.
764
-------
APPENDIX
(Abbreviations)
1C Inspiratory Capacity
FRC Functional Reserve Capacity
ERV Expiratory Reserve Volume
RV Residual Volume
TLC Total Lung Capacity
FVC Forced Vital Capacity
PEFR Peak Expiratory Flow Rate
FEF Forced Expiratory Flow
RL Resistance
Cl Compliance
L.F. Low Frequency
H.F. High Frequency
DLCO Diffusing Capacity
VA Alveolar Volume
765
-------
FUNCTIONAL AND MORPHOLOGICAL CONSEQUENCES
OF DIESEL EXHAUST INHALATION IN MICE
John J. O'Neil, Ping-Chuan Hu, Fred J. Miller,
John L. Carson, Albert M. Collier, and Donald E. Gardner
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
and
Department of Pediatrics
University of North Carolina
Chapel Hill, NC 27514
The mice used in this study were exposed for three months to
light duty diesel exhaust and were supplied by the Health
Effects Research Laboratory, USEPA, Cincinnati, OH. All of
the animals were anesthetized i.p. with sodium pentobarbital
(50 mg/kg body weight, Abbott Laboratories). We adapted
pulmonary function methods used with other small laboratory
animals to measure lung volumes and the single breath diffu-
sing capacity for carbon monoxide (2). The data were ana-
lyzed by covariance analysis using logjQ body weight as the
covariate to account for weight differences between the two
groups. Other animals were tracheostomized, the lungs were
removed and then inflation fixed with 2% gluteraldehyde at
25 cm H£0 using a low resistance system (1). The fixed lung
tissue was prepared for light and electron microscopy using
standard techniques.
The results of the pulmonary function testing are summarized
in Table 1. No statistically significant differences (p >
0.05) were observed between the two groups.
766
-------
TABLE 1. PULMONARY FUNCTION OF MICE EXPOSED
FOR 3 MONTHS TO LIGHT DUTY DIESEL EXHAUST*
Control
Number of animal s
Body
Lung
Vital
Total
weight,
weight,
capaci
g
g
ty, ml
lung capacity, ml
Residual volume, ml
5
25 _+
0.15 _+
0.89 i
1.29 +_
0.39 +
0
0
0
0
0
.8
.002
.05
.05
.03
0
0
1
0
Exposed
9
30 _+
.23 +_
.97 +_
.40 _+
.43 +
1
0
0
0
0
.1
.025
.04
.08
.07
Diffusing capacity for
carbon monoxide, and
CO absorbed/min x torr'l
0.0174 + 0.001 0.0178 + 0.003
*Mean + S.E.
Light and electron photomicrographs of parenchymal tissue
from mice following diesel exhaust inhalation for three
months are shown in Figures 1,2, and 3. The alveolar
macrophages accumulated carbonaceous material and appeared
to be larger when compared to macrophages from the control
animals (Figure 1). The macrophages became heavily laden
with this material and they appear globular with the absence
of pseudopodia (Figure 2). Diesel exhaust particles are
also present in the interstitium, presumably in the lympha-
tics (Figure 2). The macrophages accumulate this carbona-
ceous material in discrete vesicles which appear to be
membrane limited (Figure 3). Although no statistically
significant functional differences were observed between
these two groups of animals, it is clear that the diesel
exhaust particles were accumulated in the lung parenchyma.
Further investigations of pulmonary function and morphology
would provide useful information on the potential toxicity
of diesel exhaust.
767
-------
XA
I
m
',
/ »
t
*••
,- ' \ *
B
t „
^
y
Figure 1. Photomicrographs of alveoli from a control animal
(A) and an animal exposed to diesel exhaust (B). The
alveolar macrophages from the exposed animals appear to
be increased in size and to contain many carbonaceous
particles. (1.175X)
768
-------
Figure 2. Photomicrograph of alveolar tissue from an animal
exposed to light duty diesel exhaust. The alveolar macro-
phages are heavily loaded with carbonaceous particles, they
appear to be globular in shape with a notable absence of
pseudopodia. The participate material is also evident in
the interstitium, presumably in lymphatic vessels. (1,175X)
769
-------
Figure 3. Electron photomicrograph of alveolar macrophages
join the lungs of an animal exposed to diesel exhaust.
Carbonaceous material appears to be confined in discrete
membrane limited vesicles within the macrophage. The
eosinophil, frequently seen in the alveolar space, does not
appear to be phagocytizing any particles. (15.000X)
770
-------
REFERENCES
Hayatdavoudi, G., James D. Crapo, Frederick J. Miller
and John J. O'Neil. Factors which determine the degree
of inflation in intratracheally fixed rat lungs. J.
Appl. Physiol: Respirat. Environ. Exercise Physio!.
48(2), (In Press), 1980.
Takezawa, J., F. J. Miller and J. J. O'Neil. Lung
volumes and single breath diffusig capacity in small
laboratory mammals. Journal of Applied Physiology:
Respirat. Environ. Exercise Physiol. (In Press)
1980.
771
-------
ENCHANCED SUSCEPTIBILITY TO INFECTION
IN MICE AFTER EXPOSURE TO DILUTE EXHAUST
FROM LIGHT DUTY DIESEL ENGINES
K. I. Campbell, E. L. George and I. S. Washington, Jr.
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
ABSTRACT
A series of experiments was conducted in which groups of mice
were first exposed for various durations to diluted exhaust
from light duty diesel engines and then briefly to an
infectious aerosol generated by nebulizing cultures of a
bacterial pathogen (Streptococcus). Typically, post-in-
fection mortality was significantly greater in groups exposed
to exhaust than in their corresponding control groups exposed
to purified air only. Data of recent diesel and of past
diesel- and catalyst-treated gasoline engine exhaust experi-
ments suggest a somewhat greater excess mortality from
(enhanced susceptibility to) bacterial infection in mice
exposed to diesel exhaust than in those exposed to catalytic
gasoline exhaust. Limited data on acute tests of NO? and
acrolein vapor alone suggest that the infectivity-enhancing
effect of diesel exhaust could be accounted for in large part
by these components. Exposures to diesel exhaust, N02, or
acrolein did not enhance the mortality response to a viral
pathogen (A/PR8-34).
INTRODUCTION
In recent years there has developed an increasing awareness
and concern about the potential adverse effects of environ-
mental contaminants on the immune system and ability to
resist infection. A number of contaminant materials have
been demonstrated to impair various specific functions of the
immune system and, significantly, host resistance to micro-
bial pathogens, which depends on the integrity of several
772
-------
immunologic and physiologic components performing in con-
cert.
Exposure to automotive exhaust prior to the era of the
catalytic converter, and to certain of its component or
resultant pollutants such as nitrogen dioxide (NOp) and
ozone, was shown to enhance susceptibility to infection by
inhaled Streptococcus and Klebsiella pathogens.'>2>3 M0re
recently~weha vernvestigated the toxicity of catalyst-
modified automotive gasoline-engine exhaust, and diesel
engine exhausts, with respect to impairment of host resist-
ance. 4,5,6 yn-js report concerns recent similar tests of
diesel engine exhaust.
METHOD
The experimental model ("infectivity test") used was adapted
from that described by Coffin and Blomrner^ in connection with
their early auto exhaust studies and later employed by others
for evaluating various airborne contaminants. A series of
experiments was conducted in which mice were exposed to
diluted diesel-engine exhaust (test) or to purified air
(controls) for brief, intermediate, or prolonged periods,
shortly after which they were briefly exposed to an atmos-
phere containing airborne respirable pathogen in culture and
then observed for response to infection. Mortality was the
primary response by which group effects were compared. A
greater incidence of mortality after exposure to infectious
material among test mice compared to controls was considered
to represent impaired resistance to the infectious process by
whatever specific mechanism(s) responsible.
Young adult (usually 4 to 8 weeks), female, CR/CD-1*1" albino
mice were used. Purina Laboratory Chow and tap water were
provided ad libitum. The basic group size of n=20 (e.g., 20
controls, 20 test) was used in individual experiments owing
to limitations of the inhalation chamber used for infectious
challenge.
The mice were housed in groups of 10 in wire mesh cages for
their exposure to purified air (control) or diluted exhaust
(test) atmospheres. Exposures were conducted using stainless
steel and glass inhalation chambers previously described by
Hinners, et al.^,8 Chamber atmospheres were generated and
distributed also as described by Hinners, et al.8 Briefly,
* Charles River Breeding Laboratories, Inc., Wilmington,
Mass.
t Mention herein of a commercial product does not constitute
endorsement by the Environmental Protection Agency.
773
-------
exhaust was generated using alternatively one of two Nissan
CN-6 diesel engines operated on a dynamometer stand providing
for cycled operation ("Federal Short Cycle") consisting of a
repeated series of acceleration, deceleration, cruise, and
idle modes. Each day the engine was operated for exhaust
generation and distribution for 8 hours, after an initial
warm-up period during which exhaust was diverted. The fresh
exhaust was mixed and diluted (overall first-year average
dilution factor about 18:1) with carbon- and particulate-
filtered and temperature-humidity conditioned air and dis-
tributed to mixing chambers and then to the inhalation
chambers. The inhalation chambers we're ventilated at the
rate of about 15 volumes per hour, being positively exhausted
and operated under conditions maintaining chamber pressure at
about 0.3-0.4 in water, negative. Chamber atmosphere con-
ditions were controlled at very near 72°F and 50% relative
humidity. Engine performance was monitored and maintenance
and adjustments were conducted so as to maintain the design
level particulate at 6-7 mg/m^ in the test-exposure chambers.
Chambers and cages were serviced (cleaning, feeding, water-
ing) on a frequent schedule to maintain general sanitary
conditons, minimize ammonia production from animal excreta,
and observe status of animal health.
Several times daily the chamber atmospheres were auto-
matically sampled and analyzed for carbon dioxide (C02),
carbon monoxide (CO) total hydrocarbons (HC), nitrogen oxides
(NO, N02), and sulfur dioxide (S02). Chambers were also
sampled manually at least daily for determination of total
suspended particulates (TSP), and at periodic intervals for
other components of interest (e.g., aldehydes, other organ-
ics, ammonia). Methods used and other engineering details
are described elsewhere in this symposium.
A few experiments involving exposure to nitrogen dioxide
(N02)-or acrolein (AC)-contaminated atmospheres were con-
ducted. These are irritant constituents of automotive exhaust
and were used in single-contaminant experiments to invest-
igate the degree to which they might contribute to exhaust
toxicity in terms of impaired resistance.
Immediately after control or test exposure, the mice were
removed and transported. They were then placed in in-
dividually compartmented mesh cages and exposed to infectious
aerosol for a brief period in a smaller inhalation chamber of
comparable design. The infectious aerosol was generated by
nebulizing an appropriate dilution of cultured pathogen in a
DeVilbiss nebulizer (Mod. 40), the discharge of which was
passed through a section of glass or Tygon tubing (to permit
settling out of too-large droplets) and then into the
774
-------
chamber's air supply inlet. The pathogen culture strength,
dilution, nebulizer operation, and chamber operation para-
meters, and exposure duration (generally 15 and 30 minutes
for bacterial and viral aerosol challenges, respectively)
were such as to yield a low positive morality response (5-20
percent) in control subjects so as to optimize detection and
quantitation of enhanced response in test-exposed mice. In
two experiments the mice were challenged with an enteric
bacterial pathogen by intragastric intubation.
Three pathogens were used. The primary pathogen used for
challenge in most of the experiments was a B-hemolytic group
C Streptococcus pyogenes originally isolated from an animal
lesion.It was used in the studies of Coffin^ and later by
Gardner, et al.,9JO from wnom we obtained our stock culture.
Active culture in Todd-Hewitt broth was freeze-dried in
aliquots for stable storage. These were reconstituted and
diluted as needed for infectious challenge exposures. The
pathogen of next importance was a mouse-adapted A/PR8-34
influenza virus. It was passed in mice, titered, grown in
brain heart infusion broth, frozen in aliquots, and thawed
and diluted as needed for challenge exposures. The third
pathogen, used in two experiments, was a Salmonella ty-
phimurium. It was grown in trypticase soy broth, lyo-
pTTilized, and resonstituted just prior to use. The addi-
tional viral and enteric pathogens were used to investigate
whether diesel exhaust affected resistance in terms of these
infectious phenomena as it was known to in the case of
streptococcal pneumonia.
After infectious exposure the mice were housed for ob-
servation in plastic shoe-box cages in an animal room
supplied with single-pass filtered, conditioned air. Body
weights were recorded daily, and mortality and morbid signs
were monitored, for at least two weeks, or for a sufficient
additional period to accommodate recovery from the infection
response (at least a week beyond the last death, with trend of
increasing weight gain in surviving mice).
A total of 22 experiments were performed, including 13
involving acute (one 2-hour, the rest 6-hour), 7 subacute
(two 8-day, four 15-day, one 16-day, 8 hours/day), and 2
chronic (307 and 321 days, 8 hours/day) exposure periods
preceding infectious challenge. The Salmonella pathogen was
used in two subacute (15-day)diesel exhaust (DE) experiments,
the virus in four acute (one DE one N02, one with De and AC,
and one with De, N02, and AC) and one 8-day subacute (DE, NO?,
and AC) experiments, and the streptococcal pathogen was used
in all other experiments (all test exposure durations and
agents).
775
-------
Mortality following infection was the primary criterion used
to assess whether resistance to infection was impaired by
test exposure. Group mortality data was analyzed by the chi-
square (uncorrected) technique for comparison between and
among experimental groups. Other lethality data (e.g., time
to death), and body weight changes relative to controls as a
general index of morbid response to infection, were also
examined.
Protocol and design features of this work are summarized in
Appendix 1 .
RESULTS AND DISCUSSION
The average concentrations for the primary exhaust con-
stituents monitored in exposure chambers are summarized in
Appendix 2 for several typical experiments. The design
exhaust concentration for the period covered by these exper-
iments was that providing chamber atmosphere total suspended
participate loading of 6 to 7 mg/m^. The average con-
centration over the first year of operation was 6.4 mg
TSP/m3, as shown for Experiment CB-1,2. The corresponding
average concentration of N02, a respiratory irritant gas and
likely contributor to the impaired resistance to infection
effect, was 2.8 ppm, comparable to the level in the exper-
iments with NOp exposure (2.5 ppm). Concentrations of other
constituents less regularly assayed (including aldehydes)
should be available in another report in this symposium.
Data on the mortality response to infection (the primary
measure of effect) for most experiments involving exposure to
diesel exhaust are summarized in Appendix 3. Exposures of
all durations to diesel exhaust resulted in enhanced sus-
ceptibility (impaired resistance) to lethal effects of bac-
terial infection. This was qualitatively consistent thr-
oughout the experiments, was clear cut and statistically
significant in most experiments, and was highly significant
in statistical tests of pooled data. Moreover, consistent
with earlier studies of diesel and catalytic-converted petrol
exhausts, the limited acute tests clearly showed a greater
toxicity on these criteria by irradiated than by non-
irradiated diesel exhaust. This is conjectured to be due to
oxidant-oxygenated compounds (as yet unidentified) with
greater cell-damaging properties. Nonirradiated diesel
exhaust, however, was clearly toxic to the system(s) re-
sponsible for coping with infection by the inhaled strep-
tococcal pathogen, and conferred an increased sensitivity to
infection even after very brief exposure periods of 2 to 6
hours.
776
-------
It is of interest to note that the effect observed with
respect to bacterial infection (streptococcal pneumonia) was
absent in the diesel or NO? experiments in which the viral
pathogen (influenza) was used, and was clearly apparent only
in one of three experiments testing acrolein (which was very
active with bacterial infectivity). From these limited data
it is inferred that the difference is probably due to
different mechanisms, and their sensitivities, responsible
for defense against this viral than against the bacterial
pathogen.
The two subacute DE - Salmonella experiments were in-
conclusive with regard toeffects on susceptibility to
enteric infection. In both experiments mice were challenged
after about two weeks exposure to DE, but although mortality
rates in test groups were somewhat greater than those in
controls, they were not convincingly so (44 vs 41 and 100 vs
80 percent, respectively). One problem was the too high
mortality in controls. This sometime occurs as a result of
various dose or host factors and it was not feasible to
conduct further such tests in this program.
The data in Appendix 4 for tests involving acrolein and NO?
clearly indicate that exposure to either of these agents
results in increased susceptibility to bacterial infection,
(at least by the Streptococcus used here)'Acrolein was
apparently somewhat more toxic than NO?. Neither acrolein
nor NO? conclusively affected defenses against the viral
pathogen. It would appear that much of the diesel toxicity
with respect to streptococcal infectivity could be accounted
for by the NO? and/or AC constituents, but more precise
definition of this is needed.
In all but three experiments one or more other infection
lethality parameters were adversely affected by test exposure
when compared with controls, as might have been anticipated,
vis.: a) Earlier deaths (mortality beginning earlier in
exposed than in control groups), b) Lower mean number of days
to death among animals dying), and/or, c) Earlier modal day
of death. Also not surprising is the fact that in most
experiments (all but 4) there were at least some days during
post-infection observation on which the relative body weight
index (RBWI)* showed a deficit in comparison to that of
* Test group mean body weight on day n/group mean body
weight on day 0 ^___^
RBWI=Control group mean body weight on day n/group mean
body weight on day 0,
where day 0 was the day of infectious challenge.
777
-------
controls. This was considered to be a general indication of
infection morbidity reflecting malaise and unwillingness or
inability to eat and/or drink. Maximum deficits ranged up to
29%, and deficit days ranged up to 29 out of 30 days observed
after challenge (before sacrifice of survivors). Since body
weight deficits as well as overt illness occurred in many
mice that did not die from infection, this type of toxicity
(impaired resistance to infection) could perhaps also be
translated to the human situation in terms of decrement in
productivity (e.g., absenteeism and lower performance).
Appendix 5 summarizes infectivity test mortality data from
earlier catalyst-modified gasoline-engine exhaust and the
earlier diesel studies, as well as the current study. In all
cases irradiated exhaust was more toxic than nonirradiated
exhaust in the context of enhancing susceptibility to bac-
terial infection. In the diesel studies the fuel consumption
has been calculated at about 100 g/mi and contribution to
test atmosphere about 1.4 g fuel/m^. These are considerably
less than corresponding data for the catalytic exhaust
studies, about 200 g/mi and 5-7 g/m^. Daily exposures were
for 16 h in the "gasoline-catalyst study", 20 h in the "pilot
diesel study", and 8 h in the "current diesel study." Exhaust
dilution factors were roughly similar in all studies (about
14:1 in the gasoline-catalyst and pilot diesel, and 18:1 in
the current diesel). With this background, although the data
on excess mortality over controls for the 3 studies are not
strictly comparable (tests were run in same manner but not
concurrently), a general comparison among these studies
(diesels vs gasoline-catalyst) tends to indicate that diesel
exhaust may be somewhat more toxicologically potent than
gasoline-catalyst exhaust, with respect to effect on resist-
ance to bacterial infection in the acute and subacute
situations. Excess mortality in the diesel studies was
either roughly equal to or was greater than that in the
gasoline-catalyst study, yet the exposure index (hours for
acute, hours/day for subacute) in diesel studies was similar
to or less than that in the gasoline-catalyst study and the
fuel mass contributions to exposure in the diesel studies
were only about half or less of those in the gasoline-
catalyst study. The chronic test data are equivocal because
not only were exposure regimens quite different, but the
mortality in control mice in the diesel study was too high for
a meaningful comparison.
778
-------
ACKNOWLEDGMENT
The authors wish to thank the following for their con-
tribution to this project: Dr. Donald Gardner and Earl
Blommer, U.S. EPA, Research Triangle Park, for the strep-
tococcal culture; Dr. James Renters, Illinois Institute of
Technology Research Institute, Chicago, for the viral cul-
ture; Dr. Ivan Kochan, Miami University, Oxford, Ohio, for
the Salmonella culture; Mr. Dale Kraemer, U.S. EPA, Cin-
cinnati, for statistical consultation; Dr. E. Akin and
Mssrs. R. Stetler and C. Mayhew for virus-work assistance,
and Ms. J. Roe for stenographic assistance.
REFERENCES
1. Purvis, M.R., S. Miller, and R. Ehrlich. 1961. Effect
of Atmospheric Pollutants on Susceptibility to Res-
piratory Infection. I. Effect of Ozone. Jour. Inf.
Pis. 109:238-242.
2. Purvis, M.R. and R. Ehrlich. 1963. Effect of Atmos-
pheric Pollutants on Susceptibility to Respiratory
Infection. II. Effect of Nitrogen Dioxide. Jour.
Inf. Pis. 113:72-76.
3. Coffin, D.L. andE.J. Blommer. 1967. Acute Toxicity of
Irradiated Auto Exhaust Indicated by Mortality from
Streptococcal Pneumonia. Arch. Environ. Health 15:-
36-38.
4. Third Annual Catalyst Research Program Report. 1978.
Effects of Exposure on Susceptibility to Respiratory
Infection. Health Effects Research Laboratory, Of-
fice of Research and Pevelopment. U.S. Environmental
Protection Agency. EPA-600/3-78-012:148-149.
5. Campbell, K., E. George, I. Washington, and Y. Yang.
1978. Enhanced Susceptibility to Respiratory In-
fection in Mice Exposed to Automotive Diesel Em-
issions. In: Toxicological Assessment of Piesel
Emissions, Laboratory Studies Division, Health Ef-
fects Research Laboratory, U.S. Environmental Pro-
tection Agency, Cincinnati, Ohio, April 27, 1978.
6. Lee, S.D., K.I. Campbell, et al. 1978. Toxicological
Assessment of Piesel Emissions. Annual Meeting, Air
Pollution Control Association, Houston, Texas, June
25-29, 1978.
779
-------
7. Hinners, R., J. Burkart, and C. Punte. 1968. Animal
Inhalation Exposure Chambers. Arch. Environ. Health
16:194-206.
8. Hinners, R.G., J.K. Burkart, M. Malanchuk, and W.D.
Wagner 1979. Facilities for Diesel Exhaust Studies.
International Symposium on Health Effects of Diesel
Engine Emissions, Health Effects Research Laboratory,
U.S. Environmental Protection Agency, Cincinnati,
Ohio, December 3-5, 1979.
9. Gardner, D.E., F.J. Miller, J.'W. Illing, and J.M.
Kirtz. 1977. Alterations in Bacterial Defense Mech-
anisms of the Lung Induced by Inhalation of Cadmium.
Bui 1. Europ. Physiopath. Resp. 13:157-174.
10. Graham, J.A. and D.E. Gardner. 1977. Effects of
Metals on Pulmonary Defense Mechanisms Against In-
fectious Disease. In: Proc. Seventh Ann. Conf.
Environmental Toxicology, AMRL-TR-76-125, Aerospace
Medical Research Laboratory, Air Force Systems Com-
mand, Wright-Patterson Air Force Base, Ohio 45433,
October 13-15, 1976.
780
-------
APPENDIX 1. - DIESEL INFECTIVITY TESTS
DESIGN - SUMMARY
MODEL
Control/test exposure + infectious exposure —» in-
fection (Mortality, etc.)
TEST ATMOSPHERES
Diesel exhaust (non-irradiated, irradiated) - 18:1
dilution
Acrolein - 4 ppm, 2 ppm
Nitrogen dioxide - 2.5 ppm
TEST EXPOSURE DURATION ROUTE: INHALATION (WHOLE BODY)
Acute - 2 hr., 6 hr.
Subacute - 7, 15, 16 da. (8 hr/da x 7 da/wk)
Chronic - 44/46 wks (8 hr/da x 7 da/wk
INFECTIOUS MICROBIAL AGENT(S) AND EXPOSURE
Bacterial
Streptococcus pyogenes, Type c B-hemolytic (res-
piratory pathogen).Respirable aerosol of di-
luted culture, by inhalation, usually 15 min.
Salmonella typhimurium (enteric pathogen). By
intragastnc intubation of diluted culture.
Viral - Influenza, A/PR8-34, mouse adapted. Aerosol,
by inhalation, 30 min.
Exposure/dose - Factors adjusted to produce minimal
positive response in control subjects (5-20% mor-
tality).
RESPONSE TO INFECTION
Reflects whether resistance to infection is impaired.
Mortality - compare test vs. control group mortality.
Possibly other end points (e.g., time to death,
comparative body weight changes).
SUBJECT
Mouse, female, CR/CD-1, 17-25g (means), 4-6 wks for
most.
STATISTICS
Chi-Square (uncorr.) for mortality comparison.
781
-------
APPENDIX 2. - MEAN AEROMETRY FOR INFECTIVITY STUDIES
(Average concentrations in chamber atmospheres)
Exp't
AB-3
AF.-4
A8-5
fB-6
AB-7
AV-1
AV-2
AV-3
SB-2,3
SB-4
CB-1,2
RANGE
DE
CA
DE
CA
DE
CA
DE
CA
DE
CA
DE
CA
DE
CA
DE
CA
DE
CA
DE
CA
DE
DE
CA
CO
22
2
21
1
19
1
19
2
19
2
21
2
19
2
20
2
21
2
18
2
19
HC
.2
.7
.3
.7
.6
.2
.7
.7
.4
.1
.6
.3
.7
.7
.1
.5
.0
.3
.6
.1
.8
19-22
1.2-2.7
9
4
8
2
7
3
8
4
8
4
8
3
8
4
8
4
8
3
7
3
7
7
.5
.0
.8
.9
.3
.4
.0
.0
.9
.5
.7
.8
.0
.0
.5
.7
.6
.5
.3
.3
.4
-10
2.9-4.7
NO
15
13
11
9
12
11
9
12
13
9
11
NO?
.7
.17
.3
.03
.4
.12
.9
.06
.1
.05
.4
. 17
.9
.06
.2
.08
.9
.08
.8
.06
.2
10-16
0.03-0.17
1
3
2
2
1
1
2
2
3
2
2
1.8
.8
.03
.6
.03
.0
.10
.6
.05
.9
.02
.8
.13
.6
.05
.3
.02
.2
.04
.5
.04
.8
-3.6
0.02-0.13
S02
1.4
.42
2.8
.70
.91
.63
2.3
.94
1.2
.72
.91
.67
2.3
.94
1.6
.73
2.8
.66
2.1
.56
1.9
0.9-2.5
0.04-0,
TSP
7.9
5.9
6.5
6.9
5.3
7.2
6.9
7.0
6.9
7.3
6.4
5.3-
3 7.9
.94
AB=acute bacterial; AV=acute viral; SB=subacute bacterial;
CB=chronic bacterial; DE=diesel exhaust; CA=clean air.
All data as ppm by volume except TSP, which are as mg/m^.
782
-------
£2
w
8
X
H
<
R<
W
£C
H
H
CO
S
X
X
w
j
w
CO
t— 1
o
0!
O
1-1
CU
M
•o
CU
ra
•H
•o
«
^
5
•a
1-1
j-
c
o
^
Cfl
O
1-1
c
o
„
s
"^^
TO
C
00
o
U
(TJ
&J
o
a
X
w
•a
ID
0
O
o^
0
o
o
o
(N
o
in
a.
a
t/i
(N
U
3
U
\D
O
-*
X
O f-J
,_,
M yr
> >
z 3
vO (N
(S (N
O O
~3 X
O W H- <
O
3 S
m
0
(M
O
^
13
OJ c
§c
S3
in
in
£
O
O
O
o
m
O
(N
O
m
in
0
(N
O
xO
m
m
m
in
u
to
0
_,
o
o
w
3
>
cn
*
N
o
O
00
0
o
o
m
O
TO
•o
m
-*
0
o
o
m
•O
^
1
~*
*
1—1
in
O
CN
<0
"O
tn
T3
%
O
0
CU
o
o
o
o
m
3
5
c
o
o
v£>
1
O
—*
X
^D
,_,
z.
>
g
0
o
o
o
0
vO
"5
N
c
01
3
14-1
c
•H
w
H
,
kJ
"^
TO
u
O
C l <
-------
X
o
o
o;
O
Q_
X
O
CL.
l/l
UJ
Qi
O-
Q_
r— QJ
0 I—
4-1 >
On
•r- CL
4-1
4-1 OJ
CO X
•K
QJ
4—
4-
UJ
TD
n3
QJ
QJ
CL
on
O
<«
QJ
S_
on
O
CL
X
UJ
E
CL
CL
ojLn
^. OJ
E
c CL
i— CL
QJ ^J-
O
f ; —
>- c
(_) CL
=C CL
OJ
s_ on
r— I
03 O
C 4-1
QJ O
t-J • — •
c
QJ
en
o
_c
4-1
"3
D-
QJ C
1- O
Z3 •!—
on 4-1
O "3
CL t_
x z:
1 i 1 Q
0
CL CM
- CMC
•• <_> 0 O
T3 > II
0 CL
O eC =C
n c 1 f i
°
0
0
r—
o
01
1
0
X
to
c£
o
0
un
01
Ln
, —
O
o
•
o
LO
LO
O
CL
QJ
i-
4—1
-C
kO
4—
C
CD
on
O
CM
O
O
0
o
un
Ln
•
O
A
un
i —
£
o
OJ
Ln
A
o
in
'~,
O
cC
0
0
m
o
CO
on
13
^_
• i —
QJ
.4-1
Zj
U
CO •
•" — ojro
c_) o o ^a-
o > o
n n
< CLCC CL
c_j co
OJ
O
0
o
V
o
o
•-
4-1
CL
QJ
t ^
00
i —
QJ
— 2
<_>
n3
_Q
rs
1/1
O 4-1
S-
o
_c E
QJ 4-1
i — C
O QJ
i- O
(_) S_
03 QJ
CL
II
QJ
O "3
C\J 4->
784
-------
en
w
Q
D
H
33
X
W
W
CO
W
f— <
Q
Q
2
W
z
o
z
O
H
h-l
>
O
Id
ft,
X
M
a
T)
c
03
C
O
in
•H
03
(X
E
O
U
>1
-H
u
H
X
o
H
X
V)
V)
O
u
X
W
5^
t_l
• I-J
TO
*-*
O
2:
c
o
•r-<
O
o
c
r-l
03
l-i
0
e-s
Q
•o
CJ
V)
o
(X
X
W
u
4_t
•H
H
03
X
l-i
O-i
03
U
•'-'
tfl
TO
CO
,_,
M
2
*"•*
*— <
w
0
V.
c
0
u
<£
^?
c
o
r*
03
l-i
O
G
3
W
O
CL
X
m r-
o o
V 0
,
O U T3
CN
-^ sr
O O
u
A
t— I
t—l
h-t ^
2 H-.
l (
-J
m
m
m
0
CO
CO
^T
CN
^D
ON
^n
^r
CM
O
ro
_-j.
T3
r^
CJ
4-1
3
U
TO
X
W
U)
3
nj >K,
X TD
X 3 ^
CJ >^ 4-» r-H
r-l CO
--i -a i
o *-*
«X O UJ
0 —1 Q
HO -H
-> t-t >
o <; -
O --. W M
ON
0
O
<
o
m
^
1
1
0
CM
0
CM
^J
O
O
CNJ
0
in
O
CM
^
_c
CM
O
J-J
U
<
^O CN
CM O
o o
^J
o <
o o
A ,_,
1— I ^
< - A
O 2 M
m
-^
0
m
m
0
CM
^
O
CM
O
X
(-J
^
^
^s
2
CM
CN
in
06
ro
•*£>
o
OO
"^
in
in
,_ <
•—i
CO
ON
U
1C
CO
o
3
(j
TO
3
C/l
O
o
X
<£
o
A
1— i
2
^
CO
-------
Session V
MUTAGENIC AND CARCINOGENIC POTENCY OF EXTRACTS
OF DIESEL AND RELATED ENVIRONMENTAL EMISSIONS
Chairman:
Dr. Stephen Nesnow
Mutagenic and Carcinogenic Potency of Extracts of Diesel and
Related Environmental Emissions: Study Design, Sample
Generation, Collection, and Preparation.
Huisingh, J. L., R. L. Bradow, R. H. Junyers, B. U.
Harris, R. B. Zweidinger, K. M. Gushing, B. E. Gill,
and R. E. Albert.
Mutagenic and Carcinogenic Potency of Diesel and Related
Environmental Emissions: Salmonella Bioassay^
Claxton, Larry D.
Mutagenic and Carcinogenic Potency of Extracts of Diesel and
Related Environmental Emissions: In Vitro Mutagenesis and
DNA Damage.
Mitchell, Ann D., Elizabeth L. Evans, Mary Margaret
Jotz, Edward S. Riccio, Kristien E. Mortelnians, and
Vincent F. Simmon.
Mutagenic and Carcinogenic Potency of Extracts of Diesel and
Related Environmental Emissions: In Vitro Mutagenesis and
Oncogenic Transformation.
Casto, Bruce C., Georye G. Hatch, Shiu L. Huatiy,
Joellen L. Huisingh, Stephen Nesnow, and Michael
D. Waters.
786
-------
Session V
(Continued)
hutagenic and Carcinogenic Potency of Extracts from Diesel
Related Environmental Emissions: Simultaneous Morphological
Transformation and Mutagenesis in BALB/c 3T3 Cells.
Curren, R. U., R. E. Kouri , C. h. Kirii, and L. M.
Schechtman.
hutagenic and Carcinogenic Potency of Extracts of Diesel and
Related Environmental Emissions: Two-Stage Carcinogenesis
in Skin Tuiiior Sensitive Mice (SENCAR).
Slaga, T. J., L. L. Triplett, and Stephen Nesnow.
Nutagenic and Carcinogenic Potency of Extracts of Diesel and
Related Environmental Emissions: Summary and Discussion of
the Results.
Nesnow, Stephen and Joellen L. Huisingh.
787
-------
MUTAGENIC AND CARCINOGENIC POTENCY OF
EXTRACTS OF DIESEL AND RELATED ENVIRONMENTAL EMISSIONS:
STUDY DESIGN. SAMPLE GENERATION,
COLLECTION, AND PREPARATION
J. L. Huisingh1, R. L. Bradow1, R. H. Jungers1,
B. D. Harris1, R. B. Zweidinger1, K. M. Gushing2,
B. E. Gill3, R. E. Albert1,4
!0ffice of Research and Development
U. S. Environmental Protection Agency
Research Triangle Park, North Carolina
2Southern Research Institute
Birmingham, Alabama
30ak Ridge National Laboratory
Oak Ridge, Tennessee
York University Institute of Environmental Medicine
New York, New York
ABSTRACT
A major diesel emissions research program has been initiated
by the U.S. Environmental Protection Agency to assess the
human health risk associated with increased use of diesel
automobiles. This program is intended to establish the
mutagenic and carcinogenic potency of complex organics
associated with diesel particles as well as comparative
particle-bound organics from other environmental emissions
for which human epidemiological data are available.
The mobile source samples selected for this study were col-
lected from a heavy-duty diesel engine, a series of light-
duty diesel passenger cars, and a gasoline catalyst auto-
mobile. The comparative source samples incorporated into
788
-------
the study were cigarette smoke condensate, coke oven emis-
sions, roofing tar emissions, and benzo(a)pyrene. The
samples were tested using three mutagenic assays and four
carcinogenic assays as prescribed by a test matrix.
This report describes the study design, particle generation,
and sample collection and preparation. A brief summary of
the bioassays is also included.
ACKNOWLEDGMENTS
The authors wish to acknowledge J. Smith for coordinating
and managing the interlaboratory effort to obtain the com-
parative samples; S. Tejada, F. King, H. Becker, W. Ray,
and the Northrop Services staff for assistance in genera-
tion and collection of the diesel particles; J. Sturm
for generation of the VW Rabbit diesel sample; T. Pugh for
assistance in collection of coke oven samples; W. Griest
and R. Jenkins for assistance with the cigarette smoke con-
densate; A. Fowler for assistance in collection of the roof-
ing tar samples; F. Butler, C. Rogers, and F. Toth for
assistance in the samples extractions; L. King for assis-
tance in coding and shipping the samples; and the Northrop
Services' Technical Services staff for assistance in edit-
ing and preparing this report.
INTRODUCTION
The complex mixture of incomplete combustion organics
associated with diesel particle emissions has been shown
to be mutagenic in short-term bioassays (1). These organics
contain known carcinogenic polycyclic aromatic hydrocarbons,
as well as highly mutagenic polar neutral (oxygenated poly-
cyclic) compounds. A variety of combustion emission pro-
ducts have been recognized as human carcinogens, including
chimney soot (2) and coke oven emissions (3). At present,
there is no direct evidence that diesel emissions are human
carcinogens or that human exposure to environmental con-
centrations of diesel emissions will increase the incidence
of cancer in exposed populations. The expected increase
in the use of diesel-powered automobiles and resulting in-
creased ambient air concentrations of diesel emissions have
stimulated a major diesel emissions research program being
conducted by EPA's Office of Research and Development (4).
The information generated from this research will be used
to assess the human health risk associated with increased
use of the diesel engine.
Diesel exhaust is a mixture of particles (carbonaceous soot)
and gases. The principal gases (hydrocarbons, nitrogen
oxides, sulfur oxides, and carbon monoxide) are similar to
789
-------
those currently emitted by gasoline engines and other com-
bustion sources. The particles emitted from diesel engines
differ significantly in both quantity and composition from
gasoline particle emissions. Diesel cars emit 30-to-100
times more particulate matter than gasoline catalyst cars.
Gasoline particulate emissions are primarily sulfur com-
pounds, while diesel particles are composed of carbonaceous
soot with a large amount (5 - 50%) of complex organics ad-
sorbed to the surface. This study was designed to determine
the mutagenic and carcinogenic potency of these complex
organics compared to particle-bound organics from other
environmental emissions. This report describes the study
design, particle generation, and sample collection and pre-
paration for biological studies. The results and a summary
of the relative mutagenic and carcinogenic potency of these
organics are included elsewhere in this volume.
STUDY DESIGN
The objective of this study was to determine the relative
mutagenic and carcinogenic potency of the extractable or-
ganics from diesel emissions and other emissions for which
human epidemiological data are available.
Epidemiological studies have shown that a dose-response
relationship exists between exposure to coke oven emissions
and human lung cancer rates (5, 6, 7, 8). Increased in-
cidence of cancer has also been shown in epidemiological
studies of humans exposed to roofing tar emissions (9) and
cigarette smoke (10). Environmental exposure to various
hydrocarbon combustion products, measured by the benzo(a)-
pyrene (BaP) concentration, has also shown a highly con-
sistent carcinogenic potency in different epidemiological
studies for coke oven, roofing tar, and gas works emissions
(11). Coke oven emissions, roofing tar emissions, and
benzo(a)pyrene were, therefore, selected as comparative
sources for this study. Cigarette smoke condensate (CSC)
was selected as an additional comparative sample because
of the large number of animal studies that have been con-
ducted with this source (12).
The mobile source samples selected for this study included
a heavy-duty diesel engine (Caterpillar 3304), a series of
light-duty diesel passenger cars (Datsun Nissan 220C, Olds-
mobile 350, and Volkswagen turbocharged Rabbit), and a gaso-
line catalyst car (Fort Mustang II).
The design of the sample and bioassay test matrix is shown
in Table 1. The mutagenesis bioassays, selected to detect
gene mutations, chromosomal effects and DNA damage, were
all short-term in vitro assays performed in microbial or
mammalian cells. The carcinogenesis bioassays included both
790
-------
>_
Q
1—
oo
UJ
cj
Qi
O
oo
UJ
r— 1
1—
c3H
rv*
^£
r\
s:
o
o
O
2^
^
— 1
UJ
OO
LU
r— t
Q
Qi
O
U.
X
» — i
i
13
0 M-
00 O
O
0} Oi
+3
03 QJ
03 O
Q. O
0
O
00
<->
Q)
•r- Z
'o i—
CO C/1
03 Z3
CO CD ^~
QJ
O
S-
33
O CQ
CO ei;
fV
QJ 3
i — ;>
-Q
O
E". oo
Q
|
i— O
QJ
CO
QJ
•t"~ -*-"^
Q
CO
Q)
1—
XXX
x: x x
XXX
XXX
x x
x x
XXX
XXX
X X
c
o
• 1 —
4->
03
S-
S-
QJ
CO _Q
c: 03
o
•F— 1
4-> O3 QJ
ft3 E en
-t-J O
•i~ OJ 03
4-) -O r—
03 O i —
E to E ••-
O CO 03 CO
H- QJ E C
CO C O •!-
c ai c
03 Cn QJ . —
i- 0 T3 QJ
4-) C 03 03
•1 — f"~
CO CJ U >, O
•t- -t- S- S_ 03
CO C 03 03 S-
QJ QJ O C 4->
c: cn o 03
QJ O C E i-
cn u -i- r— 4->
0 C ^ =S C
C O OO D_ •— i
o
S-
03
X
X
X
>^
CD
O
1— —
O QJ
•r— i —
E -Q
QJ 03
T3 r—
Q- 03
LU >
03
E
u:
791
-------
in vitro mammalian cell assays and in vivo bioassays. The
in vivo assays included skin carcinogenesis (Senecar and
C57 black mice), pulmonary adenoma formation (strain A mice),
and intratracheal instillation (hamsters).
SAMPLE GENERATION AND COLLECTION
The mobile and comparative source samples were provided for
this study by the organizations shown in Table 2. Sample
collection techniques are summarized in Table 3.
TABLE
SAMPLE SUPPLIERS
Sample source
EPA
Gasoline MUSTANG R. Bradow/ESRL
Comparative CSC J. Smith/HERL
sources Coke B. Harris/IERL
Roof tar B. Harris/IERL
Other
Diesel
CAT
NISSAN
OLDS
VW RAB
R. Bradow/ESRL
R. Bradow/ESRL
R. Bradow/ESRL
R. Bradow/ESRL
J. Sturm/DOT
W. Griest/ORNL
K. Cushing/SRI
ESRL = Environmental Sciences Research Laboratory, DOT =
Department of Transportation, HERL = Health Effects
Research Laboratory, ORNL = Oak Ridge National Laboratory,
IERL = Industrial Environmental Research Laboratory,
SRI = Southern Research Institute
TABLE 3. SAMPLE COLLECTION TECHNIQUES
Sample source
Sampling apparatus
Media
Diesel
CAT
NISSAN
OLDS
VW RAB
Dilution tunnel
Dilution tunnel
Dilution tunnel
Dilution tunnel
Pallflex T filter
Pall flex T filter
Pallflex T filter
Pallflex T filter
Gasoline
MUSTANG Dilution tunnel
Comparative CSC
sources
Coke
Refrigerated con-
denser
Massive Air Volume
Roof tar Baghouse
Pall flex T filter
Acetone
ESP plates
Teflon filter bags
792
-------
DIESEL AND GASOLINE
The mobile source participate emission samples used in this
study were generated from the sources shown in Table 4. The
diesel engine and vehicles were all operated on the same lot
of No. 2 diesel fuel. The gasoline vehicle was operated on
unleaded gasoline at a richer-than-normal stoichiometry in
order to produce sufficient sample volume for limited
studies. The particle emission rates are shown in Table 5.
TABLE 4. MOBILE SOURCE SAMPLE DESCRIPTIONS
Sample
Diesel
source
CAT
NISSAN
OLDS
VW RAB
Description
Caterpillar 3304
Nissan Datsun
220C
Oldsmobile 350
Volkswagen
turbocharged
Rabbit
Fuel
Diesel No.
Diesel No.
Diesel No.
Diesel No.
2
2
2
2
Driving
cycle
Mode II
HWFET
HWFET
HWFET
Gasoline MUSTANG 1978 Mustang II- Unleaded HWFET
302* gasoline
*V-8, equipped with catalytic converter and exhaust gas
recirculation (EGR)
TABLE 5. PARTICLE EMISSION RATES FOR MOBILE SOURCES
Sample
Diesel
source
CAT
NISSAN
OLDS
VW RAB
Emission
9/hp/h
0.72
rate
g/nri
0.33
0.52
0.18
Gasoline MUSTANG
0.0053
To simulate actual driving patterns, the vehicles (except
for the Caterpillar) were operated on a chassis dynamometer,
using the highway fuel economy test cycle (HWFET): average
speed of 48 miles per hour in 12.75 minutes over 10.24
miles. The Caterpillar engine was mounted on an engine
dynamometer and operated at Mode II using a steady-state
operation of 2200 rpm and an 85-pound load. All of the
mobile source particulate emission samples, except those
793
-------
from the VW Rabbit, were generated in the chassis dynamo-
meter and engine dynamometer facilities of EPA's Environ-
mental Sciences Research Laboratory in Research Triangle
Park, North Carolina. The VW Rabbit samples were provided
through an interagency agreement with the Department of
Transportation.
Participate samples were collected using dilution tunnel
sampling technique (13). The emissions were diluted to
obtain realistic samples of particle-bound organics. Be-
cause of the high temperatures (>200° C) in the tailpipe
of an operating diesel engine, organic materials are gener-
ally in the gas phase; thus, soot filtered at these tempera-
tures would contain very little extractable organic material.
In the ambient air, however, particle and gaseous exhaust
is quickly cooled and diluted; as the overall temperature
is reduced, carbon particles begin to adsorb organic
material. The dilution tunnel technique simulates this
process.
The heavy-duty Caterpillar engine exhaust was split before
dilution in the tunnel to ensure that exhaust volumes did
not exceed the capacity of the dilution tunnel system. The
total exhaust from the passenger cars was diluted approxi-
mately 10-fold with filtered air prior to the sampling.
After dilution, the particulate matter was collected by
filtration in large-scale fiber samplers, using Teflon-coat-
ed glass Pallflex T 68-20 filters at a flow rate of 100 cfm.
CIGARETTE SMOKE CONDENSATE
The CSC was generated and collected by the Tobacco Smoke
Chemistry Group of the Oak Ridge National Laboratory (ORNL)
Analytical Chemistry Division for the Chemical Repository
Program. It should be noted that humans are actually ex-
posed to smoke particulate matter, not condensate, and that
there may be significant chemical differences between the
two. The cigarettes chosen for making the CSC were the
Kentucky Reference 2R1, a remake of the widely referenced
1R1. The cigarettes are 85-mm (king size), non-filter,
typical of U. S. cigarettes consumed from 1962 - 1966.
The CSC generator loads, lights, smokes, and ejects 2,000
cigarettes per hour. A maximum of 10 puffs per cigarette
is taken at a rate of 1 puff (2 sec, 35 cc) per minute. As
the cigarettes, which are in a large rotating turret, pass
smoking ports, a vacuum draws a puff from each of 40 ciga-
rettes in different stages of consumption. The smoke is
mixed and drawn through traps, which are partially filled
with acetone and immersed in a refrigerated, dry ice-
isopropanol bath.
794
-------
For this study, approximately 50,000 cigarettes were con-
sumed to produce approximately 1 kg of CSC. The CSC was
removed by rinsing the trapping system with a minimal amount
of acetone.
COKE OVEN
The coke oven emissions were collected by sampling on top of
a coke oven battery at Republic Steel in Gadsden, Alabama,
about 60 miles northeast of Birmingham. Two Massive Air
Volume Samplers (MAVS) were positioned side-by-side on top
of the north end of the coke oven battery. Due to local
wind conditions, various types of aerosols were sampled.
The prevailing wind direction was north and north-northwest
— from the direction of Gasden; thus, as unknown, but signi-
ficant, portion of the emissions sampled may have been from
the urban environment.
Initially, the electrostatic precipitator (ESP) plates coat-
ed with coke oven ambient emissions were rinsed in methylene
chloride. Later, however, the impactor and ESP plates were
dry-scraped. A relatively small amount of material (about
200 g) was recovered in more than 2100 hours of sampling,
indicating that the prevailing wind direction decreased the
concentration of suspended dust at the site.
Since the amount of ambient coke oven sample obtained was
not adequate to complete all of the in vivo bioassays, an
additional large sample of coke oven mains was obtained.
The collection and bioassay of this sample will be reported
later.
ROOFING TAR
The roofing tar emissions were generated and collected by
the Process Measurement Branch of EPA's Industrial Environ-
ment Research Laboratory in Research Triangle Park. The
North Carolina Department of Transportation provided the
tar pot and the pitch-based tar for these studies. The tar
pot was a conventional unit with an external burner. A
6-foot stack extension was added to the test unit to prevent
the burner from entering the sampling hood, which was mount-
ed over the tar pot. A Teflon-coated aluminum pipe led
from the hood to a small baghouse. Special nonreactive
Teflon filter bags were fitted in the baghouse to collect
the particulate emissions.
The sampling was conducted for 8 hours, during which the tar
temperature was held between 360° - 380° F to maintain the
tar in liquid form. Particulate matter was collected from
the Teflon filter bags, and by scraping the collector duct
and pot bottom.
795
-------
SAMPLE PREPARATION
The mobile source and comparative emission samples, except
the CSC, were soxhlet-extracted with dichloromethane (DCM)
to remove the organics. DCM was found to be the most effec-
tive of a series of solvents in removing mutagenically
active organics from diesel particles in preliminary studies
using the Salmonella typhimurium mutagenesis bioassay.
Soxhlet-extraction of the mobile sources particle-laden fil-
ters was accomplished by placing 12-16 filters, without
thimbles, in a 2.3 liter side-chamber extractor for 48 hours.
The Teflon filter bags containing roofing tar emissions were
cut into sections and soxhlet-extracted for 72 hours (until
clear). The coke oven emissions scraped from the ESP plates
were soxhlet-extracted in thimbles in a standard size extrac-
tor for 24 hours.
The soxhlet-extracted organics were filtered using Teflon
millipore filters (0.2 ym) to remove particles. Large
volumes of solvent (mobile source samples) were concentrated
to a smaller volume by rotary evaporation under reduced pres-
sure. Aliquots were evaporated to dryness under nitrogen
and stored frozen in the dark. The percent extractable and
benzo(a)pyrene analysis results are given in Table 6.
TABLE 6. RESULTS OF EXTRACTION AND BENZO(A)PYRENE ANALYSIS
Sample source Extractable Benzo(a)pyrene
matter ng BaP ng BaP
percent mg extract mg particulate
Diesel
CAT
NISSAN
OLDS
VW RAB
26-27
4-8
12-17
18
2
1173
2
26
0.5
96.2
0.4
4.6
Gasoline MUSTANG 39-43 103 44.1
Comparative CSC <1
sources Coke 5-10 478 31.5
Root tar >99 889 889
The CSC, after being drawn through the refrigerated conden-
ser, was removed from the condenser by rinsing with a minimal
amount of acetone. The acetone suspension was then placed
on a rotary evaporator and, at reduced pressure and 30° C,
the semi-volatiles were removed; evaporation was continued
until a constant weight was achieved. The tar content of
the CSC-acetone suspension was adjusted to require levels by
796
-------
the addition of appropriate amounts of acetone. The ace-
tone was removed from the necessary samples by evaporation
in a vacuum desiccator over Drierite. This treatment removed
the acetone solvent, but most of the CSC water remained with
the sample.
The samples prepared as described above, were furnished to
the Health Effects Research Laboratory in Research Triangle
Park, North Carolina, (HERL-RTP) by the suppliers noted
earlier. At HERL-RTP, the samples were coded, using a dis-
tinguishing number for each source and then assigning random
numbers to each sample, creating a blind experiment. The
coded samples were then aliquoted according to the bioassay
requirements.
All shipments were via air express, on dry ice, according
to the transportation procedures for potentially hazardous
biologicals/chemicals specified by the Department of Trans-
portation. To ensure stability, the samples were stored in
the dark at or below -30° C and bioassayed soon after re-
ceipt.
Prior to conducting the bioassay(s), the investigator(s)
added dimethyl sulfoxide (DMSO) to the samples to achieve
the desired concentration (mg sample/nil). The same DMSO
solvent was used as a control. For some specified assays
(i.e., skin carcinogenesis and oncogenic transformation),
acetone was used as the solvent and the control.
BIOASSAYS
The mutagenesis bioassays were selected to detect gene
mutations, chromosomal effects, and DNA damage. The results
of the mutagenesis test matrix reported in this volume are
for the following bioassays:
• reverse mutation in Salmonella typhimurium
• forward mutation in L5178Y mouse lymphoma cells
• forward mutation in Balb/c 3T3 mouse embryo fibro-
blasts
• forward mutation in Chinese hamster ovary cells
• mitotic recombination in Saccharomyces cerevijsiae
• DNA breakage in Syrian hamster embryo cells
0 sister chromatid exchange in Chinese hamster ovary
cells.
797
-------
The carcinogenesis bioassays were selected to provide data
as early as possible from in_ vitro and short-term TJT_ vivo
assays (e.g., skin tumor initiation). The longer term in
vivo assays (e.g., complete skin carcinogenesis and intra-
tracheal instillation) will allow scoring and pathological
examination for carcinomas. The results of the carcinogenic
testing, completed to date and reported in this volume, in-
cluding the following bioassays:
• oncogenic transformation in Balb/c 3T3 cells
• viral enhancement of transformation in Syrian
hamster embryo cells
• skin tumor initiation in Sencar mice.
The bioassays were performed, if possible, in such a way
that dose-response data would demonstrate a positive in-
crease (slope) above spontaneous levels for at least three
doses. This usually requires a preliminary toxicity test
over a wide dose range followed by a test conducted at
5-to-7 doses spaced over half-logs. Where such dose-response
data were obtained, the initial linear slope of the response
curve was determined and used to provide comparative potency
rankings of the samples. In several assays (e.g., the Balb/
c 3T3 cells), dose-response data were not obtained and the
lowest effective (positive) dose tested was determined. Dif-
ficulties in obtaining good dose-response data may be due,
in part, to difficulties in delivering increasing doses of
the sample to the cells due to solubility problems.
The short-term bioassays generally included positive and nega-
tive (solvent) controls. The assays were repeated and, in
some cases, run in more than one laboratory. Samples were
decoded after submission of the final results. The papers
which follow describe the results determined by each bioassay
laboratory. The final paper of this session summarizes all
of the results to date (14).
REFERENCES
1. Huisingh, J., R. Bradow, R. Jungers, L. Claxton, R.
Zweidinger, S. Tejada, J. Bumgarner, F. Duffield,
M. Waters, V. F. Simmon, C. Hare, C. Rodriguez, and
L. Snow. Application of Bioassays to the Charac-
terization of Diesel Particle Emissions. In: Applica-
tion of Short-Term Bioassays in the Fractionation and
Analysis of Complex Environmental Mixtures, M. Waters,
S. Nesnow, J. Huisingh, S. Sandhu and L. Claxton, eds.
Plenum Press, New York, 1978. pp. 383-418.
798
-------
2. Kipling, M. D. Soots, Tars and Oils As Causes of
Occupational Cancer. In: Chemical Carcinogens,
C. Searle, ed. ACS Monograph 173, American Chemical
Society, Washington, D.C., 1976. pp. 315-323.
3. U.S. Environmental Protection Agency. Preliminary
Report on Population Risk to Ambient Coke Oven Exposure.
Carcinogen Assessment Group, Washington, D.C., 1973.
4. U.S. Environmental Protection Agency. The Diesel Emis-
sions Research Program. EPA-625/9-74-004, Center for
Environmental Research Information, Cincinnati, Ohio,
1979.
5. Lloyd, J. W., F. E. Lundin Jr., C. K. Redmond. Long
Term Mortality Study of Steelworkers. IV. Mortality
by Work Area. J. Occup. Med., 12:151-157, 1970.
6. Lloyd, J. W. Long Term Mortality Study of Steelworkers.
V. Respiratory Cancer in Coke Plant Workers. J. Occup.
Med., 13:53-68, 1971.
7. Redmond, C. K., A. Ciocco, J. W. Lloyd, and H. W. Rush.
Long Term Mortality Study of Steelworkers. VI. Mortali-
ty from Malignant Neoplasms among Coke Oven Workers.
J. Occup. Med., 14:621-629, 1972.
8. Mazundar, S., C. K. Redmond, W. Sollecito, N. Sussman.
An Epidemiological Study of Exposure to Coal Tar Pitch
Volatiles Among Coke Oven Workers. J. Air Pol. Control
Assoc., 25:382-389, 1975.
9. Hammond, E. C., I. J. Selikoff, P. L. Lawther, and H.
Seidman. Inhalation of Benzo-a-pyrene and Cancer in
Man. Ann. N.Y. Acad. Sci., 271:161-124, 1976.
10. Wynder, E. L. and D. Hoffman. Tobacco and Tobacco Smoke.
Academic Press, New York, 1967.
11. U.S. Environmental Protection Agency. Preliminary Re-
port on POM Exposures. Carcinogen Assessment Group,
Washington, D.C. , 1978.
12. Gori, G. B., ed. Toward Less Hazardous Cigarettes,
Volumes 1-3. (NIH) 76-905, (NIH) 76-111), (NIH)
75-1280, National Cancer Institue, Washington, D.C.,
1976.
13. Bradow, R. L. Chemical Composition of Diesel Exhaust
Particles. This volume, 1980.
799
-------
14. Nesnow, S. and J. Huisingh. Mutagenic and Carcinogenic
Potency of Extracts of Diesel and Related Environmental
Emissions: Summary and Discussion of Results. This
volume, 1980.
800
-------
MUTAGENIC AND CARCINOGENIC POTENCY OF DIESEL AND RELATED
ENVIRONMENTAL EMISSIONS: SALMONELLA BIOASSAY
Larry D. Claxton
Genetic Toxicology Division
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
ABSTRACT
Due to the expected increase in the percentage of diesel
vehicles in the United States, the Environmental Protection
Agency must evaluate the health effects associated with
exposure to diesel emissions. Respirable particles from a
variety of combustion sources have the potential of being
carcinogenic and mutagenic. The objective of these studies
was to determine the relative biological activity of the
organic material adsorbed on these particles in vn vitro
mutagenesis bioassays. The organic extracts from the fol-
lowing series of emission sources were quantitative bio-
assayed in the Salmonella assay for mutagenic activity: (1)
a light-duty Oldsmobile diesel 350 engine; (2) a heavy-duty
Caterpillar diesel engine; (3) a light-duty Nissan engine;
(4) a Volkswagen Rabbit diesel engine; (5) cigarette smoke;
(6) roofing tar; (7) coke oven; and (8) a gasoline catalyst
Mustang. This paper provides a comparison of these sources
within the Salmonella bioassay and also demonstrates how
bacterial systems can be used as a quality assurance measure
in i_n vivo testing.
801
-------
INTRODUCTION
Bacterial systems have recently been used to demonstrate
that organic mixtures from combustion processes, including
diesel engines, are mutagenic (1,2). This study demonstrates
a means of comparing the mutagenic potential of various
sources with microbial test methods. Included within the
problem of source to source comparison are variables within:
(1) collection methods, (2) methods of chemical extraction,
(3) proportion of mutagenic components within the mixture,
(4) the amount of "masking" organics that are bactericidal,
(5) the metabolizing ability of the indicator organism and
exogenous activation systems, and (6) the sensitivity of the
bioassay. Although these variables do exist, they can be
controlled to some extent. In view of these considerations,
this paper demonstrates the utility of comparing organics
from various combustion sources. This study also demonstrates
how bacterial systems can aid in the identification of
active components and can be used as a quality assurance
measure in i_n vivo testing.
MATERIALS AND METHODS
The assay system used in this study was the Salmonella
typhimurium plate incorporation test as developed by Ames e_t
al. (3).Four mutant strains of Salmonella typhimurium
ITA98, T 100, TA1535, and TA98-FR1) were used in this study.
The nitroreductase deficient strain TA98-FR1 was obtained
from Dr. Herbert Rosenkranz at the New York Medical College,
Valhalla, N.Y., and the other strains were obtained from Dr.
Bruce Ames at the University of California, Berkeley. All
samples were tested in TA98, TA1535, and TA100. The diesel
and cigarette smoke samples were assayed also with TA98-FR1.
The protocol described by Ames et al_. (3) was employed with
minor modifications described below. All samples were
tested with and without Aroclor-induced S-9 as described by
Ames (3). Two minor modifications to the protocol were
used. First, the minimal level of histidine was added to
the plate media rather than to the overlay. Secondly,
plates were incubated for 72 hours rather than 48 hours.
Each plate was counted electronically using an Artek Model
880 automatic colony counter. The positive controls used
for TA98 and TA98-FR1 were 2-nitrofluorene for without
802
-------
activation and 2-anthramine for the S-9 activation control.
For TA100 and TA1535, sodium azide was substituted for the
positive control without activation. All quality assurance
procedures as described by Ames (3) and the Washington
Committee on the Ames Test (4) were followed. In addition,
all experiments were conducted in a laminar flow biological
exhaust hood under yellow lights. When sufficient sample
was present, each extract was tested in triplicate at a
minimum of five doses.
The linear portion of the dose response curve for samples
positive within the test were used to calculate a linear
regression line. The equation of that line was used to
calculate the expected response at 100 ug of organic material
(a dose located within the linear response range). This
value was termed the specific activity and was simply a
convenient method of comparing the responses of different
samples. The specific activity values for each sample are
found in Table 1. In keeping with companion papers (5), the
activity of each sample was compared to the activity of the
Nissan sample by arbitrarily assigning the Nissan sample a
relative value of 100. These comparisons are referred to as
Relative Potency and are found in Table 2.
TABLE 1. SPECIFIC ACTIVITIES AT 100 pg OF ORGANIC MATERIAL
Sample
Mustang
Cigarette
Coke Oven
Roofing Tar
B(a)P
TA98
+S9 -S9
TA100
tS9 -S9
Caterpi 1 lar
Nissan
Oldsmobile
VW
59 3
1367.1
318 7
297.5
Diesel
65 9
1225.2
614.8
399.2
115.2
881.7
169.9
426.0
167.8
1270.1
247.5
641 6
Gasoline
341.9 137.8
Comparative Samples
98.2
251.6
98.7
Neg
164.1
Neg
Control Compound
15202.3* NT
228.0
7
265 6
420.0
26438.0*
196.5
Neg
259.4
Neg
NT
*Extrapolation
NT = Not tested.
Neg = Negative.
803
-------
TABLE 2 RELATIVE POTENCY OF ORGANIC MATERIAL
Sample
Caterpillar
Nissan
Oldsmobile
VW
Mustang
Cigarette
Coke Oven
Roofing Tar
B(a)P
+S9
4.3
100.0
22.3
21.8
25.0
7.2
18.4
7.2
1112.1
TA98
-S9
Diesel
5.4
100.0
50 2
32.6
Gasol ine
11 3
Comparative Samples
Neg
13.4
Neg
Control Compound
* NT
TA100
+59
13.0
100.0
19.3
48.3
25.9
7
30.2
47.6
2997.5*
-S9
13.2
100.0
19.5
50.6
15.5
Neg
20.4
Neg
NT
*Extrapolation.
NT = Not tested.
Neg = Negative.
The description of the sample collection procedures, extraction,
and solvent exchange to DMSO is found in a companion paper (6).
TABLE 3. AVERAGE NUMBER OF REVERTANTS PER PLATE
AT 100 ug OF ORGANIC WITH OLDSMOBILE AND CATEPILLAR
QUALITY CONTROL SAMPLES
Month
Sample:
Olds
Cat
Olds
Cat
AC7:
January
February
March
April
May
June
July
August
-S9
633.33
591.00
628.33
N T
605.33
739 33
746.67
701 00
-S9
170 67
(131 80)*
168.67
N T.
203.33
218.00
200.33
169.33
+S9
318 33
393 33
362 50
N.T
305.00
421.33
366.00
422.33
+S9
115.00
(173.20)*
122 00
N.T
135 67
137.00
126.00
126.33
Calculated
804
-------
RESULTS
This study was divided into three areas: (1) the comparison
of various combustion sources using microbial mutagenesis,
(2) the monthly bioassaying of sample aliquots as a quality
assurance for concurrent mouse skin painting studies (7),
and (3) the use of a nitroreductase deficient strain to
determine if a diesel extract contained active nitrogenated
compounds.
When a mammalian activation system was added, each combustion
source sample was shown to have mutagenic activity (Tables 1
and 2). Cigarette smoke condensate, roofing tar, and benzo(a)
pyrene required metabolic activation in order to obtain a
positive response. The other samples contained direct
acting mutagens. The majority of the activity associated
with the diesel samples was direct acting. The gasoline
automobile sample had direct-acting mutagens; however, the
addition of a mammalian activation system increased the
activity of this sample. All samples were negative with
strain TA1535.
Two samples illustrate the use of microbial bioassays for
quality assurance. The Caterpillar sample and Oldsmobile
sample were tested monthly in order to determine if the
mutagenic activity of organic extracts was altered with
prolonged storage. Table 3 shows the results for the monthly
Caterpillar and Oldsmobile samples tested to date at 100 ug
of organics per plate. Each graph point represents the
average of three plates. No significance variation was
found in the month-to-month test results.
Only two samples, the diesel Nissan and cigarette smoke
condensate samples, provided sufficient material for testing
with the nitroreductase strains. TA98 and TA98-FR1 gave
very similar results with the cigarette smoke condensate;
however, TA98-FR1 gave one-half the response of TA98 with
the diesel sample.
805
-------
DISCUSSION
Each organic sample from the various combustion source
emissions was demonstrated to be mutagenic in the Salmonella
typhimurium mutagenesis assay. This study demonstrates that
emission organics can be compared within bacterial assay
systems and that microbial assays can supply important
supplemental information.
The bacterial assay demonstrated that the mutagenic sub-
stances associated with organic emissions were different
for each type of emission. For example, the cigarette
smoke condensate required activation before mutagenic
activity was seen; however, most of the activity associ-
ated with the diesel samples was direct acting. Secondly,
the range of activity found with the diesel sample spanned
the range of activities associated with the comparative
samples. Next, the lack of response in strain TA1535 sug-
gests that most, if not all, of the activity is due to
polynuclear frameshift mutagens and not to alkylating
agents. The m'troreductase deficient strain demonstrated
a difference between the two samples available for testing.
The responses of TA98 and TA98-FR1 were markedly different
with the diesel Nissan sample but quite similar with the
cigarette smoke condensate. This would suggest that the
particle bound organics from the Nissan diesel contains
inactive nitro compounds that are reduced to an active
form by m'troreductase enzymes which are found in TA98.
The calculation of relative potencies is a useful means of
understanding the range activity available from various
combustion organics. As can be easily seen from Table 2,
the maximum attainable difference within this study was seen
between the Caterpillar and Nissan samples which demonstrated
approximately 25-fold difference. The quality assurance
portion of this study provides evidence that the active
components in the microbial mutagenesis assay are stable
when stored as the extract at -70°C for at least seven
months.
806
-------
REFERENCES
1. Huisingh, 0. et aL Application of Bioassay to the
Characterization of Diesel Particle Emissions. In:
Application of Short-term Bioassays in the Fractionation
and Analysis of Complex Environmental Mixtures, Plenum
Press, N.Y., 1979.
2. Kier, L. D., E. Yamasaki, and 6. N. Ames. Detection of
Mutagenic Activity in Cigarette Smoke Condensates.
Proc. Nat. Acad. Sci. USA, 71(10):4159-4163, 1974.
3. Ames, B. N., J. McCann, and E. Yamasaki. Methods for
Detecting Carcinogens and Mutagens with the Salmonella/
Mammalian-Microsome Mutagenicity Test. Mutation Res.,
31:347-364, 1975.
4. de Serres, F. J., and M. D. Shelby. The Salmonella
Mutagenicity Assay: Recommendations. Science, 203(9):
563-565, 1979.
5. Nesnow, S. and J. Huisingh. Mutagenic and Carcinogenic
Potency of Diesel and Related Environmental Emissions:
Summary and Analysis of Results. EPA Technical Report
(this proceedings).
6. Husingh, J. Mutagenic and Carcinogenic Potency of
Diesel and Related Environmental Emissions: Preparation
and Characterization of Samples. EPA Technical Report
(this proceedings).
7. Slaga, T. S. , L. L. Triplett, and S. Nesnow. Mutagenic
and Carcinogenic Potency of Diesel and Related Environ-
mental Emissions: Mouse Skin Tumorigenesis. EPA
Technical Report (this proceedings).
General Discussion
J. SIAK: Do you have any data on the Oldsmobile or the
Volkswagen on the nitro-reductant strains?
L. CLAXTON: No, we did not have enough samples at that
particular time to go back and do those. We were just
lucky to have a little bit of sample left of those two and
we were hoping to go back with a variety of samples and use
this strain on them.
J. SIAK: You didn't show any data on characterization
of fractions from automobiles. Are you going to do some of
those studies?
J. HUISINGH: I'm sorry that it was not possible to have
all the characterization data completed before the bioassay
data, but we were pushing so hard to get all the samples
generated and all the bioassays performed. Unfortunately,
807
-------
the fractionation data had to be done under contract which
took some time to be negotiated so that those fractions are
just now being prepared. So it will be a little time before
we have all the chemical characterizations data on these
samples. I should say something that I failed to say in my
presentation. The coke oven that was selected by Republic
Steel for us to put that massive volume sampler near was a
clean coke oven and the location was oftentimes in the
opposite direction of the wind blowing, which almost always
blows.We collected that coke oven sample, since it was an
ambient sample, mixed with gas and air, and we had hoped
maybe in six months of sampling, an adequate sample could
be collected for the whole battery of testing including the
intratracheal installation. That just was not possible and
it looked like we would be there for years collecting this
sample. So they had to go back and get a coke oven sample
from the main of the coke oven which is not quite the same
as an ambient emission sample. That sample has now arrived
and will be put through this battery of tests, and it is
that sample that will be used in the intratracheal inhalation.
Later we will report on the comparison between the coke
oven mains and the ambient sample.
D. CHOUDHURY: Was your coke oven sample kept in solution
or as a dry extract?
L. CLAXTON: It was kept as a dry extract and was stored
at minus 80 degrees centrigrade in a Revco freezer.
D. CHOUDHURY: So every time it was tested, it was a
fresh solution?
L. CLAXTON: Yes.
B. BOWARD: Did I understand correctly that each of the
diesel engines you referred to, except for the caterpillar
were in transient. These were chassied down on a test
except for the caterpillar?
L. CLAXTON: Yes.
D. HOFFMANN: I saw your tobacco tar data and compared
it with other laboratories. Since the Ames Test is not 100
percent reproducible, did other laboratories test your
samples?
L. CLAXTON: Yes, we have done this on quite a few die-
sel samples. We have an on-site contractor at my lab-
oratory, Stanford Research Institute, which have run all
very similar samples, and in some cases the same sample,
and they are consistent with each other.
T. BAINES: When you take your data at the 100 microgram
point, does this not include then the spontaneous portion
of the curve?
L. CLAXTON: As we are calculating it, yes, it does.
T. BAINES: How does that influence one's ability to
make or to take the mass emissions and incorporate them
into the results to get comparative data?
L. CLAXTON: I think you would have to adjust this data
differently if you wanted to take the mass emissions and
808
-------
start to do revertents per mile or any such thing as that.
T. BAINES: How would you recommend doing that?
L. CLAXTON: You could calculate this. I think what you
are asking is that since we are including the spontaneous
revertants would not spontaneous influence our total result
in such a way as to make them not comparable. To some
extent it would, but I think you could simply do a net
revertents and then do your slope which would actually be
the same. It is not going to affect results greatly and
you would have to look at each strain individually becuase
the responses by strain are somewhat different.
T. BAINES: Your caterpillar showed remarkably low bio-
activity. Do you know of any reason why that might have
been quite low especially in light of the fact that I think
the other caterpillar, the 3208, was significantly higher
in this activity?
L. CLAXTON: Your guess is as good as mine.
809
-------
MUTAGENIC AND CARCINOGENIC POTENCY OF EXTRACTS
OF DIESEL AND RELATED ENVIRONMENTAL EMISSIONS:
IN VITRO MUTAGENESIS AND DNA DAMAGE
Ann D. Mitchell, Elizabeth L. Evans, and Mary Margaret Jotz
Biochemical Genetics Department
Edward S . Riccio, Kristie-n E . Mortelmans , and Vincent F. Simmon
Microbial Genetics Department
SRI International
333 Ravenswood Avenue
Menlo Park, CA 94025
ABSTRACT
The Saccharomvces cerevesiae D3 recombinogenic assay, the
assay for forward mutagenesis in L5178Y mouse lymphoma
cells, and the sister chromatid exchange (SCE) assay using
Chinese hamster ovary cells were used to evaluate the jin
vitro mutagenic and DNA-damaging effects of eight samples
of diesel engine emissions and related environmental
emissions. The recombinogenic assay was not sufficiently
sensitive for this evaluation, but mutagenicity was
detected in the L5178Y mutagenesis assay following expo-
sures of the cells to all of the emission samples, and DNA
damage in the SCE assay was induced by most of the emission
samples in the presence and absence of metabolic activation.
The observation of positive results in the absence of acti-
vation indicated that the samples contained substances
that were direct-acting mutagens and DNA-damaging agents.
INTRODUCTION
The potential of samples of diesel engine emissions and relat-
ed environmental emissions to induce in vitro mutagenicity
810
-------
and DNA damage were evaluated by three approaches:
the mitotic recombinogenic assay utilizing the yeast Saccharo-
myces cerevisiae D3; the assay for forward mutagenesis at the
thymidine kinase locus in L5178Y mouse lymphoma cells; and
the sister chromatid exchange (SCE) assay using Chinese
hamster ovary cells.
All testing was performed in the absence and in the presence
of an Aroclor 1254-stimulated metabolic activation system
since some carcinogenic chemicals (e.g., of the aromatic
amino type or the polycyclic hydrocarbon type) may be in-
active unless they are metabolized to active forms. In
animals and man, an enzyme system in the liver or other
organs (e.g., lung or kidney) is capable of metabolizing
a large number of these chemicals to carcinogens, and some
of these intermediate metabolites are very potent mutagens
and DNA-damaging agents.
The testing was conducted using coded samples. Upon de-
coding, the samples were identified as three mixtures
having known carclnogenicity—coke oven emission, roofing
tar erission, and cigarette snoke condensate; the emission
of a gasoline (Mustang) engine; an emission from a heavy-
duty diesel (Caterpillar) engine; and emissions from three
diesel automobile engines, identified as Nissan, VW rabbit,
and Oldsmobile.
Saccharomyces cerevisiae D3 Recombinogenic Assay
The yeast J5. cerevisiae D3 is a diploid microorganism hetero-
zygous for a mutation leading to a defective enzyme in the
adenine-metabolizing pathway (1). When grown on medium
containing adenine, cells homozygous for this mutation
produce a red pigment. These homozygous mutants can be
generated from the heterozygotes by mitotic recombination.
The frequency of this recombinational event may be increased
by incubating the organisms with various carcinogenic or
recombinogenic agents. The recombinogenic activity of a
compound or its metabolite is determined from the number of
red-pigmented colonies appearing on test plates (2) .
L5178Y Mouse Lymphoma Ilutagenesis Assay
The L5178Y mouse lymphoma assay (3-5) measures the effects
of chemicals on the forward mutation frequency of the cells
at the thymidine kinase (TK) locus. Heterozygous cells
(TK+/~) are capable of utilizing exogenous thymidine.
However, when the cells are mutated to a homozygous reces-
sive (TK~/~) condition, they utilize neither thymidine nor
thymidine analogs, the latter of which could otherwise
kill the cells competent in TK activity. One effective
analog of thymidine is trifluorothymidine (TFT) (6),
and only mutated cells can gro\r and form colonies
811
-------
in the presence of medium supplemented with TFT. Hence,
the mutagenic activity of a chemical can be determined by
the number of colonies found growing in the presence of
TFT.
Sister Chromatid Exchange Assay
The induction of DNA lesions by chemical mutagens leads to
the formation of sister chromatid exchanges (SCEs), which
may be related to recombinational or postreplicative
repair of DNA damage (7-9) . Exchanges between the two
chromatids of each chromosome are observed in metaphase
cells that have been grown in the presence of bromodeoxy-
uridine (BrdU) for two rounds of replication. Because
DNA replication is semiconservative, these chromosomes
consist of one chromatid in which both DNA strands are
BrdU-substituted and one chromatid in which BrdU substi-
tution is limited to a single strand. These chrom^tids
are stained differentially by the fluorescence-plus-
Giemsa (FPG) technique (10), which permits visualization of
SCEs by the "harlequin" pattern of darkly and lightly stained
chromatid segments produced. The SCE technique is highly
sensitive; its use permits detection of the effects of
direct-acting chemical mutagens (8,9,11) as well as those
that require metabolic activation (12,13) on mammalian
cell chromosomes. Wolff (14) and Hollstein et al. (15).
have reviewed the chemicals that have been shown to increase
SCE frequencies.
MATERIALS AND METHODS
Aroclor 1254-Stimulated Metabolic Activation System
Ames et al. (16) have described the Aroclor 1254-induced
liver metabolic activation system that we used for the re—
combinogenic assays. This reaction mixture consisted of,
for 10 ml: 1 ml of S-9 [1 g (wet weight) of liver to 3 ml
of 0.15 M KC1 buffer]; 0.2 ml of 0.4 1! IfgCl and 1.65 M
KC1; 0.05 ml of 1 II glucose-6-phosphate; 0.4 ml of 0.1 M
NADP; 5 ml of 0.2 M sodium phosphate buffer (pH 7.4); and
3.35 ml of H20. Similar reaction mixtures were used for
the L5178Y mutagenesis and SCE assays except that the S-9
fractions were prepared in sucrose-phosphate buffer;
cofactors, for 10 ml, were 24 mg of NADP and 45 mg of
sodium isocitrate; and the remaining volumes consisted of
culture medium—Fischer ' s medium with 5% heat-inactivated
serum for the L5178Y assays, or McCoy's 5a medium con-
taining 2.5% fetal calf serum,2tnM L-glutamine, and 1%
penicillin-streptomycin solution for the SCE assays.
Saccharomyces D3 Recombinogenic Assay
A stock culture of S_. cereviniae D3 is stored at 4°C.
812
-------
For each experiment, broth containing 0.057, t'gSOi., O.L57
KH2PO<,, 0.45% (NH4)2SCU, 0.35% peptone, 0.5% yeast
extract, and 2% dextrose was inoculated with a loopful
of the stock culture and incubated overnight at 30°C with
shaking.
The in vitro yeast mitotic recombination assay in suspension
was conducted as follows. The overnight culture was centri-
fuged and the cells were resuspended at a concentration of
108 cells/ml in 67 ml! phosphate buffer (pi! 7.4). To a
sterile test tube were added 1.00 ml of the resuspended
culture, 0.50 ml of either the metabolic activation mixture
or buffer, 0.20 ml of the test chemical, and 0.30 ml of
buffer. The solvent used was DI1SO. Several concentrations
of the test emission samples (up to 5X, w/v or v/v) were
tested in each experiment. For the positive control, 1,2,3-
4-diepoxvbutane vas used.
The suspension mixture was incubated at 30°C for 4 hours on
a roller drum. The sample was then diluted serially in
sterile physiological saline, and 0.2 ml of the 10 5 and
10~3 dilutions was spread on plates containing the same in-
gredients as the broth nlus 2.0% agar; five plates were
spread with the 10"5 dilution and three plates were spread with
the 10~3 dilution. The plates were incubated for 2 days at
30°C, followed by 2 days at 4°C to enhance the development
of the red pigment indicative of adenine-deficlent homozy-
gosity. Plates containing the 10~3 dilution were scanned
with a dissecting microscope at 10X magnification, and the
number of mitotic recombinants (red colonies or red sectors)
was recorded. The surviving fraction of organisms was
determined from the total number of colonies appearing on
the plates of the 10~5 dilution.
The number of mitotic recombinants was calculated per
105 survivors. A positive response in this assay is indi-
cated by a dose-related increase of more than 3-fold in the
absolute number of mitotic recombinants per milliliter and
in the relative number of mitotic recombinants per 105
survivors.
Each extract of diesel or related environmental emission
was tested at least twice on separate days. Prior to each
assay, the samples were diluted in DMSO to form a series
of concentrations that, when added to the cell/metabolic
or buffer mixture, yielded the desired set of concentrations.
The first experiment \,as a test over a wide range of concen-
trations to look for toxicity or recombinogenic activity.
If no toxicity or recombinogenicity was observed, the second
experiment was conducted at higher levels.
813
-------
L5178Y Mouse Lymphoma Mutagenesis Assay
L5178Y mouse lymphoma cells, heterozygous for thymidlne kin
ase, TK+/-, were routinely grown as a suspension culture in
Fischer's medium for leukemic cells of mice, supplemented
with 31 yg/ml penicillin (1650 units/mg), 50 pg/ml strepto-
mycin sulfate (Fo), 10% heat-inactivated horse serum, 0.1%
Pluronic F68, and 0.22 mg/kg sodium pyruvate (Flop). The
cells were cleansed of homozygous (TK~/~) cells with metho-
trexate. The cloning medium consisted of F0 with 20% heat-
inactivated horse serum, 0.22 mg/kg sodium pyruvate, and
0.35% Noble agar. The selective medium contained 2.5
TFT.
The positive controls were ethyl methanesulfonate (EMS),
which induces mutagenesis without metabolic activation,
and 3-methylcholanthrene (3--MCA), which induces mutagenesis
only with metabolic activation. The negative (solvent)
controls received 1% DMSO.
The results of dose-ranging assays were used to select
10 concentrations of each diesel emission sample for use
in the mutagenesis assays that reflected cell survival in
suspension culture ranging from 5 to 90% of the control
level. In the mutagenesis assays, duplicate samples were
used for each emission sample dilution and for the negative
and positive controls. For each concentration of emission
sample and .for the controls, 6 x 10s freshly cleansed
L5178Y TK cells were exposed to 100 yl of a 100X
concentration in 10 ml of medium in a sterile 50-ml centri-
fuge tube. The centrifuge tubes were rotated for 4 hours
in a roller drum at 37°C. After the exposure, the cells
were rinsed and then resuspended in 15 ml of F10p in 50-ml
centrifuge tubes for a final density of 4 x 105 cells/ml.
The tubes were rotated in a roller drum for 2 days for
expression of any mutations. Cell growth was monitored
daily, and the cell concentrations were maintained at
3 to 15 x 105 cells/ml during the expression period.
The cells were then seeded in soft agar mediums 3 x 106
cells in cloning medium supplemented with 2.5 yg/ml TFT,
and 600 cells in nonselective cloning medium to determine
viability.
The mutation frequency was calculated by dividing the
number of mutant cells per milliliter of original suspension
culture by the number of viable cells per milliliter of
original suspension culture. Therefore, the mutation
frequency was the ratio of mutant cells to surviving cells
at each test concentration.
814
-------
An acceptable assay was one in which (1) the compound was
tested to the level of approximately 10% total relative
growth, (2) the relative plating efficiency of the solvent
control was 60 to 110%, (3) the mutation frequency of the
solvent control was no higher than 100 x 10~6, and (4) the
mutation frequency of the positive control was within ac-
ceptable limits based on historical data in the laboratory.
The test results were considered positive if the
mutation frequency of at least one concentration was
approximately twice that of the solvent control, or higher.
A dose-response relationship was also a criterion;
however, because sample availability was limited, this
criterion was not always applied. The results were con-
sidered negative if the test was valid, but none of the
treated samples showed a mutation frequency at
least twice that of the solvent control.
Sister Chromatid Exchange Assay
Each SCE assay of the samples of diesel and related environ-
mental emissions was conducted in duplicate, using Chinese
hamster ovary (CKO) cells that were seeded at approximately
10s cells per T-75 flask and grown for 1 to 2 days before
exposure. The cells were grown in plastic T-75 flasks in an
atmosphere of 5% C02 at 37°C in McCoy's 5a complete medium,
which contains 15% fetal calf serum (FCS), 2 mM L-glutamine,
and 1% penicillin-streptomycin solution. The emission sar.:ples
were diluted serially in DMSO to form a series of concentra-
tions that were further diluted Is 100 in culture medium. The
final concentration of DMSO, therefore, was 1%, which is not
cytotoxic.
The positive controls were 10~3 M EMS, a chemical mutagen
that induces SCEs in the absence of a metabolic activation
system, and 10~3 H dimethylnitrosamine (DMtl), a compound
that induces SCEs only with metabolic activation. The nega-
tive control was DMSO diluted in culture medium.
To test the coded samples, we modified the procedures
developed by Perry and Evans (8) and Stetka and Wolff (13).
In the assays performed without metabolic activation, the
cultures were placed in McCoy's 5a complete medium con-
taining 10~5 11 BrdU, the coded sample at the desired final
concentration was added, and the cells were grown in the
dark at 37°C for 21,5 hours. Then the medium containing the
sample dilution and BrdU was aspirated, the cells were
washed two to three times with phosphate-buffered saline
(PBS) at 37°C, and McCoy's 5a complete medium containing
0.4 pg/ml colchicine was added.
In the assays performed with metabolic activation, the cell
cultures were exposed to the test sample in the metabolic
815
-------
activation mixture for 2 hours. (This shorter exposure
time was used to prevent the occurrence of cytotoxic effects
from the metabolic activation mixture.) After exposure,
the medium containing the test sample and the metabolic
activation mixture was aspirated, the cells were washed
two or three times with PBS, and McCoy's 5a complete medium
containing 10~ M BrdU was added. The cultures were then
incubated in the dark for 21.5 hours, after which colchicine
was added to a final concentration of 0.4 ij™/ml.
After 2.5 hours in colchicine, the mitotic cells were har-
vested. At harvest, medium (which might contain dividing
cells) was removed and saved. PBS was'added and the flasks
were shaken approximately 20 times until almost all the mitot-
ic cells were detached. The cell suspension in pooled
medium and PBS was centrifuged, the supernatant was aspirated,
and 4 ml of a hypotonic solution of 0.075 M ICC] was added
to the cells for 7 to 20 minutes at 37°C. The cells were
resuspended in a fixative of mcthanol:acetic acid (3:1)
three times, and air-dried slides were prepared. The cell
density on a slide was monitored using a phase-contrast
microscope, and at least two slides of appropriate cell den-
sity were made from each sample. The slides were stained
by following a modification of the fluorescence-plus-
Giemsa (FPG) technique (10), then examined to determine whether
the metaphases had divided one 'Hi), two (t!2), or three
(H3) times during exposure to Brdt'. The concentrations of
each test sample selected for cytogenetic evaluation were
the five highest concentrations that permitted enough cells
to divide twice so that SCEs could be enumerated. For each
test, two cytogeneticists each analyzed 25 cells from coded
duplicate samples for each of the five concentrations and from
the positive and negative controls. Thus, 50 cells were
evaluated per concentration.
The mean SCE frequency per cell and the mean SCE frequency
per chromosome for each concentration of the test substance
and for the controls was determined. Confidence limits were
set by determining the standard error of the mean (assuming
a Poisson distribution). These means were then evaluated by
a one-way analysis of variance. The mean SCE frequencies
per chromosome and per cell that each cytogeneticist ob-
served for the same treatment group (variance within sample)
were compared with the mean SCE frequencies observed for the
other treatment groups and for the negative control (variance
between samples). An F test was then performed to determine
whether the between-sample variance (the effect of the
chemical) was significantly greater than the within-sample
variance (the variance between observers).
An acceptable assay was one in which (1) the maximum con-
centration used yielded sufficient Ma metaphases for analy-
sis, was the solubility limit, or, 'ocause of limited
816
-------
availability of sample, was 500 ug/ml; (2) the SCE fre-
quencies in the positive controls were at least twice
the negative-control SCE frequencies, and (3) SCE fre-
quency analyses by at least two cytogeneticists were in
general agreement.
A sample was considered to have elicited a positive res-
ponse in the SCE test if the variance between treatment
groups was significantly greater (P < 0.05; two-tailed
test) than the variance within treatment groups for either
SCFs per chromosome or SCEs per cell. A chemical was
considered negative in the SCE test if the variance between
treatment groups was not significantly different from the
variance within treatment groups. Then a weak SCE response
to a sample was suspected but was not statistically sig-
nificant, one or two additional cytogeneticists were asked
to observe the same coded slides. The mean SCE freque'cies
determined by the three or four observers were then compared
by a one-way analysis of variance, and the same criteria
were applied to determine a positive or negative response.
RESULTS AMD DISCUSSION
Saccharomyces Dj Recombinogenic Assay
The heavy-duty (Caterpillar) diesel emission was not tested
in this system. For the other seven emission samples, in
several experiments a slight elevation was observed in the
number of recombinants at one or two concentrations. However
these results were neither reproducible nor dose-related.
We therefore conclude that none of the diesel and related
environmental emission samples that were tested produced a
recombinogenic effect in the yeast S. cerevisiae D3 assay.
The concentration ranges used for the different samples are
listed in Table 1.
817
-------
TABLE 1. RESULTS OF SACCHAROMYCES CEREVISIAE D3
REC011BINOGENIC ASSAY
Sample Concentration (yg) Toxicity
Coke oven emission
Roofing tar emission
Cigarette smoke condensate
Gasoline tlustanp emission
Diesel Nissan emission
VU Rabbit emission
Diesel Oldsmobile
emission
100
100
10
100
100
100
100
to
to
to
to
to
to
to
1
1
5
1
1
1
2
,000
,000
,000
,500
,000
,000
,000
Toxic
Toxic
Not toxic
!3ot toxic
Toxic
Toxic
Toxic
Because of the limited sample, a toxic dose could not le
reached.
Simmon (17) reported that the S. cerevisiae D3 mitotic re-
combination assay is about 54% accurate in detecting chemi-
cal carcinogens as mutagens. All of 20 ultimate carcinogens
tested in this study elicited a positive response, but only
18 of 49 procarcinogens increased the frequency of mitotic
recombination. It is interesting that none of 7 polycyclic
aromatic hydrocarbons (PAIIs) requiring activation, includ-
ing benzo(a)pyrene (BaP), was active in this assay (17).
Failure to detect this class of mutagens could be due to
inability of the compound to cross the yeast cell wall,
or the recommended protocol used (1) may not Le adequate
because the organisms are exposed to the test chemical for
a limited amount of time, during which no active cell growth
occurs. We are currently investigating alternate protocols
to increase the sensitivity of the assay.
L5178Y Mouse Lymphoma t'utagenesis Assay
Figures 1-8 present the results of testing the eight samples
of diesel and related environmental emissions in the L5178Y
TK "* TK mutagenesis assav. For all the reported data,
the mutation frequencies of the solvent and positive controls
were within expected ranges, based on previous experiments
conducted in this laboratory.
All of the emission samples gave positive mutagenic responses
both in the presence and in the absence of metabolic acti-
vation, using the criteria of the L5178Y mouse lymphoma TK
818
-------
3DVa3AV
SyOAIAUnS 9OL/S3INOnOO
a
<
(O) % HIMOUD 1V101 3AllV13y 30VH3AV
syoAiAyns 9oi/S3iNonoo INVIOIAI
(O) %
1V101 3AllVT3y 3DVH3AV
819
-------
(•I saoAiAans 9oi/s3iN
-------
(•) saoMAuns 9ot/S3iNcnoo
o
in
O
'o
(O) % H1MOUO 1V101 3AI1V13H
(•) syoAiAans 9oi/s3iNOioo
>
00
•&
IT
(O) % H1MOH9 TV101 3AI1V13H 39VH3AV
821
-------
SHOAIAHnS 9Ol/S3INCnOO INViniAl 30VU3AV
a E
(O) % H1MOUD "1V1O1 3AI1VT3H 3DVU3AV
(•) saoAiAans 9oi/s3tNoioo
>
00
(O)
822
-------
SHOAIAHnS gOl/SaiNCnOO INVini/M 39VH3AV
(O) % H1MOHD IViOl 3AI1VT3H 30VH3AV
30Vd3AV
I
§>
00
o
a
X
in
(O) % H1MOUO 1V101 3AllVT3b 30VU3AV
823
-------
(o) % HiMOdD nvioi aAiivnay
(•) SdOAIAdOS 9Ol/S3IN0100 INVIOIAI 39VU3AV
I
o
s
DC
10
0}
I
(O) % HIMOaO 1VJ.01 3AliV13d 3DVd3AV
824
-------
(•) saoAiAHns
INVIHIAI
-3 e
HJ.MOHO ~IV101 3AI1V13H
(•) SHOAIAHflS 9Ol/S3INOTOO INViniAl 3DVH3AV
>-
CO
$
tn
(O) % H1MOHD TV101 3AI1VH3U 39VH3AV
825
-------
(•) SHOAIAHDS 9oi/S3ii\icnoo iNvinwi SDVHSAV
a
OJ
O
o
I
O
(O) % HIMOaD "IV101 3AI1V13U 3DV93AV
39Vd3AV
SUOAIAdnS gOl/S3IN0103
>-
oo
HIMOdD TV1O1 3AI1V13H 30Vd3AV
826
-------
forward mutation assay. In comparison, mutagenicity in
response to BaP has been detected in this assay only in
the presence of activation. There were no unusual patterns
in the cytotoxic or mutagenic responses. With cell survival
levels of approximately 10%, the concentration producing the
maximum mutagenic response was different for each sample. The
four diesel engine emissions and the cigarette smoke conden-
sate were more cytotoxic in the absence of metabolic acti-
vation than in the presence of activation. Examples of
these results are illustrated by Figures 3 and 5-8. At
cell survival levels of approximately 10% or greater, the
maximum increases in mutation frequency occurred at con-
centrations ranging from 20 to 300 ug/ml. The mutation fre-
quencies observed with these samples were 2 to 3.9 times the
spontaneous frequencies.
Coke oven, roofing tar, and Mustang gasoline engine emissions
were more cytotoxic in the presence of metabolic activation
than in the absence of activation. Examples of these results
are illustrated by Figures 1, 2, and 4. With cell survival
levels of approximately 10% or greater, the maximum increases
in mutation frequency occurred at concentrations ranging from
9 to 200 yg/ml. These samples yielded mutation frequencies
3.5 to 4.3 times the spontaneous frequency. Roofing tar and
coke oven emissions were cytotoxic in the presence of meta-
bolic activation at concentrations 7 to 10 times lower than
those at which they were cytotoxic without activation.
Of the diesel engine emission samples tested, the Nissan
diesel engine emission was the most cytotoxic. Of all the
samples tested, the Mustang gasoline engine emission sample
(both with and without metabolic activation) caused the
highest increase in mutation frequency.
Sister Chromatid Exchange Assay
Figures 9-16 present the results of testing the eight samples
of diesel and related environmental emissions in the SCE
assay. Results of tests with metabolic activation on the
sample from the emission of a Mustang gasoline engine are
not presented because the CHO cultures were lost due to
contamination in the first test with activation and there was
not enough sample to repeat the test. All other samples
were tested both with and without activation.
Figure 17 compares the responses of CHO cells exposed to
diesel engine'and related environmental emissions that res-
ulted in significant increases in SCE frequencies without
metabolic activation. Since these results are based on single
experiments, minor differences in the responses to these
samples at comparable concentrations should not be considered
significant. With this caveat, we note that the coke oven
emission sample caused the greatest increase in SCE
827
-------
I
O
S
o
o
o
o
828
-------
W» (/>
CD
O
o
O
O
1H30/S30S
829
-------
T130/S30S
830
-------
a
Si
£
LU
O
831
-------
v t
0.5
8
1 3
o
t-
01
o
o
-------
o
I
(J
1133/5303
833
-------
834
-------
.a
o
E
O
o
O
I
1130/5305
835
-------
O ro
I £
CT ii
2? o
•• -O
o g
en E
836
-------
frequencies. The Mlsr-an cb'enel emission sample seemed
slightly more potent than samples of the cigarette smoke
condensate, roofing tar emission, Tustang gasoline emis-
sion, and VU Rabbit diesel emission. The Caterpillar
diesel sample was least effective in increasing SCK fre-
quencies. Cytotoxicity also varied, the most toxic being
the Diesel Nissan emission sample and the least toxic, the
Caterpillar diesel emission sample.
Figure 18 presents the responses of CHO cells to diesel and
other environmental emissions that resulted in signifi-
cant increases in SCE frequencies with metabolic acti-
vation. For comparison, the results of testing BaP, a
PAH that induces SCEs only with activation, are also
included in this figure. These results are also based
on single experiments, but BaP does appear to be more
potent (and more cytotoxic) than the other samples. The
VW Rabbit diesel emission sample was less active than the
other samples at comparable concentrations. Because the
SCE assays v/ithout metabolic activation involve a 21.5-
hour exposure to the samples and those with metabolic
activation involve only a 2-hour exposure, the relationship
between cytotoxicity and chromosome damage as measured by
SCE frequencies varies in the two assays. Direct compari-
sons of the results without attention to these relation-
ship are, therefore, inappropriate. However, we did note
that in contrast to BaP, which required metabolic activation,
no emission sample was positive only with activation.
This is consistent with the observations of Wang et al. (18),
who found that the principal mutagens in city air (speculated
to be derived from automobile exhaust), detected by the
Ames Salmonella typhimurium assay, did not require acti-
vation by liver enzymes.
CONCLUSIONS
The absence of recombinogenic responses to the diesel and
related environmental emission sample in the Saccharomyces
D3 assay may be related to the short exposure tine, during
which no cell growth occurred, or it may be due to inability
of the samples to cross the yeast cell wall.
Both the L5178Y mutagenesis and the SCE tests are _i_n vi_t_rp
mammalian assays in which the mutagenicity and/or DNA-dama-
gimr capabilities of PAHs such as BaP are detected only with
metabolic activation. The failure of the SCE test to
detect DNA-damaging effects of the cigarette smoke condensate
and the emission from the heavy-duty (Caterpillar) engine
with activation, whereas these samples induced mutagenesis in
L5178Y cells x'ith activation, may be related to the relatively
short exposure time used for SCE testing with activation.
It has been noted that all other tested samples induced SCEs
837
-------
838
-------
with activation, and all of the samples induced mutagenesis
in L5178Y cells with activation.
The observations of increases in induced mutation frequen-
cies in L5178Y cells following exposure to all of the
emission samples in the absence of activation and of in-
creases in SCE frequencies in CHO cells following exposure
to all except the Oldsmobile diesel emission sample in the
absence of activation indicate that, in contrast to BaP, the
emission samples contain one or more substances that are
direct-acting mutagens and chromosome-damaging agents.
Further testing of various fractions from the emission sam-
ples using the L5178Y mutagenesis and SCE assays could
assist in determining the identity of these direct-acting
mutagenic and DNA-damaging agents.
REFERENCES
1. Zimmermann, F. K., and R. Schwaier. 1967. A genetic
effect of symmetric dimethylhydrazine: Induction of
mitotic recombination. Naturwissenschaften, 54:251.
2. Brusick, D. J., and V. W. Mayer. 1973. New develop-
ments in mutagenicity screening techniques with yeast.
Environ. Health Perspectives, 6:83-96.
3. Clive, D., and J.F.S. Spector. 1975. Laboratory pro-
cedure for assessing specific locus mutations at the
TK locus in cultured L5178Y mouse lymphoma cells.
Mutation Res., 31:17-29.
4. Clive, D., W. G. Flamm, and J. B. Patterson. 1973.
Locus mutational assay systems for mouse lymphoma
cells. In: Chemical Mutagens: Principles and Methods
for Their Detection, Vol. 3, A. Hollaender (ed.).
Plenum Press, New York, pp. 79-103.
5. Clive, D., K. 0. Johnson, J.F.S. Spector, A. G. Batson,
and M.M.M. Brown. 1979. Validation and characteriza-
tion of the L5178Y/TK+/- mouse lymphoma mutagen assay
system. Mutation Res., 59:61-108.
6. Brown, M.M.M., and D. Clive. 1978. The utilization of
trifluorothymidine as a selective agent for TK~/~
mutants in L5178Y mouse lymphoma cells. Mutation Res.,
53:116.
7. Kato, H. 1974. Induction of sister chromatid exchanges
by chemical mutagens and its possible relevance to DNA
repair. Exp. Cell Res., 85:239-247.
839
-------
8. Perry, P., and H. J. Evans. 1975. Cytological detec-
tion of mutagen-carcinogen exposure by sister chromatid
exchange. Nature, 258:121-125.
9. Solomon, E., and M. Bobrow. 1975. Sister chromatid
exchanges—A sensitive assay of agents damaging human
chromosomes. Mutation Res., 30:273-278.
10. Wolff, S. 1976. Technique for obtaining harlequin
chromosomes. Mammalian Chromosome Newsletter, 17:26.
11. Latt, S. A. 1974. Sister chromatid exchanges, indices
of human chromosome damage and repair: Detection by
fluorescence and induction by mitomycin-C. Proc. Nat.
Acad. Sci. USA, 71:3162-3166.
12. Natarajan, A. F., A. D. Tates, P.P.W. van Buul,
M. Heijers, and N. de Vogel. 1976. Cytogenetic effects
of mutagens/carcinogens after activation in a micro-
somal system in vitro. 1. Induction of chromosome
aberrations and sister chromatid exchanges by
diethylnitrosamine (DEN) and dimethylnitrosamine (DMN)
in CHO cells in the presence of rat-liver microsomes.
Mutation Res.. 37:83-90.
13. Stetka, D. G., and S. Wolff. 1976. Sister chromatid
exchanges as an assay for genetic damage induced by
mutagens/carcinogens. Part II. In vitro test for com-
pounds requiring metabolic activation. Mutation Res.,
41:343-350.
14. Wolff, S. 1977. Sister chromatid exchange. Ann. Rev.
Genet., 11:183-201.
15. Hollstein, M., J. McCann, F. A. Angelosanto, and W. W.
Nichols. 1979. Short-term tests for carcinogens and
mutagens. Mutation Res., 65:133-226.
16. Ames, B. N., J. McCann, and E. Yamasaki. 1975. Methods
for detecting carcinogens and mutagens with the
Salmonella/mammalian-microsome mutagenicity test.
Mutation Res., 31:347-364.
17. Simmon, V. F. 1979. In vitro assays for recombino-
genic activity of chemical carcinogens and related
compounds with Saccharomyces cerevisiae D3. J. Nat.
Cancer Inst., 62:901-909.
18. Wang, Yi. Y., S. M. Rappaport, R. F. Sawyer, R. E.
Talcott, and E. T. Wei. 1978. Direct-acting mutagens
in automobile exhaust. Cancer Letters, 5:39-47.
840
-------
General Discussion
D. BRUSICK: That is very interesting data. Could you
give a relative potency ranking to the mouse lymphoma the
mouse data for various sample source?
A. MITCHELL: We feel very leery about coming up with
any sort of calculations as to relative potency for any of
these.
D. BRUSICK: Well in the mouse lymphoma it looked like
they were very close together, in contrast to SEC's and
certainly in contrast to the salmonella tests.
A. MITCHELL: Obviously the benzopyrene was the most
positive of the mouse lymphoma. The diesel Olds was not
very positive. We had to go pretty far out concentration-
wise with the caterpillar fraction. I believe our greatest
activities were in the coke oven and roofing tar complexes.
D. BRUSICK: It seemed that in the mammalian cells, in
contrast to the salmonella data that the gasoline engine
appeared to be more active than the diesel, which I find
interesting, and maybe of importance in looking at this
whole problem. If we have good solid positive data on mam-
malian cells, why are you trying to go back and regenerate
more data from the yeast system which is another microbial
assay if it isn't going to tell us anymore than we already
know from salmonella and mammalian test data.
A. MITCHELL: You are an expert on yeast, but I think
this is an EPA decision as to whether they want to keep
going on with the matrix and look at the relative activi-
ties.
W. THILLY: I was very impressed with the completeness
of the L5178 wide data. However, it seemed to me that,
after looking at the data, the diesel sample was indeed
very mutagenic to human cells in the presence of a meta-
bolic activation system but in the absence of a metabolic
activation system (up to 200 micrograms per ml) - no activ-
ity was seen. So as to your hypothesis, I would like to
turn it over; and if there were differences at the cell
level, it may be a matter of inactivation in possibly the
L578 versus the human TK6.
M. WATERS: We have noted in tests with pesticide chem-
icals a number of cases where the pesticides are picked up
in yeast systems, so I think from that point of view it is
quite interesting to continue to look at the systems.
R. BILES: You may not be prepared to give these num-
bers, but I didn't get from your presentation the mouse
lymphoma data for the diesel and the cigarette smoke con-
841
-------
densate. Could you give the approximate concentration
range with and without meatbolic activation?
A. MITCHELL: Yes, I can, but it is in the notes.
P. SABHARWAL: I think he raised a very important ques-
tion about the treatment of the S9 infraction. You can go
up to only two hours before you have toxicity. We have
resolved the toxicity problem by using a feeder layer so
that we can now go up to 24 hours. In fact the freeder
layer system is almost twice as active than that S9 in-
fraction. Our recommendation is that those who are using
S9 and are facing this problem of toxicity should consider
using the feeder layers.
A. MITCHELL: This is an interesting approach. Using
liver cells for cyto genetic assays has produced inter-
esting data. A one-to-one correlation has not been found
between the metabolic reactions in feeder layers or the
hepatocyte and the S9 fractions. I think we are getting
into another matrix of tests and suddenly this work becomes
three dimensional instead of two, although I think it would
be very interesting.
L. SCHECHTMAN: Actually one can circumvent problems
with toxicity with an S9 preparation by prescreening the S9
as one prescreens serum lots and plastic dishes whatever
you might use in a given bioassay. One can look for in-
herent toxicity of the S9 to the target cell and can screen
the S9 for aryl hydrocarbon hydroxylase activity, epoxidase
activity etc. This can a battery of tests useful in the
selection of an S9.
A. MITCHELL: Correct, we do all of these things for the
lymphoma assay. But quite frankly as to the chromatid
exchange assay we do not do these things as routinely.
842
-------
MUTAGENIC AND CARCINOGENIC POTENCY OF EXTRACTS OF
DIESEL AND RELATED ENVIRONMENTAL EMISSIONS:
IN VITRO MUTAGENESIS AND ONCOGENIC TRANSFORMATION
Bruce C. Casto
George G. Hatch
Shiu L. Huang
Northrop Services, Inc.
Research Triangle Park, NC
Joellen L. Huisingh
Stephen Nesnow
Michael D. Waters
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC
ABSTRACT
Extracts from emissions of four diesel engines, a gasoline
engine and three related environmental samples were tested
in four In vitro assay systems designed to detect carcino-
genic or mutagenic activity of chemicals. Samples from
three of four diesel extracts, the gasoline engine, and
all three related samples were positive in an enhancement
of viral transformation assay. Two diesel samples, the
gasoline engine extract and extract from coke oven emissions
were positive for mutation induction in Chinese hamster
ovary cells. Only the gasoline engine extract and the coke
oven sample were positive in a DNA fragmentation assay
using alkaline sucrose gradients. Experiments using chem-
ical transformation of Syrian hamster embryo cells as an
assay method have not been completed.
843
-------
INTRODUCTION
Several in vitro tests for the identification of potential
carcinogenic agents have been developed. Presently, the
systems that correlate best with activity iri vivo are muta-
genesis assays in microbial (1) and mammalian (2) cells,
mammalian cell transformation (3,4,5) analysis of DNA repair
(3,6,7) or fragmentation (3,7,8), and enhancement of viral
transformation (3,9,10,11). Four of the above test systems
have been employed by us to evaluate the carcinogenic/muta-
genic potential in vitro of extracts from diesel and related
environmental emissions. These tests are: mutation assays
in Chinese hamster ovary cells and the analysis of DNA frag-
mentation, chemical transformation, and enhancement of viral
transformation in Syrian hamster embryo cells.
Mutagenesis assays in microbial systems have been reported
to reliably detect 90% of known chemical carcinogens (1);
a similar correlation has been established for the Chinese
hamster ovary (CHO) mammalian cell mutation assay (2) used
in the present study. Included in the latter study (2)
were: direct acting alkylating agents, nitrosamines, het-
erocyclic nitrogen mustards, metals, polycyclic hydrocar-
bons, and ionizing and nonionizing radiation. In the group
of 80 agents tested, the carcinogenic activity of 42 was
known. The mutagenicity of this group showed a 95% corre-
lation (40/42) with their carcinogenic activity in vivo.
The activity of chemical carcinogens in transformation
assays In vitro with hamster embryo cells parallels their
activity in vivo. Pienta et. al. (4) have tested various
classes of chemicals including alkylating agents, poly-
cyclic hydrocarbons, nitrosamines, aromatic amines and
solvents and have found a 91% correlation with known car-
cinogenic activity (79/87); there were no false-positives.
Casto et al. (3) have reported on the activity of 70 chem-
icals in Syrian hamster embryo cells using two assay sys-
tems. Combined data from focus and colony assays showed
that 50/55 carcinogens were positive (91%), whereas none
of the noncarcinogens or solvent controls induced trans-
formants. Approximately 30 to 40 chemicals have been
tested in common in two separate laboratories with good
agreement between test results (12).
The close correlation between mutagenic and carcinogenic
activity, for a large number of chemical compounds, has
suggested that assays that directly or indirectly measure
damage to cell DNA might be used to screen potential car-
cinogens or mutagens. These studies are based upon the
methods used by McGrath and Williams (13) and Painter and
Cleaver (14) for studying irradiation damage to the DNA
844
-------
of E. coli and Hela cells respectively. Many chemical car-
cinogens and mutagens induce DNA damage that is demonstrable
following centrifugation alkaline sucrose gradients (10) or
alkaline elution from filters (8). Casto et al. (3) have
evaluated 77 compounds in Syrian hamster embryo cells for
the capacity to induce DNA breakage. Of 55 carcinogens
tested, 40 were positive (73%) following sedimentation in
alkaline sucrose; there were no false positives among the
remaining 22 chemicals. Using akaline elution techniques,
24 of 30 (80%) carcinogens caused breakage of DNA in V79
cells as reported by Swenberg et al. (8) and 63 of 70 (90%)
carcinogens were positive in a more comprehensive study by
Petzold et al. (15); again, there were no false-positives.
A different in vitro assay to detect potential carcinogens
and mutagens has been described by Casto (11, 16) and Casto
et al. (9,10,7) in Syrian hamster embryo cells. In this
system, treatment of cells with carcinogenic or mutagenic
chemicals results in an increased sensitivity to adenovirus
transformation. Approximately 160 chemicals have been test-
ed in this system with information available on 130 com-
pounds regarding their current carcinogenic/mutagenic status.
Of the known carcinogens or mutagens tested, 94% (83/88)
were shown to significantly increase the transformation
frequency of simian adenovirus, SA7, whereas only 1/42 non-
carcinogens was positive (caffeine). Two of the false-neg-
ative compounds, when tested with an S-9 mix, converted to
a positive response (urethane and dimethylnitrosamine).
BIOASSAY PROCEDURES
Mutagenesis Assays in CHO cells
Nearly confluent cultures of CHO cells (Clone Kl-BHY Sub-
clone DI, obtained from A. W. Hsie, ORNL) were passaged to
100 mm dishes at 1 x 106 cells/plate. After 24 hr, 1.5
to 2 million cells per 100 mm dish were treated with in-
creasing doses of test agents for 16 to 24 hours. Follow-
ing chemical treatment, cells were collected by trypsini-
zation and 105 cells were inoculated into 100 mm dishes.
A minimum of ten dishes was used for each data point.
Addition of medium to select for 6-thioguanine-resistant
cells began 48 hours and 7 days after chemical treatment
and continued for a period of 11 days. At this time, the
cultures were fixed, stained with methylene blue, and the
total number of mutant colonies enumerated.
For determination of cell survival, 100 to 200 treated or
control cells were seeded into 60 mm dishes in F10 medium.
After incubation for 9 days, the cultures were fixed,
stained, and the total number of surviving colonies deter-
mined. The CHO cloning efficiency (CE) of treated cells
845
-------
was calculated as percent survivors relative to nontreated
control cells. CE's were determined for cells replated at
increasing post-treatment periods (i.e. 24-hr intervals).
Relative survival curves as a function of post-treatment
times were calculated. Optimal chemical mutagen doses were
determined from curves generated from mutagenesis experi-
ments (vide infra).
If excess cell toxicity resulted from treatment with higher
concentrations of compounds, the cultures were pooled and
subcultured 3-to-5 days after treatment. To estimate sur-
vival and mutation frequency, treated or untreated cells
were inoculated in 60 mm and 100 mm dishes as above and
cultured for 9 or 11 days respectively. The cultures were
fixed, stained and the total number of surviving and mutant
colonies counted. A mutation frequency was calculated by
determining the probability of the number of mutants among
surviving cells.
Transformation Assays in Syrian Hamster Embryo Cell
Primary Syrian hamster embryo cell cultures (SHE) were pre-
pared by trypsinization of eviscerated and decapitated
fetuses after 13-14 days of gestation. Cells were resus-
pended in a Modified Dulbecco's medium (16) herein referred
to as MDM, supplemented with 5% heat-inactivated fetal bo-
vine serum (FBS) , and 0.22g% NaHCOa. Approximately
1.0 x 10 cells in 20 ml of medium were plated into 100 mm
plastic Petri dishes and incubated in a 5% CO2 atmosphere
at 37°C. Total cell counts after 3 days usually ranged
from l.O-to-1.2 x 107 cells per plate. Secondary SHE were
prepared by transferring 5 x 10 cells into 100 mm dishes
in 20 ml of the above medium, but with 10% FBS.
For chemical transformation, stock solutions of extracts
were prepared at concentrations of 1 or 10 mg/ml in ace-
tone. Appropriate dilutions were made in complete medium
to give the desired final concentrations. Evaluation of
transformation was conducted by the focus assay method,
adapted to SHE cells by Casto et al. (18). Secondary SHE
were plated at 5 x 10 cells/60 mm dish in 4 ml of MDM
containing 10% FBS and O.llg% NaHC03. After 24 hr, 4 ml
of chemical dilutions (as 2X concentrations) were added
to each of 10-to-20 plates per dilution and incubated for
3 days at 37°C in 5% CO2. The culture fluid was then re-
placed with fresh medium (0.22g% NaHCC>3) containing chem-
ical dilutions at IX concentrations and incubated for an
additional 3 days. After a total of 6 days' exposure,
the chemical-containing medium was removed and the cul-
tures fed at 3-to-4 day intervals with 6 ml of MDM, 10%
FBS, and 0.22% NaHC03. After 21-to-25 days from the time
846
-------
of chemical addition, the cells were formalin-fixed,
Giemsa-stained and scored for transformed foci. Verifica-
tion of the transformed morphology was made with a stereo-
microscope at 10X to 30X magnification.
Cell lethality due to chemical was determined on plates
seeded with 500 cells and treated as above. The surviving
colonies were formalin-fixed and stained with 0.02% crystal
violet approximately 8 days after seeding. The number of
surviving colonies among cells treated with chemical was
divided by the number of colonies in solvent-treated cells
to give the surviving fraction of chemically treated cells.
Analysis of DNA Strand Breaks By Sedimentation in Alkaline
Sucrose Gradients
Five to 35% sucrose gradients were prepared by 4 cycles of
freeze-thawing (19) of tubes containing 18% sucrose in 0.05%
EDTA which was adjusted to pH 12.6 with 10M NaOH. Primary
hamster cell cultures, after 3 days' incubation, were pas-
saged to 60 mm plastic dishes using 4 x 10 cells per dish.
Immediately, or after 24 hr, the cells were pulsed with
0.5 yCi/ml of 3H-labeled thymidine (3H-TdR) for 24 hr.
After 3H-TdR labeling, all cultures were changed to MDM with
0.5% FBS and 0.22 g% NaHC03, and held for an additional 24
hr. At this time each dish contained approximately 106
cells. The test chemicals were prepared as l-to-10 mg/ml
stocks in acetone and added to prewarmed medium to give the
desired final concentration. The SHE were treated for 18
hr, washed IX, and removed from the dish with EDTA. Follow-
ing centrifugation, the cells were resuspended in EDTA to
give 105 cells per 0.1 ml. One-tenth ml of the cell sus-
pension was then added to the top of a sucrose gradient
tube layered with 0.3 ml of lysing solution (1% sarkosyl
in 0.05% EDTA). The cells were then lysed at room tempera-
ture for 1 hr, placed in an SW-50 rotor, and centrifuged
for 1 hr at 30,000 rpm in a Model L5-75B ultracentrifuge
at 20°C. Three drop fractions were collected directly into
scintillation vials following bottom puncture, neutralized
with 1.0 ml of 0.2N HCl and prepared for counting by add-
ing 5 ml of Bray's scintillation fluid to each vial. Counts
were made in a Beckman LS-250 scintillation spectrometer,
10 minutes per sample. Data were plotted as percent of the
highest count in each gradient. A shift in the peak of
radioactivity greater than 3 fractions (more than 6 mm) from
the control peak was considered positive evidence for chem-
ical-induced DNA strand breaks (17).
Enhancement of Viral Transformation
Primary SHE were prepared by trypsinization of eviscerated
and decapitated embryos as described previously. Total
847
-------
cell counts after incubation for 3 days usually ranged from
3.7-to-4.5 x 106 cells per plate. Stocks of simian adeno-
virus, SA7 , were prepared in VERO cells by inoculation of
cell cultures in 100 mm dishes at an input multiplicity of
2-to-3 plaque-forming units (PFU) per cell. Virus was ad-
sorbed for 2 hr and 5 ml of medium (MEM, 5% FBS) then added
to each plate. Cytopathic effects were complete by 72 hr,
after which the infected cells were harvested, frozen-thaw-
ed 4X and virus separated from cell debris by low-speed
centrifugation. Virus stocks for transformation assays
were stored in 1 ml aliquots at -70°C.
Dilutions of test chemicals were prepared fresh each time
by dissolving in acetone and appropriate dilutions then
made in medium to give the desired final concentrations.
In each experiment, 2 plates of SHE were treated with chem-
ical for 18 hr prior to virus inoculation. Following incu-
bation with the chemical, the SHE were rinsed with complete
medium and inoculated with SA7. Transformation and clonal
assays were performed with each treatment group as describ-
ed below. Treatment of cells with chemical alone, using
these protocols, did not result in detectable transforma-
tion.
A complete description of the methodology for adenovirus
transformation has been presented elsewhere (20,21).
Briefly, the procedure was as follows: SA7 was added to
treated and control SHE (3-to-4 x 107 PFU/culture) and ab-
sorbed for 3 hr; the virus-inoculated cells were removed
with trypsin (0.25% in MDM with 0.1 mM CaCl2) then centri-
fuged and resuspended to 106 cells per ml in MDM, 10% FBS
and 0.11 g% NaHC03. Cells were then mixed and plated into
60 mm dishes using 2 x 10 cells per dish; 3-to-4 ml of
the above medium was then added to each plate. After in-
cubation for 3 days, the medium was changed to MDM with
0.1 mM CaCl2 (20,22), supplemented with 10% FBS and 0.22 g%
NaHCOa- At intervals of 4, 5 and 6 days, 3 ml of 0.5 g%
Bacto-agar medium was added, and final focus counts made
25-30 days from the beginning of the experiment. All foci
had the cellular and colony morphology representative for
SA7 virus and have been shown previously to contain virus-
specific SA7 "T" antigen. For survival assays, SHE were
resuspended to 106 cells/ml and diluted in complete medium
to give approximately 300 cells per 0.1 ml. Two-tenths ml
(600 cells) was added to each of 5 plastic dishes, follow-
ed by 3 ml of MDM with 10% FBS and 0.11 g% NaHCO3. Two
to three days later, 3 ml of medium was added to each
plate, and after 8-to-9 days' total incubation, the col-
onies were fixed in 10% buffered formalin and stained with
0.02% crystal violet. The cloning efficiency of virus-
inoculated cells under these conditions was usually 8-to-
15%. The number of colonies in five plates arising from
848
-------
chemically treated cells was divided by the number of col-
onies in five plates from control cells to give the sur-
viving fraction of chemically-treated cells.
To determine the increase in transformation frequency (en-
hancement ratio) the number of SA7 foci counted from 10
virus-inoculated control plates, each receiving 200,00 cells,
was used as the control transformation frequency and the
total number of foci expected from 106 virus-inoculated
cells was determined. On those plates receiving chemically
treated cells, the frequency of SA7 foci per 10 cells was
calculated by multiplying the actual number from five plates
by the ratio: I/surviving fraction of treated cells.
The enhancement ratio was determined by dividing the trans-
formation frequency in treated cells by that obtained from
control cells. For determination of statistical signifi-
cance, a Table of Critical Ratios (16) was constructed
from the Lorenz tables (23) which are based upon the Poisson
distribution and which were originally designed to apply
statistical tests of significance to virus plaque counts.
The ratios used in the table are based upon the actual num-
ber of foci counted and not upon survival-adjusted values.
An increased transformation frequency was considered sta-
tistically significant at the 5% or 1% confidence level if
the enhancement ratio exceeded the appropriate value obtain-
ed from the Table of Critical Ratios.
RESULTS
All eight coded samples were tested in SHE cells in two to
four separate experiments. When confluent cultures were
exposed to the test samples, extracts from Caterpillar,
Oldsmobile, Mustang and VW Rabbit were not overtly toxic
at 500 yg/ml. Extracts from roofing tar were the most
toxic (< 5% surviving cells at 100 y^/ml)i followed by
Datsun Nissan (25% survivors at 250 yg/ml), coke oven (50%
survivors at 250 yg/ml), and cigarette smoke (40% survivors
at 500 ug/ml).
When added to sparsely seeded cultures (1000 cells/dish),
the dose sufficient to reduce cell survival by greater than
80% was significantly lower; approximately 100 yg/ml for
Caterpillar, 50 yg/ml for VW Rabbit, 25 pg/ml for Olds-
mobile, roofing tar, cigarette smoke, and Mustang, 6 yg/ml
for Datsun Nissan, and 3 yg/ml for coke oven.
Data for enhancement of adenoviral transformation by the
extracts of the various test samples are shown in Table 1
and Figure 1. With exposure to each sample, an absolute
increase in the number of virus-transformed foci per plate
849
-------
TABLE 1. ENHANCEMENT OF VIRAL TRANSFORMATION BY
ORGANIC EXTRACTS FROM OLDSMOBILE AND DATSUN
NISSAN DIESEL EMISSIONS
Sample
pg/ml
Transformation
frequency
Enhancement
ratio
Oldsmobile
Datsun
Nissan
500
250
125
62.5
31.2
0
500
250
125
62.5
31.2
0
98
96
68
78
73
65
0
236
125
111
88
58
1.51*
1.48*
1.05
1.20
1.13
1.00
0
4.07
2.16
1.91
1.52*
1.00
Extracts were dissolved in acetone and added to cultures
of SHE 18 hours prior to addition of Simina adenovirus
SA7. Transformation and survival assays were conducted
as described in the text.
Number of SA7 transformed foci per 2 x 10 surviving cells.
'Enhancement was determined by dividing the transformation
frequency in sample-treated cultures bu that obtained in
solvent-treated cultures. Underlined values were signi-
ficant at the 1% level; values with asteriskt*) were
significant at the 5% level.
and a positive dose-response was induced at nonlethal con-
centrations. The Caterpillar extract, although giving a
slight positive dose response (Figure 1) and producing a
1.35- and 1.40-fold increase in foci at 500 and 1000 ug/ml
respectively, was the only sample that failed to induce a
significant enhancement at either the 5% or 1% level.
850
-------
to
E-8
J55
8 8 S
8 8
01
i/i -c
c •(->
°^
10 c
on CD
•i- in
E a)
CD S-
QL
i— CD
CD S-
CD'«
c
O en
s- CD
> u
c s_
S S
gOL x Z J8(^ Aouanbajj uoueujjo^suejj^ snj
CD CD
- >-
O >, O
01/1*4-
851
-------
The Oldsmobile extract yielded a positive dose response
(Figure 1) and a 1.48- and 1.51-fold increase in transfor-
mation frequency (Table 1); however, these increases at
250 and 500 yg/ml respectively were statistically signifi-
cant at p < 0.05, but not at p < 0.01.
For comparison, the data for the Datsun Nissan extract are
shown in Table 1. At 31 pg/ml, the increase in viral trans-
formation is similar to that observed with 500 ug/ml of
Oldsmobile extract. Dose response data for these and the
remaining extracts are shown in the composite Figure 1.
If the various extracts are ranked according to the lowest
effective concentration tested (LECT) that induces signifi-
cant enhancement (Table 2), the roofing tar extract was
found to be the most potent. Among the diesel samples
tested, Datsun Nissan appeared to be the most potent follow-
ed by VW Rabbit, Oldsmobile and Caterpillar. The extract
prepared from the gasoline-powered Mustang engine and the
diesel powered VW Rabbit both induced enhancement at 125
yg/ml, but the Mustang extract was ranked higher since a
greater increase in transformation frequency was obtained.
None of the samples were as potent as B(a)P (positive at
0.125 ug/ml), but the dose-response curves generated by
cigarette smoke and Datsun Nissan extracts were similar
to that obtained for B(a)P (Figure 1).
Each of the organic extracts was tested in two to three
separate experiments in CHO cells for the capacity to in-
duce mutation as determined by resistance to 6-thioguanine.
Cells were exposed for 24 hours and mutant selection was
begun 48 hours or 7 days after treatment. The efficiency
of mutant recovery was monitored with known 6-TG resistant
cells and was determined to be 100% under the test condi-
tions. Extracts showing a significant dose-mutation re-
sponse relationship were from Datsun Nissan, VW Rabbit,
Mustang, and coke oven; MMS was used as a positive control
(Table '3). Graphic representations of the dose-response
curves for extracts from the above emissions and the MMS
control are shown in composite Figure 2.
Induction of cellular DNA breakage by the various extracts,
proved to be the least sensitive of the three assays com-
pleted to date; only treatment with the Mustang and coke
oven extracts caused breakage of SHE DNA in alkaline su-
crose gradients. (Table 3 and Figure 3). The Mustang
852
-------
TABLE 2. POTENCY OF ORGANIC EXTRACTS OF DIESEL AND
RELATED ENVIRONMENTAL EMISSIONS FOR ENHANCEMENT OF VIRAL
TRANSFORMATION IN HAMSTER EMBRYO CULTURES
Enhancement
Sample
Roofing Tar
Coke oven
Cigarette smoke
Datsun Nissan
Mustang
V.W. Rabbit
Oldsmobile
Caterpillar
LECT
3.1
7.8
31.2
62.5
125
125
250
500
LECT
1.63
1.69
1.63
1.91
1.90
1.75
1.48*
1.35
HECT
2.58
4.04
5.60
4.07
3.06
2.40
1.51
1.35
Ranking
1
2
3
4
5
6
7
8
Controls:
B(a)P
MNNG
0.12
0.25
1.70
4.1
8.97
4.60
NA
NA
Lowest effective concentration tested that induced signifi-
cant enhancement (P<0.01) of adenovirus transformation.
Enhancement was determined by dividing the transformation
frequency of extract-treated cultures by that obtained
from solvent-treated cultures of SHE. HECT = maximal en-
hancement obtained at the highest effective concentration
tested in these experiments. Asterisk (*) indicates values
positive at the 5% level, but not at the 1% level of sig-
nificance.
"Ranking was based upon the lowest concentration causing
signigicant enhancement. Where two samples were positive
at the same concentration, the one inducing more enhance-
ment was ranked higher. NA= not included in ranking.
853
-------
TABLE 3. INDUCTION OF MUTATIONS IN CHO CELLS AND DNA
FRAGMENTATION IN SHE BY ORGANIC EXTRACTS OF
DIESEL AND RELATED ENVIRONMENTAL EMISSIONS
Sample
Test system and response
CHO Mutagenesis SHE DNA fragmentation
Caterpillar
Oldsmobile
Datsun Nissan
V. W. Rabbit
Mustang
Roofing Tar
Coke oven
Cigarette smoke
Control:
MMS
0(50-750)
0(50-300)
+(25-100)
+(100-400)
+(50-400)
0(10-80)
+ (50-200)
0(50-200)
+(50-550)
0(30-250)
0(30-250)
0(30-250)
0(30-250)
+(250)
0(30-250)
+(125)
0(30-250)
+(12.5)
A sample was considered positive in the CHO mutagenesis
system if a significant dose response curve was generated
over the dose range tested. In the DNA fragmentation
system, a sample was positive if breakage occurred shift-
ing the major peak of DNA greater than 6mm from the con-
trol DNA peak. Figures in parentheses show the dose
range tested (nonlethal to =80% cell killing).
854
-------
o C
. o
C <0
Ol
OJ
10
OJ
V
CL
Ol
S-
O (/>
O
O S-
(C -i-
s- o
4->
X T3
O) CL)
CO
V. O
O r—
4- CJ
OJ
CD
CO
C
O
Q.
CO
O)
S-
in -
O
O
CO O
C f—
O) O
O) S-
ta •<->
4-> C
o
o
855
-------
08 06 04 02
Distance Sedimented
08 06 04 02
Distance Sedimented
Figure 3. Alkaline sucrose gradients of DMA from Syrian
Hamster embryo cells treated with MMS, coke oven
or Mustang emission extracts. Sedimentation pro-
ceeded from right (top) to left (bottom).
856
-------
extract was positive only at the highest dose tested (250
yg/ml), whereas the coke oven extract was positive at two
dose levels (125 and 250 yg/ml); the positive control, MMS,
caused breakage at 12.5 yg/ml and above (Figure 3).
At this time, the extracts from the different emission
sources have been tested once for transformation of SHE
cells by the focus assay method. B(a)P, the positive con-
trol, induced focus formation at concentrations between 0.06
yg/ml and 2.0 yg/ml (8 foci/60 dishes); the organic extracts
were negative for transformation at doses from 1.5 to 100
pg/ml. Other experiments are in progress to verify these
data and to attempt to increase the sensitivity of the
system for these particular samples.
In summary, eight extracts from diesel and gasoline engine
emissions were compared to extracts from related environ-
mental emissions as coded samples in three in_ vitro assay
systems. The viral enhancement assay produced a positive
dose-response with all samples, however, the extracts pre-
pared from Caterpillar emissions failed to induce a signi-
ficant increase in the adenoviral transformation frequency.
Only four of the eight samples caused a significant dose-
response increase in 6-thioguanine resistant cells (Nissan,
VW Rabbit, Mustang and coke oven); the remaining four
(Caterpillar, Oldsmobile, roofing tar and cigarette smoke)
caused an increase in mutation frequency only at the high-
est dose where approximately 70% or more of the cells failed
to survive treatment.
Induction of cellular DNA breakage appeared to be the
least sensitive of the three assays completed to date.
Only the Mustang and coke oven emission extracts produced
a significant change in the sedimentation profile of the
DNA from treated SHE cells.
REFERENCES
1. McCann, J., E. Choi, E. Yamasaki, and B.N. Ames. 1975.
Detection of carcinogens as mutagens in the Sal-
monella/microsome test: Assay of 300 chemicals.
Proc. Natl. Acad. Science 72:5135-5139.
2. Hsie, A.W., D.B. Couch, J.R. San Sebastian, W.N.C. Sun,
J.C. Riddle, P.A. Brimer, R. Machanoff, N.L. Forbes,
F.T. Kenney, M.H. Hsie. 1978. Mutagenicity of
carcinogens as determined in a quantitative mamma-
lian cell mutation system, CHO/HGPRT: Study of
80 agents. AACR abstracts 19:93.
857
-------
3. Casto, B.C., N. Janosko, J. Meyers, J.A. DiPaolo. 1978.
Comparison of in vitro tests in Syrian hamster
cells for the detection of carcinogens. Proceed-
ings of the American Association of Cancer Re-
search 19:83.
4. Pienta, R.J., J.A. Poiley, and W.B. Lebherz III. 1977.
Morphological transformation of early passage
golden Syrian hamster embryo cells derived from
cryopreserved primary cultures as a reliable in_
vitro bioassay for identifying diverse carcinogens.
International Journal of Cancer 19:642-655.
5. Purchase, I.F.H., E. Longstaff, J. Ashby, J.A. Styles,
D. Anderson, P.A. Lefevre, and F.R. Westwood.
1979. Evaluation of six short-term tests for
detecting organic chemical carcinogens and re-
commendations for their use. Nature 264:624-627.
6. San, R.H.C. and H.F. Stich. 1975. DNA repair synthe-
sis of cultured human cells as a rapid bioassay
for chemical carcinogens. International Journal
of Cancer 16:284-291.
7. Stich, H.F., R.H.C. San, P.P.S. Lam, D.J. Koropatnick,
L.W. Lo, and B.A. Laishes. 1976. DNA fragmenta-
tion or DNA repair as an in vitro and in vivo
assay for chemical procarcinogens, carcinogens,
and carcinogenic nitrosation products. In: R.
Montesano, H. Bartsch and L. Tomatis (eds.)
Screening Tests in Chemical Carcinogenesis, Lyon,
(IARC Scientific Publications No. 12) pp. 617-636.
8. Swenberg, J.A., G.L. Petzold, and P.R. Harbach. 1976.
In vitro DNA damage/alkaline elution assay for
predicting carcinogenic potential. Biochemical
and Biophysical Research Communication 72:732-738.
9. Casto, B.C., W.J. Pieczynski, and J.A. DiPaolo. 1973.
Enhancement of adenovirus transformation by pre-
treatment of hamster cells with carcinogenic poly-
cyclic hydrocarbons. Cancer Research 33:819-824.
10. Casto, B.C., W.J. Pieczynski, and J.A. DiPaolo. 1973.
Enhancement of adenovirus transformation by treat-
ment of hamster cells with diverse chemical car-
cinogens. Cancer Research 34:72-78.
11. Casto, B.C. Dectection of chemical carcinogens and
mutagens in hamster cells by enhancement of adeno-
virus transformation. 1980. In: N. Mishra,
V. Dunkel and M. Mehlman (eds.) Mammalian Cell
Transformation by Chemical Carcinogens. Advances
in Environmental Toxicology (in press).
858
-------
12. DiPaolo, J.A. 1976. In: R. Montesano, H. Bartsch and
L. Tomatis (eds.) Screening Tests in Chemical
Carcinogenesis, Lyon, (IARC Scientific Publications
No. 12), p. 142.
13. McGrath, R.A. and R.W. Williams. 1966. Reconstruction
in vivo of irradiation Escherichia coli deoxyri-
bonucleic acid; the rejoining of broken pieces.
Nature 212:534-545.
14. Painter, R.B., and J.E, Cleaver. 1967. Repair repli-
cation in HeLa cells after large doses of X-irra-
diation. Nature 216:369-370.
15. Petzold, G. , P. Harbach, M. Bedell, and J.A. Swenberg.
1977. Screen for predicting carcinogenic potential
in vitro DNA damage/alkaline elution assay. AACR
abstracts 18:63.
16. Casto, B.C. 1973. Enhancement of adenovirus trans-
formation by treatment of hamster cells with
base analogs, and dibenz(a,h)anthracene. Cancer
Research 33:402-407.
17. Casto, B.C., W.J. Pieczynski, N. Janosko, and J.A.
DiPaolo. 1976. Significance of treatment inter-
val and DNA repair in the enhancement of viral
transformation by chemical carcinogens and muta-
gens. Chemico-Biological Interactions 13:105-125.
18. Casto, B.C., N. Janosko, and J.A. DiPaolo. 1977.
Development of a focus assay model for transfor-
mation of hamster cells in vitro by chemical car-
cinogens. Cancer Research 37:3508-3315.
19. Baxter-Gabbard, K.L. 1972. A simple method for the
large-scale preparation of sucrose gradients.
FEES Letters 20:117-119.
20. Casto, B.C. 1968. Adenovirus transformation of hamster
embryo cells. Journal of Virology 2:376-383.
21. Casto, B.C. 1973. Biologic parameters of adenovirus
transformation. Progress in Experimental Tumor
Research 18:166-198.
22. Freeman, A.E., P.H. Black, R. Wolford, and R.J. Huebner.
1967. Adenovirus type 12-rat embryo transforma-
tion system. Journal of Virology 1:362-367.
23. Lorenz, R.J. Zur statistik des plaque-Testes. 1962.
Archiv Gesamte Virusforschung 12:108-137.
859
-------
General Discussion
B. CASTO: I would like to comment that the enhancement
test actually measures the interaction of chemicals with
cellular DNA. We can show, for example, that with chemicals
there is an increased integration of viral DNA into cellular
DNA after treatment, so it does depend in some way, on the
interaction with cellular DNA. It doesn't necessarily have
to be a direct interaction.
W. THILLY: That is really interesting. This assay has
been of interest to many of us who think about the re-
lationship between mutation and the induction of cancer. A
parallel assay in prokaryotic systems is the induction of
phage from the lysogenic to the lytic phase. Apparently
some of these polycyclics or other mutagens somehow turn on
the rec A protein. The rec A protein degradates the re-
pressor on the lambd aphage. In other words, there is
normally a represser sitting on the OP binding site for the
synthesis of the first protein necessary for lytic ex-
pressions in lamba.
B. CASTO: Yes, this is an extraneously added virus, not
endogenous, but the general principle may be the same.
860
-------
MUTAGENIC AND CARCINOGENIC POTENCY OF EXTRACTS
EROM DIESEL RELATED ENVIRONMENTAL EMISSONS:
SIMULTANEOUS MORPHOLOGICAL TRANSFORMATION AND MUTAGENESIS
IN RALP/c 3T3 CELLS
R. D. Curren, R. E. Kouri, C. M. Kim, and L. M. Schechtman
Microbiological Associates
5221 River Road
Rethesda, MD 20016
ABSTRACT
Particulate extracts from six different environmental
emission sources were assayed for genotoxic activity in
mouse BALR/c 3T3 Clone A31-1 -cells. All compounds were
tested simultaneously for both transforming and mutagenic
potential with and without exogenous metabolic activation
in the form of a 9000 x g post-mitochondria! hepatic super-
natant fraction from Aroclor-1254 induced Fischer 344 rats.
Pichloromethane particulate extracts from the exhaust of two
light duty diesel engines (Oldsmobile and Nissan), one heavy
duty diesel engine (Caterpillar) and one late model gasoline
engine (Mustang II) were assayed in an identical manner to
particulate extracts from the emissions of a roofing tar pot
and a coke oven. No clear dose-dependent responses were
observed, but several of the samples showed significant
transforming and mutagenic activity. A qualitative ranking
system showed the activity of these particulate extracts for
either mutagenesis or transformation was: coke oven =
Mustang II gasoline engine > Nissan diesel engine > roofing
tar. Particulate extracts from the Oldsmobile diesel engine
and the Caterpillar diesel engine showed essentially no
activity.
861
-------
INTRODUCTION
As various economic and environmental considerations force
investigations of alternative fuels, assay systems must
become available which are capable of detecting possible
deleterious effects of exposure to the combustion products
of various organic materials. The effluents produced during
combustion of these fuels generally contain a myriad of
chemicals, some harmless, some potentially genotoxic. A
test system is therefore needed which is capable of detect-
ing potentially genotoxic compounds in such mixtures. One
assay system, which has shown its utility in detecting the
genotoxic activity both of individual chemicals and complex
mixtures (1,2,3, Schectman et al., manuscript in prepara-
tion) has as an endpoint the morphological transformation of
BALB/c 3T3 (3T3) mouse embryo cells (1,2). Mouse 3T3 cells
are said to be "transformed" when they undergo certain
morphological (1,3) and biological (1) changes that result
in a phenotype similar to that expressed in vitro by tumor-
derived cells. These in vitro transformed cells are gener-
ally assumed to be similar to malignant cells in vivo since
transformed 3T3 cells cause tumors when injected into a
syngeneic host (1). We have found that various drugs,
pesticides, dyes, food additives, pre- and post-combustion
products and industrial waste products can induce the
transformation and/or mutation of 3T3 cells. Recently we
have developed a mutagenesis assay for the BALB/c 3T3 cells
in which we monitor the ouabain-resistance (OUAr) pheno-
type (Schechtman et a!., manuscript in preparation). This
assay has enabled us to simultaneously determine both the
transforming and mutagenic activity of various compounds in
one cell system. In this paper we show the results obtained
when extracts of particulates collected from various emis-
sions sources, including diesel engines, are tested for
mutation and transformation of RALR/c 3T3 cells in the
presence and absence of exogenous metabolic activation.
MATERIALS AND METHODS
CELL LINE
The BALB/c 3T3 (3T3) cells used in this study were from
passages 12-18 of clone A31-1 established by Dr. Takeo
Kakunaga (1).
862
-------
CHEMICALS
Benzo(a)pyrene (Ba^), NAPH, NAPPH, NADP, and ouahain were
obtained from the Sigma Chemical Company St. Louis, MO.
Acetone was obtained from the J. T. Raker Chemical Company,
Phil 1ipshurq, MJ, and dimethylsulfoxide (DMSO) was obtained
from Fisher Scientific Company, Fair Lav/n, MJ. N-methyl-N'-
nitro-N-nitrosoguanidine (MNNG) was purchased from the
Aldrich Chemical Company, Milwaukee, WI. The various test
articles were provided by the Environmental Protection
Agency, Research Triangle Park, NC. Methods of collection
and extraction of the test articles are presented elsewhere
in this publication.
TRANSFORMATION AND MUTAGENESIS ASSAYS
Growth medium was Fagle's minimum essential medium (MA
Bioproducts, Walkersville, MD) supplemented with L-glutamine
(2.0 mM), non-essential amino acids (0.1 mM), penicillin
(100 units/ml), streptomycin (100 ug/ml) and 10% heat-
inactivated fetal bovine serum (Lot number 2910118, Flow
Laboratories, McLean, VA). This lot of fetal bovine serum
was previously checked for its ability to support growth
of 3T3 cells in terms of cell generation time, plating
efficiency, saturation density and maintenance of an even
cell monolayer for at least five weeks.
In the mutagenesis and transformation assays with metabolic
activation, 1-2 x 10^ 3T3 cells were suspended in a 4 ml
reaction pool (pH 7.4) which consisted of phosphate buffered
saline, 1.8 mM NADPH, 1.4 mM NADH, 7.4 mM glucose-6-phos-
phate and 1.9 mM NADP. To this was added 100 ul of a 9,000
x g supernatant (S-9) from the homogenized livers of Aroclor-
1254 induced Fischer 344 rats (2), plus the appropriate
amount of chemical to be tested. For assays performed
without exogenous metabolic activation the reaction mixture
consisted of 3T3 cells in 4 ml growth medium plus the
chemical to be tested. The reaction mixture for either
assay was incubated with continuous gentle agitation at 37°
C for 2 hours in a humidified atmosphere of 5% CO;? in air.
The reaction was stopped by centrifugation (1000 x g, 4° C,
10 min). The cell pellet was washed once in Hank's balanced
salt solution (MA Bioproducts, Walkersville, MD) and then
suspended in growth medium. This final suspension was used
as the source of cells for the cytotoxicity, transformation
and mutagenesis assays.
For the transformation assays, cells were seeded at 1 x
10^ cells/60mm dish and incubated for 4 weeks with medium
863
-------
changes twice a week. The cells were then fixed (methanol),
stained (Giemsa), and examined for Type II and Type III
morphologically transformed foci using the criteria estab-
lished by Reznikoff _et_al- > (3) for C3H 10T !/2 cells. For
the cytotoxicity assay, cells were seeded at a density of
250 cells per 60 mm dish. The cytotoxicity data was then
used to calculate the number of cells at risk in the trans-
formation assay.
For the mutagenesis assay, cells from the post-reaction
suspension were seeded into 75 cm^ culture flasks and
grown for 48-96 hours to allow for expression of the muta-
genic events. Cells were then collected by trypsinization
and plated at a density of 2 x 10^ cells/60 mm dish for
selection of mutant colonies resistant to 1 mM ouabain
(supplied fresh each week) according to the procedures of
Raker et al., (4). Plating efficiency assays were run con-
comitantly with the mutagenesis assays so that mutation
frequencies could be determined. Cells were fixed and
stained after 7-9 days for determination of cloning effi-
ciency and after 21 days for determination of the number
of OUAr cells.
RESULTS
In order to establish a baseline for the sensitivity of the
test system in the absence or presence of exogenous metabolic
activation, two model compounds were assayed. MNNG, a
compound which is genotoxic for most micro-organisms and
which does not require metabolic activation, was tested at
seven different concentrations in the absence of metabolic
activation (Table 1). MNNG induced Type II foci, Type III
foci, and ouabain-resistant (Qua1") mutants at several
dose levels. Type III foci were observed at lower doses
(0.03 ug/ml) than Ouar mutants (0.3 ug/ml). A second
chemical (BaP) was also assayed to determine the sensitivity
of the system in detecting a compound which requires meta-
bolic activation for its mutagenic/carcinogenic activity.
Table 2 shows that both the transformation and mutation
assays coupled with a source of exogenous metabolic acti-
vation could detect BaP at a level of 0.3 ug/ml. Data
obtained with either chemical demonstrate a linear dose
response. Linear correlation coefficients for MNNG were
0.92 and 0.97 for transformation and mutagenesis, respec-
tively. Correlation coefficients were 0.95 for the trans-
formation and 0.92 for the mutagenesis induced by doses of
BaP between 0.01 and 3 ug/ml. The transformation and
mutation frequencies induced by 10 ug BaP/ml were lower than
expected and may reflect either the solubility of the
compound or a differential response of the 3T3 cells with
864
-------
rH
W
J
CO
<^
EH
O
Z
z >-<
S EH
M
[ti ^>
O M
EH
>H CJ
EH <
M
> O
M H
EH r4
U 0
i
Z r^
< CO
^3
-o
U Z
H W
z o
U O
U X
< w
EH
S 0
- w
0 U
H Z
X W
o en
EH CO
0 <
EH
>< W
u n:
EH
a
x z
EH M
Cu CO
O J
r^
Z W
o u
M
EH ' '
< 1
Z rH
M ro
S ^
rj ,
H rH
W U
Q
ro
CO EH
^ ro
o
w o
z \
l^^^HOO (N
O> fN V V v
^rr~^t*DOoO' — i (N
•^ i— i
m in iri if) u~i m •*& *r
i— IfHi — 1 i — IP— li — 1 i — 1 i— 1
^x "x, "x, \, ^x \^ \^ \^
iHCNrNOi— IrHO i— 1
r— 1
coou~i^o>i/^o> CN
(Ti^ro^r-^ro m
O-H.HOOOO O
i — 1 V V V
^rrNjrsJorsjrHo o
rH
OO^Oi — lOCNJtH
in fO rH IT)
oinrorHooo (N
rHOOOOOO O
u o
:s co
-z s
S Q
Ul
rH C
-a o
— , | | -iH C
0) TJ O r\
O O C O ^ W
O O _n 0)
r^ UH o Vi 75 xl
ro O 0) .^ CO
VH C^ V --H
4-1 in 0) -O
n) 0) di-H ^
x; 3 o WH a4-i
c o .H o
(N UH M £ 4-1
^ M 0 ,
)-l 0) H H ^ (H
o an ai
MH 0) M d
CO 0) d. M
C CD QJ >i i-i --H /-
O -rH >,EH 0 n
•H C EH rj 0 .^
Wl 0 -0 Q.UH ^
C rH -O (U -^ ^
0) O CD E (CrH rg
a o E >H ^ H c
3 H_i o <2 ^ M °
ra o ^ co ^ o)
ra c L, O4 s-<
C >H C (0 JJ >i Q
XI rH 4-) p
rH £ 4-1 ~2 .C rn
Hi •S >, d 4J .,
O C >.rH C .H .^
•H >lrH rH __ J j-
0) »CO(T3UI>iCOJl
.c ojtnu-HuDo
y *(d*iHCPCXIn
•H tJ1 O 0) W
4-1 >1 0 *-H rJ -H ,,
ra '4JrHOtTrat-
o) >,-.H o x; 0) Is
C '"H CLl r4 4H O r-.
+ 1 (U X rj O g
T3 0 4-1 E O 0)^
4^j il2uH>oli'i§
(T3 **H >i O (T3 3 o
01 0) O rj E Z
M ^ CJ l-t ^
4J"CnD^O)^OII O
djr^.^.^'c I'ra-a V-i
r4Oc^3zc4-io)
01XIOCZ (O^-Xl
01 CJ O. II EH M 0
MS e n HZ
rH II O M M || |
rH -O O M M EH II
01 C W U 1 It,?
O ^0 C_) ^J E™* E^ E~* 'O "^
m «
a -*-1
"^ 0
ra 4J
_J r4
'c a
rH
p ri
U -H
c
^L /^
S „
o n
i^M
° 0
GJ _£-
"c 4-1
fc -r-t
*7 3
"^
.* W
QJ
>^ rC
u to
C -H
D
CTM-i
II
S T3
4-1
II \
S
s -o
*s; 'V
865
-------
ft >>
a EH
CQ M
^
Cn rH
O EH
O
EH
> H
M J
EH O
U CQ
£H
O W
z £
2J O
ffi H
O H
CO ft
Z ft
« CO
EH
><
D tJ
Z CO
< 3
O
•• Z
U W
W X
O W
a o
s
M
- o
U Z
M W
X CO
0 If
EH K
O ft
EH
> W
O K
W
X Z
EH 1-H
6-1 en
O J
z w
o u
M
EH rH
< 1
Z rH
H r^i
ti "*
W •
EH rH
W U
Q
CO EH
D m
O
W O
2 ^v
< CQ
EH <-}
I-H1 iVDLnr- to^rHo
i-H
vDT-r^m^ro^r-oai
^ororgr-inMr-HCOrHo
Cs| CD CM V t-^ \J V
-ior-^inoo^ooo
M m I-H
^TrininLnmLnTLnin
\V\\"X\S\X.\'X\
(Nj
oroi-Hooooooo
rH i— t
1)
c
o
a< -u
O^ fO (L)
1 CQ U
CO < (D
+ c
+ + o
d< 1 CTi 01
TJ OD 1 O
CQ < CO i<
c
- — -H
a% f3
MH 1 XI
O C/l fO
•~^ C T) D
0) H QJ O
(J jJ T3
C C TJ OJ O
OJ nj cu O -U
W -P T3 C W
OJ fti 0) O T5
Qj M U) O -P rH U)
-H C U
4J O Q) t3 O Wl
to o JH (i-H JH -H (U 4-1 MO D rH
O CXS 0. M OJ )H 0) -P
MHtU DJ QJ rH CLi QJ C, O
x T3 o txn a. 4-)
C CU] Qj^H-rH U]
•H - CO
(X -HO S>H MrHOQJ
D^TO) O4HO U C
tn-^r^jiM MHUI o r-iO
C :£ dT3QJ>iCr^O
0)0)0, i-J-Pfc -P >-*
rtJOU)6 >,C4->£OO
O CO — '3 >rH -H O
^ 4J •WU'HCQJ4-JE -H
O 4-1 t/J QJ fO -H tT^ QJ ,C CO ^ J
(C >i • tr-o^wHC
-PHcn-rH>iO-HCr-Htn to
Ul 4J^HO.O ESO^M
SS ^"a! ° ^ OJO^MH cxi
CUtT1CXl6^ H^
Q) in D C ~H £ 3 C ^ in -P S
qj ,— t ^ c C 2 ^"* ~\ ^ 4-J ||
SO On3 IIEHMS3
OCrHCUll MdSTJ
WMftJUE M N H 21 -P
rH ri] M O H M | || \_
rH TJOMM £H || S
QJCCWOI Irx.^ CM^
CS ~O O ~*3 Cj *-<= CT)~S^ 'V
866
-------
z
o
M
2 EH
O <
H >
CO H
CO EH
H U
£ <
td
u
iJ H
< hJ
EH 0
Z CO
td <
S EH
Z U
0 S
tx
M CO
> 3
Z 0
Id Z
CO O
3 O
O X
H td
C£
0
S td
O O
tX 2
In Id
CO
CO Id
EH Ci
CJ Qj
<
K P*!
EH O
X
Id Id
CJ
fc. Z
O td
CO
>H PQ
EH <
H
U W
M I
X EH
0
H Z
M
Id
a co
EH Id
U
a
o
CO
Id
o
w
en
Id
OH
-O
W
CO
EH O —
Id ^
O M U"1
.J X 3-
o —
EH
W
C5
Z « ~
< a -H
2 W E
Id CO tn
en id 3
O EH ~
Q
Id
Z
w
ca
•o
td
CO — .
EH O r-H
CO D E
td \
S U Cn
0 M 3
J X —
o
EH
w
u
2 « '—
< O •-)
2 w E
U W tP
W W P-
O H ^
Q
IK
o z
O EH
Id H U
cj en rf
D H EH
O S X
M td Id
0) QJ QJ
o o c o c c
O CO O -^ O O
CN <~\ 2; 22
O O
O O O o 0
o co o o in o
ro oj ro to ( j — I
1 | | i-H IT) 1
O O O 1 O
i-H rH •— 1 LO (N r— I
0) OJ 0)
o o c tfi c c
o n- O O O
IN 2 CM 2: 2
o o o
o o o o o o
o co o r- ^ in o
ro CN <^i I i i— i
( i i in in i
o o o -o
r-i i-H i— ( rsj (N (-H
Q)
c
a» H
C C C CP
M H -H ^t H C
fdtji QJCT1 T3 tr^W
•HC-HCH CM C
•-i W -H cq DJ oj
a--" o M c: c ^H c -H o
V-lQJ Efl) -H (^(U (ti (— i
1 — 1
T3 O
OJ en
-p
rO QJ
-P -C
w -P
(U O
X U
4-)
O
C >
•H -H
rC ^
4-J IT]
S CD
•u
CD ^H
-P cq
U] ^
CD H
^
CD D
M CO
CD
S oy>
o
01 CO
CD
CO C
O fO
T3 J3
4J
C
OJ 0)
> CO
CD CD
en ^
td -Q
867
-------
increasing BaP concentration. Such a decrease in the
transformation frequency induced by high doses of polycyclic
aromatic hydrocarbons has been observed by others (1).
Dichloromethane (DCM) extracts of particulates from diesel
and related environmental emissions were tested at seven
doses with and without a source of exogenous metabolic
activation. Dose selection was determined by preliminary
cytotoxicity assays. If the test article was toxic, a dose
range was selected which included at least one toxic dose.
If the compound was not toxic in the preliminary assay, it
was tested through a range of 7 doses with an upper limit of
1 mg/ml. A further consideration used for dose selection
was to have as many common doses as possible among the test
chemicals. Table 3 lists the doses tested for each sample
as well as the relative toxicity of the sample to 3T3 cells
either in the presence or absence of exogenous metabolic
activation. The three diesel DCM particulate extracts were
all toxic at a minimum of one dose, while the three compara-
tive samples showed no toxicity during the treatment period
for the simultaneous mutagenesis/transformation assay. The
addition of exogenous metabolic activation had no effect on
the toxicity of the Caterpillar sample but reduced the
toxicity of the Oldsmobile and Nissan diesel samples.
The results from the transformation and mutation assays with
the six particulate extracts were less easy to interpret due
to the lack of a clear dose response. In general, even
though several individual doses of a particular DCM extract
did induce a significant increase in either the frequency
of transformed foci or OWJr colonies, a dose-dependent
increase in either transformation or mutation frequency
was not found (see Discussion for further comments).
The majority of the PCM extracts were above their limit of
solubility in the aqueous growth medium at all seven test
doses. Therefore it may be reasonable to assume that all
seven dose levels were, in fact, quite similar. With this
assumption the data for all doses of a given DCM extract
were combined, e.g., the total number of Type III foci
observed for an individual extract were divided by the
total number of surviving cells. This frequency was then
considered as the resultant of the dose range as specified
in Table 3. At the same time the data from the appropriate
negative (solvent) control (one negative control as assayed
concomitantly with each extract) were combined to give the
historic background transformation and mutation frequencies
for this assay series. A positive control of either 1.0 ug
MMNG/ml for assays in the absence of exogenous metabolic
activation or 12.5 ug RaP/ml for assays in the presence of
exogenous metabolic activation was tested simultaneously
868
-------
TABLE 4
TOTAL GENOTOXIC EVENTS OCCURING AFTER
EXPOSURE TO VARIOUS DOSES OF EMISSION EXTRACTS
IN THE ABSENCE OF EXOGENOUS METABOLIC ACTIVATION1*
SOURCE OF
EMISSION
EXTRACT OR
TYPE OF CONTROL
TRANSFORMATION
MUTATION
TYPE III TF , OUAr MF
FOCI (xlO~5) COLONIES (xlO~6)C
CATERPILLAR
DIESEL ENGINE
OLDSMOBILE
DIESEL ENGINE
ROOFING TAR
NISSAN
DIESEL ENGINE
COKE OVEN
MUSTANG II
0
1
2
14
6
11
<0
0.
0.
3.
2.
4.
68
33
61
43d
13d
10d
1
2
20
5
45
27
0.
0.
3.
1.
8.
4.
18
53
14rf
06d
27d
49d
GASOLINE ENGINE
POSITIVE CONTROL 29
(1 ug MNNG/ml)
NEGATIVE CONTROL Q
(0.25% solvent)
12.34a 139 35. 5a
<0.34 1 0.18
a Doses for the various emission extracts were as
discussed in Table 3, however, data from doses
giving <30% survival were not included.
b TF = Transformation frequency for Type III foci only.
c MF = Mutation frequency.
d Data significantly different from combined negative
controls at p <0.05.
869
-------
with each extract. These data were also combined to give an
historic positive control for this assay series. Tables 4
and 5 list these resultant frequencies obtained from the
combined data and compare each test article to the appro-
priate historic solvent control. It can be seen that in the
absence of exogenous metabolic activation (Table 4) the
Oldsmobile and Caterpillar diesel samples did not induce a
significant number of mutants or transformed foci. However
the Nissan, coke oven and Mustang II samples were quite
active in inducing both a significant number of transformed
foci and OUAr colonies. The roofing tar pot sample
induced a high frequency of OUAr colonies but not a
significant number of transformed foci.
When the same samples were tested in the presence of exo-
genous metabolic activation similar results were found
(Table 5). The Caterpillar and Oldsmobile samples again
showed no genotoxic activity in either the transformation or
mutation assays. The roofing tar and Nissan diesel samples
were mutagenic, but the frequencies of transformed foci
which they induced were not statistically different from the
combined negative controls. Both the coke oven and Mustang
II samples induced a significantly high transformation and
mutation frequency. Even though a large number of mutants
were observed after treatment with the coke oven sample, the
combination of the high spontaneous frequency which occurred
during this specific assay and the lack of a dose response
made the determination of significance difficult.
DISCUSSION
An assay system utilizing BALR/c 3T3 cells can function
very well for routinely detecting both the transforming
and mutagenic potential of many compounds. Historical
data from our laboratory show remarkably low spontaneous
transformation and mutation frequencies, allowing the
system to detect chemicals with fairly weak biological
activity. The simultaneous use of both the transformation
and mutation endpoint greatly enhances the utility of the
test system. The in vitro conditions that effect this assay
system will be described at length elsewhere (manuscript in
preparation).
Preliminary studies with MNNG and BaP, chemicals also used
as positive controls in the absence (MNNG) and presence
(BaP) of exogenous activation for each individual DCM
extract assay, showed that the 3T3 cells can respond in
a dose-dependent fashion for both mutagenesis and trans-
formation (see Tables 1 and 2). However, when the complex
870
-------
TABLE 5
TOTAL GENOTOXIC EVENTS OCCURING AFTER
EXPOSURE TO VARIOUS DOSES OF EMISSION EXTRACTS
IN THE PRESENCE OF EXOGENOUS METABOLIC ACTIVATIONa
SOURCE OF TRANSFORMATION MUTATION
EMISSION
EXTRACT OR TYPE III TF , OUAr MF
TYPE OF CONTROL FOCI (xlO~5) COLONIES (xlO~6)c
CATERPILLAR Q
DIESEL ENGINE
OLDSMOBILE Q
DIESEL ENGINE
ROOFING TAR 4
NISSAN 3
DIESEL ENGINE
COKE OVEN 6
MUSTANG II 7
GASOLINE ENGINE
POSITIVE CONTROL 19
(12.5 ug BaP/ml)
NEGATIVE CONTROL 0
(0.25% solvent)
<0.30 1 0.20
<0.55 2 0.49
1.07 11 1.73d
1.04 10 l.8ld
2.41d - - e
2.92d 19 3.97rf
10.3d 48 14. 2d
<0.36 1 0.26
a Doses for the various emission extracts were as
discussed in Table 3, however, data from doses
giving <30% survival were not included.
b TF = Transformation frequency from Type III foci only.
c MF = Mutation frequency.
d Data significantly different from combined negative
controls at p < 0.05.
e Spontaneous MF for this assay was significantly
higher than the historical controls for this assay
series.
871
-------
emissions extracts were assayed, none yielded a dose-depen-
dent increase in either mutation frequency or transforma-
tion frequency. This lack of dose response, even though
individual doses of several of the extracts showed consi-
derable genotoxic activity, could be due to a combination
of reasons. The insoluble nature of the extracts in the
aqueous growth media, the short (2 hours) exposure time,
or the limited number of cells at risk, could all have
been contributing factors.
The absence of a dose response makes quantitative compari-
sons among the extracts difficult. However, if the total
number of genotoxic events (Type 111 foci or OUAr clones)
for the seven assay doses (five doses in two cases since
only data obtained at survivals > 30% was used for these
calculations) of each DCM extract tested was divided by the
total number of cells at risk in that assay, a frequency
could be obtained which was reflective of the activity of
the extract over the range of concentrations assayed. The
level of significance of each of these frequencies was
determined by comparison to historic negative (solvent)
control frequencies obtained by combining the individual
negative controls from each OCM extract assay. Data from
controls in the absence or presence of exogenous metabolic
activation were combined separately. Using this type of
analysis, Tables 4 and 5 show that DCM particulate extracts
from the six sources can be rated as follows: Mustang II
gasoline engine = coke oven > Nissan diesel engine > roofing
tar. The Caterpillar and Oldsmobilo diesel engine emission
particulate extracts showed no significant genotoxic acti-
vity for 3T3 cells. However, when drawing any further
conclusions, e.g., on the relative health hazards of the
various emissions sources, the total amount of emission that
each source normally emits to the atmosphere as well as the
quantitative relationship of each extract to the actual
emission (see Huisingh, et a!., this symposium) must be
considered.
REFERENCES
1. Kakunaga, T. 1973. A quantitative system for assay
of malignant transformation by chemical carcinogens
using a clone derived from RALB/c 3T3. Int. J. Cancer
12:463-473.
2. Schechtman, L. M., and R. E. Kouri. 1977. Control of
benzo(a)pyrene-induced mammalian cell cytotoxicity,
mutagenesis and transformation by exogenous enzyme
fractions. Progress in Genetic Toxicology. D. Scott,
872
-------
B. A. Bridges, and F. H. Sobels, eds., Elsevier/North-
Holland Biomedical Press, New York. pp. 307-316.
3. Reznikoff, C. A., D. W. Brankow, and C. Heidelberger.
1973. Establishment and characterization of a cloned
line of C3H mouse embryo cells sensitive to post-con-
fluence inhibition of cell division. Cancer Research
33:3231-3238.
4. Baker, R. M., D. M. Brunette, P. Mankovitz, L. H.
Thompson, G. F. Whitmore, L. Siminovitch, and J. F.
Till. 1974. Ouabain-resistant mutants of mouse and
hamster cells in culture. Cell 1:9-21.
General Discussion
D. BRUSICK: Did you say that all of the samples were
positive?
R. CURREN: 1 said we got a positive response in the
form of type three foci with all the samples that we had.
We would consider the Oldsmobile extract and the cater-
pillar extract moderately active, Benzopyrene, for ex-
ample, tested in this system gives nothing.
W. THILLY: Many mutagenicity assays did show that
benzopyrene whether in the alpha or the epsilon form,
have activities within the same order of magnitude. For
instance with the Salmonel, 1 a typhimurium benzo(e)pyrene
test,has about ten" percentfof the activity on a molar
basis with benzo(a)pyrene.
R. CURREN: Actually I didn't show the slide, but we
do see mutagenicity with benzo(e)pyrene.
W. THILLY: I think you might have been slightly mis-
leading during your presentation when you said benzo(e)-
pyrene has been shown negative in carcinogenicity assays
and mutagenicity assays everywhere.
R. CURREN: I thought I said in general, low activity
is found with benzo(e)pyrene. Of course, this activity
could result from contamination with benzo(e)pyrene.
873
-------
MUTAGENIC AND CARCINOGENIC POTENCY OF EXTRACTS OF
DIESEL AND RELATED ENVIRONMENTAL EMISSIONS: TWO-
STAGE CARCINOGENESIS IN SKIN TUMOR SENSITIVE MICE
(SENCAR)1
2
T. J. Slaga and L. L. Triplett
Biology Division, Oak Ridge National Laboratory, Oak Ridge,
Tennessee 37830
and
Stephen Nesnow
Health Effects Research Laboratory, U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina,
27711
ABSTRACT
Skin tumors can be induced by the sequential application of a sub-
threshold dose of a carcinogen (initiation phase), followed by re-
petitive treatment with a noncarcinogenic tumor promoter. There
is a very good dose-response relationship between the induction
of the number of papillomas per mouse at early times (10 to 20
weeks) by either tumor initiators and promoters and the final
carcinoma incidence after a longer latency (20 to 50 weeks) in
SENCAR mice. This system not only can be used to determine
the tumor initiating and promoting activities of a compound but
if the agent is given repeatedly by itself one can also determine
if it is a complete carcinogen, i. e., if it has both tumor
initiating and promoting activity. In addition, if the agent is
given concurrently with a known complete carcinogen or a turn or
initiator one can also determine if the agent has co-carcino-
genic or co-tumor initiating activity or even possibly anti-
carcinogenic activity. Likewise, if the agent is given concur-
rently with a known tumor promoter one can determine if the
agent has co-promoting or anti-promoting activity. Using the
Research sponsored by U.S. EPA contract Number 79-D-X0526
under Interagency Agreement DOE Number 40-728-78 and the
Office of Health and Environmental Research, U.S. Department
of Energy, under contract W-7405-eng-26 with the Union Car-
bide Corporation.
2
To whom requests for reprints should be addressed at Biology
Division, Oak Ridge National Laboratory, P.O. Box Y, Oak
Ridge, TN 37830.
874
-------
SENCAR skin carcinogenesis system we have undertaken the
determination of the skin carcinogenic, co-carcinogenic, tumor
initiating and promoting activities of various diesel emission
particle extracts as well as for comparative purposes, stand-
ards such as benzo(a)pyrene and 12-0-tetradecanoylphorbol-13-
acetate and extracts of emissions from a gasoline engine,
roofing tar, coke oven and cigarette smoke condensate. Most
of the studies are still in progress but some preliminary data is
available on the comparative tumor initiating activities of the
various samples at 14 weeks. Caterpillar Diesel and cigarette
smoke condensate were essentially without activity, whereas the
Mustang gasoline Catalyst and Olds Diesel gave extremely low
values although they were slightly above background. Roofing
tar, coke oven and Nissan Diesel all gave moderate activity at
high doses (1 to 10 mg) but when compared to benzo(a)pyrene
were relatively low values.
INTRODUCTION
Based on knowledge derived from epidemiological studies, it is
currently thought that the majority of cancers in humans are
caused by environmental factors. Studies with experimental
animals along with the epidemiological data have provided evi-
dence that some chemicals in our environment are responsible
for a significant proportion of such cancers.
Current information suggest that chemical carcinogenesis is a
multistep process with one of the best studied models in this
regard being the two-stage carcinogenesis system using mouse
skin. Table 1 summarizes some aspects of carcinogenesis in
experimental animals. Skin tumors can be induced by the se-
quential application of a subthreshold dose of a carcinogen
(initiation phase) followed by repetitive treatment with a non-
carcinogenic tumor promoter. The initiation phase requires
only a single application of either a direct or indirect car-
cinogen at a subthreshold dose and is essentially irreversible,
while the promotion phase is brought about by repetitive treat-
ments after initiation and is initially reversible, later be-
coming irreversible. This system not only can be used to
determine the tumor initiating and promoting activities of a
compound but if the agent is given repeatedly by itself one
can also determine if it is a complete carcinogen, i. e., if
it has both tumor initiating and promoting activity. In addition,
if the agent is given concurrently with a known complete car-
875
-------
TABLE 1
CARCINOGENESIS IN MAN AND EXPERIMENTAL ANIMALS
1. Complete Carcinogenesis
2. Cocarcinogenesis
3. Tumor Initiation
4. Tumor Promotion
5. Additive and Synergistic Effects of Carcinogens, Tumor
Initiators and Tumor Promoters
6. Co-initiating and Co-promoting Agents
7. Anti-Carcinogenesis
8. Anti-initiating and Anti-promoting Agents
cinogen or a tumor initiator one can also determine if the agent
has co-carcinogenic or co-tumor initiating activity or even pos-
sibly anti-carcinogenic activity. Likewise, if the agent is given
concurrently with a known tumor promoter one can determine
if the agent has co-promoting or anti-promoting activity. Fur-
thermore, like most Carcinogenesis systems, skin carcinogens
may have additive or synergistic effects. This system has pro-
vided an important model for studying Carcinogenesis and for
bioassaying carcinogenic agents.
Recently, the generality of the two-stage system or multistage
Carcinogenesis has been shown to exist in a number of systems
besides the skin such as the liver, bladder, colon, esophagus,
stomach, mammary, displacental as well as cells in culture
(1). The various two-stage systems known and the initiating
and promoting agents involved in each system are shown in
Table 2. As is apparent, quite a diversity of initiating and pro-
moting agents exists among the various two-stage systems.
Tumor Initiation
As stated earlier, the tumor initiation phase appears to be an
876
-------
irreversible step which probably involves a somatic cell muta-
tion as evidenced by a good correlation between the carcino-
genicity of many chemical carcinogens and their mutagenic
activities (2, 3). Most tumor initiating agents either generate
or are metabolically converted to electrophilic reactants,
which bind covalently to cellular DNA and other macromole-
cules (4, 5). Previous studies have demonstrated a good cor-
relation between the carcinogenicity of several polycyclic
hydrocarbons and their ability to bind covalently to DNA (6, 7).
The Millers have proposed a significant general theory to ex-
plain the initial event in chemical carcinogenesis which states
that all chemical carcinogens that are not electrophilic re-
actants must be converted metabolically into a chemically
reactive electrophilic form which then reacts with some
critical macromolecule to initiate the carcinogenic process
(4,5).
Cocarcinogenesis and Tumor Promotion
The term cocarcinogenic action means the augmentation of
carcinogenesis by a noncarcinogenic agent applied con-
comitantly with a carcinogen (8). A promoting agent is one
applied repeatedly after a single dose of a tumor initiating
agent that results in tumors (8). Tumor promoters can be
either weak carcinogens or noncarcinogens. Some cocar-
cinogens such as croton oil, phorbol esters and anthralin
are also tumor promoters. A cocarcinogen primarily has
an influence on the initiation of the carcinogenic response
through an effect on the carcinogen (permeability, meta-
bolism or repairability of the carcinogenic damage) or the
target tissue (for example, hormonal stimulation of mam-
mary gland growth and croton oil stimulation of epidermal
growth). A wide variety of agents have been found to be
cocarcinogenic in experimental animals such as aliphatic
hydrocarbons, aromatic hydrocarbons, aromatic hydro-
carbons, phenols, long-chain acids, alcohols, hormones,
radiation, viruses, food additives and diet, itself (1).
Table 3 lists some of the more important cocarcinogenic
agents that have been tested (1, 8). Due primarily to the
ease of determining cocarcinogens as well as promoters
in the skin system, most of the cocarcinogens have been
discovered using that system.
Refer to Table 2 which summarizes the various two-stage
877
-------
TABLE 2
TWO-STAGE CARCINOGENESIS SYSTEMS
Organ
System
Mouse
skin
Rat 81
mouse
liver
Mouse
lung
Rat
colon
Polycyclic aromatic hydro-
carbons, urethane, direct-
acting electrophiles
[(epoxides, (3-propiolactone,
Bis chloromethylether, N-
acetyoxyacetylamino -
fluorene and N-methyl-N'
-nitro-N-nitrosoguanidine
(MNNG)]
2-acetamidofluorene,diethyl-
nitrosamine, 2-methyl-N,
N' -dimethyl-4-aminoazo-
benzene
LJrethan, poly cyclic aro-
matic hydrocarbons
Dimethylhydrazine
croton oil, phorbol
esters, fatty acids and
fatty acid esters, sur-
face-active agents,
linear long chain al-
kanes, tobacco smoke
condensate and ex-
tracts of unburned to-
bacco, certain eu-
phorbia macrocyclic
diterpenes, citrus oil,
anthralin and other
phenols
phenobarbital, DDT;
butylated hydroxy-
toluene, (BHT); poly-
chlorobiphenyls,(PCB).
BHT, phorbol
Bile acids, high fat
diet, high cholesterol
diet
Rat N-methyl-N-nitrosourea
bladder
Rat & Polycyclic aromatic
mouse
mammary
gland
Rat N-methyl-N'-nitro-N-
stomach soguanidine
Rat Diethylnitrosamine
esophagus
Saccharin, cyclamate
Hormones, high fat
diet, phorbol
Surfactants
Diet
878
-------
TABLE 2 (continued)
Organ
System
Mouse cell Polycyclic aromatic hydro- Phorbol esters, sac-
culture carbons, radiation charin
systems
Rat cell N-methyl-N' -nitro-N- phorbol esters
culture nitrosoguanidine
system
o
See reference 1 for details
systems used and the various promoting agents associated with
them. In many cases it is very difficult to distinguish between a
cocarcinogen or a promoter because many of the agents have not
been critically tested.
The phorbol ester tumor promoters are the most potent of the
known tumor promoters, having been shown to have many cellular
and biochemical effects (1,9,10). Some of the most important
effects which seem to be intimately associated with the phorbol
ester tumor promoters' action are as follows: (a) They induce
changes in the phenotype of normal cells that mimic features of
transformed cells, (b) They induce dedifferentiation of adult epi-
dermal cells to embryonic looking cells, (c) They induce a 200-
400 fold increase in epidermal ornithine decarboxylase activity
which specifically correlates with tumor promotion, (d) They in-
duce protease activity, (e) They stimulate phospholipid synthesis
and DNA, RNA and protein synthesis, (f) They stimulate cellular
proliferation, (g) They inhibit terminal differentiation not only in
the skin but in other cellular systems such as Friend erythro-
leukemic cells, chicken myoblasts and neuroblastoma cells (1).
Only a limited number of studies have been performed con-
cerning the mechanism by which other tumor promoters act.
Both phenobarbitol and BHT stimulate DNA synthesis followed
by cell proliferation in the liver and lung, respectively (1).
879
-------
TABLE 3
COCARCINOGENS IN VARIOUS TISSUES*
Skin
Catchol
pyrogallol
lauryl alcohol
decane
undecane
tetradecane
n-dodecane
pyrene
benzo(e)pyrene, B(e)P
fluoranthene
benzo(g, h,i)perylene
anthralin
croton oil
phorbol esters
phenols
nicotine
surfactants
radiation
viruses
Lung
asbestos
radiation
n-dodecane
hormones
estradiol
Mammary gland
ferric oxide
magnesium oxide
hypoxia
prolactin
Bladder
L-tryptophan
saccharin
cyclamate
Check Pouch
X-radiation
croton oil
Liver
cyclopropenoid fatty acids
alcohol
See reference 1 for details.
880
-------
MATERIALS AND METHODS
Animals
SENCAR mice, a mouse stock selected for its increased sensi-
tivity to carcinogenesis as described by Boutwell were used in
this study (9). These mice were derived from breeding Charles
River CD-I mice with male STS Skin Tumor Sensitive originally
derived from Rockland mice by Boutwell (9) and selecting for
eight generations for sensitivity to DMBA-TPA two-stage system
of tumorigenesis. These mice were initially obtained from Dr.
Boutwell, McArdle Laboratory for Cancer Research, University
of Wisconsin, Madison, Wisconsin, and now are being raised at
the Oak Ridge National Laboratory, Oak Ridge, Tennessee, In
addition, C57 black mice obtained from Jackson Laboratory,
Jackson, Maine are also being used for these studies. Since the
sensitivity to skin cancer is much less than the SENCAR mice
tliis report will only deal with the results from the SENCAR
mice.
Chemicals
TPA was obtained from Dr. P. Borchert, University of Minnesota,
Minneapolis, MN. DMBA was purchased from Sigma Chemical
Co., St. Louis, MO; BP and pyrene from Aldrich Chemical Co. ,
Milwaukee, WT. The various diesel emission particle extracts
as well as the other environmental emission samples were ob-
tained from EPA under a code designation. All the agents were
consistently prepared under yellow light immediately before
use. In addition, all samples were applied topically in 0. 2 ml
of spectral quality acetone.
Tumor Experiments
The initial studies related to comparing tumor initiation and
complete carcinogenesis of many carcinogens and the dose-
response studies using DMBA, BP and TPA involved the use of
only SENCAR mice with 30 to 40 females per group. The other •
studies related to the various emission samples obtained from
EPA utilized two strains of mice (SENCAR and C57 black) with
80 mice per treatment group (40 males and 40 females). In
addition, 5 dose levels were used for the carcinogenic, tumor
initiating, cocarcinogenic and tumor promoting activities of
881
-------
the various samples. BP was used as the standard for the com-
plete carcitiogenesis and tumor initiation studies and TPA for
the tumor promotion studies. Pyrene was used as the positive
control for cocarcinogenesis studies with BP as the carcinogen.
See Table 4 for a summary of objective and protocol. All the
TABLE 4
OBJECTIVES OF THE DIESEL RESEARCH PROGRAM
Determine; Complete carcinogenesis
Cocarcinogenesis
Tumor Initiation
Tumor Promotion
Samples: Olds Diesel
Nissan Diesel
U.W. Rabbit Diesel
Caterpillar Diesel
Mustang gasoline catalyst
Coke Oven
Roofing Tar
Cigarette Smoke Condensate
Protocol: SENCAR and C57 Black Mice
40 males and 40 females
5 dose levels
Standards: Benzo(a)pyrene for complete carcinogenesis
and tumor initiation
Pyrene for cocarcinogenesis with benzo(a)-
pyrene
Phorbol ester (TPA) for tumor promotion
mice were shaved with surgical clippers 2 days before the initial
treatment and only those mice in the resting phase of the hair cy-
cle were used. Skin tumor formation was recorded weekly and
papillomas greater than 2 mm in diameter were included in the
cumulative total if they persisted for 1 week or longer. Both the
number of mice with tumors and the number of tumors per
mouse was determined and recorded weekly. At random, papil-
lomas and carcinomas were removed for histological verification.
882
-------
RESULTS AND DISCUSSION
Complete Carcinogenesis vs. Tumor Initiation
Whenever a known skin carcinogen has been appropriately tested,
it has shown skin tumor initiating activity (9,11-24). In a two-
stage mouse skin system, initiation is the only stage that re-
quires the presence of the carcinogen, and the measured car-
cinogenic potency of a chemical reflects its capacity for tumor
initiation. The results in Table 5 show that there is both a good
qualitative and quantitative correlation between the complete
carcinogenic and tumor initiating activities of several chemical
carcinogens in mouse skin. This is true when one considers the
number of papillomas per mouse at early times (10 to 20 weeks)
or the final carcinomas incidence after tumor initiation.
TABLE 5
Comparison of complete carcinogenesis
and tumor initiation in mouse skina
Relative Potency'-'
Complete Tumor
Carcinogenesis Initiation
Compound
DMBA
MC
BP
2-OHDP
7BrMe-12MeBA
BP- 7, 8 -oxide
DB(a, h)A
BA
DB(a, c)A
Pyrene
BP-4, 5 -oxide
Anthracene
(Carcinomas)
100
50
30
30
20
20
20
5 ± 5
0
0
0
0
(Papillomas)
100
50
30
30
20
20
20
5
3
0
0
0
This is a summary of over 100 compounds wliich shows that an
excellent qualitative and quantitative correlation exists between
complete carcinogenesis and tumor initiation in mouse skin.
Relative potency was determined from dose-response data.
DMBA was given a maximum value of 100.
883
-------
It is possible that a carcinogen lacking promoting ability would
not be detected when tested as a complete carcinogen. In this
regard, however, we have found a number of chemical com-
pounds that have tumor initiating activity but lack complete car-
cinogenic activity (12,17-19, 22). These pure skin tumor
initiators are listed in Table 6. Due to these considerations,
we feel that it is important to test a compound as a tumor
initiator as well as a complete carcinogen. The major dis-
advantage of the skin system which is also true of other car-
cinogenesis systems is that many carcinogens are tissue
specific.
There is a good dose-response relationship of many carcino-
gens used as tumor initiators in the two-stage carcitiogenesis
system using SENCAR mice. This is illustrated in Table 7. A
good dose response relationship exists for DMBA and BP to
initiate skin tumors in SENCAR mice. As can be seen a good
correlation exists between the number of papillomas per mouse
at 15 weeks and the final carcinoma incidence at 50 weeks. The
percent of mice with papillomas has also a reasonable cor-
relation but the dose response is very narrow. Table 8 shows
the dose-response ability of TPA to promote tumors after DMBA
initiation. As was the case for tumor initiation, there is also a
very good dose-response relationship for tumor promotion when
considering either the number of papillomas per mouse at 15
weeks or the percent of mice with squamous cell carcinomas at
50 weeks. Similar results have been reported using Charles
River CD-I mice (13,16) or ICR/Ha Swiss mice (25, 26).
TABLE 6
Agents that are possibly pure tumor initiators
Skin Tumor Relative3
Initiators Potency
BP-7,8-diol-9,10-epoxide 25
MNNG 15
BA-3,4-diol-l,2-epoxide 2
BA 0.5
DB(a,c)A 0.2
Chrysene 0.1
Urethan 0.1
aRelative potency was determined from dose-response data.
DMBA was given a maximum value of 100.
884
-------
W
w
5
^s
1— 1
H
I— 1
s
o
H
£
^
<
H U
(•— t I-T-
d w
CO 0)
< z
w "
T r/1
J-i ^
H *
z i
°l
M H
W
11
CH W
~
CO
W
on
Z
0
OH
f^V
t/J
W
cc;
w
W3
O
Q
<
PM
H
"o
bO
3
UO
UH
O
CO
e
o
• I-H
y
eg
o
"R
&
>>
3
0)
CD
5
0)
o
'?
j_>
_d
•u>
• TH
^
£
o
*4->
nj
•jd
•
CD
JH
p •£ -y
o^- -s cfl
^>
>
CO
CtJ
£ m
O CD ^
,-H CO (U
•-J 3 CD
a 2 &
S S in
^ !-, -H
<*J S
° ci^
#
CO
« CD
co < — i
O 0
Q S
IS
^^"
M
o
Is
;d
p<
i — i
O Q CN ID lO o OO
O ^ CS UO CO i— 1
rH
O O CO O O 00 Q
o o ON cs o r^ >S
1— 1 I— 1 i— 1
cxj oo cs 10 in cs Tf
N v£3 CO O f~- CO •-!
o o — < --H o o o
O "H • O O "5
—1 O CS| rH
<<:<;
-------
oo
cd
J
ea
<£*
H
CO
O
W
H
0
0
a,
O
H
y^
fc o
H
i £~ t
H s
**
Q)
4— >
ai
4->
•g
0)
? H
0 'o
E m
n) QJ
X >
Hd)
^H
co
n)
| m
0^ ^ "nl
." '^ ^
'E rt ^
^ o m
° r-
~*
s
S|3
E-rH ->
D-> ">•
O OH
!-° J^
'>
W
cd
E aj w
r2 w 'a;
i — i ^ QJ
•a i s
M ' — i
t^— i QJ
Tt
4— >
CD
•r-l _
^ 6
o o ^
" -3 -^
c ^J
• i-H P_|
O ^^
co bo
0 3
o
s
E
o
On
CN O ^ LO CO
CO ^ -^t* CN
o o o o m
C C O oo
^H i— 1 i — 1
O CM i^ ^O "^t1
CO t^ ^O CO O
OO vD C~~ CO i— 1
0 •« N -i ^
OH CM BH CM P-,
H H H H H
-------
Studies on the Carcinogenesis of Diesel Emission Particle Extracts
We were interested in determining the skin carcinogenic, cocar-
cinogenic, tumor initiating and promoting activities of diesel
emission particle extracts. In addition, we wanted to compare
the above results to other environmental emission particle ex-
tracts (coke oven, roofing tar, cigarette smoke and gasoline
engine), as well as positive controls such as BP for tumor
initiation and complete carcinogenesis, TPA'for promotion and
pyrene for cocarcinogenesis with BP. Presently, we can only
present a preliminary comparative study on the skin tumor
initiating activities of all the samples. Figure 1 shows that an
excellent dose-response relationsliip exists for the initiating
ability of BP at 14 weeks. A similar dose-response relation-
ship exists for the tumor initiating activity of the Nissan Diesel
sample but using mg quantities for initiation (Figure 2). The
coke oven sample also gave a reasonable dose-response curve
(Figure 3). Table 9 summarizes the skin tumor initiating
activities of the various samples. The comparison of their
tumor initiating activities was made at 14 weeks. The tumor
response is expressed as the mean number of papillomas per
mouse per mg. The values represent the slope from a linear
regression analysis when plotting data from dose-response
studies. Caterpillar Diesel and cigarette smoke condensate were
for all practical purposes zero whereas Mustang gasoline and
Olds Diesel gave extremely low values, although they were above
background. Roofing tar, Nissan Diesel and coke oven all gave
moderate activity at liigh doses (1 to 10 mg) but when compared
to benzo(a)pyrene were relatively low values. 10 Milligrams of
the roofing tar sample had a tumor initiating activity com-
parable to 50 )Ug of BP, whereas 10 mg of the Nissan Diesel was
comparable to 60 ug of BP, and 10 mg of the coke oven sample
was comparable to 80 jug of BP.
887
-------
TABLE 9
Comparative Skin Tumor Initiating Activities of Extracts
of Diesel and Related Environmental Emissionsa
Sample
# Papillomas/mouse/mg
(14 weeks)
R
Caterpillar Diesel
Nissan Diesel
Olds Diesel
Mustang Gasoline
Cigarette Smoke Condensate
Roofing Tar
Coke Oven Emissions
Benzo(a)pyrene
0
0.258
0.115
0.09
0
0.182
0.307
46.2
0.996
0.95
--
--
0.999
0.876
0.984
o
samples and promoted one week later by twice weekly appli-
cations of 2 lag of TPA.
The values represent the slope from the linear regression
analysis of the dose-response studies and measure of fit (R ).
4.50
4.00
3.50
3.00
2.50
2.00
1.50
1.00
0.50
0.00
I
I
I
I
10 20 30 40 50 60 70
BENZO (A) PYRENE (14) , M9
80
90 100
Figure 1. Dose-response relationship for BP applied topically to
male and female SENCAR mice once and followed 1 week later
by twice weekly application of 2 «g of TPA for 14 weeks.
888
-------
250 -
2.00 -
o
1.50 -
z
<
1 00 -
050 -
0.00
000
2.00
4.00
6.00
8.00
10.00
NISSAN DCM (14), mg
Figure 2. Dose-response relationship for the Nissan Diesel
sample (Nissan DCM) applied topically to male and female
SENCAR mice and followed 1 week later by twice weekly appli-
cations of 2 jag of TPA for 14 weeks.
350
3.00
2.50
2.00
1.50
1 00
0.50
000
0.00 200 4.00 6.00
COKE OVEN 114), nig
8.00
10.00
Figure 3. Dose-response relationship for the coke oven sample
applied topically to male and female SENCAR mice and followed
1 week later by twice weekly applications of 2 jag of TPA for 14
weeks.
889
-------
Abbreviations used: TPA, 12-0-tetradecanoylphorbol-13-ace-
tate; DMBA, 7,12-dimcthylbenz(a)anthracene; BP, benzo(a)-
pyrenc; MC, 3-mcthylcholantlircnc; 2-OHBP, 2-hydroxybenzo(a)-
pyrenc; 7BrMc-12iMcBA, 7-Bromom ethyl-12-methylbenz( a)anth-
racene; BP-7, 8-oxide, benzo(a)pyrene-7, S-oxide; DB(a, h)A,
dibenz(a, h)anthracene; BA, benz(a)aiithracene; DlXa,c)A,
dibenz(a, c)anthracene; BP-4, 5-oxide, benzcXa)pyrcne 4, 5-oxide;
BP-7, 8-diol-9, lO-epcmde, (±)-7f>, 8a-dihydroxy-9a-10a-epoxy-
7, 8,9,10-tetrahydrobenzo(a)pyrene; MNNG, N-methyl-N' -nitro-
N-nitrosoguanidinc; BA-3, 4-diol-l, 2-epoxide; (±)-3a, 4p-di-
hydroxy-la, 2a-epoxy-l, 2, 3, 4-tctrahydrobenz(a)anthracene.
890
-------
REFERENCES
1. Carcinogenesis: A Comprehensive Survey. Vol. 2, Mech-
anisms of Tumor Promotion and Cocarcinogenesis, edited
by T. J.Slaga, A. Sivak and R. K. Boutwell. 1978. Raven
Press, N.Y. 2: 1-588.
2. Ames, B. N., J. McCann, and E. Yamasaki. 1975. Methods
for detecting carcinogens and mutagens with the Salmonella/
mammalian-microsome mutagenicity test. Mutation Res.
31: 347-364.
3. McCann, J., and B. N. Ames. 1976. Detection of car-
cinogens as mutagens in Salmonella microsome test: Assay
of 300 chemicals. Discussion. Proc. Natl. Acad. Sci.U.S.,
73: 950-954.
4. Miller, B.C., andJ.A. Miller. 1976. The metabolism of
chemical carcinogenesis to reactive electrophiles and
their possible mechanism of action in carcinogenesis. In:
Chemical Carcinogens, C. E. Searle ed., ACS, Washington,
D.C., p. 732.
5. Miller, E.G. 1978. Some current perspectives on chemi-
cal carcinogenesis in humans and experimental animals:
Presidential Address. Cancer Res., 38: 1479-1496.
6. Brookes, P. and P. D. Lawley. 1964. Evidence for the
binding of polynuclear aromatic hydrocarbons to the nu-
cleic acids of mouse skin; relation between carcinogenic
power of hydrocarbons and their binding to deoxyribo-
nucleic acid. Nature 202: 781-784.
7. Slaga, T.J., S.G. Buty, S. Thompson, W.M. Bracken and
A. Viaje. 1977. A kinetic study on the in vitro covalent
binding of polycyclic hydrocarbons to nucleic acids using
epidermal homogenates as the activating system. Cancer
5f!-,37: 3126-3131.
891
-------
8. VanDuuren, B. L., G. Wity, and B. M. Goldschmidt. 1978.
Structure-activity relationships of tumor promoters and co-
carcinogens and interaction of phorbol myristate acetate and
related esters with plasma membranes. In: Mechanisms of
Tumor Promotion and Cocarcinogenesis. edited by T. J.
Slaga, A. Sivak and R. K. Boutwell. Raven Press, New York
2: 491-507.
9. Boutwell, R. K. 1964. Some biological aspects of skin car -
cinogenesis. Progr. Exptl. Tumor Res., 4: 207-250.
10. Boutwell, R.K. 1974. The function and mechanism of pro-
moters of carcinogenesis. CRC Critical Rev. Toxicol., 2:
419-443.
11. Slaga, T.J., G. T. Bowden, B. G. Shapas and R. K.
Boutwell. 1974. Studies on macromolecular synthesis after
some carcinogenic polycyclic hydrocarbons used as
initiators of skin tumorigenesis, Cancer Res. 34: 771-777.
12. Slaga, T.J., G. T. Bowden, B. G. Shapas and R. K.
Boutwell. 1973. Macromolecular synthesis following a
single application of alkylating agents used as initiators of
mouse skin tumorigenesis. Cancer Res. 33: 769-776.
13. Slaga, T.J., J. D. Scribner, S. Thompson and A. Viaje.
1976. Epidermal cell proliferation and promoting ability
of phorbol esters. J. Natl. Cancer Inst. 57: 1145-1149.
14. Slaga, T.J., A. Viaje, D. L. Berry, W. Bracken, S.G.
Buty and J.D. Scribner. 1976. Skin tumor initiating
ability of benzo(a)pyrene 4, 5-, 7, 8-, and 7, 8-diol-9,10-
epoxides and 7, 8-diol. Cancer Letters 2: 115-122.
892
-------
15. Bowden, G.T., T. J. Slaga, B.C. Shapas and R. K. Boutwell.
1974. The role of aryl hydrocarbon hydroxylase in skin
tumor initiation by 7,12-dimethylbenz(a)anthracene and 1,2,-
3
5,6-dibenzanthracene using DNA binding and thymidine- H-
incorporation into DNA as criteria. Cancer Res. 34: 2634-
2642.
16. Slaga, T.J., G. T. Bowden, J.D. Scribner and R. K.
Boutwell. 1974. Dose-response studies on the skin tumor
initiators, 7,12-dimethylbenz(a)antliracene and benz(a)-
anthracene. J. Natl. Cancer Inst. 53: 1337-1340.
17. Scribner, J.D. and T. J Slaga. 1975. Tumor initiation by
N-acetoxy derivatives of piperidine and N-acrylacetamides.
J. Natl. Cancer Inst. 54: 491-493.
18. Slaga, T.J., W. M. Bracken, A. Viaje, D. L. Berry, S.M.
Fischer, D.R. Miller, W. Levin, A. H. Conney, H. Yagi
andD. M. Jerina. 1978. Tumor initiating and promoting
activities of various benzo(a)pyrene metabolites in mouse
skin, jn; Polynuclear Aromatic Hydrocarbon. Eds. R. I.
Freudenthal and W. Jones, Raven Press, New York, pp.
371-382.
19. Slaga, T.J., G. L. Gleason, J. DiGiovanni, D.L. Berry,
M. R. Juchau and R. G. Harvey. 1979. Tumor initiating
activities of various derivatives of benz(a)anthracene and
7,12-dimethylbenz(a)anthracene in mouse skin. Proc. Ill
International Battelle Conference on Polynuclear Aromatic
Hydrocarbons, Eds. P.W. Jones and P. Leber, Ann Arbor
Press, Michigan, pp. 753-764.
893
-------
20. Slaga, T.J., R. P. Iyer, W. Lyga, A. Secrist. G. H. Daub and
R. G. Harvey. Comparison of the skin tumor-initiating
activities of dihydrodiols, diol-epoxides and methylated deri-
vatives of various polycyclic aromatic hydrocarbons. Proc.
Fourth International Symposium on Polynuclear Aromatic
Hydrocarbons (In press).
21. Slaga, T.J., W. M. Bracken, A. Viaje, D. L. Berry, S.M.
Fischer and D. R. Miller. 1978. Lack of involvement of
6-hydroxymethylation in benzo(a)pyrene skin tumor initiation
in mice. J. Natl. Cancer Inst., 61: 451-455.
22. Scribner, J.D. 1973. Tumor initiation by apparently non-
carcinogenic polycyclic aromatic hydrocarbons. J. Natl.
Cancer Inst. 50: 1717-1719.
23. Hecht, S.S., E. LaVole, R. Mayzarese, N. Herota, T.
OhmoriandD. Hoffmann 1979. Comparative mutagenicity,
tumor-initiating activity, carcinogenicity and in vitro meta-
bolism of fluorinatecl 5-methylchrysenes. J. Natl. Cancer
Inst. 63: 855-861.
24. Cavalieri, E., R. Roth, and E. Rogan. 1979. Hydroxy-
lation and conjugation at the benzylic carbon atom: A pos-
sible mechanism of carcinogenic activation for some methyl-
substituted aromatic hydrocarbons, In; Polynuclear Aro-
matic Hydrocarbons, Eds. P. W. Jones and P. Leber, Ann
Arbor Press, Michigan, pp. 517-529.
25. Van Duuren, B. L. 1969. Tumor-promoting agents in two-
stage carcinogenesis. Progr. Exptl. Tumor Res. 11: 31-68.
26. Van Duuren, B. L., A. Sivak, A. Segal, I. Seidman and C.
Katz. 1973. Dose-response studies with a pure tumor-pro-
moting agent, phorbol myristate acetate. Cancer Res. 33;
2166-2172.
894
-------
General Discussion
R. ALBERT: Could you discuss the negative result of
cigarette condensate which is kind of peculiar since cig-
arette condensate certainly has been shown to be a complete
skin carcinogen. I wonder whether or not, for example, you
use enough of it so that the benzopyrene content of the
cigarette tar should have produced a response.
T. SLAGA: The maximum dose level' that we used was ten
milligrams. The maximum we could get into solution and put
on the skin was two milligrams at a time, we used five
applications of 2 mg each to get our maximum dose of ten
milligrams. The data that has already been published in
terms of cigarette smoke condensate by Hoffmann's group and
a few others were at higher levels than we are dealing
with. I think that at increased concentrations, we would
achieve tumor initiation. These are short exposures too,
only 14 weeks. It is possible that the latency period may
be much longer for the cigarette smoke condensate. The
tumor initiation studies as well as the complete carcino-
genesis studies are planned to last at least 60 weeks. So
then, these results are preliminary and might change as the
experiments progresses.
SPEAKER: Tom, most of us feel that the skin carcino-
genesis assay, is one of the best, if not the best and most
quantitative in vivo carcinogenesis assay that we have at
the present time. I wonder whether you could address the
question of what type of carcinogen we can hope to pick up
with the skin carcinogenesis assay. I think that is be-
coming an extremely crucial question since in complex mix-
tures, such as diesel exhaust, we are obviously dealing
with a great variety of carcinogens and initiators. It
becomes very essential for us to know what we can expect to
pick up with that assay and what kinds of compounds we
might miss with that assay.
T. SLAGA: I tried to emphasize in the beginning of the
talk that there is tissue specificity to a large number of
carcinogens, but in another slide I tried to point out a
number of the various different agents that are positie when
tested on the skin. These include not only the polycyclic
hydrocarbons but a number of alkylatic agents, urethane,etc.
It probably picks up the widest range of carcinogens as any
in vivo system that I know. In general, most in vivo systems
are a lot more specific then skin. A disadvantage of the
skin system tht we are presently utilizing is that the skin
does not always metabolize every procarcinogen to the active
metabolite. We are modifying the system by injecting the
test substance. IP, in order to allow the liver or any
other tissue to metabolize it. We have some preliminary
data with hydrocarbons that we can inject them IP, promote
895
-------
on the skin, and get a nice tumor response. We have some
data with acetoaminoflurene which is a liver carcinogen
showing that the skin doesn't appear to metabolize it. If
you give it repetitively to the skin you will not get skin
cancer. If you give it to IP, at a reasonable dose, then
follow topically with a tumor promoter, then you get skin
cancer. This may be true for a number of other agents which
we are pursuing right now.
R. BILES: Can you tell me the number of animals you used
per group and also the number of animals with papillomas.
Was your data presented as the numbers of mean papillomas
per animal.
T. SLAGA: I tried to give some idea on a couple of the
dose response studies that I presented, showing the number
of papillomas per mouse and how that correlates to the per-
cent of animals with tumors which further correlates with
the final carcinoma yield.
R. BILES: What about animals with papillomas? How many
animals per group did you use?
T. SLAGA: Eighty. On one of my slides I showed that we
used 40 males and 40 females per dose. In the last couple
of slides, males and females were averaged together since
there was essentially no difference. In general once we
reach a certain level of number of benign tumors per mmouse,
all of our animals get papillomas. The dose response re-
lationship when presented on the percent of animals with
tumors, only goes from a zero to 100 percent. Therefore,
once all of your animals have tumors, be it only one tumor
or ten tumors, you get the same type of results. This re-
sults in a very narrow dose response relationship. If you
look at the number of papillomas per mouse, then you extend
that dose response relationship.
R. BILES: So all of your animals had papillomas.
T. SLAGA: At low doses there may be fewer papillomas per
animal but all of them had papillomas.
D. HOFFMANN: I do differ from you based on literature
reports which have very clearly shown when combustion pro-
ducts, tobacco tar, gasoline engine exhaust are fraction-
ated, mutation activity is observed only in the PAH fraction.
Since we know from all the mutagenicity studies that the
major activity is not in the PAH fraction. I doubt that you
can pick up the carcinogenicity of those agents which are
carcinogenic but not mutagenic. There is no evidence that
neutral components, in combustion products which are not PAH
produce tumors.
T. SLAGA: Your question is a little misleading. When I
stated that you get a tumor response with those various
mixtures, I wans't trying to say that it was related to the
896
-------
hydrocarbons though I gave it in terms of PAH equivalent
response. The fact that some of our data has come out negative,
for tumor initiation in these relatively sensitive animals,
could be related to a dose relationship. In which we are
not using a high enough dose. When you are dealing with a
complex mixture, and are looking at the whole mixture, you
have a number of factors that not only could cause initiation
of tumors, but you have a number of agents that could also
negate that response. I think your studies showed that by
fractionation one can get the specific fractions that possess
more tumor initiating and promoting, actiity.
L. DEPASS: Have you seen any kind of dose response
relationship in terms of time to papilloma in your study?
T. SLA6A: We haven't really analyzed our data in terms
of which compound or mixture gave rise to tumors the soonest.
We have that data but I haven't really tried to rank them
according to time to tumor data. Basically I would say that
the samples that are responsive the nissan, the coke oven,
and the roofing tar, all had about the same latency which is
very close to what benz(a)pyrene had. However, the number of
tumors was much less compared to benz(a)pyrene.
897
-------
MUTAGENIC AND CARCINOGENIC POTENCY OF EXTRACTS OF
DIESEL AND RELATED ENVIRONMENTAL EMISSIONS:
SUMMARY AND DISCUSSION OF THE RESULTS
Stephen Nesnow* and Joellen L. Huisingh
Carcinogenesis and Metabolism Branch* (MD-68)
and
Genetic Bioassay Branch
Genetic Toxicology Division
Health Effects Research Laboratory
U. S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
ABSTRACT
The proposed conversion from gasoline powered automobiles
to diesel powered vehicles has prompted the Environmental
Protection Agency to evaluate the potential health effects
associated with exposure to diesel emissions. At present,
there is no direct epidemiological link between this
exposure and human health. Therefore, a research program
was constructed to compare the health effects associated
with diesel emissions with those from other emission
sources for which epidemiological information was avail-
able. The emission sources chosen were cigarette smoke,
roofing tar, and coke oven. An additional comparative
emission source which was evaluated was a gasoline catalyst
engine.
Respirable particles from a variety of combustion sources
have the potential of being carcinogenic and mutagenic.
The objective of these studies was to determine the rela-
tive biological activity of the organic material adsorbed
on these particles in both j_n vitro mutagenesis and i_n
To whom reprint requests should be addressed.
898
-------
vitro and j_n vivo carcinogenesis bioassays. The organic
extracts from the following series of emission sources
were quantitatively bioassayed in a matrix of tests for
their carcinogenic and mutagenic activity: (1) a light-
duty Oldsmobile diesel 350 engine; (2) a heavy-duty Cater-
pillar diesel engine; (3) a light-duty Nissan engine; (4)
a Volkswagen Rabbit diesel engine; (5) cigarette smoke;
(6) roofing tar; (7) coke oven; and (8) a gasoline catalyst
Mustang.
The test matrix consisted of the following bioassays:
reverse mutation in Salmonella typhimurium; mitotic recombi-
nation in Saccharomyces cerevisiae; DNA damage in Syrian
hamster embryo cells (SHF51sTsTer chromatid exchange in
CHO cells; gene mutation in L5178Y mouse lymphoma cells,
Balb/c 3T3 mouse embryo fibroblasts and CHO cells; viral
enhancement of SHE cells; oncogenic transformation in
Balb/c 3T3 cells; and skin tumor initiation in Sencar and
C57 black mice.
The results of this test matrix are discussed.
INTRODUCTION
The proposed conversion from gasoline powered automobiles
to diesel powered vehicles has prompted the Environmental
Protection Agency to evaluate the potential health effects
associated with exposure to diesel emissions (1). At
present, there is no direct epidemiological link between
this exposure and human health (2). Therefore, a research
program was constructed to compare the health effects
associated with diesel emissions with those from other
emission sources for which epidemiological information was
available. The emission sources chosen were cigarette
smoke, roofing tar, and coke oven. An additional compara-
tive emission source which was evaluated was a gasoline
catalyst vehicle.
Respirable particles from a variety of combustion sources
have the potential of being carcinogenic and mutagenic
(3). The objective of these studies was to determine the
relative biological activity of the organic material
adsorbed on these particles in both i_n vitro and i_n vivo
mutagenesis and carcinogenesis bioassays. The organic
extracts from the following series of emission sources
were quantitatively bioassayed in a matrix of tests for
their carcinogenic and mutagenic activity: (1) a light-duty
Oldsmobile diesel 350 engine, (2) a heavy-duty Caterpillar
3304 diesel engine, (3) a light-duty Datsun Nissan 220C
engine, (4) a Volkswagen turbocharged Rabbit diesel engine,
(5) cigarette smoke, (6) roofing tar, (7) coke oven, and
(8) an unleaded gasoline catalyst Mustang II.
899
-------
The collection, characterization, and description of these
samples are described by Huisingh et al. (4).
The test matrix consisted of the following bioassays:
reverse mutation in Salmonella typhimurium; mitotic recom-
bination in Saccharomyces cerevisiae;D~NA breakage in
Syrian hamster embryo celTs(SHE); sister chromatid
exchange in Chinese hamster ovary (CHO) cells; gene muta-
tion in L5178Y mouse lymphoma cells, Balb/c 3T3 mouse
embryo fibroblasts, and CHO cells; viral enhancement in
SHE cells; oncogenic transformation in Balb/c 3T3 cells;
and skin tumor initiation in Sencar and C57 black mice.
The potency of complex mixtures of organic compounds in
biological systems will depend on the uptake, distribu-
tion, metabolism, and binding of each component of the
mixture in each biological system; the specific biological
endpoint (e.g., gene mutation); the sensitivity of each
biological system to the individual components of the
mixture; the possible interactions, both synergistic and
antagonistic, which arise from the exposure to multiple
agents; and the type of quantitative method which is
applied to the data. Differences in these parameters will
result in altered potency.
Aware of these problems and shortcomings, we have attempted
to correlate biological end effects from a variety of
carcinogenesis and mutagenesis bioassays with the nature
or source of a series of combustion samples. The purpose
of this comparative study is to evaluate each of the j_n
vitro bioassay systems in comparison with each other and
with the j_n vivo system. This paper will summarize and
discuss the results of the bioassays reported in the
session on the Mutagenic and Carcinogenic Potency of
Extracts of Diesel and Related Environmental Emissions
(5-9).
RESULTS AND DISCUSSION
QUANTITATION METHOD
General Procedures
The method for relating effect with dose depends on the
bioassay being considered. The bacterial mutation system
of Ames (AMES), the sister chromatid exchange bioassay in
Chinese hamster ovary cells (SCE), the mammalian cell
mutagenesis bioassay in L5178Y mouse lymphoma cells,
(L5178Y) and the mouse skin tumor initiation bioassay
(TUMOR INITIAT) gave good dose-response relationships and
the slope of the linear portion of the dose response curve
was chosen for potency estimation (Table 1). The Balb/c
900
-------
TABLE 1. COMPARATIVE POTENCY: QUANTITATION METHOD
MUTAGENESIS CARCINOGENESIS
VIRAL TUMOR
BIOASSAYS: AMES SCE L5178Y BALB ENHANCE BALB INITIAT
QUANTITATION
METHOD: Slope Slope Slope LECT LECT LECT Slope
LECT = Lowest effective concentration tested.
Slope = Slope of the linear portion of the dose response
curve.
3T3 mutagenesis and oncogenic transformation bioassays
(BALB) did not give good dose response with the samples
tested and doses chosen. A quantitation method was
therefore applied which utilized the lowest effective
concentration tested (LECT) which exerted a biological
response. It is to be emphasized, however, that with pure
agents and more closely spaced doses the Balb assay does
respond in a dose-related fashion. The viral enhancement
bioassay in Syrian hamster embryo cells (VIRAL ENHANCE)
gave dose related response information in terms of absolute
frequency but for comparison to the Balb system and due to
the kind of statistics applied to the assay, the LECT
method was chosen. It is possible that the two methods,
slope and LECT, when applied to the same data could give
dissimilar relative rankings. Bioassay results in
Saccharomyces cerevisiae, DNA damage in SHE cells, gene
mutation in CHO cells, and skin tumor initiation in C57
black mice were not utilized in this analysis due to
marginal results.
Normalization of Data
In order to reduce or normalize the potencies in the
bioassays to a common denominator, the following system
was applied: the activity of the Nissan diesel sample
regardless of bioassay type or quantitation method used
was given a value of 100. All the results of other emis-
sion samples were then related to the Nissan sample. For
example, in the Ames bioassay, Strain TA-98 less metabolic
activation, the potency of Nissan and Oldsmobile by the
slope method was 1225 and 615 revertants/100 |jg, respec-
tively. Assuming the Nissan potency equals 100 by simple
mathematical relationship, the Oldsmobile sample equals 50
or one-half the potency of the Nissan sample. In the case
of a LECT potency, the inverse relationship is applied.
For example if the LECT for Nissan is 75 ug/ml and that
901
-------
for CAT is 300 pg/ml, assuming Nissan equals 100, then CAT
equals 75/300 x 100 or 25, one-quarter the potency of the
Nissan sample. Using this kind of analysis, the absolute
data was converted to normalized values. The reason the
Nissan was chosen as the sample to normalize to was that
it was the only diesel sample tested in all bioassays
which gave positive results.
COMPARATIVE RANKINGS
Gene Mutation Assays
A comparison of the test results of the eight samples plus
the positive control in the three gene mutation assays is
found in Table 2. The three gene mutation assays compared
TABLE 2. COMPARATIVE RANKINGS: GENE MUTATION ASSAYS
SAMPLE
AMESC
-MA
L-5178Y1
-MA
+MA
BALB/3T31
-MA
+MA
DIESEL:
Cat
Ni ssan
Olds
VW Rab.
5.4
100
50
33
4.3
100
23
22
16
100
58
21
Ie
100
64
50
75
100
94
I
25
100
58
I
GASOLINE: Mustang 11 25 32
COMPARATIVE
SOURCES: Cigarette 0 7 42
Coke 13 18 26
Roof Tar 0 7 16
36
300 7500
21 300 300
339 300 15
850 150 7500
STANDARDS:
B(a)P
MNNG
0 1112
189 25000
25000
^Salmonella typhimurium strain TA-98.
\-5178Y mouse lymphoma cells (TK locus).
"BALB 3T3 mouse embryo fibroblasts (ATPase locus).
j
Metabolic activation by an Aroclor-1254 induced rat
hepatic S-9.
eTesting incomplete at this time.
902
-------
were Ames strain TA-98, L5178Y at the TK locus, and
BALB 3T3 at the ATPase locus. Each system was performed
without (-MA) and with (+MA) metabolic activation which
consisted of an Aroclor-1254 induced rat hepatic S-9. The
emission samples were previously described and are abbre-
viated as Cat for heavy duty Caterpillar; Nissan; Olds for
Oldsmobile 350; VW Rabbit for the turbocharged diesel
Rabbit; Mustang for the unleaded catalyst gasoline engine;
Cigarette for standard cigarette smoke condensate; Coke
for coke oven emissions; Roof Tar for roofing tar emis-
sions; B(a)P for benzo(a)pyrene; and MNNG for N-methyl-
N' -nitro-N-nitrosoguanidine.
The Nissan, as per our definition, has a value of 100 in
all assays. Both the Ames and lymphoma assay (without
activation) show the same relative potency within the
diesel samples with Nissan > Olds > Rabbit > Cat. When
the gasoline sample is included, the rankings are: Ames:
Nissan > Olds > Rabbit > Mustang > Cat; mouse lymphoma:
Nissan > Olds > Mustang > Rabbit > Cat. Both of these
cell types lack the oxidative enzymes required for the
activation of mutagens or carcinogens. The diesel and
gasoline samples show primarily direct acting activity.
Balb 3T3 cells possess the enzymes required for the activa-
tion of carcinogens especially polycyclic aromatics (8)
and therefore one cannot compare all these systems across
the board and expect similar results. Comparison of the
three systems with metabolic activation showed the relative
ranking (with some exceptions) to be: Nissan > Olds >
Rabbit > Cat.
The comparative sources samples generally show more meta-
bolic activation dependence for maximal effect and less
quantitative correlation in the gene mutation assays.
DNA Damage Assay
The mobile source samples tested for sister chromatid
exchange in CHO cells, without activation, (Table 3) gave
the following ranking: Nissan > Rabbit, Mustang » Cat,
Olds.
In the presence of metabolic activation, the SCE results
with the comparative sources ranked in similar order with
the mouse lymphoma results: Roof Tar > Coke > Cigarette.
903
-------
TABLE 3. COMPARATIVE RANKINGS: DNA DAMAGE ASSAY
SCE (CHO)a
SAMPLE -MA +MAL
DIESEL: Cat 4 0
Nissan 100 100
Olds 0 0
VW Rab 30 50
GASOLINE: Mustang 29 Ic
COMPARATIVE
SOURCES: Cigarette 47 0
Coke 171 44
Roof Tar 60 291
STANDARDS: B(a)P 0 1750
Sister chromatid exchange in Chinese hamster ovary cells.
Metabolic activation by an Aroclor-1254 induced rat
hepatic S-9.
CTesting incomplete at this time.
Oncogenic Transformation
The results from assays which relate to the transformation
of cells in culture are found in Table 4. Both of these
systems were evaluated by the LECT method. The viral
enhancement assay measures the ability of a chemical to
enhance viral transformation in mammalian cells while the
BALB assay measures the ability of the chemical to directly
transform mammalian cells. SHE cells like BALB cells
contain the enzymes needed for the metabolic activiation
904
-------
TABLE 4. COMPARATIVE RANKINGS:
TRANSFORMATION ASSAYS
ONCOGENIC
SAMPLE
VIRAL
ENHANCEMENT (SHE)
BALB 3T3
-MA
+MA
DIESEL:
GASOLINE:
Cat
Nissan
Olds
VW Rab
Mustang
0
100
0
50
50
1.3
100
1.9
NTb
100
0
100
0
NT
2000
COMPARATIVE
SOURCES: Cigarette
Coke
Roof Tar
STANDARDS: B(a)P
MNNG
200
800
1040
50000
31250
Ic
10
I
8300
I
I
500
16700
Metabolic activation by an Aroclor-1254 induced rat hepatic
S-9
bNot tested.
Testing incomplete at this time.
of carcinogens especially polycyclic aromatics (8). At
this time not all the data points are complete in the Balb
3T3 assay. The mobile source samples, presently rank for
viral enhancement in the absence of metabolic activation:
Nissan > Rabbit > Mustang » Cat, Olds, while the compara-
tive samples rank: Roof Tar > Coke > Cigarette.
Mouse Skin Tumorigenesis
The results obtained in the Sencar mouse skin tumorigenesis
experiments for tumor initiators have been compared at 14
weeks after treatment (interim score) for all the samples
and at 27 weeks of treatment (final score) for three
samples (Table 5). The interim score is represented here
in both absolute terms (papillomas/mouse/mg) and in the
905
-------
TABLE 5. COMPARATIVE RANKINGS: SENCAR
MOUSE SKIN TUMORIGENESIS
SAMPLE
GASOLINE: Mustang
TUMOR INITIATION
INTERIM SCORE*
RANKING POTENCY1-
35
COMPARATIVE
SOURCES: Cigarette 0
Coke 119
Roof Tar 71
STANDARDS: B(a)P
17900
0.090
0.00
0.307
0.182
46.2
FINAL SCORE
POTENCY5
DIESEL:
Cat
Nissan
Olds
VW Rab
0
100
45
Ic
0.00
0.258
0.115
I
0.00
0.145
71.6
Interim score 14 weeks after treatment.
Papi11omas/mouse/mg.
Testing incomplete at this time.
normalized rankings while the final scoring for comparison
is in absolute terms. There are, as expected, differences
in the slope potencies from incomplete y_s completed experi-
ments but those differences do not exceed 2-fold in the
Olds and B(a)P samples. The Cat and Cigarette samples
were negative in this assay system up to 10,000 pg/mouse.
The lack of activity of the cigarette smoke condensate may
be due in part to the method of sample collection. This
sample is not an organic extract of particles as are the
other samples and is therefore comparatively much less
concentrated. This may explain the very low percentage of
benzo(a)pyrene found in the cigarette smoke condensate
sample (Table 6). The Olds, Mustang, and Roofing Tar were
less than Nissan while the Coke oven sample was slightly
higher. The ranking of the mobile source samples was:
Nissan > Olds > Mustang » Cat.
906
-------
TABLE 6. CORRELATION BETWEEN ORGANICS AND TUMOR INITIATION
PERCENT ORGANIC
EXTRACTABLE
ng B(a)P/
mg EXT
MOUSE SKIN
TUMOR
INITIATION"
DIESEL:
Cat
Nissan
Olds
VW Rab
GASOLINE: Mustang
27
8
17
18
43
COMPARATIVE
SOURCES: Cigarette
Coke 7
Roof Tar > 99
STANDARDS: B(a)P
2
1173
2
26
103
< 1
478
10C
0.00
0.258
0.115,
r
0.09
0.00
0.307
0.182
46.2
Interim score 14 weeks after treatment (papillomas/
mouse/mg).
3Testing incomplete at this time.
A comparison of the amount of B(a)P per milligram extract
and the potencies of those samples as skin tumor initiators
is found in Table 6. There is a semiquantitative correla-
tion observed, except for the Olds sample, between ng
B(a)P/mg extract and papi1lomas/mouse/mg extract.
COMPARISON ACROSS TEST SYSTEMS
The relative rankings of the mobile source samples are
listed in Table 7. These results are from those bioassays
performed in the presence of exogenous metabolic activation.
There is a consistency in these results with the Nissan
sample the most potent and the Cat sample the least potent
in all bioassays.
The relative rankings of the comparative source samples is
found in Table 8. In three of five systems, the identical
rank order of Roof Tar > Coke > Cigarette is found, and in
all five systems, the Cigarette sample was the least
potent.
907
-------
TABLE 7. RELATIVE RANKINGS OF MOBILE SOURCE SAMPLES
BIOASSAY
RANK ORDER
AMES
SCEC
Nissan > Mustang, Olds, Rabbit » Cat
Nissan > Olds > Rabbit > Mustang
Nissan > Rabbit » Cat, Olds
VIRAL ENHANCEMENT Nissan > Mustang, Rabbit » Cat, Olds
TUMOR INITIATION
Nissan > Olds > Mustang » Cat
In the presence of an Aroclor-1254 induced rat hepatic S-9
Mouse skin tumor initiation in Sencar mice 14 weeks
after treatment
TABLE 8. RELATIVE RANKINGS OF COMPARATIVE SOURCE SAMPLES
BIOASSAY
RANK ORDER
AMES Coke > Roof Tar, Cigarette
L5178Ya Roof Tar > Coke > Cigarette
SCEa Roof Tar > Coke » Cigarette
VIRAL ENHANCEMENT Roof Tar > Coke > Cigarette
TUMOR INITIATION13 Coke > Roof Tar » Cigarette
In the presence of an Aroclor-1254 induced rat hepatic S-9
:>Mouse skin tumor initiation in Sencar mice 14 weeks
after treatment
908
-------
co
C3
Z
t— *
**s
Z
Q£
>-
O
z
LU
O
LU
>
t—H
^
O£
«^
a.
s:
o
o
en
LU
1
CO
Z
LU
ro
co
i — i
CO
LU
Z
LU
f n
_
00
r*^
r-H
m
i
— i
ro
LU
c_>
CO
ro
CO
LU
V
•4
CD CD LT> i— i
o «*•
rH
O CD CD r-
O 'Z.
rH
O O O O
CD m
i — i
-a
m CD co i —
c\j o in z
rH
U
i—(d*)-O
O CD LO
rH
CD O O CD
CD in
rH
r"O
«3- CD ro CM
CD CM CM
rH
C JD.
ro ro
in in ac
-»-> in -a
ro "- r- ^
(_) Z O >•
. .
— 1
LU
CO
LU
V— 1
O
m CD en rH CD
rO rH F~~ O
r- 1 en
i--
i — i
CD i — ii — i O CD
O CD CD
o in r^
CM CD
rH
O O O CD CD
in CD o *d" CD
CM OO CD CD
rH CD
un
CD O LO O CD
O O rH CD CD
in ro in o
r*-~ r^- in
CM
CD i — ( cn CD en
ro CM ro in OO
ro oo T— i
1 — t CD ro
c: cu r-
ro i- O-
+J ro ID il- /— N
in LU O)J^ O ro
3 >..••- O O • • -— '
. • S i— i CO C_) CJ CC COOQ
LU (— LU CD
z < o ce
HH o; oi
C
CU
E
4^
ro
01
&_
^_>
>4—
• O
cn
i in
CO -^
ai
(j a;
•i- 3
rO *3"
Q-i-H
CU
JZ S-
OJ
ra H-
S- ro
-o ai
CD u
(J "-
-rj
£= S-
• ,- ro
U
•=t c
m cu
CM CO
rH
1 C
S- T-
o
, — c
u o
O •!-
S- -(->
ro T-
M — •( —
O
1-
CU O
u e
c ^
CU -)->
in
cu c
S- •<-
pi *s
in
cu
^: cu
+-> in
3
c: o
>-H s:
CU
E
4_)
U>
•i —
J~
4_)
4_)
(tJ
CU
±^
cu
a.
£
0
o •
c -o
• ,- OJ
4_>
C35 l/>
c a>
•i- +J
^
in 4->
cu o
h- Z
rO -Q U "O
909
-------
The normalized rankings for 7 bioassays are compared in
Table 9. The results presented are from those bioassays
performed in the presence of exogenous metabolic activa-
tion including Ames, SCE, L5178Y and Balb-mutagenesis and
transformation. The quantitative results from the mobile
source samples show a general overall consistency. Cat, a
very weak sample in Ames is inactive in SCE, viral enhance-
ment, Balb transformation, and mouse skin tumor initiation.
Olds, weak in Ames, L5178Y, Balb mutation, and mouse skin
tumor initiation is inactive in SCE, viral enhancement,
and BALB transformation. In three of the four assays for
which test data is available, SCE, L5178Y, and viral
enhancement, the Rabbit sample gave almost identical
quantitative results. The Mustang sample gave similar
results in Ames, L5178Y, and mouse skin tumor initiation,
while markedly dissimilar results were obtained in the
Balb assay.
With the comparative source samples, there is little
agreement between the quantitative results from these
bioassays with the exception o'f the Roofing Tar sample in
L5178Y, viral enhancement and Balb transformation.
In theory, gene mutation and skin tumor initiation arise
from similar one hit, single process, irreversible mechan-
isms and should give similar results assuming equal toxi-
city and mutagen/carcinogen transport/activation by the
various cell types. A comparison of the results of the
mobile source samples in Ames and L5178Y gene mutation
with those results in mouse skin tumor initiation seem to
support this hypothesis.
In conclusion, a series of extracts from diesel and gaso-
line emission samples were evaluated in a battery of
bioassays and a broad general agreement was found among
most of the bioassays. Similar experimentation with the
Cigarette, Coke oven, and Roofing Tar samples produced
dissimilar results. Additional experimentation and analy-
sis will continue in this important area of environmental
mutagenesis and carcinogenesis.
REFERENCES
1. Ember, L 1979. The diesel dilemma: EPA's diffi-
cult decision. Environment, 21:15-41.
2. Health Effects Associated with Diesel Exhaust Emis-
sions: Literature Review and Evaluation. 1978. EPA
600/1-78-063.
910
-------
3. Waters, M. D., S. Nesnow, J. L. Huisingh, S. S. Sandhu,
and L. Claxton, eds. 1979. Application of short-term
bioassays in the fractionation and analysis of complex
environmental mixtures. Plenum Press, N.Y.
4. Huisingh, J. L. 1980. Mutagenic and carcinogenic
potency of extracts of diesel and related environ-
mental emissions: Preparation and characterization
of the samples. In: Proceedings of the Interna-
tional Symposium on Health Effects of Diesel Engine
Emissions. This volume.
5. Claxton, L. 1980. Mutagenic and carcinogenic potency
of extracts of diesel and related environmental
emissions: Salmonel la typhimurium assay. J_n:
Proceedings of the International Symposium on Health
Effects of Diesel Engine Emissions. This volume.
6. Mitchell, A. D., V. F. Simmon, K. E. Mortelmans, E.
S. Riccio, M. M. Jotz and E. L. Evans. 1980. Muta-
genic and carcinogenic potency of extracts of diesel
and related environmental emissions: In vitro muta-
genesis and DNA damage. In: Proceedings of the
International Symposium on Health Effects of Diesel
Engine Emission. This volume.
7. Casto, B. C., G. G. Hatch, S. L. Huang, J. L. Huisingh,
S. Nesnow, and M. D. Waters. 1980. Mutagenic and
carcinogenic potency of extracts of diesel and related
environmental emissions: I_n vitro mutagenesis and
oncogenic transformation. In: Proceedings of the
International Symposium on Health Effects of Diesel
Engine Emissions. This volume.
8. Curren, R. D., L. M. Schechtman, C. M. Kim, and R. E.
Kouri. 1980. Mutagenic and carcinogenic potency of
extracts of diesel and related environmental emissions:
Simultaneous transformation and mutagenesis in BALB/c
3T3 cells. In: Proceedings of the International
Symposium on Health Effects of Diesel Engine Emissions.
This volume.
9. Slaga, T. J. , L. L. Triplett, and S. Nesnow. 1980.
Mutagenic and carcinogenic potency of extracts of
diesel and related environmental emissions: Two
stage carcinogenesis in skin tumor sensitive mice
(Sencar). In: Proceedings of the International
Symposium on Health Effects of Diesel Engine Emissions.
This volume.
911
-------
General Discussion
SPEAKER: As the data stands now, number one; what is the
bottom line in terms of mutagenesis; and number two; in
terms of carcinogenesis?
S. NESNOW: I think as the data stands now, we have ev-
idence that certain of these samples do not exhibit any
activity. Then there are a range of samples that exert
activity which are fairly comparable with the Nissan and
there are other samples, that are very high. I think the
bottom line is that given a multitude of these systems we
have confirmed again and again the carcinogenicity and mu-
tagenicity of most of these samples. We are sort of break-
ing new ground trying to test potencies and trying to eval-
uate and compare samples.
912
-------
Session VI
MUTAGENICITY OF INHALED DIESEL EMISSIONS
Chairman:
Dr. David Brusick
A Study of Diesel Emissions on Drosophila.
Schuler, Ronald L. and Richard W. Niemeier.
Metaphase Analysis, Micronuclei Assay and Urinary Muta-
genicity Assay of Mice Exposed to Diesel Emissions.
Pereira, M. A., T. H. Connor, J. Meyne, and M. S.
Legator.
In-Vivo Detection of Mutagenic Effects of Diesel Exhaust by
Short-Term Mammalian Bioassays.
Pereira, M. A., P. S. Sabharwal, P. Kaur, C. B. Ross,
A. Choi, and T. Dixon.
Sister Chromatid Exchange Analysis of Syrian Hamster Lung
Cells Treated In Vivo with Diesel Exhaust Particulates.
Guerrero, Robert R., Donald E. Rounds, and John
Orthoefer.
Test of Diesel Exhaust Emissions in the Rat Liver FOCI
Assay.
Pereira, M. A., H. Shinozuka, and B. Lombardi.
The Effect of Diesel Exhaust on Sperm-Shape Abnormalities in
Mice.
Pereira, M. A., P. S. Sabharwal, L. Gordon, and A. J.
Wyrobek.
Testing for the Ability of Marine Diesel Fuel Vapors to
Induce Micronuclei or Sister Chromatid Exchanges in Peri-
pheral Lymphocytes Taken From Dogs Exposed Continuously by
Inhalation for Thirteen Weeks.
Benz, R. Daniel and Patricia A. Beltz.
913
-------
A STUDY OF DIESEL EMISSIONS ON DROSOPHILA
Ronald L. Schuler
Richard W. Niemeier
National Institute for Occupational Safety and Health
Experimental Toxicology Branch
Division of Biomedical and Behavioral Science
4676 Columbia Parkway
Cincinnati, Ohio A5226
ABSTRACT
A sex-linked recessive lethal test was performed on male
fruit flies of the species Drosophila melanogaster, (Oregon-R
strain), exposed to an approximate five-fold dilution of
exhaust from a diesel engine. The eight hour exposure was
achieved by drawing diluted diesel exhaust from a three cubic
meter stainless steel exposure chamber housing laboratory
animals through a two liter reaction flask modified for use
with Drosophila. A preconditioned sampling bag was used to
collect the emissions after passing through the exposure
chamber containing the flies. Results of analyses performed
on the diesel exhaust mixture showed: carbon dioxide - 0.17Z,
carbon monoxide - 12.2 ppm, hydrocarbons - 11.6 ppm, nitrogen
oxide - 3.8 ppm, nitrogen dioxide - 2.9 ppm, sulphur dioxide -
1.0 ppm, and particulates - 2.18 mg/m^.
Two broods of the ~?2 generation were investigated for the
occurrence of recessive lethal events. These broods approx-
imated the developing gametogenic stages of mature sperm (Pj
matings on days 2 and 3 post-exposure) and spermatocytes (P-^
matings on days 8 and 9). Additionally, the F^ generation
was evaluated for the occurrence of mosaic recessive lethal
events which might escape detection in the ?2 generation. An
equal number of F2 and F-j flies for both broods served as
concurrent controls.
Results indicate that, under the conditions tested, the die-
sel exhaust did not increase the mutation frequency of the
914
-------
exposed flies ($2 rate = 0.30%, F^ rate = 0%) when compared
to the concurrent controls (F? rate = 0.37%, F-j rate = 0.15%).
INTRODUCTION
Within the past few years energy conservation has become a
major issue throughout the world. While new sources of en-
ergy are being sought and developed, old sources are being
modified for more efficient and greater economy. As a result,
new emphasis may be placed on the diesel engine and its role
in mass transportation and various industrial applications.
Therefore, because of the large population potentially ex-
posed to the emissions of such engines, knowledge of any
associated hazards is vital. Of particular importance is
knowledge of any mutagenic effects of diesel emissions. The
purpose of the study described in this paper was therefore to
use a standard bioassay to evaluate the potential mutagenic
effects of diesel exhaust. Specifically, the Drosophila sex-
linked recessive lethal bioassay (1), an excellent screen for
genetic hazards, was used in this investigation. This bio-
assay is a useful system for detecting most types of genie
damage and has been used to study mutagenic effects of a
number of gases and aerosols.
At the request of the United States Environmental Protection
Agency, this Drosophila bioassay was performed at the E.P.A.
facility (Center Hill Laboratories, Cincinnati, Ohio) utiliz-
ing the available resources of ongoing diesel exhaust inhala-
tion experiments. This study was an effort to extract addi-
tional data from these chronic diesel studies and at the same
time, the intent was not to affect the integrity of the studies
in progress by altering such variables as exposure time or
concentration. Thus the current study was limited to an 8
hour exposure at the concentration being tested.
EXPERIMENTAL PROCEDURE
Exposure Regimen
Approximately 200 two to five day old wild type male fruit
flies (Oregon-R) were exposed to an air stream consisting
of diesel engine exhaust gases diluted five-fold with filter-
ed ambient air. The diesel engine used in this study was a
6 cylinder Chrysler-Nissan engine (198 cu. in. displacement)
with automatic transmission and dynamometer. A complete de-
scription of the operational parameters of this engine is
found in Hinners et_ jd. (2). Gross particulates were
removed from the ambient dilution air by an MSA-CBR Filter*
unit containing high efficiency (99.9% >0.3 microns) HEPA
filters. The flies were held in a specially fabricated
stainless steel cage (Wire Cloth Company, Cambridge, Maryland),
915
-------
Mating and Scoring
Randomly selected 100 exposed and 100 control flies were
individually mated to two 3 to 5 day old virgin Muller-5
(In[l] scSIL sc8R + S, scsl sc8 wa B) females (3) on the sec-
ond day post-exposure. All flies, throughout the study, were
kept incubated at a constant 25°C within 8-dram (^30 ml) shell
vials containing an instant Drosophila medium, Formula 4-24
without dye (Carolina Biological Supply Co., Burlington, MC,
27215). Carbon dioxide was used to anesthetize the flies.
Pure C02 gas (99.9%) regulated from a cylinder, passed through
a cotton-filled box (3X5X1 inch) having a platform con-
structed of fine wire mesh. The flies were allowed to deposit
their eggs for the next two days. On the fourth day post-
exposure, the females were transferred to new vials contain-
ing fresh media and allowed to continue egg deposition. Con-
currently, the males were individually placed into fresh
vials in order to maintain a record of parental lines. These
two "sub-broods" represented brood 1, i.e., the sperm used
in these matings were mature spermatozoa (post-meiotic) at
the time of exposure (4).
The males were again individually mated to fresh virgins on
the eighth day post-exposure. As before, the females were
permitted to deposit their eggs for the next two days before
being transferred to fresh vials. These two "sub-broods"
represented brood III, i.e., the sperm used in these matings
were in the spermatocyte (pre-meiotic) stage of spermato-
genesis at the time of exposure (4).
The parents (PI) of both second "sub-broods" were removed
before emergence of the F^ generation. Ten F^ females were
selected at random from each PI vial (from the original
matings) up to 70 vials/brood and individually mated to two
of her Muller-5 brothers. Therefore, a total of 700 F^
matings/Brood/control and exposed groups were performed.
The resultant F2 cultures were scored for a sex-linked re-
cessive lethal event according to standard procedures (1).
An F3 test was performed by selecting one female from every
other F2 culture vial and mating her with two of her Muller-
5 brothers. The resultant F3 cultures were scored in the
same manner as the F2 cultures.
RESULTS
Comparative analyses of the atmospheres of the Drosophila
chamber and the 3 cubic meter chamber (Table 1) revealed that
the concentration values for most atmospheric components were
lower in the Drosophila chamber with the exception of the
hydrocarbon fraction. Specifically, the Drosophila chamber
showed 54.8% of the C02 level of the 3 cubic meter chamber,
55.2% of the CO level, 124.7% of the HC level, 29.5% of the
NO level, 76.3% of the NO, level, 43.5% of the SO, level and
30.1% of the particulate level of the three cubic meter
chamber.
916
-------
The fly cage (Figure 1) was cylindrical in shape (8^ X
3^ inches) with one end terminating into a cone with a quick
connect/disconnect port allowing rapid ingress and egress to
and from the cage interior; the opposite end, together with
the remainder of the structure, was enclosed with 24 by 24
mesh, 0.04-inch wire diameter, stainless steel screen. The
cage was placed within a glass chamber (2000 ml Reaction
Flask, Ace Glass Inc.). The fly cage was secured within the
chamber by a preformed band of silicone rubber which made an
occlusive seal between the glass wall and the screen, thus
allowing for all the diesel exhaust mixture to pass through
the cage. This air was pulled by vacuum pump through this
system at a rate of 1.06 liters per minute. The source of
the engine exhaust mixture was a 3 cubic meter stainless
steel exposure chamber (Figure 2) that was being used for re-
peated exposures of laboratory animals to diesel exhaust.
These animals were present in the chamber during the 8 hour
exposure period. The diesel engine was alternately turned on
and off at 15-minute intervals throughout the 8 hour exposure
period. An identical exposure system was used for an equal
number of male flies. However, these flies served as con-
current negative controls exposed only to the filtered am-
bient air. Both the control and the exposed flies were
randomly collected from the same stock of 2 to 5 day old
males.
Figure 1
Fly Chamber with Cage
917
-------
Figure 2
Fly Chamber Connected to 3 Cubic Meter Chamber
Sampling and Analysis
Characterization of the atmosphere within the 3 cubic meter
chamber was compared to the atmosphere sampled after passing
through the Drosophila chamber (Table 1). Analyses were per-
formed by E.P.A. personnel of the Center Hill Laboratories,
with the exception of the gravimetric particulate sample. An
integrated air sample of the fly chamber, collected in a pre-
conditioned mylar plastic sampling bag, was taken over the
entire exposure period. The CO and C(>2 levels were measured
by a Beckman Infrared Analyzer, hydrocarbons by hydrogen flame
ionization on a Bec'-nan Total Hydrocarbon Analyzer,S02 by pul-
sed fluorescence on a Thermo-electron SC>2 Analyzer, and NO/N02
by a chemiluminescent (ozone reaction) method using a Thermo-
electron NO/N02 Analyzer. Particulate concentrations were
determined gravimetrically using 37 mm glass fiber filters
mounted between the fly chamber and the sampling bag. Tem-
perature and percent relative humidity were recorded hourly.
Temperature was measured with a thermometer mounted within
the Drosophila chamber. Relative humidity of the Drosophila
chamber was measured using a Weather Measure Corp. Relative
Humidity Indicator, Model HMI-1A. Temperature and humidity
readings were not available from the 3 cubic meter chamber.
918
-------
Scoring of the F2 cultures resulted in three lethals found in
678 vials scored from brood I in the control group compared
with three lethals in 670 vials of the exposed group (see
Table 2). Controls from brood III yielded two lethals in 676
vials compared to one lethal in 680 vials of the brood III ex-
posed group. No lethals were found in the F3 cultures of
brood I of both the control (334 vials) and the exposed (also
334 vials) groups; only one lethal was found in the brood III
control group (346 vials) while no lethals were found in the
exposed group (336 vials).
The combined F2 (broods I and III) results of five lethals/
1354 vials scored for the control group gives a "background"
mutation rate of 0.37%. In comparison, the combined F2
results in the exposed group of four lethals/1350 vials yield
a rate of 0.30%. The combined F3 (broods I and III) results
of one lethal/680 vials gives a rate of 0.15% for the control
group while the exposed group had no lethals recorded in 670
vials scored.
When the F2 and F3 totals are combined, the control group
shows a rate of 0.29%, six lethals/2034 vials, versus a rate
of 0.20%, four lethals/2020 vials, for the exposed group.
TABLE 1
CHARACTERIZATION OF CHAMBER ATMOSPHERES
Component
Measured
C02
CO
HC
\T0
NOT
SO 2
Part iculate
Relative Humidity
Temperature
3 cu m
Chamber
0
22
9
12
3
2
7
.317
.1
.3
.9
.8
.3
.3
-
-
ppm
ppm
ppm
ppm
ppm
ing/m3
Drosophila
Chamber
0
12
11
3
2
1
2
29
22
.17
.2
.6
.8
.9
.0
.2
.17
. 3°
7
ppm
ppm
ppm
ppm
ppm
mg/m3
+ 1 .5
C + 0 .2
919
-------
w
to
H
,J
to
w
(V,
•a
0)
V-4
o
o
10
si
1"
w jS
4-1
0)
rH
O
S3
H
to
w
EH T3
0)
<£ O
W cj
E-I CO
w
iH r^
w ge
> £ -^
M t; M
s ^1
O 4-1
W (U
ft! rH
w o
£*J p^i
S3
M
1
e
to
•o
o
0
LJ
CO
CN
[T!
6*5
O
CO
0
II
0
CM oo ^o
-------
DISCUSSION
The number of recessive lethals found in the ?2 and the ?3
treatment groups of both broods was smaller than in the
control groups, indicating a lack of mutagenic activity ex-
hibited by the diesel exhaust under the conditions of this
study.
The effluent concentrations of some of the diesel exhaust
components of the fly chamber differed substantially from
those measured in the 3 cubic meter chamber. Much of the
particulate loss may have been a result of adsorption along
the network of fine wire mesh of the cage or to the walls of
the glass chamber. A set of particulate samples was taken
concurrently at a later date to examine this possibility,
however, no significant difference in concentrations was
found between the sample taken directly from the inhalation
chamber and the sample taken immediately after the fly cham-
ber and cage. Other diesel exhaust components may have been
adsorbed onto the walls of the chamber or scavenged by the
rubber tubing connections before reaching the sampling de-
Two factors must be considered for the proper interpretation
of the results of this study. First, the number of flies
used in this test is capable of detecting moderate or stron-
ger mutagens exhibiting 3 to 5 or more times the background
or control mutation rate (albeit a low rate of approximately
0.1% to 0.6%-many authors); a weaker activity of twice the
background rate or less could have eluded detection. How-
ever, there is no evidence in the data that suggests that
weak activity might have been present, i.e., more mutations
were found in the control groups than in the treated groups.
A second, and perhaps more relevant, fact is that the objec-
tive of the recessive lethal test is more to reveal qualita-
tive information rather than quantitative or dose-response
type information (5). Therefore, a higher concentration
should be administered to the flies for a more thorough as-
sessment of the mutagenic potential of the diesel exhaust.
Specifically, the concentration level used should result in
a significant degree of mortality or sterility in the flies
after exposure or, if this is mechanically unachievable, the
exposure concentration should represent the maximum level
that can be consistently generated and maintained over the
duration of the desired exposure period. The duration of
exposure can be extended up to approximately 24 hours and
still allow for brood analysis with minimal overlapping of
the separate stages of spermatogenesis. In addition, more
brood types could be investigated to aid in the detection of
any stage specificity effect that might otherwise escape
detection.
921
-------
Furthermore, it has been demonstrated that some of the com-
ponents of diesel exhaust have been shown to be mutagenic
either alone or synergistically in the combined treatment
with other substances. Carbon monoxide alone (957 CO,
5% 02/6 hours) did not increase sex-linked recessive lethals
in Drosophila, but when combined with azide or potassium
cyanide, a significant increase in recessive lethals was
found (6). Azide and potassium cyanide, either alone or in
combination, did not show such an increase. Three to four
percent nitric oxide (NO) delayed spermatogenesis, increased
dominant lethality in Drosophila, and increased the percent-
age of sex-linked recessive lethals in X-irradiated Oroso-
phila (7). Sulfur dioxide (SO-}) alone exhibited a signifi-
cant increase in the mutation rate above spontaneous levels
in Tradescantia (8). These facts, coupled with the fact that
the hydrocarbon fraction of the exhaust may contain potential
carcinogens and/or mutagens, reinforce the need for an addi-
tional, more challenging, dose for administration to Droso-
phila.
REFERENCES
1. Wiirgler, F. E. , F. H. Sobles and E. Vogel. 1977.
Drosophila as an assay system for detecting genetic
changes. In: Handbook of Mutagenicity Test Procedures,
B. J. Kilbey, M. Legator, W. Nichols and C. Ramel, eds.
Elsevier/North-Holland, Amsterdam, Holland, 1977.
pp. 335-373.
2. Hinners, R. G., J. K. Burkhart, M. Malanchuk, and W. D.
Wagner. 1979. Animal exposure facility for diesel ex-
haust studies. Presented at Biological Studies of En-
vironmental Pollutants: I. Symposium on Aerosol Genera-
tion and Exposure Facilities, April 1-6, Honolulu, Hawaii.
3. Lindsley, D. L., and E. H. Grell. 1968. Genetic varia-
tions of Drosophila melanogaster. Carnegie Inst. Wash.
Publ., No. 627, Washington, DC, 1968. 472 pp.
4. Chandley, A. C. and A. J. Bateman. 1962. Timing of
Spermatogenesis in Drosophila melanogaster using Tri-
tiated Thymidine. Nature, 193:299-300.
5. Lee, W. R. Chemical mutagenesis. 1976. In: The
Genetics and Biology of Drosophila, Vol. 1C, M. Ashburner
and E. Novitski, eds. Academic Press, New York, NY,
1976. pp. 1299-1341.
6. Clark, A. M. 1958. Genetic effects of carbon monoxide,
cyanide and azide on Drosophila. Nature, 181:500-501.
922
-------
7. Rinehart, R. R. 1963. Some effects of nitric oxide and
oxygen on dominant lethal production in X-irradiated
Drosophila virilis males. Genetics, 48:1673-1681.
8. Sparrow, A. H. and L. A. Schairer. 1974. Mutagenic
response of Tradescantia to treatment with X-rays, EMS,
DBE, Ozone, 502, ^2^ anc^ several insecticides. Mutation
Res., 26:445.
* Mention of a company or product name does not constitute
endorsement by NIOSH.
General Discussion
S. SODERHOLM: It is rather off the main point of your
talk, but there is a puzzling difference in concentration
especially gas concentration between the inhalation chamber
and the Drasophila chamber. The fact that it didn't occur a
second time suggests that you had a leak in the line some-
where and a small negative pressure in the main chamber.
This can occur fairly simply.
R. SCHULER: Yes, one of our disappointments was the
inability to pinpoint the exact reasons why we had dif-
ferences .
R. SOHRECK: Could you tell us the route of entry of the
particle into the fly? Do you feel it comes in through
their tiny respiratory passageways in the side, and what do
you expect the dose is per animal?
R. SCHULER: I don't know the exact dose, however, the
route of entry is through the spiracles on the side of the
abdomen of the fly. They can selectively open and close
these spiracles. However, they only do this when they are
challenged with a very irritating type of exposure. Normal-
ly, this is demonstrated with erythromycin. However, the
fact that they are readily put under anesthesia by carbon
dioxide in a fraction of a second indicates that the entry
is rapid.
D. BRUSICK: It might be worth noting that the Drosophila
can also ingest the particles. It has been well documented
mutations can be induced in Drosophila by ingestion. Do you
feel that relates in any way to what you observed?
R. SCHULER: Again as I emphasized, I would prefer to be
doing a much higher and more challenging dose in terms of
concentration and exposure time before I make any definite
statements as to why we got these results. We were limited
by the conditions of an ongoing experiment.
923
-------
METAPHASE ANALYSIS. MICRONUCLEI ASSAY AND
URINARY MUTAGENICITY ASSAY OF MICE
EXPOSED TO DIESEL EMISSIONS
M.A. Pereira
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
T.H. Connor, J. Meyne and M.S. Legator
University of Texas Medical Branch
Galveston, Texas
ABSTRACT
Female Swiss mice were exposed 8 hours per day to diesel
exhaust for 1, 3 and 7 weeks. Urine was collected overnight
for 4 days prior to sacrifice while the mice continued to be
exposed for eight hours during the day. The presence of
mutagens was determined by the Ames Salmonella test. One
hour prior to sacrifice each mouse received 1 mg/kg colce-
mide. After sacrifice, the marrow from each femur was ob-
tained. The marrow from one femur was used to prepare slides
for metaphase analysis and the other for micronuclei assay.
Other mice received i.p. 50 mg/kg cyclophosphamide 24 hours
prior to sacrifice or 1 umole/kg benz(a)pyrene in each of
four daily doses prior to sacrifice and served as positive
controls. The Ames Salmonella assay of the unconcentrated
urine after 1, 3 and 7 weeks and concentrated urine after 7
weeks exposure to diesel exhaust did not significantly vary
from clean air controls. Results In the micronucleus test,
and metaphase analysis cyclophosphamide produced a strong
positive response and the 7 week diesel exposure was not
different from clean air controls.
The genotoxic hazard of exposure to diesel exhaust was in-
vestigated in mice. Diesel exhaust contains thousands of
924
-------
organic chemicals some of which are known carcinogens and
mutagens including polycyclic aromatic hydrocarbons. The
high particle concentration of diesel exhaust results in the
sequestering on the particles of much of the polycyclic
aromatic hydrocarbons. In order for the genotoxic agents in
diesel exhaust to exert their activity they must be deposited
in the lungs, eluted from the particles and distributed
throughout the body.
Genotoxic agents can produce two types of deleterious alter-
ations, i.e. tnutagenic and clastogenic effects. Agents that
cause mutations can be detected in the urine of animals by
the Ames Salmonella assay (Legator, et al. 1977). Clas-
trogens can be detected by metaphase analysis for breaks in
chromatids and chromosomes (Cohen and Hirschhorn, 1971) and
by the micronuclei assay (Schmid, 1977). Micronuclei result
from the membrane encapsulation of chromosomal material that
did not segregate with the spindle apparatus as the result
of chromosome breakage or disruption of the spindle apparatus.
In this communication, we report investigations that at-
tempted to detect the absorption and systemic distribution
of genotoxic agents in mice exposed to diesel exhaust by a
combined protocol consisting of 1) metaphase analysis, 2)
micronuclei and 3) urinary mutagenicity.
MATERIALS AND METHODS
Ten female Swiss mice (Charles River) were assigned to the
various treatment groups. Those exposed to diesel emissions
were exposed 8 hours per day, 5 days per week, at the U.S.
EPA Health Effects Research Laboratory, Cincinnati, Ohio.
The diesel emission was produced by a Nissan 6 cylinder run
on the Federal Short Cycle. The exhaust exposure contained
6-7 mg/m^ particles at a 1:18 dilution (Hinners, 1980).
Control groups were exposed under similar conditions without
the diesel emission. Urine samples on the 4 days prior to
sacrifice were collected overnight in a brown bottle embedded
in salt-ice and kept at (-) 15 to (-) 10°C. On the day of
sacrifice, colcemide (1 mg/kg) was injected i.p. 30 minutes
before the animals were killed by cervical dislocation.
The marrow from one femur was used for the preparation of
slides for micronuclei and the other for slides for metaphase
analysis according to Connor et al. (1979). The urine sam-
ples were either (a) tested directly with 5 strains of Sal-
monella typhimurium using method described by Connor et al
(1979) or (b) concentrated on an XAD-2 column according to
the procedure of Yamasaki and Ames (1977). The samples were
concentrated 20-fold and 50 ul of concentrate were assayed
with and without B-glucuronidase (100 units per plate) using
£L typhimurium TA98 and TA100. The unconcentrated urine
samples were tested at 200 ul per plate in the presence of
925
-------
B-glucuronidase. Additional groups of control mice received
i.p. 1 cyclophosphamide at 50 mg/kg, one time; 2) benz(a)-
pyrene at 1 umole/kg four times; 3) water (vehicle for cyclo-
phosphamide), one time; and 4) corn oil (vehicle for benz(a)-
pyrene), four times. The slides and the frozen urine samples
were transported to the University of Texas Medical Branch
for analysis. Figure 1 outlines the treatment and analysis
protocol described in this paper.
RESULTS
Metaphase analysis was performed on only the 7-week diesel
exhaust exposures and controls. There was no indication of
chromosomal or chromatid damage (Table 1). Cyclophosphamide
controls produced many damaged cells and benz(a)pyrene at
the low dose used was inactive. The low dose of benz(a)-
pyrene was chosen as an approximation of the dose of benz(a)-
pyrene expected to result from the diesel exhaust exposure.
The micronuclei bioassay was performed on mice exposed
for 1, 3 and 7 weeks to diesel exhaust. There was no
significant difference in the number of micronuclei in poly-
chromatic erythrocyctes from controls and mice exposed to
diesel exhaust (Table 2). Cyclophosphamide did induce micro-
nuclei while benz(a)pyrene was inactive. With both the
unconcentrated and the concentrated urine samples, all the
clean air and diesel exhaust exposure values were within the
expected ranges for the various strains (Tables 3 and 4).
Where possible, urine was collected sequentially at 1, 3 and
7 weeks from the same mice exposed to either clean air or
diesel exhaust. Again, the cyclophosphamide and not the
benz(a)pyrene was detected in both the unconcentrated and
the concentrated samples with strains TA1535 and TA100.
DISCUSSION
We attempted to determine the absorption and systemic dis-
tribution of the genotoxic agents, mutagens and clastogens,
presence in diesel exhaust emissions. Mice were exposed to
diesel exhaust and evidence of systemic distribution examined
by the combined protocol of 1) metaphase analysis, 2) micro-
nuclei assay and 3) urinary mutagenesis. The metaphase
analysis and the micronuclei assay detect chemicals capable
of breaking chromosomes and chromatids and disrupting the
spindle apparatus. The urinary mutagenicity assay detects
the presence of chemical mutagens. These assays have the
advantage of being in vivo systems that allow for the meta-
bolic and pharmacokinetic factors which control the effective-
ness of the mutagens present in diesel exhaust.
The combined protocol used was unable to detect the systemic
distribution of genotoxic agents resulting from diesel
926
-------
clean air
1 week
3 weeks
7 weeks
diesel emission
urine
metaphase
micronucleus
urine
metaphase
*micronucleus
urine
metaphase
micronucleus
Figure 1. Treatment and Testing Protocol
927
-------
TABLE 1
METAPHASE ANALYSIS IN MICE EXPOSED TO DIESEL EXHAUST
No
E xposure T i me
Control 7 weeks
Diesel 7 weeks
Cycl ophosphami de
cBenzo(a)pyrene
Hetaphases
Analyzed
750
950
250
750
Chroma_ti'd Chromosome Concentric
Breaks Breaks Ti^ures
3 02
1 0 2d
200(cells grossly damaged)
7 2 4
Fragrients
]
0
3
annce were exposed 8 hours per day, 5 days per week, 10 mice per group
50 ng/kg, i.p , 4 times
1 „ mole/kg, i. o., 4 times
Robinsonian Translocations
TABLE 2
Micronucleus Test in Mice Exposed to Diesel Exhaust
Micronuclei per .
Exposure Time
Diesel 1 week3
Control 1 week
Diesel 3 weeks
Control 3 weeks
Diesel 7 weeks
Control 7 weeks
Cyclophosphamide
Benzo(a)Pyrene
1000 Polychromatic Erythrocytes
Mean S.D.
1.0 0.82
0.4 0 . 70
0.9 1.20
1.1 0.74
0.9 0.99
1.7 1 . 34
31.6 18.79
0.8 0.79
amice were exposed 8 hrs per day, 5 days a week for numbers of
weeks specified, 10 mice per group.
50 mg/kg, i.p., 1 time
clymole/kg, i.p., 4 times
1000 cells counted per animal
928
-------
TABLE 3
Mutagenicity of Urine Samples of Nice Exposed to Diesel Exhaust
and Assayed with Salmonella typhlmuri'um
Histidine revertants per plate
Exposure
Clean air
Diesel
Cyclophosphamide
Control (water)
B(a)P
Time
1 week (4)b
3 weeks (6)
7 weeks (3)
1 week (2)
3 weeks (6)
7 weeks (3)
(1)
(1)
(1)
Control (corn oil) (1 )
TA1535
5C
9
2
4
10
6
370
7
0
3
TA1537
2
1
2
0
1
4
3
0
1
0
TA1538
2
2
10
0
8
9
9
14
6
10
TA100
44
43
22
44
45
23
345
27
10
26
TA98
6
9
15
11
8
13
16
5
17
0
Control values for each strain have been substracted, 200 ul of urine added
per plate, 100 units of g-glucuronidase added per plate, 2 plates per sample
Number of samples tested
Mean of samples
929
-------
TABLE 4
MUTAGENICITY OF CONCENTRATED URINE SAMPLES
OF MICE
ASSAYED
Exposure
Clean air
Diesel
Cyc lophosphamide
B(a)P
EXPOSED TO DIESEL EXHAUST AND
WITH SALMONELLA TYPHIMURIUM
Histidine revertants
Time B-glu TA100
1 week - 10
+ 19
3 weeks - 9
+ 11
7 weeks - 44
+ 24
1 week - 19
+ 25
3 weeks - 0
+ 26
7 weeks(3)c - 33
(3) + 27
308
+ 303
10
+ 0
per plate3
TA98
15
10
13
9
21
24
14
11
16
11
10
19
4
H
0
0
Control values for each strain have been subtracted, 100
units B-glucuronidase added, as indicated.
Urine was concentrated 20-fold by XAD-2 column and 50 ul of
concentrated urine added per plate.
Number of samples tested is one if not given, in which
case the values are the average of the samples tested.
930
-------
exhaust emissions. The failure to detect genotoxic agents
could have resulted from (a) the level of genotoxic agents
in diesel exhaust was too low to be detected or (b) the
route of exposure (inhalation) and (c) the absorption of the
genotoxic agents onto particles did not allow them to be
absorbed, distributed and metabolized. The combined protocol
is sensitive to many genotoxic agents including cyclophos-
phamide -(Connor, et al., 1979; Connor et al., 1980; Legator
et al. unpublished results). However, the combined protocol
was not sensitive to benz(a)pyrene at 1 umole/kg for 4 days.
This dose of benz(a)pyrene was chosen to approximate the
dose of benz(a)pyrene resulting from the diesel exhaust
exposure. Benson et al, 1978 have reported that 100 mg/kg
(400 umole/kg) i.m. of benz(a)pyrene could be detected in
the urine. Siou et al, 1977 have reported that 100-500
mg/kg (400-2000 umole/kg of benz(a)pyrene induced micro-
nuclei and Balser and Rohrborn, 1976 reported the induction
in Chinese hamsters of chromosomal aberrations.
It would appear that the bioassays employed in the combined
protocol are not sensitive enough to benz(a)pyrene to detect
the amount of benz(a)pyrene present in diesel exhaust. How-
ever, it can not be ruled out that genotoxic agents are
present in diesel exhaust in sufficient concentration to be
detected by the employed bioassays if they are not seques-
tered on particles. We plan to further investigate this
possibility by testing under the combined protocol diesel
exhaust particles and their extracts.
Acknowledgement
The work upon which this publication is based was per-
formed pursuant to Grant No. R806119401 with the Environ-
mental Protection Agency, HERL - Cincinnati.
931
-------
REFERENCES
1. Balser, A. and G. Rohrborn. 1976. Chromosome
aberrations in Oocytes of NMRI mice and bone marrow cells of
Chinese hamsters induced with e,4-benzpyrene. Mutation Res.
318:327-332.
2. Benson, A.M., R.P. Batzinger, S-Y.L. Ou, E. Bueding, Y-
N Cha and P. Talalay. 1978. Elevation of hepatic gluta-
thione S-transferase activities and protection against muta-
genic metabolites of benzo(a)pyrene by dietary antioxidants.
Cancer Res. 38:4486-4495.
3. Cohen, M.M. and K. Hirschhorn. 1971. Cytogenetic
studies in animals, IN: Chemical Mutagens. Principles
and Methods for Their Detection, A. Hollaender (Ed.) Vol.
2, Plenum Press, N.Y.
4. Connor, -T.H., J. Meyne and M.S. Legator. 1980. The
mutagenic evaluation of tetrakis (hydroxymethyl) phosphonium
sulfate using a combined testing protocol approach. Jour.
Env. Pathol. and Toxicol. In Press.
5. Connor, T.H., J. Meyne and M.S. Legator. 1979. A
combined testing protocol approach for mutagenicity testing.
Mutation Res. 64:19-26.
6. Hinners, R.G., Burkart, J.K., Malanchuk, M. and Wagner,
W.D. (1980). Facilities for diesel exhaust studies. Pro-
ceedings of the International Symposium on Health Effects of
Diesel Engine Emission, Dec. 1979.
7. Legator, M.S., T.G. Pullin and T.H. Connor. 1977. The
isolation and detection of mutagenic substances in body
fluid and tissues of animal and body fluid of human subjects.
IN: Handbook of Mutagenicity Test Procedures. B.J. Kilbey,
M. Legator, W. Nichols and C. Ramel (Eds.) Elsevier Sci.
Pub. Co., Amsterdam.
8. Schmid, W. 1977. Micronucleus Test. IN: Handbook of
Mutagenicity Test Procedures. B.J. Kilbey, M. Legator, W.
Nichols and C. Ramel (Eds.) Elsevier Sci. Pub. Co.,
Amsterdam.
9. Siou, G., L. Conan and A. Doinel. 1977. Mutagenic
effects of benzene and benzo(a)pyrene as revealed by the
Howell-Jolley bodies test. Cah. Notes Doc. 8^:433-444.
10. Yamasaki, E. and B.N. Ames. 1977. Concentration of
mutagens from urine by adsorption with a nonpolar resin XAD-
2: Cigarette smokers have mutagenic urine. Proc. Natl.
Acad. Sci. USA, 74:3555-3559.
932
-------
General Discussion
D. ROUNDS: Can you tell me how the benzo(a)pyrene was
administered in the SCE Study?
M. PEREIRA: It was done by IP injection.
D. ROUNDS: Have you tried doing an IP injection of
the diesel particles?
M. PEREIRA: Not yet; we are planning to do that.
A. BROOKS: I am a little curious as to how you are
going to increase the concentration very much higher than
what you already have?
M. PEREIRA: The dilution ratio of diesel exhaust in
the studies we have reported was one to 18. This con-
centration has since been increased at the EPA Center
Hill Facility.
W. PEPELKO: Yes, a dilution ratio of about one to 18
was used. We are presently using a dilution ratio of one
to nine which doubles our particulate concentration to 12
millograms per cubic meter.
A. BROOKS: Have you tried any other cell systems
where chromosome aberrations persist chromosomic, since
both the systems that you are using are rapidly pro-
liferating and the amount of chromosome damage you see
would be dependent on this rate? In other words, at each
cell division, the abnormal chromosomes are eliminated so
that you are never going to achieve a very high total
chromosome aberration frequency in that system.
M. PEREIRA: That is why we believe more in the hyper-
kinetic system.
933
-------
IN-VIVO DETECTION PF MUTAHFMIC EFFECTS OF DIESEL
FXHAUST BY SHORT-TEPM MAMMAL I AM PIOASSAYS
M.A. Pereira1, P.S. Sabharwal?, P. Kaur3, C.P. Ross3
A. Choi3 and T. Dixon3. Environmental Protection
Agency1, Cincinnati, OH 45219; T. H. Morgan School
of Biological Sciences?, University of Kentucky,
Lexington, KY 40506; EHRT3, Cincinnati, OH 45220
APSTPACT
Male Chinese hamsters were exposed to diesel exhaust and
clean air for six months at the Center Hill Facility of
the U.S. Environmental Protection Agency in Cincinnati,
Ohio. The animals were kept in specially constructed
inhalation chambers and exposed to clean air or diesel
exhaust for eight hours daily. The animals were sacri-
ficed and slides prepared to study the mutagenic effects
of diesel exhaust by four in vivo short term mammalian
bioassays. Sperm morphology bioassay revealed a 2.67 fold
increase in sperm abnormalities in the animals exposed to
diesel exhaust as compared to those exposed to fresh air.
Micronucleus bioassay revealed a 50% increase in the number
of micronuclei in polychromatic erythrocytes obtained from
animals exposed to diesel exhaust. However, no increase
in sister chromatid exchange or chromosomal abnormalities
was observed in bone marrow cells of animals treated with
diesel exhaust. During these studies a decrease in mitotic
index was observed in animals treated with diesel exhaust.
Introduction
The projected increase in the number of diesel cars to be
manufactured make it imperative to study the mutagenic
effects of diesel exhaust. Diesel exhaust is an extremely
934
-------
complex mixture of inorganic and organic compounds. Some
of the organic compounds in diesel exhaust are known muta-
gens and carcinogens in the free form. However, in diesel
exhaust the chemicals are sequestered on particulate matter.
The genotoxic effects of these particle-bound chemicals
are unknown.
The mutagenic potential of these chemicals in diesel ex-
haust is influenced by pharmacokinetic and metabolic
factors that can only be addressed by in vivo studies.
Synergistic effects on biotransformation pathways, phar-
macokinetics and interactions at the target organs may
be important in determining the overall effects on the
animals. In vivo studies of the genotoxicity of diesel
exhaust reported in this communication were designed
to permit full interplay of these factors.
It has been shown that there is an increase in the fre-
quency of abnormal sperm in mice exposed to various car-
cinogenic and mutagenic chemicals (1,2). Bruce et al.
(3) reported increase in the frequency of abnormal sperms
when animals were exposed to X-ray at a very low dose of
30 rads. Since carcinogens and mutagens cause genetic
damage, it is likely that the observed sperm abnormali-
ties are a result of genotoxic effects (4).
It has been observed that there is an increase in the
frequency of micronucleated polychromatic erythrocytes
in animals exposed to various carcinogenic and mutagenic
chemicals (5, 6). Wild (7) also reported that the induc-
tion of micronuclei for many common chemical mutagens and
carcinogens correlated with results from other mutagenicity
and carcinogenicity tests. The micro-nucleus test is a
procedure to detect chromosome fragments that did not
segregate with the spindle apparatus. The fragments are
not incorporated into the cell nuclei but rather form
smaller nuclei called micronuclei. Micronuclei are found
in a variety of cells; however, quantitation of micronuclei
is most easily carried out in erythrocytes since the micro-
nuclei remain in the red blood cells even after the expul-
sion of nucleus by the erythroblasts. Consequently, the
micronuclei are easily detected.
Various known carcinogens and mutagens have been reported
to increase significantly the frequency of sister chromatid
exchange (SCE) both in vitro (8) and in vivo. The use of
5-bromodeoxyuridine (BUdr) and 33258 Hoechst staining tech-
niques for the detection of sister chromatid exchange
induction in cultured cells has proven useful for highly
accurate analysis (9). The relevance and flexibility of
SCE analysis was greatly enchanced, however, when the
in vivo induction of SCE in chick embryos (10), mouse
935
-------
spermatogonia (11), and mouse bone marrow (12) was des-
cribed. Since in vivo induction of SCE is an extremely
sensitive bioassay (13), we have studied the mutagenic
effects of diesel exhaust on male Chinese hamsters using
this test.
The induction of chromosomal aberrations has been used as
an indicator for mutagenic agents in the environment (14).
This bioassay shows excellent correlation when used with
known mutagens and carcinogens (15). It has been clearly
shown by many investigators (16) that chemicals which
induce mutations at specific loci in eukaryotic cells
invariably result in cytologically observable chromosomal
damage expressed as aberrations. In the present investi-
gations we have determined the effect of diesel exhaust in
Chinese hamster on 1) sperm morphology; 2) the induction
of micronuclei in bone marrow cells; 3) the induction of
sister chromatid exchange in bone marrow cedlls; and 4)
the cytology of bone marrow cells.
Materials and Methods
Male Chinese hamsters were exposed to diesel exhaust daily
(8 hrs) for 6 months at the Center Hill Facility of U.S.
Environmental Protection Agency at Cincinnati, Ohio. Con-
trol animals were exposed to fresh air. The procedurs for
various bioassays are described below:
1. Sperm Morphology Bioassay
The animals v/ere sacrified by cervical dislocation and the
cauda epididymes removed. A sperm suspension was prepared
by mincing the cauda in 2 ml of phosphate buffered physio-
logical saline. It was then pipetted a few times to insure
uniform distribution of the sperm cells in the suspension.
A fraction of each suspension was mixed (10:1) with 1%
Eosin Y in alcohol. After 30 minutes, a few drops of this
suspension were used to make a smear. Five slides were
prepared for each animal. The slides were allowed to air
dry and mounted with a covers!ip using Permount. For each
animal 500 sperm cells were observed at 400-fold magnifi-
cation with blue-green filter and the percentage of abnor-
mal sperm/animal scored from coded slides.
2. Micronucleus Rioassay
Ten animals exposed to diesel exhaust and ten animals ex-
posed to fresh air were used for the experiment. After
sacrifice of the animals by cervical dislocation, the
femur was removed by cutting through tibia and pelvis;
the bone was freed from muscle. Then, the epiphyseal part
of the bone was torn from the rest of tibia. The proximal
936
-------
end of the femur was then cut off so that a small opening
to the bone marrow canal could be seen.
Centrifuge tubes were filled with fetal bovine serum up
to the rim. A 1 cc syringe with mounted needle was used
to aspirate about 0.2 ml of serum. Then, the needle was
inserted a few mil 1 injeters deep into the proximal part of
bone marrow. With the femur completely submerged, the
syringe was gently aspirated and flushed several times.
The process was repeated for the distal part of the bone
marrow.
The centrifuge tubes were spun at 1000 rpm for 5 minutes.
Supernatant was removed from the tubes using a Pasteur
pipette and the serum was completely removed. The sediment
was gently aspirated into the capillary portion of a sili-
conized Pasteur pipette, and a small drop was transferred
to a slide and spread out using a cover slip held at an
angle of 45 degrees. These preparations were air dried for
24 hours.
The preparations were stained for 3 minutes in undiluted
May-Gruenwald solution, and 2 minutes in May-Gruenwald
diluted with double distilled water 1:1, and rinsed in
distilled water and stained again for 10 minutes in Giemsa,
diluted with water 1:6. Then, the slides were rinsed
under tap water and blot dried with filter paper. The
dried slides are then cleared in Xylene for 5 minutes and
mounted in Permount.
The slides were coded and screened to determine the number
of micronuclei per 1000 polychromatic erythrocytes per
animal.
3. j_n vivo SCE Rioassay
Six animals were exposed to diesel exhaust, six animals to
clean air and four animals to benz(a)pyrene (RaP) (Sigma) as
a positive control. A pellet of 70-75 mg 5-bromodeoxyuri-
dine (Sigma) was implanted suhcutaneously in the abdominal
area (17). Care was taken not to disturb the peritoneal
membrane. Ether was used to anesthetize the animals and #4
Ethicon nylon sutures were used to close the incision made
for implanting the pellet. Animals in the experimental
group were given thirty minutes to recover from the ether.
The animals were then returned to diesel exhaust chamber.
Animals in the positive control group were injected intra-
peritoneally with R(a)P at 100 mg/kg body weight two hours
following implantation. Animals in the negative control
group were given thirty minutes to recover from the ether
and then returned to their clean air chambers.
937
-------
Twenty-four hours following implantation, each animal was
injected with colchicine (Sigma) at 10 mg/kg body weight.
Two hours following treatment with colchicine, each animal
was sacrificed by cervical dislocation. After sacrifice,
the femur was removed from each leg by cutting through the
tibia and pelvis. Using tissue paper, all muscle was
removed from the bone. Then the epiphyseal portions of
the femur v/ere clipped leaving the diaphysis with the bone
marrow canal exposed at both ends. The bone marrow cells
were flushed from the canal using physiological saline
solution warmed to 37°C. This suspension was centrifuged
for 10 minutes at 1000 rpm. The saline was then discarded,
and the cells were resuspended in 0.075M KC1 for 30 minutes
at 37°C. The cells were fixed in methanol-glacial acetic
acid (3:1) and were allowed to stand in the first fix for 20
minutes and in the second fix overnight. After one change
of the fixative, on the following day, the cells were placed
on cold slides and were allowed to air dry.
After air drying for 24 hours, differential staining was
performed by (8) allowing the slides to stand in the absence
of light for 15-18 minutes in a solution of 33258 Hoechst
(5mg/100 ml H20), American Hoechst Corporation, (2) dip-
ping in distilled water and blotting dry (3) applying six
drops of Mcllvaines buffer solution (ph=8.0), coverslipping,
and placing on a slide tray warmed to 50°C with a 20 watt
black light mounted 5 cm above the slides, and exposing
the slides to the black light for 22 minutes, (4) floating
the coverslip off by dipping in distilled water (5) blot-
ting the slides dry and allowing them to stand for 2 hours,
and (6) placing the slides in 2% Giemsa in Gurr buffer
solution (pH=6.8) for 10 minutes. Slides were coded for
observations.
Twenty-five metaphase cells from each animal were evaluated
for SCE. The mitotic index was determined for each animal
by counting the number of metaphase cells observed per 1000
cells. Five thousand cells were counted from each animal.
4. Cytological Aberration Bioassay
For these studies, the slides prepared for SCE were used.
Twenty-five metaphase cells from each control and diesel
exhaust treated animal were evaluated for polyploidy,
various chromosomal and chromatid aberrations.
Results and Discussion
1. Sperm Morphology Rioassay
Various abnormalities in sperm shape were observed both in
control and diesel exhaust treated animals (Fig. 1) The
938
-------
NORMAL SPERM (-
CELL
A. B. C. D. E.
ABNORMAL SPERM CELLS
F.
FIGURE 1. Normal and abnormal shape of Chinese hamster
sperm eel Is.
939
-------
observed abnormalities included curled head, amorphous
head, non-sickled shaped head and abnormally small size of
head usually accompanied by curled head. Table I shows the
results for each animal and the pooled data. The percen-
tage of abnormal sperm significantly increased in animals
exposed to diesel emissions. Both the diesel exhaust
exposed animals and controls had a very low frequency of
abnormal sperm. The frequency of sperm abnormalities,
while varying among different inbred strains, is rela-
tively constant within any particular animal strain (18).
Thus, the differences in frequency of abnormal sperm can
be significant even in such small percentages as observed
in this study; however due to the small sample size these
results should be considered preliminary until reported.
In a larger sample of mice exposed to diesel exhaust for
7 weeks, no increase in the frequency of abnormal sperm
was observed (19).
2. Micronucleus Bioassay
There was some increase in the frequency of micronuclei
in animals exposed to diesel exhaust for 6 months. The
treated animals showed 0.3% to 1.3% cells with micronuclei
as compared to controls which had 0.2% to 0.6% cells with
micronuclei (Table II). The mean for the untreated animals
was 0.4% while it was 0.69% for the animals exposed to
diesel. According to the Student t test procedure the data
showed significant increase in micronuclei in hamsters
exposed to dieset exhaust at the P=0.2 level. In mice,
exposure to diesel exhaust for only 7 weeks did not result
in an increase in micronuclei (20). Whether the lack of an
effect of diesel exhaust exposure in mice on the induction
of micronuclei is a species variation or a result of the
shorter duration of exposure is under further investigation.
3. In vivo SCF Rioassay
The mitotic index of animals exposed to diesel exhaust was
clearly decreased, but no effect was observed on the fre-
quency of SCE (Table III and IV). The evaluation of meta-
phase cells from animals in this group was hampered by the
clumping of metaphase chromosomes. It has been noted that
these animals weighed more than the other animals used in
this evaluation. We have previously observed a similar
clumping effect in obese animals. Further investigation
will be required to determine if both the decreased mitotic
index and the problem of clumped metaphases was due to the
weight of the animals or the prolonged exposure to diesel
emissions. The possibility of a "combined effect" must
also be entertained.
940
-------
TABLE I. INDUCTION OF SPERM ABNORMALITIES IN CHINESE
HAMSTERS EXPOSED TO DIESEL EXHAUST
TREATMENT
CLEAN AIR
(CONTROL)
DIESEL
EXHAUST
DURATION OF ANIMAL
TREATMENT
6 MONTHS 1
2
3
4
5
6
7
8
9
10
6 MONTHS 1
2
3
4
5
6
7
8
9
10
NUMBER # OF ABNORMAL
SPERM (%)
0.4
0.4
0.2
0.2
0.2
0.4
0.6
0.6
0.2
0.4
Average = 0.36 ± .05
1.8
1.2
1.0
1.4
1.2
0.6
0.0
1.0
0.4
1.0
2
Average = 0.96 ± .16
Average ± Standard Error of the Mean.
The Increase was significant with P = 0.0024 by the
Student t Test.
941
-------
TABLE II. INDUCTION OF MICRONUCLEI BY DIESEL EXHAUST
IN CHINESE HAMSTER BONE MARROW CELLS
TREATMENT DURATION OF
TREATMENT
CLEAN AIR 6 MONTHS
(CONTROL)
DIESEL 6 MONTHS
EXHAUST
ANIMAL #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
# POLYCHROMATIC % MICRO-
ERYTHROCYTES NUCLEI
OBSERVED
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
Average
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
Average
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
= 0.40 ±
0.
0.
0.
0.
1.
0.
0.
0.
0.
0.
= 0.69 ±
5
3
6
3
2
5
5
4
4
3
1
.04
5
3
9
5
3
7
6
3
9
9
2
.10
Average ± Standard Error of the Mean.
The Increase was significant with P = 0.02 by the Student t
Test.
942
-------
TABLE III. EVALUATION OF THE SCE/CELL AND THE MITOTTC INDEX
TREATMENT ANIMAL #
CLEAN AIR 1
2
3
WEIGHT
(GM)
30.2
31.0
34.0
30.1
31.7
29.0
MITOTIC
INDEX
1
44
39
39
39
41
29
CELLS
OBSERVED
25
25
25
25
25
25
SCE/CELL
2
5.12 ± 0.5
4.36 ± 0.4
4.84 i 0.4
5.80 + 0.4
4. 72 ± 0.3
4.28 + 0.4
B(a)P 1 32.2 25 25 10.04 + 1.2
2 35.4 32 25 9.80 ± 0.7
3 31.0 27 25 10.04 ± 0.8
4 29.1 18 25 12.40 + 1.1
HESEL
EXHAUST
1
2
3
4
5
6
39.
35.
35.
38.
46.
38.
7
2
8
8
2
2
30
15
33
30
34
34
25
25
25
25
25
25
4
4
4
4
4
4
.40 +
.24 ±
.68 ±
. 36 +
.36 ±
.04 ±
0.
0.
0.
0.
0.
0.
5
5
4
4
4
4
The number of metaphase cells per 1000 cells observed was
determined five times for each animal and the number
appearing in this column represents the average.
2
Average ± Standard Error of the Mean.
Diesel exhaust did not cause an increase in the frequency
of SCE in Chinese hamster bone marrow cells. The B(a)P
positive control group responded with an increase in the
number of SCE/cell from 4.82 in controls to 10.57 (Table
IV). It would appear that diesel exhaust exposure did
not result in an increase in sister chromatid exchange in
Chinese hamsters even though there was an apparent increase
in the number of micronuclei.
943
-------
TABLE IV. SISTER CHROMATIC EXHANGE IN BONE MARROW
CELLS OF CHINESE HAMSTERS EXPOSED TO
DIESEL EXHAUST (POOLED RESULTS)
TREATMENT
CLEAN AIR
(CONTROL)
NUMBER OF
ANIMALS
6
MITOTIC
INDEX
391
CELLS
OBSERVED
150
SCE/CELL
4.82 +_ 0.22
B(a)P, 100 mg/kg 4
DIESEL EXHAUST 6
25 100 10.57 + 0.5
29 150 4.34 + 0.2
1-This number represents the average mitotic index obtained
by observing 5,000 cells per animal.
^Average + Standard Error of the Mean.
4. Chromosomal Aberration Bioassay
The bone marrow cells were studied for chromosomal aberra-
tions such as chromatid gaps, chromatid breaks, chromosome
gaps, chromosome breaks, chromatid deletion, fragmentation,
acentric fragments, translocations, triradials, quadrira-
dials, ring chromosomes, dicentric chromosomes, double
minutes, polyploidy and aneuploidy. For each animal, 25
metaphase cells were observed (Table V). The number of
polypioid cells in both control and diesel treated animals
was the same. No aberrations in chromosomes or chromatids
were observed in either controls or treated animals. It is
apparent that the diesel exhaust does not cause any cyto-
logical abnormalities, however, as stated above, there was
a decrease in the mitotic index. This investigation is now
being extended to study the genotoxic effects of higher
concentration of diesel exhaust, diesel particulate matter
and its extract.
The work upon which this publication is based was performed
pursuant to contract no. 68-03-1460 with the Environmental
Protection Agency, HERL- Cincinnati.
944
-------
TABLE V. POLYPLOIDY AND CHROMOSOMAL ABERRATIONS IN BONE
MARROW OF CHINESE HAMSTERS EXPOSED TO DIESEL
EXHAUST
ANIMAL
NUMBER
1
2
3
4
5
6
7
8
9
10
11
12
TREATMENT # OF
POLYPLOID
CELLS
CLEAN AIR (CONTROL) 1
CLEAN AIR (CONTROL) -
CLEAN AIR (CONTROL) 1
CLEAN AIR (CONTROL) 1
CLEAN AIR (CONTROL) -
CLEAN AIR (CONTROL) -
DIESEL EXHAUST
DIESEL EXHAUST 1
DIESEL EXHAUST
DIESEL EXHAUST 1
DIESEL EXHAUST
DIESEL EXHAUST 1
ABERRATIONS* TOTAL # OF
G B T R F METAPHASES
OBSERVED
----- 25
----- 25
----- 25
----- 25
----- 25
----- 25
----- 25
----- 25
----- 25
----- 25
----- 25
----- 25
ABERRATIONS *
G = GAPS
B = BREAKS
T = TRANSLOCATIONS
R = RING FORMATION
F = FRAGMENTATION
945
-------
REFERENCES
1. Wyrobek, A.J., and W.R. Bruce. 1975. Chemical induction
of sperm abnormalities in mice. Proc. Nat. Acad. Sci.
USA, 72:4425-4429.
2. Heddle, J.A., and W.R. Bruce. 1976. Comparison of
tests for mutagenicity and carcinogenicity using assays
for sperm abnormalities, formation of micronuclei and
mutations in Salmonella. In: Cold Spring Symposium on
Origins of Human Cancer, pp. 1549-1557.
3. Bruce, W.R., R. Furrer, and A.J. Wyrobek. 1974.
Abnormalities in the shape of murine sperm after acute
testicular X-irradiation. Mut. Res., 23:381-386.
4. Wyrobek, A.J., J.A. Heddle, and W.R. Bruce. 1975.
Chromosomal abnormalities and the morphology of mouse
sperm heads. Can. ,J. Genet. Cytol., 17:675-681.
5. Schmid, W. 1975. The micronucleus test. Mut. Res.,
31:9-15.
6. Maier, p., and W. Schmid. 1976. Ten model mutagens
evaluated by micronucleus test. Mut. Res., 40:325-338.
7. Wild, D. 1978. Cytogenetic effects in the mouse of 17
chemical mutagens and carcinogens evaluated by micro-
nucleus test. Mut. Res., 56:319-327.
8. Perry, P. and H.J. Evans. 1975. Cytological detection
of mutagen-carcinogen exposure by sister chromatid
exchange. Nature, 258:121-124.
9. Latt, S.A. 1974. Sister chromatid exchanges, indices
946
-------
of human chromosome damage and repair: detection by
fluorescence and induction by mitomycin C. Proc. Nat.
Acad. Sci., 71:3162-3166.
10. Bloom, S.E., and T.C. Hsu. 1975. Differential
fluorescence of sister chromatids in chicken embryos
exposed to 5-bromodeoxyuridine. Chromosoma, 51:261-267.
11. Allen, J.W., and S.A. Latt. 1976. Analysis of sister
chromatid exchange formation in-vivo in mouse
spermatogonia as a new test system for environmental
mutagens. Nature,260:449-451.
12. Vogel, W., and T. Bauknecht- 1976. Differential
chromatid staining by in-vivo treatment as a muta-
genicity test system. Nature, 260:448-449.
13. Bauknecht, T., W. Vogel, U. Bayer, and D. Wild. 1977.
Comparative in-vivo mutagenicity testing by SCE and
micronucleus induction in mouse bone marrow. Human
Gene., 35:299-307.
14. Sabharwal, P.S., and Lockard, J. 1979. Induction of
sister chromatid exchange and polyploidy by carbaryl in
V-79 cells. rn Vitro, 15:172.
15. Legator, M.S., K.A. Palmer, S. Green, and K.W. Peterson.
1969. Cytogenetic studies in rats of cyclohexalamine,
a metabolite of cyclamate. Science, 165:1139-1140.
16. Evans, H.J. 1976. Cytological methods for detection
of chemical mutagens. In: Chemical Mutagens-
Principles and Methods for Their Detection, (A.
947
-------
Hollaender, ed.). Vol. 4, Plenum Press, New York, pp.
1-29).
17. Allen, J.W., C.F. Shuller, R.W. Mendes, and S.A. Latt.
1977. A simplified technique for in-vivo analysis of
sister chromatid exchanges using 5-bromodeoxyuridine
tablets. Cytogenet. Cell Genet., 18:231-237.
18. Scares, E.R., W. Sheridan, J.K. Baseman, and M. Segall.
1979. Increased frequencies of aberrant sperm as
indicators of mutagenic damage in mice. Mut. Res., 64:
27-35.
19. Pereira, M.A., P.S. Sabharwal, A.J. Wyrobek. 1980.
Sperm abnormality bioassay of mice exposed to diesel
exhaust. In this volume.
20. Pereira, M.A., T.H. Connor, J. Meyne, M.S. Legator. 1980.
Metaphase analysis, micronuclei assay and urinary muta-
genesis assay of mice exposed to diesel emissions. In
this volume.
948
-------
General Discussion
LENBERG: With your sperm morphology testing, it seems
to me that the techniques involved were somewhat sub-
jective and you could introduce error just by handling.
Was the technician that did the analysis running the
operation using the blind analysis method?
P. SABHARWAL: Yes, all these studies are by the blind
method. The technicians do not know what samples they
are given. They were also scored by the same techni-
cians.
S. KAPLAN: Are some slides submitted to the tech-
nician a second time to see if he can replicate the
previous reading?
P. SABHARWAL: We have evaluated that variaton by
another procedure. Slides that had Iready been evaluated
were given to us by another laboratory. Our evaluation
of the slides agreed with the other laoratory.
S. KAPLAN: Yes, you need reliability as well as va-
1idity checks.
J. VOSTAL: In view of the fact that you are stuyding
systemic effects and you are using unbelievable high
doses, would you be willing to speculate whether much of
the observed effects are due to ingestion from grooming
and not from inhalation?
A. SABHARWAL: There is always that possibility since
the whole animal is placed in the chamber. Even skin
absorption could be taking place. This is similar to
what happen to us in nature.
J. VOSTAL: I would probably take a very serious ex-
ception to your statement. Obviously the exposure route
for a person being exposed to concentrations may be ten
thousand times lower than those used in the in vivo ex-
posure of animals will be relevant. We are more in-
terested in establishing the relevance of our data to the
field situation. We are more interested in finding out
the role of potential absorption from a deposit in the
respiratory system.
A. SABHARWAL: Would you kindly tell me how you do
that?
J. VOSTAL: I suppose that you don't do it unless you
are really sure that the dose results in a biological
effect. It is not that I would object to the data you
can produce by the injection of the diesel particles but
it is a question of how shall I interpret those data in
view of the question which we all face, namely, will
there be any potential artifacts generated by the diesel
exposure on our roads?
949
-------
A. SABHARWAL: I am looking for the same kind of an-
swer so if anybody has an answer we would like to know
it.
D. BRUSICK: Were the sperm abnormalities also done in
hamsters or just the mosue?
A. SABHARWAL: All the work reprted was in Chinese
hamsters. A report on sperm abnormalities is reported in
another presentation at this symposium.
D. BRUSICK: Since the effects observed in sperm ab-
normalities were significant, but not greatly different
than controls, do you have a reasonable historical data
base in this species which would indicate the sort of
variance and background that you would expect.
A. SABHARWAL: We need to perform more experiments
before we answer that question.
950
-------
SISTER CHROMATID EXCHANGE ANALYSIS OF
SYRIAN HAMSTER LUNG CELLS TREATED IN VIVO UITH
DIESEL EXHAUST PARTICULATES
Robert R. Guerrero, Donald E. Rounds
Pasadena Foundation for Medical Research
Pasadena, California
John Orthoefer
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio
Abstract
Polycyclic aromatic hydrocarbons extracted and concentrated
from diesel exhaust particulates have been shown to be
mutagenic and carcinogenic, but attempts to induce pulmonary
tumors through chronic inhalation of diesel exhaust by ex-
perimental animals have failed. We have attempted to resolve
this incongruity by measuring chromosomal damage in lung
tissue of chronically exposed hamsters, using the highly
sensitive test for chemical mutagens, sister chromatid ex-
change (SCE) analysis.
To determine the degree of responsiveness of the test system
to both diesel exhaust particulates and benzo(a)pyrene (BaP),
these agents were instilled intratracheally into anesthetized
hamsters as suspensions in 0.25 ml volumes of Hank's balanced
salt solution (HBSS). Lung tissues from these animals were
subsequently cultured J_n vitro and chromosomes from the
resulting cell divisions were scored for exchanges of chroma-
tin between sister chromatids.
Control animals, treated weekly with 0.25 ml of BSS for 10
weeks, showed an average value of 12 SCE's per cell, while
animals treated weekly with 200 ng BaP over a 10-week period
showed an average of 17 SCE's per cell. HBSS, given as a
single treatment also produced an average of 12 SCE's per cell
in control animals, but animals treated with a single instil-
951
-------
lation of 12.5 ug BaP showed an average SCE value of 19. These
data confirmed that the procarcinogen BaP can be metabolical ly
activated by lung cells jji vivo and also demonstrated the
efficacy of using this technical approach to study the effect
of chemical mutagens that enter the lungs.
Diesel exhaust particulates, administered in a range from 0 to
20 mg per hamster over a 24-hour exposure period, produced a
linear SCE dose-response ranging from 12 to 26 SCE's per
metaphase. This curve suggested that a concentration of 3 mg
of diesel particulates per hamster would not produce a
statistically significant increase in SCE's above control
values. One group of 8 hamsters, chronically exposed to
diesel exhaust particulates for 3 months showed an average of
12 SCE's per cell. This was equivalent to a set of 5 control
animals which also showed an average of 12 SCE's per cell.
Although the scope of this study was limited, the data
demonstrated that a 3-month exposure to 6 mg/m^ of diesel
exhaust particulates was insufficient to produce measurable
mutagenic changes in lung cells. This negative response is
consistent with the results from other studies in which
similar exposures failed to produce pulmonary tumors.
INTRODUCTION
Diesel exhaust contains a wide variety of polycyclic aromatic
hydrocarbons (PAH) adsorbed onto carbon particulatesH).
These agents, when extracted and concentrated, have been shown
to be mutagenic in the Ames test(2) and carcinogenic when
painted on the skin of C57 Black and A strain mice(3).
However, chronic exposure studies at concentrations and expo-
sure times which produce fibrosis and emphysema in experi-
mental animals, have failed to induce tumors in the respira-
tory tracts of these animals^).
Development of a mutagenesis/carcinogenesis response requires
an exposure to appropriate concentrations of carcinogenic
agents over suitable lengths of time. Since chronic exposure
studies have been conducted for half the lifetime of the
experimental animals and have caused respiratory pathology
other than tumor induction(4), it is most probable that the
concentration of carcinogens in diesel exhaust have been too
low to produce the carcinogenic response, even when diesel
engine performance was adjusted to maximize the production of
PAH. An extrapolation of the dose-response curve, published
by Saffiotti, et al(^) showing a linear relationship between
benzo(a)pyrene (BaP) concentration and lung tumor formation in
hamsters, suggests that a minimum of 2.5 mg BaP, accumulated
over a 30-week exposure period, would be necessary to produce
952
-------
a significant number of lung tumors. However, in diesel
exhaust studies with hamsters, one can calculate with respir-
atory total flow data that if all carbon particles from a
chronic exposure (10 mg particulates/m^ for 5 hrs/day, 5
days/week for 1 year) were retained in the lung, hamsters
would accumulate no more than 50 mg of particulates. Huis-
ingh, et a! (2) have reported that diesel exhaust can contain up
to 25 ng BaP/mg of particulate. Therefore, the maximum BaP
exposure per animal would be 1.25 ug or about 2,000 times less
than the concentration necessary to cause lung cancer. Addi-
tional factors such as plating of particles on the nasomucosal
membranes^), mucociliary clearance and inefficient elution
of the carcinogens from the particulates by lung tissue
fluids^) would make the differential even more extreme. The
accuracy of this extrapolation can be tested by performing
sister chromatid exchange (SCE) analysis on primary cultures
of hamster lung cells following J_n vivo exposure to diesel
exhaust or any diesel constituent. SCE analysis is a very
sensitive short-term bioassay for measuring genetic damage
caused by mutagens/carcinogensW. SCE analysis can detect
mutagenic changes following exposure to nanogram or microgram
concentrations of mutagen(^).
Polycyclic aromatic hydrocarbons eluted from inhaled carbon
particles can be metabolically activated by aryl hydrocarbon
hydroxylase contained in trachea! 0^) and bronchial tis-
sues!1"'). The active products bind with cellular DNAH2) ancj
the resultant genetic alteration can then be visualized as an
increase in SCE'sHS). j^is experimental approach, though
technically difficult, is superior to other systems for
assessing the actual mutagenic risk incurred by animals
following J_n vivo exposure to diesel exhaust and other
chemicals with Tow" mutagenic potential, due to the inherent
sensitivity in the system.
MATERIALS AND METHODS
Preparation of the Diesel Exhaust Particulates. The particu-
late fraction from diesel exhaust was collected on Gelman A/E
glass fiber filters by personnel in the Health Effects
Research Laboratory, Cincinnati, Ohio, then frozen and shipped
to our facility in Pasadena, California, for testing. The
particulate fraction was brushed from the filters with a
camel's hair brush. This process produced a heavy contamin-
ation with glass filter fibers. Therefore, the diesel
particulate-fiber mixture was suspended in 2% Emulphor EL-620
(GAP Corp., New York, N.Y.), a nonionic wetting agent. This
allowed the particles to dissociate from the fibers and made
it possible to separate the two components with differential
centrifugation (5 minutes at 1000 rpm in a clinical centri-
953
-------
fuge). The supernatant, containing the participate suspen-
sion, was collected and pooled from 15-20 washings of the
glass fiber pellet. A residual contamination of short glass
fibers was removed by filtering the pooled supernatant through
number one Whatman filter paper.
The carbon particles were concentrated by centrifugation at
10,000 rpm for 30 min. in a refrigerated International
centrifuge. The pellet was then washed twice with distilled
water to remove the Emulphor. The pelleted particulates were
spread on the floor of glass 100 mm diameter petri dishes and
air dried in a dessicator. The dried residue was scraped off
the glass-with a stainless steel spatula, then was weighed on
an analytical balance and resuspended in Hank's balanced salt
solution just prior to intratracheal instillation.
Preparation of Benzpyrene/Hematite Suspensions. Equal
weights of BaP (Sigma Chemical Co., St. Louis, Missouri) and
hematite (Fisher Scientific Co., Philadelphia, Pennsylvania)
were placed in a mortar and ground together for an hour. This
preparation gave a fine, homogeneously distributed dust con-
taining 50% BaP and 50% hematite by weight. In preparing the
dust, care was taken to ensure that both samples were dry to
avoid formation of large clumps. The dust was stored at 0° C
in aluminum foil wrapped tubes to exclude light. Just prior to
use, the dust was weighed on an analytical balance and
dispersed in the BSS by using ultrasound and a vortex mixer
alternately for 30 minutes.
Treatment of Experimental Animals. The experimental animals
used in the intratracheal instillation portion of this study
were 4-6 week old male Syrian hamsters obtained from Simonsen
Laboratories, Inc., Gilroy, California. A separate group of
control and experimental animals were chronically exposed to
diesel exhaust in Cincinnati, Ohio. The animals were exposed
in environmental chambers supplied with clean air or with air
containing 6.39 +_ 0.78 mg/m^ diesel exhaust particulate (DEP)
for 8 hrs/day, 7 days/week for a period of .about 3 months
(Table 1). The animals were flown to our laboratory at the end
of the 3 months for SCE analysis.
The intratracheal ly treated animals were given 0.25ml volumes
of either HBSS solution, DEP suspended in HBSS or BaP/hematite
suspended in HBSS, using the intratracheal instillation pro-
cedure described by Saffiotti, et al.O^). The animals were
anesthetized with an intraperitoneal injection of 6.7 ml/kg of
a 1% solution of sodium brevitol (Eli Lilly & co., Indian-
apolis, Indiana). Each anesthetized animal was placed on a
slanted board. With its back on the board, its mouth was kept
open by hanging the lower incisors on a wire hook while the
954
-------
Table 1
Component Concentrations in Exposure Chamber
Components
Monitored
C02, %
CO, ppm
HC, ppm as C
NO, ppm
NOg, ppm
S02, ppm
Participate, mg/m^
Reference
Chambers
0.05 + 0.01
2.07 + 0.53
3.49 + 0.38
0.08 + 0.03
0.05 + 0.03
0.05 + 0.01
-
Exhaust
Emission
Chambers
0.29 + 0.03
19.72 + 2.13
7.84 + 0.99
11.23 + 1.53
2.65 -+• 0.55
2.06 + 0.43
6.39 + 0.78
955
-------
upper incisors were retained by a tight rubber band. All of
the solutions and particulate suspensions were maintained at
37° C in a water bath. Just prior to instillation, each
suspension was vortex mixed for 5 seconds then was drawn in and
out of a 1.0 ml tuberculin syringe fitted with a blunt 19 gauge
needle several times to further insure uniform distribution of
the particulates, then the syringe was filled with 0.25 ml of
the test suspension. The blunt needle was about 60 mm long and
bent at a 135° angle 45 mm from the tip. A direct focusing lamp
from a dissecting microscope provided a view of the pharynx
after the tongue of the hamster was gently pulled outward and
laterally with forceps. The tip of the needle was inserted
under the epiglottis to uncover the vocal cords, then gently
inserted into the trachea. The needle was pushed almost to the
bottom of the trachea, then the suspension was gently injected
and the needle withdrawn. Following instillation of the
suspension into the lungs, the animals showed a brief apnea,
after which they rapidly recovered and resumed regular res-
piration.
Tissues Culture Procedure. The animals were sacrificed by
dislocation of the cervical vertebrae and then heart and lungs
were quickly excised and transferred into sterile lOOmmpetri
dishes. The heart, trachea and bronchi were removed and the
remaining lung tissue was finely minced with sterile scissors
in 3 ml McCoy's 5A medium supplemented with 10% fetal calf
serum, 100 units of penicillin, and 100 ug streptomycin/ml.
The dishes were allowed to incubate overnight in a 5% CO;?
incubator at 37° C. The following day unattached cells and
tissue fragments were removed, washed in NBSS to remove
erythrocytes and cellular debris, then the viable cells and
tissue fragments were distributed into 3 additional petri
dishes for each lung. The attached cells in the original dish
were washed once with HBSS, to remove cellular debris, then
the HBSS was replaced with complete McCoy's medium supple-
mented with 10% fetal bovine serum.
All four culture dishes for each animal were incubated in a C02
incubator at 37° C. The earliest mitotic figures were
observed at about 60 hrs of incubation and colonies of 50-200
cells were formed at about 5-8 days (Fig. 1). Cultures
containing lung tissue previously treated with either benzy-
pyrene or low concentrations of DEP were often seen to
progress faster than cultures of control lung. However,
cultures from animals given a total of 100 mg DAP over a span
of 10 weeks showed cells which were engorged with DEP (Fig. 2).
These cultures showed no mitotic activity over several weeks
of incubation.
Sister Chromatid Exchange Analysis. When the cultures showed
colonies containing 50 or more cells, they were treated with
956
-------
.•* ** * • *
\
•s;' li '%~ i*>' * -
V ,*% \* I .^%ifcV<*Vi\h * /* t ^ .,
* * V > „« - * f W w« » fc* T A*.-,".'« » - ,
\ * . -!^ V4-,k *•••' A* „ *
K
*
*•)
*V/W^:^«Sff*^.
••*\~~*»
*f"
*c
Figure 1. Photomicrograph of epithelial cells from 8-day
culture of Syrian hamster lung tissue.
957
-------
* . *
» .:%,: •>>
Figure 2. Carb on laden lung cells from a hamster which had
received 10 mg DEP/week for 10 weeks. Top, Phase
contrast image showing cellular detail. Bottom,
Bright field image of the same field showing
opaque diesel particles in the cytoplasm.
958
-------
10 ug bromodeoxyuridine (BrdU) in subdued illumination. The
dishes were wrapped irv aluminum foil and incubated in the dark
for an additional 40 hrs, during which time the cells
completed 2 cell cycles. The cultures were then treated with
0.05 ug/ml colcemid (Grand Island Biological Co.) for 3 hrs to
block cell division and accumulate as many mitotic figures as
possible. The cells were harvested with trypsin (0.05%
trypsin in Ca++, MG++ Free HBSS) and treated for 30 minutes at
37° C in 0.075 M KC1. The cells were then fixed in cold (4° C)
Carnoy's solution (1 part glacial acetic acid: 3 parts
methanol), washed 2 times in cold Carnoy's solution and
dropped on clean glass microscope slides to form chromosome
spreads.
When the slide preparations were thoroughly dry, they were
stained for 20 minutes at room temperature in a 5 ug/ml
solution of Hoechst 33258 (American Hoechst Corp., Somerville,
New Jersey) in 1/15 M Sorensen's buffer (pH 6.8). The slides
were placed in a pyrex baking tray and covered with 1/15 M
Sorensen's buffer. The buffer was overlayed with Saran wrap
to minimize water evaporation. Then the slides were exposed
to light from a bank of Gro-lights (General Electric Co.,
Cleveland, Ohio) overnight (18 hrs). The cells were then
stained with freshly prepared 3% Giemsa (Scientific Products,
Irvine, California) in 1/15 M Sorensen's buffer (pH 6.8). The
sister chromatid exchange frequency was scored in 20-30
spreads for each animal using oil immersion bright field
optics.
RESULTS AND DISCUSSION
Appraisal of the Test System. Preliminary studies with Lab-
Tek culture vessels (Miles Laboratories, Inc., Naperville,
Illinois) demonstrated that chromosome spreads, suitable for
SCE analysis, could be obtained after 4-5 days of culture
(Fig. 3). However the yield of these spreads was low,
especially for treated lung tissue. Therefore, we elected to
harvest cells after 8-10 days of culture. It is not likely
that the SCE frequency was reduced due to DNA repair during
this incubation period because the DEP treated lung cells
contained visible carbon particle inclusions (Fig. 4) which
persisted throughout the culture period. Therefore, the cells
were essentially subjected to a continuous DEP treatment until
they were fixed and stained for SCE analysis.
Mean SCE values for 12 control animals are shown in Table 2 and
Fig. 5. These mean values ranged from 9.3 to 14.1 and showed
an average of 11.8 for the 12 animals. By way of comparison,
4 animals sacrificed 24 hrs after single instillations of 12.5
ug BaP, showed a range from 18.0 to 20.7 SCE's, with an average
of 19.5. Five positive control animals given weekly instil-
959
-------
**-»^
A
v •n''
>.
Xjyj' <
^••MMl' "1. ^ .«
\\§1<
^XO^O
\
A " B
Figure 3. Differentially stained chromosome spreads from 96
hr. cultures of hamster lung. A. Spread from a
control hamster showing few SCE's. B. Spread
from a BaP treated hamster showing several
oL L S *
•$m
"T . -*"1 ** . * t * .1
v "*" &^. ^^B^ x*?
:;n^ wl
^ x:
Figure 4. Epithelial cells from a 7-day hamster lung cul-
ture. Left, Phase contrast image showing cellu-
lar detail. Right, Bright field image showing
carbon particle inclusions in the cytoplasm.
960
-------
Table 2
Mean SCE Values for Control and BaP Treated Hamsters
Control Benzo(a)pyrene treated animals
Animals _ 1 x 12.5 tig _ 10 x 0.2 ug
10.1 18.0
13.1 19.9
11.1 19.4
12.3 20.7
9.3
11.4
13.7 - 16.3
12.0 - 16.5
11.9 - 18.7
11.9 - 16.8
11.1 - 17.4
11.83+0.42* 19.50+0.65* 17.14+0.48*
*Average SCE values + SE for the sets.
961
-------
C.C.
20-
18-
K
IS
i '*-
X
K 14-
Q_
•9
&
(E 10-
u
ID
X
3 .
Z ""
Ul
<9
i 6-
4-
2-
0
o
o
o
o
<&
A
0
o
0
o
0
0
§
0 Control 1 X 12.5 q.g. 10 X 0.2 M.g.
Hamsters Benzpyrene Benzpyrwie
Figure 5. Histogram of the SCE frequency in control and BaP
treated hamster lung tissues. The circles
represent the mean SCE values for each animal in
the set. The height of the bar represents the
average value for each set.
962
-------
lations of 200 ng BaP per week for 10 weeks had an average of
17.1 SCE's with a SCE range of 16.3 to 18.7. Although these
data are not optimized for maximum responses (i.e., activation
times vs DMA repair rates) they demonstrate (a) that it is
technically feasible to perform SCE analysis on primary
cultures from control and treated hamster lungs, and (b) BaP
can be metabol ical ly activated in vivo HO* ^ ) , and can produce
a significant increase in SCE's above control values. The
accumulated dose of BaP over the 10-week interval was only 2
ug. This sensitivity level appears to be 3 orders of magnitude
greater than pulmonary tumor production in Syrian hamsters
intratracheal ly instilled with milligram quantities of
Dose-Response Curve for PEP. Primary lung cultures can be
unsuitable for SCE analysis for several reasons: (a) harvested
cells can fail to propagate, (b) mitotic figures can be lost
during the chromosome spreading procedure, (c) the spreads
that are found can be either in the first or third cell cycle
and therefore are not suitable for scoring, and (d) lungs
treated with relatively high DEP concentrations yield spreads
with so much carbon that the chromosomes are obscured (Fig. 6)
which makes SCE scoring unreliable or impossible. Because of
these factors, we were only able to collect adequate data from
one hamster for each of the DEP concentrations tested (0, 6, 11
and 20 mg of DEP). The mean SCE values and the standard errors
of the means for these 4 animals were 1 1 .4 _+ 0.2, 16.7^0.1,
19.8 + 0.1 and 26.4 + 0.3, respectively. Although these data
form a linear dose-response curve (Fig. 7), caution should be
exercised in interpreting its significance. We can conclude,
however, that the primary lung cells do respond to DEP as a
mutagen over a 24 hr treatment period.
Chronic Exposures to DEP in the Atmosphere. Lung cell
cultures provided suitable numbers of differentially stained
chromosome spreads for 5 control animals. The range of mean
SCE values for these hamsters was 10.1 to 13.1 with an average
of 11.52 _+ 0.58 (Table 3). Eight animals chronically exposed
to 6.39 mg DEP/mg^ for 3 months gave suitable cultures for
analysis. The range for this set was from 10.0 ot 14.8, with
an average of 11.86 +_ 0.47.
has calculated that a 92 g Syrian hamster inhales
an average of 0.06 liters of air per minute. Using this value,
we can calculate that animals given a chronic exposure of 6 ug
DEP/m3, 8 hrs/day, 7 days/week, for 3 months, would accumulate
a maximum of 15.6 mg of DEP. However, losses due to DEP
plating on the nasal mucosa, mucociliary clearance, and
pulmonary macrophase phagocytosis would reduce the DEP con-
centration to an estimated 1.5 to 3.0 mg/animal. The
preliminary dose-response data in Fig. 7 suggest that no
significant increase in SCE's would be expected for this level
963
-------
B
Figure 6. Differentially stained chromosome spreads showing
varying amounts of carbon particulate inclusions.
A. Spread from a control animal. B. Spread
from an animal receiving 6 mg DEP. C. Spread
from an animal receiving 11 mg DEP. D. Spread
from an animal receiving 20 mg DEP.
964
-------
18-
16-
5 10 15
DIESEL EXHAUST PARTICULATE CONCENTRATRATION (mg)
Figure 7. Dose-response curve of the SCE frequency re-
sulting from DEP administered J_n_ vivo over a
range from 0 to 20 mg per hamster by intra-
tracheal instillation.
965
-------
Table 3
Mean SCE Values for Hamsters Exposed for
Three Months to Control Air or Diesel Exhaust Emissions
Control Air Diesel Exhaust
10.1 11.6
13.1 14.8
10.9
11.9
11.1 12.2
10.0
12.3 11.1
11.0 12.4
11.52+0.58* 11.86+0.47*
*Average SCE value + SE for the sets.
966
-------
of DEP exposure. The data in Table 3 support that prediction.
It is reasonable to assume that if the sensitive SCE test shows
measurable response following 3 months chronic exposure,
pulmonary tumor production would be highly un
Since SCE increases were induced with intratracheal instilla-
tion of DEP (Fig. 7), we must assume that longer exposures and
higher concentrations of DEP would show a statistically
signficiant increase in the SCE response. Such data would
provide an important basis from which risk assessments for DEP
exposures could be made.
REFERENCES
1. Menster, M. and A.G. Sharkey, Jr. (1977). Chemical
Characterization of Diesel Exhaust Particulates.
NITS, PERC/RI-77/5.
2. Huisingh, J., R. Bradow, R. Jungers, L. Claxton, R.
Zweidinger, S. Tejada, J. Bumgarner, F. Duffield and M.
Waters. (1978). Application of Bioassay to the
Characterization of Diesel Particle Emissions. EPA-
600/9-78-027, U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina. 32 pp.
3. Kotin, P., H.L. Falk and M. Thomas. (1955). Aromatic
Hydrocarbons III. Presence in the Particulate Phase of
Diesel-Engine Exhausts and the Carcinogenicity of
Exhaust Extracts. AMA Arch. Ind. Health, 11:113-120.
4. Stuart, B.D., R.J. Palmer, R.E. Filipy, K. Mapstead and
D. Teats. (1978). Biological Effects of Chronic
Inhalation of Coal Mine Dust and/or Diesel Engine
Exhaust in Rodents. Batelle Pacific Northwest Labora-
tory Annual Report for 1977 to the DOE Assistant
Secretary for Environment. Part 1. Biomedical Sci-
ences.
5. Saffiotti, U., R. Montesano, A.R. Sellakumar and D.G.
Kaufman. (1972) Respiratory Tract Carcinogenesis
Induced in Hamsters by Different Dose Levels of Benzo-
(a)pyrene and Ferric Oxide. J. Natl. Cancer Inst.
49:1199-1204.
6. Kendrick, J., I.B. Rubin, D.A. Creasia, W.L. Maddox,
M.R. Guerin and P. Nettesheim. (1976). Respiratory
Tract Deposition of Smoke Particles Using a Nasal
Bypass Device. Arch. Environ. Health 31 (3) :131-135.
7. Creasia, D.A., J.K. Pogenburg, Jr. and P. Nettesheim.
(1976). Elution of Benzo(a)pyrene from Carbon Parti -
967
-------
cles in the Respiratory Tract of Mice. J. Toxicol.
Environ. Health., 1(16):967-975.
8. Carrano, A.V., L.H. Thompson, P.A. Lindl and J.L.
Minkler. (1978) Sister Chromatid Exchange as an
Indicator of Mutagenesis. Nature 271:551-553.
9. Perry, P. and H.J. Evans. (1975). Cytological
Detection of Mutagen-Carcinogen Exposure by Sister
Chromatid Exchange. Nature 258:121-125.
10. Mass, M.J. and D.G. Kaufman. (1978). (3H)Benzo-
(a)pyrene Metabolism in Tracheal Epithelial Microsomes
and Tracheal Organ Cultures. Cancer Res. 38:3861-3866.
11. Hsu, I.C., G.D. Stoner, H. Autrup, B.F. Trump, O.K.
Selkirk, and C.C. Harris. (1978). Human Bronchus -
Mediated Mutagenesis of Mammalian Cells by Carcino-
genic Polynuclear Aromatic Hydrocarbons. Proc. Natl.
Acad. Sci., USA. 75(4):2003-2007.
12. Harris, C.C., H. Autrup, R. Connor, L.A. Barrett, E.M.
McDowell and B.F. Trump. (1976). Interindividual
Variation in Binding of Benzo(a)pyrene to DNA in
Cultured Human Bronchi. Science 194:1067-1069.
13. Potescu, N.C., D. Turnbull and J.A. Dipaolo. (1977).
Sister Chromatid Exchange and Chromosome Aberration
Analysis with the Use of Several Carcinogens and
Noncarcinogens: Brief Communication. J. Natl. Cancer
59:289-293.
14. Saffiotti, U., F. Cefis and L.H. Kolb. (1968). A
Method for the Experimental Induction of Bronchogenic
Carcinoma. Cancer Res. 28:104-124.
15. Paintal, A. (1970). Breathing. In: Proceedings of
the Hering-Breuer Centennial Symposium, pp 59-71.
General Discussion
D. BRUSICK: In one of the preceeding paper we saw some
data using extracts in Chinese hamster cells in vitro which
indicate that sister Chromatid exchange could be induced.
Now we have data in hamsters, for both the particles and
extracts. Do you think that the dose received in vivo can
account for the difference observed under in vitro situation
versus what we observed in vivo?
D. ROUNDS: Basically, I believe it is a matter of dose
but we have to consider that the whole animal is a very
complex system. We have to think about the method by which
968
-------
acting components can be diluted off the carbon and dis-
tributed throughout the system.
D. BRUSICK: It might be interesting, and maybe not even
too far fetched, to consider an experiment of the type that
you have done with a combination of diesel inhalation ex-
posure at low concentration and other environmental ma-
terials such as benzene. In other words, the combination of
the particulates and solvents be worth looking at?
D. ROUNDS: Yes, we had hoped that we could use the tech-
nique for any kind of air pollutant. We have previously
done some studies with gene cells which were derived from
the human body when exposed to ozone in vitro sister chroma-
tid exchange was induced. This would be a good way to ex-
amine
ozones, benzene, or any kind of ir pollutant in order to
evaluate the kind of erosol and the kind of mutagenic re-
sponses that can occur in te pulmonary tract of the exposed
animal. S. SODERHOLM
S. SODERHOLM: I don't believe you mentioned in your
calculation of the inhaled dose a deposition efficiency.
There is actually an experimental number that says a depo-
sition efficiency in the deep lung, meaning that which was
retained after 24 hours, is roughly 20 percent for the size
of particles used in your experiments. Was that taken into
account?
D. ROUNDS: We estimated 10 to 20 percent for the one and
a half to three milligrams that I quoted. Three millograms
would result from a 20 percent deposition efficiency.
D. NETTESHIM: I believe it would be wise in the future,
in such experiments, to incorporate some controls with some
other dust. I wonder if by instilling some innocusus ma-
terial such as ferric oxide, whether a similar SCE response
would be obtained? It might also be worthwhile to look at
the lymphocytes in the lymph nodes, since the material is
going to be trapped there for a long time. I don't know
whether there are enough cells there for you to use the SCE
test, but I think this is another possibility to consider.
D. ROUNDS: We had thought about using a powder of acti-
vated charcoal for the intratracheal instillation but de-
scribed against it because there are all kinds of organic
material absorbed to charcoal. The lymphocyte approach
could be a very logical extension to the SCE test.
969
-------
TEST OF DIESEL EXHAUST EMISSIONS IN THE RAT LIVER FOCI ASSAY
M. A. Pereira
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
H. Shinozuka and B. Lombardi
Department of Pathology
University of Pittsburgh School of Medicine
Pittsburgh, Pennsylvania 15261
ABSTRACT
An initiation/promotion assay in rat liver has been
developed by Sells et al (1). Partial hepatectomy was
used to enhance initiation, and a choline-devoid diet as
promoter. The induction of carcinogenesis was determined
by the focal appearance of gamma glutamyl transpeptidase
(GGT) positive hepatocytes. We adopted this assay to
diesel exhaust emission by performing a 2/3 partial
hepatectomy, and exposing the rats to either clean air or
diesel exhaust emission. The rats were fed either a
choline-devoid or a choline-supplemented diet for three
or six months. The animals were sacrificed and liver
sections stained for GGT were examined for the presence
of foci of GGT(+) hepatocytes. The results indicate that
diesel exhaust exposure does not result in a systemic
dose of carcinogens sufficient to be detected in the
liver foci assay.
INTRODUCTION
Diesel exhaust emissions contain thoudands of organic
chemicals, some of which are carcinogenic polycyclic
aromatic hydrocarbons. The high particle concentration
of diesel exhaust results in sequestration of much of the
polycyclic aromatic hydrocarbons in the particles. The
sequestered carcinogens must be eluted in order to be
active. Therefore, the assessment of the health hazard
970
-------
of exposure to diesel exhaust requires the demonstration
JH vivo of its carcinogenic activity in a bioassay
impacted by pharmacokinetic and metabolic factors unique
to particle-bound carcinogens.
The two stage evolution of carcinogenesis, consisting of
an initiation and a promotion stage, originally described
by Berenblum for mouse skin (2, 3) has been demonstrated
in many other tissues (4-8). Promoters are agents that
reduce the latent period required for the appearance of
tumors, but have no ability to initiate target cells.
Tumor promoters have been described for skin, liver,
lung, bladder and colon carcinogenesis (4-9).
Initiation/promotion bioassays are being developed, for
testing the potential carcinogenicity of environmental
factors, that take advantage of the reduction in the time
required for the appearance of either preneoplastic
lesions or of tumors afforded by the use of promoters.
Partial hepatectomy has been shown to enhance the initi-
ation of rat liver cells by chemical carcinogens (10-11).
Thus, administration of non-liver carcinogens, including
7-12- dimethylbenz(a)anthracene (12), within twenty-four
hours after a partial hepatectomy, results in liver
cancer. Initiated liver cells can be detected, histo-
chemically, from the development of foci of GGT(+)
hepatocytes (13-14), and the evolution of initiated cells
to foci of GGT(+) hepatocytes is efficiently promoted in
rats fed a choline-devoid (CD) diet (15). Therefore, the
bioassay in rats outlined in figure 1 involves 1) A 2/3
partial hepatectomy, to enhance initiation; 2) feeding a
CD diet as a promoter; and 3) induction of foci of GGT(+)
hepatocytes as the end point. Using this bioassay, we
have attempted to assess the carcinogenicity of diesel
exhaust.
MATERIALS AND METHODS
Male Sprague-Dawley rats (Charles River), weighing 150-
175 gm, were used. The choline supplemented (CS) and CD
diets were prepared as described by Shinozuka et al (15).
The rats received a 2/3 partial hepatectomy or a sham
operation, and exposure to either clean air or diesel
exhaust was begun on the same day. The rats were fed the
CS or CD diet jid libitum.
The diesel engine and exposure conditions at the U.S.
Environmental Protection Agency, Health Effects Research
Laboratory facilities in Cincinnati, Ohio, are described
by Hinners (16). Briefly, the six cylinder Nissan engine
971
-------
used was run on the Federal Short Cycle. The animals were
exposed to diesel exhaust for eight hours daily at 6
mg/m3 particles (1:18 dilution).
After three and six months of exposure, and feeding
either the control CS diet or the CD diet, groups of rats
were sacrificed by decapitation and the liver excised.
Small blocks of liver tissue were fixed in 95% ethanol -
1% acetic acid, dehydrated and embedded in paraffin.
Sections (8 urn) were stained for gamma glutamyl trans-
peptidase activity by the method of Rutenberg et al (17)
or with hematoxylin-eosin. Approximately 2-3 cm2 of
liver sections from each rat were examined for the
presence of foci of GGT(+) hepatocytes (Figure 1).
RESULTS AND DISCUSSION
The rats were divided into six groups of 25 each and were
treated as described in Table 1. Table 1 contains also
the experimental results. Significant numbers of foci of
GGT(+) hepatocytes developed in none of the groups of
rats, in particular in those exposed to diesel exhaust
for three or six months. There was also no remarkable
liver toxicity after six months of exposure. It is
evident, therefore, that the bioassay used in these
studies, is unable to reveal the carcinogens and hepato-
toxins which are present in diesel exhaust. This
inability could be due to: 1) the sequestration of the
carcinogens and hepatoxins in the diesel exhaust par-
ticles, so that they are not distributed systemically;
and 2) the rat liver foci bioassay, as employed in these
studies, is not sensitive enough.
ACKNOWLEDGEMENT
The work upon which this publication is based was
performed pursuant to Contract No. 68-03-2793 with the
Environmental Protection Agency, HERL-Cincinnati.
REFERENCES
1. Sells, M.A., S.L. Katyal, S. Sell, H. Shinozuka and
B. Lombardi. 1979. Induction of foci of altered
gamma-glutamyl transpeptidase positive hepatocytes
in carcinogen-treated rats fed a choline-deficient
diet. Br. J. Cancer, 40:274-283.
2. Berenblum, I. 1941. The co-carcinogenic action of
croton resin. Cancer Res., 1:44.50.
972
-------
m^m i ^siRp^^ais
Figure 1.
. ISP?
4<,*fJ^-l&%%"'-.?" . •
>* *'^l! J: <'«»&"&''. .
Foci of gamma glutamyl transpeptidase posi-
tive hepatocytes. Rats were administered
100 mg/kg diethylnitrosamine by intraperi-
toneal injection. The rats were fed lab-
oratory chow for a week and then fed a
choline-devoid diet for six weeks. The dark
areas are GGTase positive islands (70X).
973
-------
Table 1. Rat Liver Bioassay of Diesel Exhaust
Treatment No. of GGT(+) Foci/cm29
1. P.H.b - Diesel Exhaust - CD diet
3 months 0
6 months 0.20 + 0.20
2. P.H. - Diesel Exhaust - CS diet
3 months 0
6 months 0
3. S.0.c - Diesel Exhaust CD diet
3 months 0
6 months 2.28+2.28
4. P.H. - Clean Air - CD diet
3 months 0
6 months 5.68 + 3.07
5. P.H. - Clean Air - CS diet
3 months 0
6 months 0.29 + 0.29
6. S.O. - Clean Air - CD diet
3 months 0
6 months 2.82 + 1.91
a - At least 2 square centimeters were examined for the
presence of GGT(+) foci.
b - P.H. = 2/3 partial hepatectomy.
c - S.O. = Sham Operation
974
-------
3. Berenblum, I. 1941. The mechanism of carcino-
genesis: A study of the significance of co-
carcinogenic action and related phenomena. Cancer
Res., 1 :807-811. ~
4. Solt, D. and E. Farber. 1976. New principle for the
analysis of chemical carcinogenesis. Nature, 263:
701-703.
5. Peraino C., R.J.M. Fry and D.D. Grube. 1978. Drug-
induced enhancement of hepatic tumorigenesis. In:
Carcinogenesis: A Comprehensive Survey. T.J.
Slaga, A. Sivak and R.K. Boutwell (Eds.) Vol. 2,
421-432, Raven Press, N.Y.
6. Pitot, H.C., L. Barness and T. Kitagawa. 1978.
Stages in the process of hepatocarcinogenesis in rat
liver. _In: Carcinogenesis: A Comprehensive Study.
T.J. Slaga, A. Sivak and R.K. Boutwell (Eds.) Vol.
2, 433-442, Raven Press, N.Y.
7. Reddy, B.S., J.H. Weisburger and E.L. Synder. 1978.
Colon cancer: bile salts as tumor promoters. In:
Carcinogenesis: A Comprehensive Survey. TT3.
Slaga, A. Sivak and R.K. Boutwell (Eds.) Vol. 2,
453-464, Haven Press, N.Y.
8. Witschi, H. and S. Lock. 1978. Butylated hydroxy-
toluene: A possible promoter of adenoma formation
in mouse lung. _!TK Carcinogenesis: A Com-
prehensive Survey. T.J. Slaga, A. Sivak and R.K.
Boutwell (Eds.) Vol. 2, 465-474, Raven Press, N.Y.
9. Hicks. R.M., J. Chowaniec and J. St. J Wakefield.
1978. Experimental induction of bladder tumors by a
two-stage system. In: Carcinogenesis: A Com-
prehensive Survey, T.J. Slaga, A. Sivak and R.K.
Boutwell (Eds.) Vol. 2, 475-487, Raven Press, N.Y.
10. Scherer, E. and P. Emmelot. 1976. Kinetics of
induction and growth of enzyme-deficient islands
involved in hepatocarcinogenesis. Cancer Res.,
36:2544-2554.
11. Cradtiock, V.M. 1976. Cell proliferation and ex-
perimental liver cancer. In: Liver Cell Cancer.
H.M. Cameron, D.A. Linsell~and G.P. Warwick (Eds.)
153-201, Elsevier Sci. Publ Co., Amsterdam.
975
-------
12. Marquidt, H., S.S. Sternberg and F.S. Philips. 1970.
7,12-Dimethylbenz(a)anthracene and hepatic neopla-
sia in regenerating rat liver. Chem.-Biol. Inter-
actions , 2:401-403.
13. Kalengay, M.M.R., G. Ronchi and V.J. Desmet. 1975.
Histochemistry of gamma-glutamyl transpeptidase in
rat liver during aflatoxin B^-induced carcinogen-
esis. J. Natl. Cancer Inst., 55:579-588.
14. Pugh, T.D. and S. Goldfarb. 1978. Quantitative
histochemical and autoradiographic studies of hepa-
tocarcinogenesis in rat fed 2-acetylaminofluorene
followed by phenobarbital. Cancer Res., 38:4450-
4457.
15. Shinozuka, H., B. Lombardi, S. Sell and R.R. lam-
marino. 1978. Early histological and functional
alterations of ethionine liver carcinogenesis in
rats fed a choline deficient diet. Cancer Res.,
38:1092-1098.
16. Hinners, R.G., J.K. Burkart, M. Malanchuk, and W.D.
Wagner. 1980. Facilities for diesel exhaust stud-
ies. Proceedings for the International Symposium on
Health Effects of Diesel Engine Emission, December
1979.
17. Rutenberg, A.M., H. Kim., J.W. Fischbein, J.S.
Hankers, H.L. Wasserkrug and A.M. Seligman. 1969.
Histochemical and ultrastructural demonstration of
gamma-glutamyl transpeptidase activity. J. Histo-
chem Cytochem., 17:517-526.
976
-------
THE EFFECT OF DIESEL EXHAUST
ON SPERM-SHAPE ABNORMALITIES IN MICE
M. A. Pereira
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
P. S. Sabharwal
T. H. Morgan School of Biological Sciences
University of Kentucky
Lexington, Kentucky
L. Gordon and A. J. Wyrobek
Biomedical Sciences Division
Lawrence Livermore Laboratory
University of California
Livermore, California 94550
ABSTRACT
The sperm-shape abnormality bioassay in mice was used to
determine whether chemical mutagens in diesel exhaust reach
the testes. Strain A male mice (30 per group from 4 to 6
weeks of age) were exposed for 31 or 39 weeks to either
diesel exhaust or clean air. After exposure, Eosin y-
stained, air-dried smears of cauda epididymal sperm were
scored for changes in sperm-head abnormalities in three
different laboratories. There was no difference in the
proportion of abnormally shaped sperm in controls and mice
exposed to diesel exhaust.
977
-------
Diesel exhaust emissions contain thosuands of organic
chemicals, including known carcinogens and mutagens.
Because of the high particle concentration of diesel
exhaust a large proportion of the organics, especially
the polycyclic aromatic hydrocarbons, are adsorbed onto
particles (1). For genotoxic agents in diesel exhaust to
affect the mammalian testes they must be deposited in the
lungs, eluted from the particles, and distributed in the
body so that active metabolites can reach the testes.
The murine sperm-shape abnormality bioassay provides a way
of assessing possible spermatoxic and genotoxic effects of
chemical agents in 'vivo (2). Studies in mice have shown
that a) sper-head shape is under genetic control, b) chemi-
cal induction of sperm abnormalities is correlated with
germ-cell mutagenicity, and c) mutations associated with
induced sperm-abnormalities can be inherited in the off-
spring of treated mice (3). Several other factors, such
as x-rays, heat, and severe fevers involving the testes
can also induce sperm-head abnormalities (3). Besides
identifying testicular toxins and mutagens the sperm abnor-
mality assay may also be an effective indicator for car-
cinogens. For example, sixty-five (65%) percent of 38 known
carcinogens tested by this bioassay assay showed positive
results, a correlation very similar to that found using the
Salmonella bioassay, which is associated with carcinogenic
potential (4).
In this communication, we report the effect of inhalation
exposure to diesel exhaust on the murine testes using the
sperm-head abnormality assay.
METHODS AND MATERIALS
Male mice (strain A/Strong, Strong Research Foundation,
Calif.) at 4 to 6 weeks of age were divided into 4 groups of
30 mice each and exposed to either clean air or diesel
exhaust emissions for 31 or 39 weeks. Thirty-one and 39
weeks were chosen because they represent approximately 6 and
8 complete spermatogenic durations. This experiment design
permitted 2 independent assessments of the effects of
diesel. The diesel exhaust engine and exposure conditions
at the U.S. Environmental Protection Agency, Health Effects
Research Laboratory facilities in Cincinnati, Ohio, are
described elsewhere by Hinners (5). Briefly, a 6-cylinder
Nissan diesel engine was run on the Federal Short Cycle and
animals were exposed to exhaust for 8 hours daily at 6
particles concentration using a 1:18 dilution.
978
-------
At the end of the exposure interval, the mice were sacrified
by cervical dislocation, the cauda epididymides excised, and
minced in 4 ml of phosphate buffered physiological saline.
The suspension from each mouse was mixed by repeated pipet-
ting and filtered through 74 ym nylon mesh. An aliquot was
mixed (10:1) with 1% Eosin-Y prepared in water and incubated
for 30 min. Smears were made, air-dried, and mounted with
Permount. For each mouse in the control and exposed groups
receiving 31 week exposures, 500 sperm were examined in each
of three different laboratories. The control and exposed
mice receiving 39 week exposures were examined at only one
of these laboratories. Examples of normal and abnormal
sperm shapes are presented in Figure 1.
RESULTS AND DISCUSSION
The effect of diesel exhaust emission on sperm shape is
summarized in Table 1. Although there were some differences
in the scoring criteria used in the 3 labroatories, none of
the laboratories detected a difference in the percentage of
abnormally shaped sperm in mice exposed for 31 weeks to
diesel exhaust when compared to clean air controls. Labor-
atory III scored an additional 2 groups of 30 mice each,
that were exposed for 39 weeks to either clean air or diesel
exhaust. As shown in Table I, the percentages obtained for
these animals were not statistically different from the
values obtained for the 31 week exposure, and also showed
no difference between control and exposed mice.
These results show that inhalation exposure to diesel
exhaust did not induce a detectable increase in sperm
head abnormalities in mice used. The failre to detect
an increase suggests that the testes is not affected
by inhalation of diesel exhaust. The negative finding,
however, may be due in part to the genotype used in this
study. Although the A/Strong strain is genetically able
to activate and metablize hydrocarbons, it has a very
high spontaneous rate of sperm shape abnormalities in
unexposed males (approximately 27%, see Lab III, Table 1)
when compared to the 1 to 2% found in B6C3F1 hybrid males
which are normally used in this bioassay (6). This high
background rate and the unusually high variance in the
unexposed mice could have masked any small positive effect.
The failure to detect increases in sperrn shape abnormalities
might also have been due to 1) the low levels of genotoxic
material sin diesel exhaust, 2) the strong adsorption of
such chemicals onto the diesel particles which may have
prohibited their release in vivo, or 3) the inability of the
released materials to be transported from the lung, to be
metabolically activated, or to reach the testes.
979
-------
SPERM-SHAPE ABNORMALITIES
NORMAL
ABNORMAL
Figure 1. Shapes of normal and abnormal sperm from mice.
980
-------
0
M
*
£2
M
>>
8
o
a
OS
o
a
Jjl
2
w
pL|
C/i
§
H
co
O
s
3
w
CO
M
Q
0
H
O
&4
[V,
TT-I
g
CO
(X
co
U-l
0 -0
CO
CO Ou
no to
tO XI
4-1 CO
c
tu >-,
O r-J
J-l r-H
0) to
PM S
o
&
XI
14-1
O to
r-4
Vi tO
• •
CN CN
r^ CN
ON M
r-l
+ 1 +1
r~ oo
• •
f^ r^
CM CM
m CM
• •
r^ m
+ 1 +1
ON C^
• •
m 0
CO
M
01
CO
e
cu
ex
10
0)
p.
X!
to
s^
r— J
rH
CO
to co S
0) O) S
M M o
3 3 C
to co Xi
O O CO
0. CX
X X >n
0) 01 O
w, y ^j
0) CO P
CO CO
S E* u
)-i
rH CT\ 4)
CO ro 0^
CO J3 0
981
-------
Further studies are planned to a) identify, characterize,
and quantitate the amount of genotoxic agents in diesel
exhaust particles and their purified extracts, b) to eval-
uate the role of particle adsoprtion, metabolic transport,
and activation, and c) to compare the genotoxic activities
of purified agents in vitro and in vivo.
ACKNOWLEDGEMENT
The work upon which this publication is based was performed
pursuant to Contract No. C 3305NANX with the Environmental
Protection Agency, HERL-Cincinnati, and under the auspices
of the U.S. Department of Energy by the Lawrence Livermore
Laboratory under Contract No. W-7405-ENG-48. Reference to
a company or product name does not imply approval or recom-
mendation of the product by the University of California or
the U.S. Department of Energy to the exlusion of others that
'may be suitable.
REFERENCES
1. Huisingh, J., R. Bradow, R. Jungers, L. Claxton, R.
Zweidinger, S. Tejada, J. Bumyarner, F. Duffield, M.
Waters, V. P. Simmon, C. Hare, C. Rodriguez, and L.
Snow, 1978. Application of Bioassay to the Characteri-
zation of Diesel Particle Emissions. In: Application
of short-term bioassays in the fractionation and analy-
ses of complex environmental mixtures, pp. 381-418. EPA
publication 600/9-78-027.
2. Wyrobek, A. J. and W. R. Bruce, 1975. Chemical Induc-
tion of Sperm Abnormalities in Mice. Proc. Nat. Acad.
Sci. USA, 72:4425-4429.
3. Wyrobek, A. J. and W. R. Bruce, 1978. The nduction of
sperm-shape abnormalities in mice and humans. In:
Chemical Mutagens: Principles and Methods for Their
Detection, A. Hollaender and F. J. de Serres (Eds.),
Vol. 5, 257-284, Plenum Press, NY.
4. Heddle, J. A. and W. R. Bruce, 1977. Comparison of
tests for mutagenicity or carcinogenicity using assays
for sperm abnormalities, formation of micronuclei and
mutations in Salmonella. In: Origin of Human Cancer,
H. H. Hiatt, J. D. Watson and J. A. Winstein (Eds.) Vol.
3, 1549-1557, Cold Spr. Harb. Lab., NY.
982
-------
5. Hinners, R. G., J. K. Burkhart, M. Malanchuk, and W. D.
Wagner, 1980. Facilities for diesel exhaust studies.
Proceedings of the International Syposium on Health
Effects of Diesel Engine Emissions, Dec. 1979.
6. Wyrobek, A. J., 1979. Changes in mammalian sperm mor-
phology after x-ray and chemical exposures. Genetic
Suppl. 92:sl05-sll9.
983
-------
TESTING FOR THE ABILITY OF MARINE DIESEL FUEL VAPORS
TO INDUCE MICRONUCLEI OR SISTER CHRQMATID EXCHANGES
IN PERIPHERAL LYMPHOCYTES TAKEN FROM DOGS
EXPOSED CONTINUOUSLY BY INHALATION FOR THIRTEEN WEEKS
R. Daniel Benz and Patricia A. Beltz
University of California, Irvine
Toxic Hazards Research Unit
P.O. Box 3067, Overlook Branch
Dayton, Ohio 45431
ABSTRACT
Purebred beagles were continuously exposed by inhalation to
0.05 mg/1 or 0.30 mg/1 marine diesel fuel (DFM) for thirteen
weeks. Peripheral blood samples were taken during this ex-
posure and after being stimulated to grow in vitro, the
lymphocytes were examined for the presence of micronuclei,
an indication of chromosome breakage, or sister chromatid
exchanges (SCEs), an indication of chromosome rearrangement.
No micronuclei above control levels were found in the ex-
posed animals after one, nine or thirteen weeks exposure and
no SCEs above control level were found after thirteen weeks
of exposure to DFM.
INTRODUCTION
Marine diesel fuel (DFM) is used by a large number of ships
operated by the United States Navy. It is prepared from
natural crude oil and is composed of a mixture of a large
number of different branched and cyclic hydrocarbons, in-
cluding a small amount of benzene. To simulate the experi-
ence of ship crews who might be exposed to DFM in confined
spaces for the duration of cruises, animals were exposed in
inhalation chambers to DFM continuously for thirteen weeks.
984
-------
Two groups each of six purebred dogs (three of each sex),
150 Fischer 344 rats (75 of each sex) and 140 C57B1/6 mice
(all female) were exposed continuously to 0.05 mg/1 and
0.30 mg/1 DFM, respectively, for thirteen weeks in Thomas
Dome inhalation chambers. Another identical group of ani-
mals, housed separately, served as controls. During and at
the end of the exposure period, various parameters were
studied to determine the effect of DFM on the exposed ani-
mals. These parameters included hourly observations of the
animals for signs of stress, counts of animal deaths, peri-
odic animal weighings, periodic hematological and blood
chemistry studies including red blood cell osmotic fragility
measurements, and organ weighings and gross and histopatho-
logic examinations of tissues of animals who died or were
sacrificed immediately after the end of the exposure period,
nineteen months postexposure or after the end of the ani-
mals' natural lifespan. The particulars of these observa-
tions are recorded in MacEwen and Vernot (1,2).
We report here the results of two other tests that were
done. These were included to determine if DFM could induce
cytogenetic damage to the dogs exposed by inhalation. Dogs
Figure 1. Thomas Dome Inhalation Chamber.
985
-------
were chosen for the testing because we wished to develop
general cytogenetic toxicologic testing methods that could
be applied directly to humans. Dogs are large mammals as
are humans and it is very easy to obtain a blood sample
from a dog without otherwise harming the animal and so per-
iodic samples can be taken from the same animals during a
chronic exposure.
Periodic blood samples were taken from the dogs and lympho-
cytes stimulated to divide in vitro were examined for the
presence of micronuclei, a very quickly and easily detected
indication of induced chromosome breakage, and sister chrom-
atid exchanges, an indication of induced chromosome rear-
rangement and the most sensitive test for mammalian cyto-
genetic damage known.
HOIST KINO-
Figure 2, Cross Sectional View of a
Thomas Dome and Access Air Lock
986
-------
MATERIALS AND METHODS
Details of the methods of generation and monitoring of DFM
vapors can be found in MacEwen and Vernot (1). One of the
Thomas Dome inhalation chambers used for this study is shown
in Figure 1 with a cross sectional diagram of it shown in
Figure 2. There are eight of these domes at the Toxic Haz-
ards Research Unit. This research was sponsored by the
Aerospace Medical Research Laboratory, Aerospace Medical
Division, Air Force Systems Command, Wright-Patterson Air
Force Base, Ohio under Contract No. F33615-76-C-5005 with
the University of California, Irvine. The experiments re-
ported herein were conducted according to the "Guide for the
Care and Use of Laboratory Animals," Institute of Laboratory
Animal Resources, National Research Council.
To detect induced chromosome breakage in this study, the
micronucleus test was used (3-7). Blood samples were taken
from all the dogs involved in the study two weeks before
exposure began and after one, nine and thirteen weeks of
continuous exposure to DFM vapors. After each bleeding,
the blood was immediately incubated for 75 hours (four cell
doubling times) in McCoy's 5A cell culture medium at 37° C
with a 5% C02 atmosphere in the presence of phytohemaglu-
tinin to stimulate lymphocyte growth. The cells were then
fixed onto microscope slides and stained with Giemsa blood
stain. At least 500 cell nuclei were scored for the pre-
sence of accompanying micronuclei on each of two slides for
each animal. Figure 3 shows a canine peripheral lymphocyte
nucleus with accompanying micronucleus.
lOum
Figure 3. Canine Peripheral Lymphocyte
Nucleus and Associated Micronucleus
987
-------
To detect induced chromosome rearrangement, the sister
chromatid exchange (SCE) test was used (8-11). Blood
samples were taken from dogs after thirteen weeks of con-
tinuous exposure to DFM and incubated as described for
the micronucleus test but for 49 hours and in the presence
of bromodeoxyuridine. At that point colchicine was added
and the cultures were incubated for four additional hours
to trap the cells in metaphase. After this time the cells
were swollen and fixed onto microscope slides and stained
with Hoechst 33258 and Giemsa blood stains. Cells in
metaphase were then examined for the presence of induced
sister chromatid exchanges. Figure 4 shows a peripheral
lymphocyte cell in metaphase that had been taken from an
unexposed dog. The metaphase cell has been spread and
stained to show the presence of sister chromatid exchanges.
lOym
Figure 4. Canine Peripheral Lymphocyte Metaphase Chrom-
osomes Stained to Reveal Sister Chromatid Exchanges.
988
-------
Leo
5
u.
O
O
1/1
O
to
CO
OJ
u
o
O- Q)
lO CO
S- QJ
OJ -i-
.c Q
Q-
•i- O)
S- E
OJ -r-
Q. i-
O IT!
S_ i—
O rO
CO
O
Q.
x
CO
cr,
• O
un o
OJ q-
s- o
+> S113D OOO'l 83d
jo
989
-------
RESULTS
The results of our tests for the presence of induced micro-
nuclei in the peripheral lymphocytes of the dogs exposed by
inhalation to DFM are shown in Figure 5. Each point shown
is the average count for all six animals exposed at each
dose level. No differences were seen between males and fe-
males. No micronuclei above background counts were found
to be induced by the dose levels of DFM vapors used in this
study. All counts were unusually high at the thirteen week
sampling. These slides contained an unusually high amount
of cell debris which led to this result.
The results of our test for the presence of induced sister
chromatid exchanges in lymphocytes of some of the male
dogs involved in this study are shown in Table 1. Values
are shown for five individual animals. No SCEs above back-
ground were found to be induced by the dose levels of DFM
tested.
Table 1. Induction of Sister Chromatid Exchanges
in Peripheral Lymphocytes of Dogs Exposed
by Inhalation to Marine Diesel Fuel
Animal
Identification
Number
X 48
X 88
X 52
Y 10
X 14
Exposure
Level
0.00
0.00
0.05
0.05
0.30
Number of
Metaphases
Examined
56
53
12
2
20
Number of
Exchanges/
Chromosome
0.0471
0.0573
0.0442
0.0147
0.0600
Number of
Exchanges/
Cell
3.67
4.47
3.45
1.15
4.68
DISCUSSION
The major purpose in doing this study was to establish pro-
cedures and techniques for performing cytogenetic tests with
peripheral lymphocytes obtained from dogs. We wished to de-
velop an animal model for future direct cytogenetic testing
of humans (12,13) using dogs which are also large mammals.
Stetka et al (14) have done similar work with rabbits.
Some direct testing of humans using the sister chromatid
exchange test has already been done (15-17), but we feel
that carefully controlled animal exposure studies are
needed to validate such testing. This study of the effect
of DFM on dogs is a beginning in doing such a validation.
REFERENCES
1. MacEwen, J.D. and E.H. Vernot. 1978. A subchronic in-
990
-------
halation toxicity study of 90 day continuous exposure to
diesel fuel marine. In: Toxic Hazards Research Unit
Annual Technical Report: 1978, AMRL-TR-78-55, Aerospace
Medical Research Laboratory, Wright-Patterson Air Force
Base, Ohio, August, 1978.
2. MacEwen, J.D. and E.H. Vernot. 1979. A subchronic tox-
icity study of 90-day continuous inhalation exposure to
diesel fuel marine. In: Toxic Hazards Research Unit
Annual Technical Report: 1979, AMRL-TR-79-56, Aerospace
Medical Research Laboratory, Wright-Patterson Air Force
Base, Ohio, August, 1979.
3. Heddle, J.A. 1973. A rapid in vivo test for chromosom-
al damage. Mut. Res. 18:652-654.
4. Countryman, P.I. and O.A. Heddle. 1976. The production
of micronuclei from chromosome aberrations in irradiated
cultures of human lymphocytes. Mut. Res. 41:321-332.
5. Countryman, P.I. and J.A. Heddle. 1977. A true micro-
culture technique for human lymphocytes. Hum. Gen. 35:
197-200.
6. Heddle, J.A., R.D. Benz and P.I. Countryman. 1978.
Measurements of chromosomal breakage in cultured cells
by the micronucleus technique. In: Mutagen Induced
Chromosome Damage in Man, H.J. Evans and D.C. Lloyd
eds., Edinburgh University Press, pp. 191-198.
7. Heddle, J.A., C.B. Lue, E.F. Saunders and R.D. Benz.
1978. Sensitivity to five mutagens in Fanconi's anemia
as measured by the micronucleus method. Cane. Res. 38:
2983-2988.
8. Latt, S.A. 1973. Microfluorometric detection of deoxy-
ribonucleic acid replication in human metaphase chromo-
somes. Proc. Nat. Acad. Sci. 70:3395-3399.
9. Perry, P. and S. Wolff. 1974. A new giemsa method for
the differential staining of sister chromatids. Nature
251:156-158.
10. Chaganti, R.S.K., S. Schonberg and J. German. 1974. A
many fold increase in sister chromatid exchanges in
Bloom's syndrome lymphocytes. Proc. Nat. Acad. Sci.
71:4508-4512.
11. Perry, P. and H.J. Evans. 1975. Cytological detection
of mutagen/carcinogen exposure by sister chromatid ex-
change. Nature 258:121-125.
991
-------
12. Crocker, T.T., R.D. Benz and R.E. Rasmussen. 1978.
Cytologic effects of Air Force chemicals. AMRL-TR-78-
77, Aerospace Medical Research Laboratory, Wright-Pat-
terson Air Force Base, Ohio, September, 1978.
13. Benz, R.D., R.E. Rasmussen, A.M. Rogers, P.A. Beltz and
T.T. Crocker. 1979. Cytologic effects of Air Force
chemicals (second of a series). AMRl-TR-79-55, Aero-
space Medical Research Laboratory, Wright-Patterson Air
Force Base, Ohio, August, 1979.
14. Stetka, D.G., J. Minkler and A.V. Carrano. 1978. In-
duction of long lived chromosome damage, as manifested
by sister chromatid exchange, in lymphocytes of animals
exposed to mitomycin C. Mut. Res. 51:383-396.
15. Carrano, A.V. 1978. Sister chromatid exchanges: A
rapid quantitative measure of genetic damage. Energy
and Technology Review, UCRL-52000-78-10.-1-8, Lawrence
Livermore Laboratories, Livermore, California, October,
1978.
16. Lambert, B., A Linblad, M. Nordenskjtild and B. Werelius.
1979. Increased frequency of sister chromatid exchanges
in cigarette smokers. Hereditas 88:147-149.
17. Kucerova, M., A. Polikova and J. Batora. 1979. Compar-
ative evaluation of the frequency of chromosome aberra-
tions and sister chromatid exchange numbers in periph-
eral lymphocytes of workers occupationally exposed to
vinyl chloride monomer. Mut. Res. 67:97-100.
992
-------
Session VII
CARCINOGENIC EFFECTS OF EXPOSURE TO DIESEL EMISSIONS
Chairman:
Dr. Michael Pereira
Carcinpgenicity of Diesel Exhaust Particles By Intratracheal
Instillation-Dose Range Study.
Shefner, Alan M., Lawrence Dooley, Arsen Fiks, C. J.
Grubbs, John H. Rust, and Ward R. Richter.
The Tumor-Producing Effects of Automobile Exhaust Condensate
and of Diesel Exhaust Condensate.
Misfeld, J.
Long-Term Diesel Exhaust Inhalation Studies with Hamsters.
Heinrich, I)., W. Stober, and F. Pott.
Carcinogenicity of Diesel Exhaust as Tested in Strain 'A'
Mice.
Orthoefer, John G., Wellington Moore, Dale Kraemer,
Freda Truman, Walden Crocker, and You Yen Yang.
993
-------
CARCINOGENICITY OF DIESEL EXHAUST PARTICLES BY
INTRATRACHEAL INSTILLATION-DOSE RANGE STUDY1
Alan M. Shefner, Lawrence Dooley, Arsen Fiks,
C. J. Grubbs2, and John H. Rust
Life Sciences Research Division, IIT Research Institute
Chicago, Illinois
Ward R. Richter
Department of Pathology, University of Chicago
Chicago, Illinois
ABSTRACT
A study to determine the carcinogenicity of diesel exhaust
particles, extracts of these particles, and other materials
included for risk assessment purposes has been initiated.
All materials under study will be administered to hamsters
by intratracheal instillation. Prechronic dose range
studies of diesel exhaust particles suspended in gelatine-
saline and of particles bound to hematite (Fe203) have been
carried out. Hamsters were treated for 15 weeks with test
materials and killed 5 weeks later for histopathologic
evaluation. Effects of treatment on weight gain, survival
and induction of pathologic changes have been determined.
The results will be used to select doses for lifetime
carcinogenicity trials.
JThis study was initiated under EPA Grant R806326-01-1 and
continued under Grant R806929-01-0. Dr. Donald E. Gardner,
Inhalation Toxicology Branch, Health Effects Research Lab-
oratory, EPA, Research Triangle Park, serves as Project
Officer.
2Current address - Southern Research Institute,
Birmingham, Alabama.
994
-------
The study we are reporting was carried out as part of EPA's
program effort to assess the health effects of diesel ex-
haust emissions. Our overall program plans include the
evaluation in hamsters of some of the matrix of materials of
interest to EPA; namely, diesel exhaust emission extract,
coke oven extract, roofing tar extract, and cigarette smoke
condensate. However, our specific test system, intra-
tracheal instillation in the hamster, is well adapted for
studies of particulates and affords us the opportunity to
also evaluate the carcinogenic properties of diesel exhaust
emission particles as such.
This report describes results obtained in a dose range study
with diesel exhaust particles (DP). Similar studies on the
four extracts and condensates previously mentioned are in
progress.
The diesel exhaust particles from a 5.7L Oldsmobile engine
were provided to us by the Environmental Sciences Research
Laboratory of EPA at Research Triangle Park. Their method
of generation and collection and their chemical composition
has been described. The preparation of our dosage forms,
analysis of particle size distribution and quantification of
dose levels was carried out at our institution and is des-
cribed in a separate paper (See J. Graf, this volume).
Our bioassay protocol calls for the administration of test
substances by intratracheal instillation in hamsters for
15 weeks, sacrifice of individual animals when they appear
moribund, complete necropsies of each animal, and final
diagnoses by histopathologic examination of tissue sections.
A preliminary dose-range study was carried out in order to
provide an estimate of a maximum tolerated dose to be used
for dose setting in the subsequent lifetime carcinogenicity
assay. Since we had no good estimate of the inherent tox-
icity of the diesel exhaust particles in this model and
anticipated the possibility of delayed or cummulative
toxicity on the weekly dose schedule, we decided to carry
out the dose range test utilizing the same 15-week dose
period and to sacrifice surviving hamsters five weeks
following the last weekly dose.
Initial test groups were large, consisting of 50 male
hamsters per dose. Hamsters were received from Engle
Animal Laboratories at 5 weeks of age and placed on test
at 15 weeks of age. Animals were observed twice daily
during the week and once daily during week-ends. They were
housed three per cage in polycarbonate cages which were
changed twice per week. Hamsters were fed Wayne Lab Blox
995
-------
and AbSorbDri was used as bedding material. All animals
were weighed once per week during the period of the experi-
ment.
The design of the dose-range study is shown in Figure 1.
All groups were treated once weekly by a 0.2 ml intra-
tracheal instillation containing the indicated quantity of
test or control substance with the exception of the 10 mg/wk
diesel particulate and diesel particulate plus ferric oxide
groups which were treated twice weekly.
There were few signs of systemic toxicity in any of the
diesel exhaust particle test groups. Deaths during the
treatment and holding periods were generally 6% or less per
group, and this is typical for this method of administration.
There were no deaths in Groups 1-4 during the 5-week post
treatment holding period. There were only 2 deaths in the
diesel particulate-ferric oxide groups in the same period,
1 in Group 5 and 1 in Group 7. Mortality in the benzo(a)-
pyrene positive controls was higher throughout, totaling
38% in the high dose B(a)P group (Group 9) and 12% in Group
10. Total mortality in the ferric oxide control and vehicle
control groups was 6% and 10%, but no animals in either
group died during the post treatment period. None of the
untreated shelf control animals died.
Weight gains during the period of the experiment were
similarly unaffected by treatment. All groups gained
weight, though Group 1 and Group 9 animals gained slightly
less than animals in other groups. . High dose animals
frequently lost weight during the first 2-3 weeks of treat-
ment and then appeared to recover and gain weight.
Ten animals from each group were processed for complete
histopathologic examination. There was little non-respir-
atory tract pathology observed. A liver lesion of moderate
degree was found in five of the high dose B(a)P hamsters
(perioholangitis) and this may be reflective of the other
signs of toxicity found in this group.
Lungs and tracheas of from 10-25 hamsters per group have
also been examined. A variety of lesions have been found
which is probably not surprising in view of the large doses
administered and of the apparently long retention time of
even a single dose of intratracheally administered diesel
exhaust particles. A gross indication of prolonged reten-
tion was obtained by administering single doses of 5 mg of
diesel particles to a group of hamsters. Individual
animals were killed at 1 hour, 8 days, 30 days, and 60 days
996
-------
o
d
01
j3
C7
0)
!H
PM
-)C
d
o
;s -H
O 4->
H crj
CO SH
W 4-J
p d
i— 1 QJ
rJ 0
w <; d
Pi H O
e> w
i-i ^
PM M
Pi
w
PM
X
PJ
4-1
d
01
e
4-J
CO
QJ
rM
H
fX
j-j
O
^
O
A!
QJ
QJ
£g
•^
X
CM
d
o
•rl
4J
Cfl
1— 1
i-H
•H
4-1
CO
d
M
^^
O
LO
co
QJ
rH
O
•H
4J
i-l
CO
PM
i— 1
0)
co
QJ
•H
P
A! AJ
QJ QJ
QJ QJ
r*z l~*>
— .. -^.
X X
1— 1 !— 1
d d
0 0
•H -r-l
4-1 4-1
CO cfl
i-H i-H
i-H r-l
•H -H
4-J 4-1
CO CO
d d
M M
— ^. ^^
e> o
LO co
co co
QJ 0)
rH rH
O O
•H -H
4-J 4-1
iM iM
Cfl CO
PM PM
i-H i-H
QJ QJ
co co
QJ Q)
•rl -H
P P
A: A;
QJ 0)
QJ QJ
£2 ***
•> — ^ —
X X
I— 1 Csl
d d
0 O
•H -H
4-1 4-1
Cfl CO
rH i— 1
rH 1-H
•H -H
4-1 4J
CO CO
d d
M rH
^. —
O 0
I-H in
CO
Q)
rH
O
•H 0
4-1 O
SH 0
Cfl QJ
p_j rT |
1 — ' ~h
QJ
CO •
QJ PM
•H •
P P
A! Ai
QJ QJ
QJ QJ
kS I!S
\-^
X X
i-H i—l
d d
0 0
•H -rl
4-1 4-1
Cfl Cfl
r-l i— 1
i-H rH
•H -iH
4-1 4-1
CO CO
d d
I-H M
^^^^
O O
U~) CO
1 CO 0
O O
^J. CM C
QJ QJ
PH PH
-4^- -4^
. .
PM PM
• •
P P
A: A:
QJ 0)
0) QJ
>S *"*"
^^^
X X
i — 1 -H
d d
0 0
•rl -H
4-1 4-1
Cfl Cfl
1 — 1 1 — 1
rH 1— 1
•rl -H
4-1 4-1
CO CO
d d
H M
— — ^
O 0
rH CO
O CO C1
0 0
si CN r
QJ QJ
PM PM
-f- -(-
PM
• X~N
PM cfl
. ^ — '
P PQ
A;
QJ
0)
£2
^^,
X
i-H
d
o
•rH
4-1
Cfl
i-H
rH
•iH
4-1
CO
d
H
—
O
s
LO
i— i
O CO
O
^ CM
QJ
_j_
PM
'rt
\^S
«
A:
QJ
QJ
*~y^
^
X
T-l
d
o
•H
4J
cO
rH
rH
•H
4->
CO
d
H
^ —
e>
in
Oi
rH
O
•H
^
QJ
^>
d
•H
CO
o
CM
QJ
A:
QJ •
QJ
|^
^-^
X •
1— 1
CO
^
O
M
H
«^
i ~j
i ]
M
H
CO
ET
M
O
Iz;
.
.
rH
O co
M i-H
4-1 O
d !H
O 4-1
u d
O
QJ C_>
i— 1
o MH
•H i-H
£!> QJ
QJ ft
> C/3
CO
O
CN
QJ
P-.
LH 01
0 >
•rl
rH M)
O
QJ
** M
QJ QJ
d !3
QJ
SH co
^,1 — I
DJ Cfl
/•™\ T-l
Cfl rM
^— - QJ
O 4-J
N Cfl
d e
QJ
43 4-1
CO
- QJ
CO 4-1
QJ
i — 1 d
O QJ
•H 4:
4J 3:
H
Cfl
CL, •
TJ
i— 1 01
QJ }H
co QJ
QJ 4-J
•iH CO
'O -H
d
<4H -rl
0 6
T)
>. Cfl
4-J
•rl QJ
4-) M
d QJ
cfl £
^j
D^ CO
rH
O CO
4-1 -H
rH
CO QJ
rH 4-1
QJ Cfl
LIH ^
QJ
!H QJ
i— 1
d M
o d
•H -H
4-1 CO
cfl
JH >,
co
0)
Jj
£3
4J
X
0
rH
1— 1
CO
CO
T3
QJ
V-i
QJ
4J
CO
•rl
fj
•iH
F;
T3
Cfl
4-1
rj
QJ
d
o
a-
6
o
CJ
42
CJ
cfl
0)
UH
O
4-1
cfl
42
4-1
CO
•rl
d
o
•rl
4-1
ctf
rH
4-1
rj
QJ
O
pj
O
O
QJ
4J
CO
O
.
4-1
4-1 rH CN43
i-H
CM CO
-4" LO
^o r^
CO CTi
O
rH
r-H
1— 1
CM CO
i-H rH
d d
QJ O
O
c d
O 0)
U 42
QJ
PM
43
4J
•H
W)
rl
QJ
u
n
997
-------
post treatment. At 1 hour [Figure 2] the trachea and all
lobes of the lung show the presence of particles and this
can better be seen after trimming [Figure 3]. At 30 days
there appears to be some clearance from the lungs [Figure 4]
and a good deal from the trachea [Figure 5].
Histologic sections of the lungs of hamsters at these time
points show that at one hour the particles are mostly local-
ized with little material in alveolar spaces [Figure 6].
By 30 days [Figure 7] and at 60 days [Figure 8] there are
particles in alveolar spaces and a chronic cellular response
including hyperplasia in areas where particles are present.
In the dose response study where animals received either 15
or 30 intratracheal instillations, a variety of types of
lung pathology was seen in some animals of all dose groups.
A shelf control hamster [Figure 9] can be compared to a
vehicle control hamster [Figure 10]. Treatment with vehicle
produced multifocal areas of mild adenomatous hyperplasia
but no inflammatory reaction. Ferric oxide in vehicle
produced pneumonia in addition to adenomatous hyperplasia
[Figure 11].
Benzo(a)pyrene and ferric oxide showed a more localized
distribution [Figure 12] of the ferric oxide in bronchial
regions as compared with diesel particles. A granulomatous
lesion can also be seen. Tumors were found in some of the
B(a)P-ferric oxide animals upon microscopic examination.
There included a carcinoma in situ in a main bronchus
[Figure 13], a sarcoma in the trachea [Figure 14] which ex-
tended around cartilage, and a squamous cell carcinoma of
the trachea [Figure 15].
Two responses were present in the lungs of animals treated
with diesel particles alone. First, adenomatous hyperplasia
[Figure 16] was present in the region of terminal bronch-
ioles and adjacent alveoli. This was more severe and ex-
tensive than in vehicle controls and was often associated
with metaplasia to ciliated epithelium [Figure 17] and/or
with squamous metaplasia [Figure 18], which was not present
in the vehicle controls. Adenomas [Figure 19] were more
numerous and larger than adenomas in vehicle controls.
The second lesion of DP-treated animals, a multifocal re-
active response of alveolar walls, was not found in vehicle
controls. This consisted of chronic pneumonitis [Figure 20],
cellular hyperplasia [Figure 21], papillomatosis [Figure 22],
macrophage infiltration [Figure 23], increased mitotic
activity [Figure 24], and changes characteristic of atypia
998
-------
[Figure 25] in individual cells. The significance of this
lesion is still under discussion.
It should be kept in mind that this study was conducted for
dose setting purposes preliminary to the conduct of a life-
time study. Conclusions as to the consequences that might
be expected from long term holding of treated animals best
awaits the conduct of the chronic study. However, the
results of this dose-range experiment can be summarized as
follows:
1. Diesel particles and diesel particles plus ferric oxide
at the doses tested did not increase mortality above
their respective control levels.
2. High doses of both preparations decreased weight gain
in the first few weeks of treatment but had no
appreciable effect on weight gain by the end of treat-
ment.
3. There was no significant non-respiratory tract path-
ology in treated animals.
4. Lesions of the lungs were common in treated animals
and included these findings:
• Adenomatous hyperplasia is more severe and
extensive in DP treated animals than in vehicle
control animals.
• Adenomas are more numerous and larger in the DP
treated animals than in the vehicle control.
• Metaplasia to ciliated epithelium and squamous
metaplasia occurred in some DP treated animals
but not in vehicle controls.
• Severe multifocal reactive pneumonitis with hyper-
plasia and evidence of atypia occurred in DP treated
animals but did not occur in vehicle controls.
999
-------
s_
o s-
D.
O)
I/) 10
CD O)
C •!-
3 Q
03 C
O) O
.C -i-
O -P
CO (O
S_ .—
T3 +->
0) 10^-^
4J C •
O >—i
n
|
en
O •
D- l/>
Ol
i. i—
3 O
O T-
i — I fl3
--- D.
(O i —
S- Ol
O)
U -r-
10 4-J
4-> 10
3
O)
1000
-------
00
J-,
O to
ro
rei to
S. r—
-a 4->
a) to,
+J c
O 1-1
O)
to
CO 00 CJ
•<- O •!-
Q Q. -t->
0)
OJ
CO
O^-
D-
00
in OJ
o s.
ro to
• — a.
tD i —
S- 01
CD if
o a,1
CO ••—
•r- O
:=>
cr
o E
•r- LO
O
ea e
s- o
o •'--
_c +J
I— to
U •!-
to -M
01
1001
-------
1002
-------
\f' fv i—"r> H< T ,}, >,^ 1
:'^ ' ^
• >*^
1003
-------
-*
**.&&&/*.$
* * •. ~ - r¥%
Jni5iSi^i|t&!^
S-
O)
1T3
O
s-
c
O
O
(11
O
01
OJ
1004
-------
I
ro
~~0
c
Ol
OQ
OJ
S- +->
T3
CD
fO S.
6 h-
O
c ,
O D.
Ol
s-
CO
O
UJ OJ
u. a_
u
aj ro
c LU
£0,
>> E
a. ro
O -
N
O>
CO
13
en
1005
-------
I
nj
*
t * ' -
f i
** « s
.r ,'
<*
¥
%f ,
*
*,?
> -' <-
V f« *
* i
* ' » „
if V*«
s
<
*
u
Nl
c
-------
1007
-------
V) CT>
OJ E
•r- LO
Q- --
O <1>
S- •(->
t/1
10 -O
r- 0)
rO IS
+-> QJ
0> S-
01 O)
3 r—
O O
O" «5
t/1 O.
CO
D>
1008
-------
1009
-------
-------
ion
-------
THE TUMOR-PRODUCING EFFECTS OF AUTOMOBILE EXHAUST
CONDENSATE AND OF DIESEL EXHAUST CONDENSATE
Mathematical-Statistical Evaluation of the Test Results*
J. Misfeld
Institut fur Mathematik
Universitat Hannover
Welfengarten 1
3000 Hannover, FRG2
INTRODUCTION
This paper deals with the mathematical-statistical eval-
uation of animal experiments devoted to the following
questions:
- To what extent is the carcinogenic or tumorigenic potency
of automobile exhaust condensate (this means gasoline
engine exhaust condensate) and diesel exhaust condensate
dependent on the dose?
- How do the activities of these two condensates differ?
In other words: What is the relative potency of these
two substances?
the support of the Bundesinnenministerium (Ministry
of Internal Affairs) und des Umweltbundesamtes.
cooperation with H. Brune, Biologisches Laboratorium,
c/o Vaselinwerk, Worthdamr.i 23-27, 2000 Hamburg 11, FRG, and
G. Grimmer, Biochemisches Institut fur Umweltcarcinogene,
Sieker Landstr. 19, 2070 Ahrensburg, FRG.
1012
-------
What fractions (groups of substances) of the condensates
are mainly responsible for the carcinogenic potency?
To what extent can the potency of these mixtures be
explained by fractions or single substances (as for
example benzo(a)pyrene)?
In order to answer these and other questions, a working
group^ planned animal experiments (5), prepared AEC, DEC
and subfractions of these condensates. In the first part
of this note I shall discuss some fundamental mathematical-
statistical aspects of analyzing these experiments, and then
I give the results.
MATHEMATICAL-STATISTICAL METHODS OF THE EVALUATION
First of all one has to consider the dose-response-relation-
ship of substances being tested. The response of a tumor-
genie or carcinogenic substance can be measured by frequen-
cies and by the tumor-induction period. So we have to
consider two values, a qualitative one (= is there a tumor:
yes - no) and a quantitative one (= time until a tumor is
developed). These two values are not independent: Higher
tumor-rates are correlated with lower tumor-induction times.
In the following let us consider the parameter "tumor-
rates".
One of the best tested carcinogenic substance is benzo(a)-
pyrene (BaP). As an example for a typical dose-response
curve look at Figure 1, which shows the relation between
the 4 doses of BaP applied to 1200 mice and the tumor-
frequencies (3). In 1860 G. T. Fechner (1) made use of a
law which was first stated by E. H. Weber, in order to
describe the relation between a stimulant and sensation.
This law says that one gets a normal distribution function,
if one transfonns the doses into logarithms. So one gets
the typical sigmoid dose-response curve, being symmetric
about the 50%-point, as shown in Figure 2. This curve
change over to a straight line if one transfonns the fre-
quencies P into probits Y with the help of the well known
formula
.Y-5 -±u2
P = -i- . I e du
.j_. f
V2~7r -•'
[See Figure 3 and (2)].
1013
-------
Animals With
Tunors
100"
50- —
0 0,01 0,02 appi. dose
I nig !
Figure 1. Tumor-Induction in Mice With Benzo(a)pyrene (3).
^ ^Animals With
Tumors (%)
100--
50--
5-10"
Figure 2. Sigmoid Dose-Response Curve for the Tumor Induc-
tion in Mice With Benzo(a)pyrene (3).
1014
-------
70-
50-L
30-
10-
Animals With
Tumors (%)
Appl. Dose [mg]
10
-3
10
-2
Figure 3. Linear Dose-Response Relationship for the Tumor-
Induction in Mice With Benzo(a)pyrene (3).
1015
-------
Next we examine the possibility of giving a numerical
comparison of the arcinogenic potency of different agents
and of deducing the proportion of the potency of a mixture,
which is deducible from the potency of its components.
Assuming that the dose-response lines for carcinogens A and
B are parallel in the transformed plane, we have a constant
quotient dg/d/\, where d/\ = dose of carcinogen A and dg =
dose of carcinogen B, which result in equal number of
animals with tumors. The following definition establishes
a measure of relative carcinogenic potency:
Definition: Let dose d/\ of substance A and dn of substance
B cause the same response. Then
PA/B := dB/dA
is the relative potency of A with respect to B. We say that
A is PA/B~ times as potent as B.
Figure 4 gives an example of substances A and E, where B is
10-times as potent as A.
Our next problem is to predict the potency of a mixture on
the basis of potencies and amounts of its components. There
are various theoretical possibilities (4), the simple case
being that of similar action. In this case, the relative
potency P of the mixture (with respect to any chosen stan-
dard) may be computed by
p = z "vPv
v=l
from the amounts TV and relative potencies pv (with respect
to the same standard of the components v = 1,2, 3,...,n (the
amounts TV are taken as fractions of 1, i.e.,
,
z TTV = 1).
v=l
It is then possible to compute quantitatively the proportion
of the potency of the mixture that is deducible from that of
a component acting in a similar way.
1016
-------
Definition. Let A be a component of mixture M, the weight
percentage of A in M be u, the relative potency of A (with
respect to a standard) be P/\ and the relative potency of M
(with respect to the standard) be p^. Then,
R = IT . P/\/PM
is called the percentage of the potency of M deducible from
A (acting in a similar way).
Animals
With
Tumors
50
B
A
1
10
Dose
Figure 4. Definition of Relative Potency.
B is 10 times as active as A.
1017
-------
In our last example let us assume that A is a mixture with
component B and that the weight-percentage of B in A is 5%.
Application of the definition results in:
R = 10-5 = 50% of the activity of A is deducible from B.
The results of the evaluation, which are based on the
proportions of tumor-bearing animals, may be misleading,
since differences in mortality may occur as a result of,
e.g., different toxicities of substances or of doses.
Animals in experimental groups that are treated with more
toxic substances tend to die earlier and are thus at minor
risk of developing tumors. This fact may cause an under-
estimation of the tumorgenic potency of higher doses and
of substances with higher toxicity.
In order to avoid such bias the analysis should be based on
age-specific death and tumor rates. By considering these
competing risks, the analysis should be able to separate the
tumorgenic process from death due to other causes (5,7).
RESULTS OF THE EXPERIMENT
In order to answer the above questions 995 female CFLP-mice
were dropped twice a week on the skin with AEC, DEC, BaP and
a mixture of 15 carcinogenic PAH's, the weight proportion
of which was chosen as in the AEC. The design of the
experiment is contained in Table 1. We cannot discuss the
methods of planning this design, especially hypotheses
leading to the applied doses [see (4) and (5)]. The results
of the experiment are shown in Table 2. Figure 5 shows that
there are differences in death rates between the groups.
Therefore the raw tumor-rates were transformed into age-
standardized rates. These new rates are shown in Table 3.
The test results are summarized in Figure 6. Differences
in parellelism are not significant, therefore relative
potencies could be calculated (see Table 4).
It is very interesting, that AEC is 42-tinies as active as
DEC. (This means: In order to get the same tumor-frequen-
cies one has to take a 42-times higher dose from DEC than
from AEC.) Table 5 shows, what proportions of the conden-
sates AEC and DEC can be explained by BaP and the 15 selec-
ted PAH's.
We may summarize: The mathematical-statistical evaluation
shows:
1. The tumor-producing effect of AEC is 42-times as high as
that of DEC.
1018
-------
Dead Animals
(in «)
16 24 32 40 48 56 64 72
O Untreated Controls
-• BaP 15.4 pg
-* AEC 2.6 mg
-+ DEC 17.1 mg
96 104 1 12 120 128 136
Figure 5. Development of the Death Rate for Some Test
Groups.
90-
50 -
10-
Animals
with Tu
(*)
O,OO1 O,O1 0,1 1
Figure 6. Age-Standardized Test Results.
1019
-------
TABLE 1. EXPERIMENTAL DESIGN TESTING THE TUMOR-PROMOTING
ACTIVITY OF AEC, DEC AND SUBFRACTIONS
Test Group
No. /Substance
00 Control Without Treatment
01 Control Solvent
02 BaP
03 BaP
04 BaP
11 AEC
12 AEC
13 AEC
21 DEC
22 DEC
24 DEC
31 15 PAH's of AEC*
32 15 RAM's of AEC
Applied
Individual
Dose in mg
.
-
3.85 • ID'3
7.69 • 10-3
15.38 • ID'3
0.29
0.88
2.63
4.30
8.60
17.15
3.5 • ID'3
10.5 • ID'3
Animals in
Test Group
80
80
65
65
65
80
80
80
80
80
80
80
80
AEC: Autombile Exhaust Condensate (Gasoline engine exhaust
condensate)
DEC: Diesel Exhaust Condensate
*15 PAH's of AEC (weight proportion,as in AEC): Benzo(c)-
phenanthren (0.08 pg), Cyclopentenopyren (1.85 pg), Benzo-
(a)anthracen (0.09 pg), Chrysen (0.21 pg), Benzofbjfluoran-
then (0.17 pg), Benzo(k)fluoranthen (0.06 pg), Benzo(j)-
fluoranthen (0.09 pg), Benzo(a)pyren (0.30 pg), 1,12-
Methylen-benzo(e)pyren (0.14 pg), 10,ll-Methylenbenzo(a)-
pyren (0.05 pg), Dibenz(a,j)anthracen (0.10 pg), Indeno-
(l,2,3-cd)pyren (0.21 pg), Dibenz(a,h)anthracen (0.02 pg),
M 300A (0.07 pg), M 300B (0.06 pg).
1020
-------
t/1)
1
=
LU
1—
co
LJ
t—
^£
CY'
CM
LU
CQ
e
o -— i
s- T- -a to
O -P O -V
E 0 -r- 0)
=3 =3 1- GJ
1 — "O OJ 3
C Q_ — -
*— 1
s_ °
00 O
r- E OJ
E ^ "=3
c _c "o
oo
•r- .Q
3 <
•(-> i — +->
Ol TO OO
c: > a> • — •
i > _^
S- en OJ
QJ :3 CZ C1J
cnoo -i— 3
'O ^ *— *
5- 4- Z5
ai o Q
C Q.
c ai
LU <: i—
r— 0)
ro E
a> TJ c:
•i- -r- -r-
CX-i— OJ
ex ~o to
< c o
i— i 0
ai
u
n 03
O +J
S_ to
o .a
•4-J C/0
to -^
0) •
1— O
^
O CD ^J- f-H [^-
^^ *-O t:^"
o o co a*, r-i
O CD CM O CTv
CO i-D CO
o o i — i cr> r—
CM CO LT>
LO tO CO CO LO
co r-^ r^ to LO
co co •* «sr «*•
r^. r^. to to to
CO CO CO
1 1 1
o o o
ii ...
U-) CTl CO
co to co
...
co r^ UD
1 — 1
c
OJ
E
rC
a>
£_
+J 4->
3 C
o ai
^: >
-t-> i —
•r- O
3 00
, ,
0 0
S_ S-
4_> 4_J
c c o. a. Q-
O O t3 rt3 O3
C_3 O CO CD m
O r— I CM CO «3"
o o o o o
CM OJ T\J O CM 'sO rH CO
r~-» r^ LO r~^ r^. en r*-*
^H
co co co otoi — cor-^
o^i-co ocMrvi t—ioo
.— I •=!- CO *—< CO
COLDLO OCMCTv ^-lOl
CO to CM
OCO'sJ' O' — ir^ OOtO
ooi^-to cocor — r^-(^
cocnco Lntoi— c r~-Ln
i-^r~-r-~. i^(~^i~^ r~.r^
co co
i i
0 0
en co co o o LT>
CNJCOtO CO to •— 1 ••
OOCM *d-cor^ LOLO
i — i . .
CO O
1 — 1
0 0
LU LU
4- M-
O O
j/i yi
•j- -y*
^ ^
0. o o o
UJLULU LULULU LOLO
<:
-------
oo
h-
_J
ID
oo
o:
h-
oo
LU
H-
Q
LU
M
(_— (
O
QC
o -o c
£- t- GJ
3 ra 3
r— "a CT
sr cu
jC ro S_
+->+-> LU
•r- oo
co a>
rO U
e s c
•i— ra QJ
C OS 13
< o-
QJ
1_
LU
c a.
O) •!- 13
i— O
JD CO >-
ra i — CD
3 ra
.— E +J
ra -i— CO
> C QJ
LU ct | —
i — ai
ra E
ai -o c
•1- -1- -1—
Q-T- QJ
Q_~o CO
OO
CO \
QJ •
(— O
O O *^" CM CO LO CO '-O
OO COOLO OO^-OO
CM r^. co oo en
o o co ci ^n oo oo oo
oo CMocn o«*oo
OO to CO t-H ^f CO
coco ^i-^-^- cocnco
r^^ ^^ to to ^D r~^ r^ r^^
00 OO 00
1 1 1
o o o
01 CO 00
II • • • CM CO <£>
LO Cn CO O O CM
CO ^O OO
• • •
ro r-» LO
i — i
^
c
QJ
E
ra
QJ
i-
h-
4-> -I-J
13 C
O OJ
J^ >
•i- O
3 OO
i — , —
O O
s_ s_
4J 4->
CC CLCL.CL CJCJO
OO rarora LULULU
OC_> CQCHCQ -cC^C^
O^H CMOOs^- ,— ICMOO
OO OOO i — It— Ir-H
o cn LO I-H o
O i — t , — 1 t-H OO
T 1 OO
O UD 1 — . OO 1 —
O CM CM <—< CO
r— I OO
LO ^O t — 1 1 — LO
r^ r~^ [ — ^ r~^- I — ^
00 OO
i i
o o
O CD LO
O") t^3 i — 1 • »
M- CO C^ LO LO
i — t • •
OO O
I — 1
O (_>
LU LU
cC e£
M — H—
0 0
CO CO
_ *
T" "1~
^ s^
a. a.
CJ O O
LU LU LU LO LO
Q Q O >~~( < — '
.— 1 OvJ OO ^H CM
CM CM CM OO OO
1022
-------
TABLE 4. RELATIVE POTENCIES OF DEC, SUBFRACTIONS AND BAP
RELATED TO AEC AND OF BAP RELATED TO DEC
Substance Relative Potency 95% Confidence Limits
AEC 1
15 PAH's of AEC
BaP
DEC
DEC
BaP
68
187
0.024
1
7950
46
127
0.015
5210
- 101
- 286
- 0.035
-
- 13300
TABLE 5. PROPORTION OF THE POTENCY OF AEC AND DEC, WHICH
CAN BE EXPLAINED BY SELECTED PAH'S.
Explained Proportion95% Confidence
Substance of the Potency [%J Limits
AEC BaP 9.6 6.5 - 14.7
15 PAH's of AEC 41.2 27.5 - 60.8
Total-PAh-Fraction (6) 100 80.0 - 100
DEC BaP 16.7 11.U - 28.0
1023
-------
2. A synthetic mixture of 15 RAM's, known to be carcino-
genic, shows 41% of the activity of the whole condensate
AEC.
3. BaP can explain 9,6% of the activity of AEC and 16,7% of
the activity of DEC.
Let me conclude with the remark, that further research is
needed in order to show that these results are valid for
other models, especially for inhalation studies.
REFERENCES
1. Fechner, G. T. Elemente der Psychophysik. Breitkopf
und Hartel, Leipzig, 1860.
2. Finney, D. J. Probit Analysis. Cambridge University
Press, 3rd ed., 1972.
3. Lee, P. N., and J. A. O'Neill. The effect of time and
dose applied on tumour incidence rate in benzopyrene
skin painting experiments. Brit. J. Cancer, 25:759-770,
1971.
4. Misfeld, J., and J. Tinim. Mathematische Aspekte der
Planung und Auswertung von Experimenten zur Krebsfor-
schung. In: Methodik der Untorsuchung von Rauchkonden-
saten im Tierversuch, Wissenschaftliche Forschungsstelle
in VdC, Hamburg, 1969. pp. 85-119.
5. Misfeld, J. and J. Timm. Mathematical planning and
evaluation of experiments. In: Air pollution and
cancer in man (IARC Scientific Publications No. 16),
LYon, 1977. pp. 11-28.
6. Misfeld, J. and J. Timm. The tumor-producing effect of
automobile exhaust condensate and fractions thereof,
Part III: Mathematical-statistical evaluation of the
test results. J. of Env. Path, and Toxic., 1:747-772,
1972.
7. Misfeld, J. and K. H. Ueber. Tierexerpeirnente mit
Tabakrauchkondensaten und ihre statistische Beurteilung.
Planta Medica, 22:281-292, 1972.
8. Misfeld, J. Matheniatisch-statistische Aspekte der
Wirkungs-Bilanz-Analyse. In: VDI-Bericht Nr. 358
"Luftverunreinigung durch polycyclische aromatische
Kohlenwasserstoffe - Erfassung und Bewertung", Essen,
1980, to appear.
1024
-------
General Discussion
P. LEE: What is the biological basis for your fre-
quency transformation? Is that a form of gamma function?
J. MISFELD: Yes, it is the interference of the gamma
function.
P. LEE: What is the biological basis for this fre-
quency transformation?
J. MISFELD: I don't know, the data used for this
transformation. This calculation is valid when we have a
single transformation in other experiments and not only
in biological experiments.
SPEAKER: We have heard that both gasoline and diesel
exhaust contain compounds which are not the conventional
PAH's in that they do not require, for example, micro-
somal activation to express mutagenicity. Have you any
data that the samples which you painted on the skin still
contain such compounds or have they been lost in the
extraction.
J. MISFELD: No, there may be other substances but we
don't believe that there are other toxics in these ex-
tracts.
SPEAKER: Do you suggest that gasoline exhaust is more
carcinogenic in this assay than diesel exhaust? Is this
true also if you express it per mileage?
J. MISFELD: Yes.
J. HAFELE: You may have described the engines used,
but I don't recall it. With respect to the gasoline
engine was unleaded fuel used?
J. MISFELD: The engine was a Mercedes. The fuel in
the gasoline engine was leaded.
J. HALEY: I understand that the latency period for
tumors is a very strong function of the dose and then in
the low dose ranges the latency period can exceed the
life of the animal. Would you comment?
J. MISFELD: Yes, but if you take the two month selec-
tion period and conduct other inhalation exposures you
get nearly the same results.
D. HOFFMANN: It appears that in mouse skin both the
carcinogenic and tumorgenic response can be accounted in
this study by the PAH or similar polar components. Based
on your diesel data, do you think that what you are see-
ing in mouse skin is primarily due to the PAH's?
J. MISFELD: I an mot a biologist.
R. SCHRECK: Could you give us some description of the
collection technique, particularly for the gasoline. Are
you filter-collecting this, or is this strictly a conden-
sate?
J. MISFELD: This was described earlier.
1025
-------
LONG TERM DIESEL EXHAUST INHALATION
STUDIES WITH HAMSTERS
U. Heinrich, W. Stober
Fraunhofer Institute of
Toxicology and Aerosol Research
Nottulner Landweg 1o2
44oo Munster
Federal Republic of Germany
F. Pott
Medical Institute of
Environmental Hygiene at the
University of Dusseldorf
Gurlittstr. 53
4ooo Dusseldorf
Federal Republic of Germany
ABSTRACT
Preceding a life-time inhalation exposure test,
two range finding inhalation studies over three
weeks and five months with diesel exhaust in
varying concentrations were conducted to deter-
mine its acute and subacute toxicity. The ani-
mals (Syrian golden hamsters) received total
exhaust and exhaust without particles 7-8 h/d
and 5d/w. In order to evaluate the toxic effects,
measurements and observations were made including
some clinico-chemical and haematological para-
meters, body weight, organ weights and the
histopathology of lung, trachea, liver and kidney.
The evaluations indicated effects due to ex-
posure, particularly evident in the lung. The
range finding studies were carried out to deter-
mine an exhaust concentration which, in long-term
exposures, would not lead to a significant shorte-
ning of the normal life expectancy of the hamsters,
1026
-------
Since then, a life-time experiment was started at
an exhaust concentration of 1/7 (total exhaust
and exhaust without particles) and has been going
on for about 1o months. In addition to the ex-
posure to diluted exhaust, some exposure groups
receive in addition certain doses of a well-known
carcinogen (diethylnitrosamine or dibenz(a,h)-
anthracene), which, is expected to induce a
tumor rate of 15 to 2o % in the respiratory tract.
This animal model increases the probability of
detecting slight carcinogenic or cocarcinogenic
effects of inhaled diesel exhaust. Besides regular
body weight checks, 22 different clinico-chemical
parameters were measured after 8 months into the
life-time experiment.
This paper deals with the first results of a small-
scale inhalation study with Syrian golden hamsters
on the biological effects of diesel exhaust
emissions.
Our research project was initiated in 1977 as a
pilot study at what is now the Fraunhofer
Institute of Toxicology and Aerosol Research in
Miinster, West Germany. The study is part of a
programm of a Task Group on the Airborne Carcino-
genic Burden on Man, which was founded in 197O
and is now funded by the West German Federal
Environmental Agency. To a small extent our
particular research project was also supported by
the Daimler-Benz Company.
The basic objective of our experimental study is
to find an answer to the question as to whether
and to what extent diesel exhaust is able to pro-
mote the development of lung cancer by way of
inhalation. But other long-term biological
effects are considered as well. Although the
exhaust condensates of diesel and gasoline engines
have already been proven to be carcinogenic in
various non-inhalative animal models, no con-
clusive results are available on the inhalation
hazard. Neither epidemiological studies nor
animal inhalation exposure data exist up till now
which would permit a definite statement on the
carcinogenic potential of chronic inhalation
exposures to man or animal.
1027
-------
Some results of in-vitro experiments using diesel
exhaust condensate were published by EPA in late
1977 (EPA, Environmental News, Nov. 12, 1977).
These data have given weight to the suspicion that
the mutagenic effects seen in the Ames test are
caused chiefly by the soot particles contained in
the exhaust. Since then, our long-term inhalation,
program having been designed at that time,
focusses on the special role the soot particles
of diesel exhaust may play in producing the
effects under investigation.
Therefore, our pilot study employs diesel exhaust
in two ways side by side. One line of experi-
ments makes use of the original diesel exhaust
diluted to suitable concentration levels, while
the other line utilizes diesel exhaust void of any
particulate matter.
The actual experimental exposures started in early
1978 and are made up of a successive series of
three experiments. Each of these experiments
determines the experimental conditions of the sub-
sequent one. The first experiment was a prelimi-
nary investigation and lasted only 3 weeks. It
simply served to provide a rough estimate of the
acute toxicity of the total exhaust. Furthermore,
it suited as a trial test run of the whole experi-
mental set-up.
A variety of exhaust dilutions was used at this
stage. Given as volume ratios of exhaust to
clean air, mixtures of 1/2, 1/3, 1/5, 1/10, 1/15,
and 1/20 were passed through the exposure chambers.
The 1/2 dilution made it impossible to maintain
representative temperature and humidity conditions
in the chambers, so that this exhaust concen-
tration was eliminated from subsequent investi-
gations. In these preliminary exposures, only
body weight checks and the general condition of
the animals were taken as parameters to determine
a toxic effect, and on that basis, no significant
differences between exposed and control animals
could be established.
In the subsequent investigation, which ran for
over 5 months, an exhaust dilution was to be
found that would give the highest concentration
possible without impairing significantly the
life expectancy of the animals during the third
experiment which is currently under way as the
final life-time study. Life span shortening
1028
-------
could be caused by the effects of NO, N02, SO2,
CO and soot particles, all of which were
administered in concentrations known to be toxic.
The dilution ratios used here were 1/3, 1/5, and
1/10. The evaluation of the toxic effects from
these exhaust dilutions was based on regular body
weight checks, and on the results of some clinico-
chemical and haematological analyses. Furthermore,
certain organ weights were checked and some impor-
tant organs were inspected by histopathological
examinations.
In order to obtain a carcinogenic effect in inhala-
tion experiments, it is necessary to exploit the
relatively short life span of the rodent to the
full. In view of the generally long induction
period of carcinogens in humans, animal experi-
ments must be conducted at high exposure concen-
trations just below the level of acute toxicity
in order to compensate at least to some extent for
the short life expectancy in comparison to man.
Since the concentration of known carcinogens in
diesel exhaust is relatively low, only weak dilu-
tions of the exhaust emissions may have a chance
to produce experimental lung cancer by way of
inhalation. For this very reason, an additional
approach was designed for our inhalation experi-
ments that may give us a better probability of
finding at least a cocarcinogenic effect, if there
is any.
It may well be that the effect of straightforward
exhaust inhalation will remain within the flat
range of the general sigmoidal dose-response
relationship for tumor incidences. Even after
exposure to the maximum permissible long-term
concentration, the increase of the tumor inci-
dence rate may still be insignificant for
suitable sizes of animal test groups. However,
by administering known carcinogens in addition,
the basic tumor incidence rate may be shifted
and moved into the steep slope of the dose-
response relationship. By pre-inducing tumor
rates of some 15 - 20 %, the additional exhaust
exposure may then create significant changes in
test groups of feasible size. As shown schema-
tically on the graph (Fig.1), these changes need
not be necessarily coergistic. One of us (F. Pott)
1029
-------
TUMOR
RATE
1 SPONTANEOUS TUMOR RATE
2 TUMOR RATE AFTER DIESEL
EXHAUST INHALATION
3 TUMOR RATE AFTER OBIaflJA
OR DEN APPLICATION
4 POSSIBLE SYNERCISTIC OR INHIBITORS
EFFECT OF DB(a,h)A/DEN * DIESEL EXHAUST
DOSE
Figure 1.
has shown in subcutaneous experiments with car
exhaust condensate and benzo(a)pyrene that the
exhaust condensate had a slight inhibitory
effect on the known carcinogen.
Two different carcinogens are used to induce the
15-2O % tumor rate in our inhalation studies (Fig.15).
Some additional groups receive the polycyclic
aromatic hydrocarbon dibenz(a,h)anthracene. It is
administered intratracheally once a week for 2O
weeks. Other exposure groups receive a single sub-
cutaneous injection of diethylnitrosamine (DENA). In
contrast to the local effect of dibenz(a,h)anthra-
cene, DENA has a systemic effect. In our parti-
cular case of Syrian golden hamsters, the effect
is chiefly organ-specific and involves the res-
piratory tract. The application of DENA avoids
minor lesions to the trachea, which cannot be
completely ruled out in spite of the special intra-
tracheal method used here.
The two carcinogens were actually administered to
the exposed animals in the third experiment, the
life-time exposure currently under way. In the
5-month experiment, half of the exposed animals
received an intratracheal instillation of the non-
carcinogenic PAH pyrene once a week for 20 weeks.
The purpose of this procedure was to determine
whether the application method as such causes a
different reaction under diesel exhaust exposure
1030
-------
in comparison with animals without instillations.
The 5-month experiment was completed by the end of
1978 and the subsequent life-time experiment has
now been in operation for ca. 10 months. Through-
out the project, Syrian golden hamsters are used
as the test animals. This animal model is
especially suited for the testing of the carcino-
genic effects on the respiratory tract for the
following reasons:
1. The spontaneous lung tumour rate is
very low;
2. Infections in the respiratory tract are
infrequent; and
3. The morphology of induced neoplasms
seems to be very similar to tumors
in the human respiratory tract.
A disadvantage of this animal model should not
be ignored here: the incidence of kidney amyloi-
dosis at the age of 1.5 to 2 years is rather
common. This causes death in a. percentage of the
animals in this age group.
The exhaust used in the exposures is produced by a
2400 ccm Daimler Benz diesel engine. The engine,
which is linked to an electrical dynanometer brake
of the eddy current type, is mounted permanently
on a spring-supported foundation. Its output and
rotation are held constant at 16 kilowatts and
2400 r.p.m. The diesel fuel used is a European
Reference Fuel and contains 0.36 % sulphur. The
exhaust required for the exposures is drawn from
a normal exhaust pipe into the mixing chambers.
Part of it goes there directly, but another part
passes first through a centrifuge, where the parti-
culate matter is removed. Before the exhaust reaches
the mixing chambers, its temperature is maintained
above dew point, thus avoiding any condensation
build-up and subsequent loss of certain substances
in the exhaust. In the mixing chambers, the total
exhaust and the exhaust without particles is diluted
with cooled and dried clean air to the concentration
required for the exposures. When the exhaust cools
off in the mixing chambers, the dried diluting air
takes care of the surplus moisture, so that even at
this stage the dew point is not exceeded and water
condensation is avoided. For safety reasons, the
exhaust, which has passed through the centrifuge
and has been diluted in the mixing box, will pass
in addition through a fine filter. Thus, in case
1031
-------
of an unforeseen breakdown of the centrifuge, the
exhaust particles will then be removed by the
filter, so that the exclusive exposure to the gas
phase of the exhaust is maintained. The filter
does not significantly influence the concentra-
tions of CO, CO2, SO-, NO, NO and total hydro-
carbons. Measurements did not differ much from
each other before and after the filter. Using
supplementary mixing boxes, placed directly in
front of the inhalation chambers the exhaust-air
mixture can be further diluted to suit the expo-
sure conditions.
A uniform horizontal flow is passed through the
inhalation chambers. Special baffles and perfo-
rated diaphragms at the entrance and the exit
openings guarantee an even distribution of the
aerosol inside the chambers. In each chamber,
there are 24 animals housed in wire cages on each
of two levels. In each cage compartment there are
three animals. The exhaust flow passes through
the chamber at a rate of 80 litres per minute.
Under this condition, the chamber atmosphere will
be renewed about 20 times per hour. The chamber
temperature is kept at 24-25 C and the relative
humidity at 5O-6O %. The animals are exposed
7-8 h/d and 5 d/w. During the exposure period,
there is no food in the cages; this is to keep the
oral intake of diesel soot to a minimum.
As already mentioned, three exhaust dilutions were
used in the 5-month inhalation study - the ratios
were 1/3, 1/5, and 1/10 (Fig.2). One half of each
dose group received the total exhaust, and the other
half was exposed to the exhaust without particles.
Of the 48 animals per dose group, 24 received an
intratracheal instillation of 0.5 mg pyrene in
0.15 ml physiological saline solution. This
mixture was administered once a week for 20 weeks.
In order to obtain a suspension as homogeneous
and finely distributed as possible, some Tween
20 - an emulsifier - was added to the saline solu-
tion. After adding the pyrene, the mixing was
effected ultrasonically. The same method of
emulsification is used in the current life-time
exposures for making the dibenz(a,h)anthracene
suspensions. An appropriate number of controls,
some receiving instillations and some not, were
kept in inhalation chambers ventilated by clean
air.
The concentrations of the following gaseous com-
1032
-------
5 MONTH INHALATION STUDY WITH DIESEL EXHAUST
EXPERIMENTAL GROUPS
HAMSTER + PYRENE
INSTILLATION (iTfi)
HAMSTERS
TOTAL EXHAUST
1/3
21
21
1/5
21
21
1/10
21
21
EXHAUST WITHOUT
PART
1/3
21
21
CLES
1/5
21
21
1/10
21
21
CLEAN AIR
21
21
Figure 2.
ponents in the exhaust were measured continuously
in the inhalation chambers using a compact com-
mercial measuring system: CO, CO2, NO, NO-, NO ,
total hydrocarbons, methane, non-methane Hydro-
carbons and SO-. In the following figures, the
concentration of these gaseous components is shown
separately for total exhaust and exhaust without
particles. However, the differences between the
two exposure atmospheres are very small. The con-
centrations in the highest and lowest dose group
were as follows:
and 11 ppm for CO
40-45 ppm
1.5 volume %
7O ppm
3 ppm
1 3 ppm
and 0.5 volume % for CO2 (Fig.3)
and 2O ppm for NO
and 0.5 ppm for NO2 (Fig. 4)
and 7 ppm for total hydrocarbons
Furthermore, about 5 ppm were measured in all
inhalation chambers for methane. Since the
methane portion in the exhaust itself is under
1 ppm, the methane concentration in the chambers
is mainly generated by the animals (Fig. 5). The
SO,, concentration was ca. 13 ppm in the highest
dose group and 2-3 ppm in the lowest (Fig.6). The
gravimetrically determined particle mass concen-
tration in the inhalation chambers was about 17 mg.,
per m for 1/3 dilution, approximately 11 mg per m
1033
-------
MEASUREMENT OF CO AND COg IN THE CHAMBERS
pprr
50 4
m
EXHAUST / CLEAN AIR
1 / 3 JS»
M
8
n
ii
In
I
I
1
1
1
1
1
m
1
I
I
-
1
i
i
VOL% 1 ' 5
-1,5 1 / 10 \\N
CLEAN AIR I 1
TOTAL EXHAUST - 1
EXHAUST WITHOUT PAR-
TICLES - 2
-10
•0,5
Figure 3.
MEASUREMENT OF NO,NOX AND N02 IN THE CHAMBERS
EXHAUST / CLEAN AIR
1 / 3 8888
1 / 5
1 / 10 '//.
CLEAN AIR CD
TOTAL EXHAUST-1
EXHAUST WITHOUT PARTICLES - 2
1 NO
1 NOa
Figure 4.
1034
-------
MEASUREMENT OF CnHm;CHt AND CnHm-CHj IN THE CHAMBERS
ppm
15 •
EXHAUST / CLEAN AIR
1 / 3 S8SS
1 / 5
1 / 10 ,\V
CLEAN AIR I 1
TOTAL EXHAUST - 1
EXHAUST WITHOUT PARTICLES-
CnHm-CHa
1 2
Figure 5.
for the 1/5 dilution and about 4 mg per m for
the 1/1O dilution (Fig.6). The particle size
distribution of soot particles in exhaust was
MEASUREMENT OF S02 AND PARTICLES
soa
EXHAUST / CLEAN AIR
1 / 3 VfSf
1 / 5
1 / 10 XV-
CLEAN AIR CTl
TOTAL EXHAUST « 1
EXHAUST WITHOUT PARTICLES • 2
Figure 6.
1035
-------
determined just before the entrance to the
inhalation chambers using an aerosol spiral
duct centrifuge. The modal peak of the particle
mass distribution had an aerodynamic diameter
of 0.1 urn (Fig.7).
mg
um-mo ,
30 -
MASS DISTRIBUTION OF DIESEL EXHAUST PARTICLES
1,0
Dae (jim)
Figure 7.
After regular body weight checks every 2 weeks,
1O animals from each group were examined at the
end of the 5-month study for the following para-
meters : erythrocyte and leucocyte counts, mean
corpuscular volume of erythrocytes, total haemo-
globin, CO-haemoglobin and metnemoglobin, haemato-
crit, some enzymes in the plasma, such as two
transaminases, GOT and GPT, alkaline phosphatase,
serum urea, and finally, weights of lung, heart,
liver, kidney and spleen. Lung and trachea as
well as liver and kidney were also examined histo-
pathologically.
In the following figures, the results of these
examinations are separated into those groups which
received intratracheal instillations and those
groups which did not. The caption "itr" under-
neath the column diagram denotes those exposure
groups which received instillations with pyrene;
the number 1 represents those groups which were
1036
-------
exposed to the total exhaust, and the number 2
represents those groups, which only inhaled the
gaseous phase of the exhaust.
In comparison to the controls, the number of
erythrocytes is somewhat decreased in all expo-
sure groups (Fig. 8). However, this difference is
significant only for those groups, which received
NUMBER OF RED BLOOD CELLS
fjf
EXHAUST/CLEAN AIR
1 / 3 X®
1 / 5
1 / 10 \\\
CLEAN Al RE3
TOTAL EXHAUST = 1
EXHAUST WITHOUT PARTICLES =
Figure 8.
the gaseous phase of the exhaust without the
particles. In contrast, the mean corpuscular
volume of the erythrocytes, when compared to the
controls, is significantly higher for almost all
exposed groups (Fig. 9). The haematocrit value
for exposed animals, which should be low due to
the small number of erythrocytes, shows this sig-
nificantly only in the highest dose group (Fig.
10). The absence of this effect in the other dose
groups is probably due to the fact that the re-
duced number of erythrocytes is compensated by
the significantly higher mean corpuscular volumes.
Since the haematocrit reproduces the cellular vol-
ume proportion of the blood, the influence the
smaller number of cells should have, must have
been neutralized by the larger volume of the cells.
The haematocrit value of one exposure group could
not be measured, because the measuring instru-
ment had broken down and there was no blood
1037
-------
MEAN CORPUSCULAR VOLUME
i-h
i
EXHAUST/CLEAN AIR
1 / 3 88
1 / 5
1 / 1D \V
CLEAN AIR I—I
TOTAL EXHAUST. 1
EXHAUST WITHOUT PARTICLES-2
Figure 9.
Vol%
60-
HEMATOCRIT
Tl
EXHAUST / CLEAN AIR
1 / 3 >X«H
1 / 5
1 / 10 \\N
CLEAN AIR C3
TOTAL EXHAUST - 1
EXHAUST WITHOUT PAR-
TICLES - 2
Figure 1o.
available for a repeat.
No significant difference between any exposure
group and controls could be determined for the
1038
-------
haemoglobin concentrations.
The methemoglobin and CO-haemoglobin values of the
exposed animals appear to be higher than those of
the controls; in the highest dose group they are
considerably higher (Fig.11). However, all of these
MEASUREMENT OF METHEMOGLOBIN AND CARBOXYHEMOGLOBIN
EXHAUST / CLEAN AIR
1 / 3 £**
1 / 5
1 / 10 \\x
CLEAN AIR 1 I
TOTAL EXHAUST- 1
EXHAUST WITHOUT PARTICLES-2
ii
Figure 11.
values still lie in a range where there is no ser-
ious risk to the transport of oxygen. The rela-
tively large methemoglobin formation in each of
the highest dose groups reflects the high NO-con-
centration in these exposures, as NO must be seen
as an originating factor in the formation of
methemoglobin. CO-haemoglobin could only be de-
tected in the highest dose groups, and only then
when the measurements were made immediately after
the exposure. This is because CO-haemoglobin
decomposes very rapidly.
In contrast, the increased methemoglobin values
could be measured even after a weekend when no
exposure took place.
The leucocyte count was apparently slightly lower
in all exposure groups, in some cases even signi-
ficantly, but no dose-dependent effect could be
observed. Instead, the values for the different
exposure groups varied randomly.
1039
-------
The results of the enzyme measurements, GOT, GPT,
and alkaline phosphatase, show no clear indication
of organ lesions. However, the blood urea content
points to a kidney insufficiency. Animals, which
were exposed to total exhaust, showed a significant
increase in blood urea concentration (Fig.12). The
histopathological examination of the kidneys could
UREA' IN PLASMA
mg/dl
EXHAUST / CLEAN AIR
1 / 3 tm
1 / 5
1 / 10 X\\
CLEAN AIR I I
TOTAL EXHAUST - 1
EXHAUST WITHOUT PAR-
TICLES - 2
Figure 12.
not, however, definitely confirm this insufficiency.
Even the histological examination of the liver, the
weights of heart, kidney, liver and spleen, and
body weight checks indicated no difference to the
controls.
An effect, which was clearly caused by the expo-
sure to the diesel particles was found in the lung,
the primary target organ of inhaled exhaust. A
dose dependency could be observed even in a simple
macroscopic examination of the lungs. They dif-
fered by their grey or black colouring, in relation
to the increasing concentration of the soot par-
ticles. Even the lymph nodes in the lung, which
were prepared for microscopic inspection, were
clearly colored black. This findings indicate
that the removal of the inhaled soot particles
occurs to a noticeable extent by transport the
lymphatic system. It appears possible that soot
particles or various substances, which may desorb
1040
-------
from the particles, may cause a burden on other
organs by further lymphatic transport.
In the 5-month experiment, the wet and dry weight
of the lungs was determined from 10 animals per
exposure group. In all 3 dose groups, the lung
wet weight expressed as a percentage of the body
weight of animals exposed to total exhaust, was
significantly increased in the two highest dose
groups. This was also the case for those animals,
%bw WET WEIGHT OF THE LUNGS
1.5-
ri
EXHAUST/CLEAN AIR
1 / 3 K*
1 / 5
1 / 10 \\N
CLEAN AIB C3
TOTAL EXHAUST.1
EXHAUST WITHOUT PARTICLES. 2
Figure 13.
which in addition received intratracheal instil-
lations (Fig. 13) .
On the other hand, those animals, which only in-
haled the gaseous phase of the exhaust, had lung
wet weights almost identical to the controls.
Furthermore, no difference could be detected
between the animals with and without instillations,
The same dose-dependent effect can be seen in the
lung dry weight of those animals exposed to the
total exhaust. Again, all animals, which inhaled
exhaust without particles, showed no significant
difference to the controls (Fig.14).
Since the increase in the wet weight and the dry
weight of the lungs was of the same rate, and
1041
-------
DRY WEIGHT OF THE LUNGS
EXHAUST/CLEAN AIR
1 / 3 »»
1 / 5
1 / 10 \\V
CLEAN AIR I—]
TOTAL EXHAUST-1
EXHAUST WITHOUT PARTICLES«2
Figure 14.
since the water content of the lungs of these
animals did not differ from that of the controls,
it can be assumed that the increase in lung weight
cannot significantly come from an oedema formation,
but must be chiefly due to a tissue proliferation.
This can be concluded, because the soot particles
inhaled and deposited in the lung would account
only to a small extent for the weight differences
actually observed. Assuming a breathing volume
of 1OO ml per minute for the,hamster and a particle
concentration of 17 mg per m in the highest dose
group, then, after 5 months, the inhaled particle
mass per animal would be about 7O mg without taking
into account the very effective particle removal
by the nose of the hamster. Assuming further a
lung deposition rate of 50 %, which in reality is
an excessive rate, then the particle mass deposited
in the lung would only account for a maximum of
5-10 % of the increase of the absolute wet weight,
and a maximum of 20-30 % in case of the dry weight.
In view of the histopathological examinations, the
main cause of the increase in lung weight must be
the inflamed, granulomatous, and proliferative
alterations of the lungs of the exposed animals.
However, only by means of a morphometric analysis
of the lung tissue is it possible to make a dose-
1042
-------
dependent quantification of the histopathological
findings. Since this elaborate examination has
not yet been carried out, only a general statement
can be made at this time: the lesions of the lung
tissue are much more pronounced in the highest
dose groups than in the lowest. These lung al-
terations were not found in the animals, which
inhaled only the gaseous phase of the exhaust,
and, of course, not in the controls.
At the end of the 5-month experiment, a few
animals from each group were saved. For another
9-1O months, they were kept under normal clean
air conditions. After this period, the macro-
scopical and histological examination carried out
on the animals which were exposed to total exhaust
showed distinct particle depositions in the lung,
including the animals exposed to the lowest dose.
In particular, the histological picture of the
lungs in the lowest dose group is partly marked
by a strong proliferation of the terminal bro-
chiolar epithelium into the alveolar spaces. It
was not possible to detect this effect in the
examinations directly after the exposure period
of 5 months. It must be pointed out, though, that
there were very few animals available for this
late examination, and that it is not possible at
this time to furnish proof of a causal relation-
ship between the particle deposits and the adeno-
matous proliferation.
The results of this 5-month inhalation study
constituted the basis for the design of the life-
time exposure test. In that investigation, there
is a total of 18 exposure groups, with 48 animals
each (Fig.15). The study is now into its tenth month.
Six of these groups receive total exhaust, 6 other
groups receive exhaust without particles. The
remaining 6 groups serve as controls and receive
clean air. The exhaust is given to all groups at
the same dilution ratio of 1/7, exhaust in clean
air. The details regarding the engine, inhalation
chambers and exposure periods are identical to
those employed in the 5-month experiment. After
an exposure period of 2 months, two groups of each
exposure condition, that is a total of 6 groups,
received a single injection of diethylnitrosamine
in two different doses. Another 6 groups received
intratracheal instillations of a suspension of 2
different doses of dibenz(a,h)anthracene once a
week for 20 weeks. Thus, tumor induction rates
will be produced at two different dose levels. As
1043
-------
LIFETIME INHALATION STUDY H1TH DIESEL EXHAUST
ITR, INST. / S.C,
WITHOUT
DB(a,h)A 6 MG
(20 x 0.3 MG in 0.02 ML)
DB(a,h)A 2 MG
(20 x 0. 1 MG IN 0. 15 ML)
PYRENE 2 MG
(20 x 0.1 MG IN 0.15 ML)
DEN 1. 5 MG/KG By
DEN 4. 5 MG/KG B«
CLEAN AIR
48
48
48
48
48
48
TOTAL EXHAUST
48
48
48
48
48
48
EXHAUST
WITHOUT
PARTICLES
48
48
48
48
48
48
DB(a,h)A = DIBENZ(a,h)ANTHRACEN
DEN = DIETHVLNITROSAMINE
Figure 15.
mentioned earlier, with animals treated in this
way, there is an increased probability of detecting
a carcinogenic or cocarcinogenic effect of inhaled
diesel exhaust with statistical significance. In
order to exclude any effects caused simply by the
method of intratracheal instillation, one group of
each exposure condition was instilled with the
non-carcinogenic PAH, pyrene. There is also one
group in each of the exposure arrangements which
does not receive supplementary treatment.
The following table lists the concentrations of
some exhaust components. They are measured in the
inhalation chambers and at a distance of a few
metres in the ducts feeding the exhaust into the
chambers; the values are subdivided into total
exhaust and exhaust without particles (Fig.16).
In comparing the measurement values in front of
and inside the chambers, it can be seen that the
S02 concentration decreases on the way to the
chambers; this is probably due to adsorption
effects. On the other hand, the NO2 concentration
increases along the flow, which is due to the
oxidation of NO to N02. The total hydrocarbon
and methane concentration is higher in the chambers
than in front of them. This effect is caused by
1044
-------
CO [ppm]
C02 [vol «]
S02 [ppm]
SHM ["""]
CH4 [ppm]
CNHM - CH, [ppm]
NO [ppm]
NOX [ppm]
N02 [ppm]
02 [vol «]
PARTICLES [mg/m3]
MEASUREMENT
IN THE CHAMBERS
/>
16,9 (+3,0)
0,63 (tO, 14)
4,7 (+1,6)
8,5 (+3,4)
2,1 (±0,8)
6,3 (+3,0)
15,8 (+7,1)
16,3 (+7,2)
0,45 (+0,42)
20,0 (+0,7)
-
B
17,9 (+3,3)
0,67 (+0,14)
4,9 (±0,9)
8,9 (+2,8)
2,5 (+0,9)
6,3 (+2,6)
15,6 (+6,9)
15,9 (±6,5)
0,40 (±0,28)
19,5 (+0,6)
4,2 (±0,3)
BEFORE THE CHAMBERS
A
16,3 (±3,2)
0,52 (+0,14)
7,9 (±3,8)
5,3 (+3,8)
1,0 (±0,33)
4,3 (±3,7)
17,6 (±8,5)
17,7 (±10,2)
0,10 (±0,09)
20,0 (±0,7)
-
B
18,0 (±3,4)
0,57 (+0,16)
7,0 (+1,4)
4,2 (±1,7)
1,0 (+0,32)
3,3 (±1,4)
16,0 (±8,8)
16,4 (±10,3)
0,21 (±0,27)
19,1 (+0,5)
-
A EXHAUST WITHOUT PARTICLES
B TOTAL EXHAUST
Figure 16.
the animals. The particle concentration of
4.2 mg/m corresponds to that of the 1/1O dilution
in the 5-month experiment. A comparison with the
data of the different exhaust dilutions in the
5-month experiment is impaired, in particular with
regard to the soot concentration, because the
duct system to the exposure chambers in the
life-time experiment is of a different, much larger
design which involves additional branching. This
latter aspect causes an increased loss of particles
on the way to the exposure chambers. It became
obvious in the 5-month study, that the particle
concentration of the life-time test should not be
increased beyond this level. Otherwise effects
will become prominent, which are purely caused by
the mechanical burden of particles on the lung.
In turn, for the dilution chosen, the concen-
trations of some harmful gaseous substances in the
exposure chambers are so low, that they could well
be compared with extreme situations of actual
environmental conditions.
About 8 months into the life-time study, 22 dif-
ferent clinico-chemical and haematological para-
meters were measured on those exposure groups, which
did not receive supplementary treatment (Fig.17).
1045
-------
INVESTIGATIONS IN CLINICAL CHFH1STRY
ESSAY
GOT [u/l]
GPT [u/l]
Y - GT [u/l]
CHE [u/l]
LDH [u/l]
AP [u/l]
« - HBDH [u/l]
a - AMYIASE [u/l]
UREA [mg/dl]
CHOLESTERIN [mg/dl]
TRIGLYCERIDE [mg/dl]
CREATININE [mg/dl]
PROTEIN [g/dl]
CONTROL
27.9 (+5.7)
52.6 (±9.4)
1.9 (+0 9)
289.1 (+33.8)
170.3 (±43.0)
206.2 (+43.6)
96.1 (±22.4)
244.0 (±57.2)
50 7 (±16.3)
147 7 (+26.2)*
127.8 (±49.7)
0.41 (+0.09)*
5.7 (+0.35)
GASEOUS EXHAUST
32.7 (±6.9)*
55.3 (±11.0)
2.0 (±0.68)
299.2 (+39.9)
178.8 (±38.7)
264.2 (+46.1)*
86.1 (±22.1)
277.1 (±60.6)
60.2 (±17.4)*
143.6 (±21.6)
179.0 (±39.8)*
0 32 (±0.06)
6.0 (+0.33)*
TOTAL EXHAUST
34.8 (±5.5)*
53.6 (±8 1)
4.2 (±1.2)*
293.9 (±39.8)
226.6 (±51.8)*
248.6 (±56.5)*
106.7 (±21.0)
260.6 (±93.2)
58.5 (±9.7)*
127 4 (±20 7)
133.6 (+37.3)
0 30 (±0.05)
5.94 (±0.47)*
Figure 17.
From each group, 25 animals were examined. Enzymes
such as GOT, -GT, LDH and alkaline phosphatase
were significantly increased, especially in animals
exposed to the total exhaust. As in the 5-month
experiment, the concentration of blood urea was
markedly higher. The total protein content in the
serum was also significantly increased in exposed
animals. However, the various electrophoretically
separated fractions of the serum protein did not
differ significantly from the controls.
As in the 5-month experiment, the haematological
data showed again that the mean corpuscular volume
was significantly higher in the exposed animals,
while the erythrocyte and leucocyte counts were
again lower in the exposed animals. Whereas the
CO-haemoglobin content was still significantly
enhanced in exposed animals, the methemoglobin
1046
-------
INVESTIGATIONS IN HEMATOLOGY
ESSAY
ERV [l06/mm3]
LEUCO [103/mm3j
HCT [vol *]
HB [g/dl]
MCV [a3]
HBE [pg]
MCHC [gHb/100 ml Ery]
CO - HB [*]
MET - HB [%]
CONTROL
7.39 (+0.60)
5.2 (+1.2)
51 0 (+3.7)
16.3 (+0.9)
68.7 (+0.8)
22.1 (+1.3)
32.1 (+1 7)
0.08 (+0 13)
0.26 (+0.31)
GASEOUS EXHAUST
7 04 (+_0.48)*
4.94 (+1.4)
49.7 (+3.3)
16.8 (+0.76)*
70 3 (+0.74)*
23.9 (+1.41)*
33.8 (+1.97)*
2 02 (+0.81)*
0.35 (+0.38)
TOTAL EXHAUST
7.07 (+0 62)
4 25 (+1.1)*
49.7 (+4.11)
16.0 (+1.27)
70.1 (±0.91)*
22.6 (+1.13)
32.2 (+1 60)
1.59 (+0.4)*
0. 27 (+0 44)
Figure 18.
values in exposed and control animals showed no
difference so far (Fig.18).
These examinations will be repeated every 4-5
months during the experiment, so that at the end
of the experiment a statement can be made not only
on the histological alterations in the respiratory
tract, but also on the general chronic toxicity
of diesel exhaust under these exposure conditions.
Acknowledgment:
The work published in this paper is part of the
research activities of the Working Group "Unter-
suchungen iiber die carcinogene Belastung des Men-
schen durch Luftverunreinigungen" of the Umwelt-
bundesamt of the Federal Republic of Germany.
1047
-------
CARCINOGENICITY OF DIESEL EXHAUST AS TESTED
IN STRAIN 'A1 MICE
John G. Orthoefer, Wellington Moore, Dale Kraemer,
Freda Truman, Walden Crocker, and You Yen Yang
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
ABSTRACT
Groups of Strain 'A' mice were exposed to diesel exhaust by
inhalation and diesel particulate by intraperitoneal in-
jection. The animals were exposed from seven to eight weeks
and then sacrificed 26-30 weeks postexposure. Other animals
were exposed for up to seven months by the inhalation route.
Some animals were promoted using urethane at a dose below
which tumors would occur. There was no increase in incidence
of pulmonary adenomas in the animals exposed to either diesel
exhaust or diesel particulate over the control animals. In
the animals which were promoted using urethane at a low dose,
there was a significant increase in pulmonary adenomas.
Diesel pariculate was found in the lungs and bronchial lymph
nodes of animals exposed to diesel exhaust 26-30 weeks after
cessation of exposure.
INTRODUCTION
The proposed increase in the use of automotive diesel engines
in automobiles has caused concern over the potential health
effects of exposure to diesel emissions. For the regulated
gaseous emissions, automotive diesel engines do not rep-
resent an unusual hazard when compared with automotive
gasoline engines. However, the particulate. emissions from
diesel engines are considerably higher than those from
automobiles run on unleaded gasoline, and concern has been
expressed about the possible hazard of inhalation of the
particulate material. Adsorbed onto the surface of the
1048
-------
carbonaceous participate are a number of potentially toxic
substances. Among these substances are a number of poly-
cyclic hydrocarbons including known carcinogens such as
benzo(a)pyrene (BaP). Extracts of the particulate have been
positive for mutagenesis in vn vitro screening tests. In
higher animals additional factors such as particle size,
mucociliary clearance, physiological availability of the
adsorbed material, residence time and translocation of the
particulate from the deep lung all impact upon the toxicity
and possible carcinogenicity of diesel particulate. Studies
conducted in this laboratory indicate that diesel particulate
may remain for a considerable period of time in the lungs of
experimental animals following cessation of inhalation ex-
posure (Moore et al, 1978).
In an attempt to assess the carcinogenic potential of diesel
exhaust and diesel particulate alone, we chose to use the
Strain 'A' mouse lung tumor model first described by Ander-
vont and Shimkin (1940). This model system has been used as
a carcinogenic bioassay for numerous chemicals including
polycyclic hydrocarbons (Shimkin and Stoner, 1975). The
approach used in this study consisted of exposing part of the
animals to diesel exhaust by inhalation and the remaining
animals to intraperitoneally in-
jected diesel particulate.
METHODS
Diesel-Exhaust Generation System. The diesel exhaust ex-
posure facility is composed of engine system, air dilution
system, and animal exposure chambers. The engine system
consisted of a six-cylinder (198 cu. in.) Nissan automotive
diesel engine and a Chrysler Torque-Flite transmission cou-
pled to an absorption dynamometer. All the exhaust from the
engine was mixed with CBR filtered and conditioned air in a
dilution tube at a dilution rate of one to 13. From the
dilution tube, diluted exhaust entered a large volume mixing
chamber from which a portion in Study 1 passes through
dynamic flow irradiation chambers (to simulate sunlight) and
is then conducted to animal exposure chambers. In Study 2, no
irradiated exhaust was used and the exhaust was all diverted
away from the irradiation chambers. The remaining exhaust
was piped directly from the mixing chamber to the non-
irradiated animal exposure chambers. The engine was cycled
continuously and run for 20 hours daily in Study 1 and eight
hours daily in Study 2. The fuel used was No. 2 diesel fuel.
Continuous cyclic monitoring of the chamber atmospheres was
carried out for carbon dioxide, carbon monoxide, total
hydrocarbons, and nitrogen oxides. Particulate samples were
1049
-------
collected daily. The average concentrations of the measured
exhaust components during the exposure period are presented
in Table 1 for Study 1 and Table 2 for Study 2. Particulate
samples were also collected on Nucleopore membranes for
sizing by scanning electron microscopy (SEM). A repre-
sentative (SEM) photomicrograph of the diesel particulate
from both studies is shown in Figure 1.
The particulate material for intraperitoneal injection was
collected on Fluoropore filters during Study 1 from the
exposure chambers which housed the mice that were being
exposed by inhalation. The filters were weighed, scraped,
and then sonicated in distilled water to remove the diesel
particulate. The suspension was refrigerated, protected from
light and resonicated prior to intraperitoneal injection.
The particulate was collected and processed one to two days
prior to injection. A scanning electron micrograph of the
diesel particulate contained in the suspension is shown in
Figure 2. For the control, filter samples were taken from the
clean air exposure chamber and processed in the same manner
as the diesel particulate filters. The suspensions were
cultured periodically for bacterial contamination.
Animal Exposure. In Study 1, Strain "A" mice, weighing 19-20
gm each, were obtained from the Strong Research Foundation
and the Jackson Laboratory. Male A/Strong mice, were
randomly divided into four groups; 25 were placed in the
irradiated exhaust exposure chamber, 25 in the nonirradiated
exhaust exposure chamber, 25 in the control air chamber, and
25 (positive control) were injected intraperitoneally with
urethan (20 mg/mouse) and placed in the control air chamber.
The mice were exposed for seven weeks to diesel exhaust and
then held in control air chambers for 26 weeks until
sacrifice.
For the diesel particulate exposure, the A/Strong mice were
divided into six groups of 30 mice each. Groups I, II, III
received 235 ug, 117 ug, and 47 ug of diesel particulate per
injection, respectively. Group IV (clean air control)
received injections of material from the clean air filters.
The mice were injected with 0.1 ml of the appropriate
solution intraperitoneally three times weekly for eight
weeks. Group V (positive control) received an intraperi-
toneal injection of 20 mg of urethan. Group VI received no
injections. The animals were held in control air chambers
and sacrificed at the same time as those exposed to diesel
exhaust.
The strain A/Jackson mice were randomly divided into three
treatment groups of 20 females and 20 male each. One group
was exposed to nonirradiated diesel exhaust; one group was
1050
-------
- — ^
1 —
>>
-o
13
1 *
CO
'
00 ^.
c: cu
O OJ
••- 3
-4->
n3 t-
S- CU
-l-> D-
C
O) oo
<_> >>
C 03
O Q
CO
r — •
CD
i- >1
O) 03
-C Q
f~)
00 S-
o o>
E Q-
+->
et oo
S_
S- :3
O) O
_o n;
rO O
JT CXI
C_)
OJ
S-
3
oo
O
Q.
X
t±J
i-
<
c:
03
ID
c3
00
S-
O)
JD
E X3
03 OJ
JZ -l->
^ 5 r^3
Ol T3
S- 03
Z3 i-
00 i-
o >->
Q.
X
* I 1
"O
OJ
J_)
03
-o
03
S-
S-
1
£=
O
•^.
+->
c
Ol
c
o
Q
E
O
<_>
o
00
<1) X3 S_ Li_ U_ -1—0)
"O-r- O3OO "XTD
•i- X U LO O CD O •!-
xo OI^LO -o ••- x
O C !- ro -r- -0 O
•i— o -a x
-a sr >> o c c
ni cu cu
c: c o en ai
o o i— -r- o o
JD J3 03 S_ S- S_
i- i— -I—' 4—^ 4—^ +-1
03 rt3 O •!-•!— -t-
o c_) I— 2: ^: ^
c O
1
I O
i — r-^
01 co
, — ID
ro cxi
i — ro
CXI U3
E
O-
o.
^ — ,
CXI
o
OO
„
QJ OO
-o E
X CD C7
O 4-> E
• i— 03
•O ' — OO
3 00
S- O O3
=3 -r- SI
M- +->
i — i_
^ O3
LO CL.
1051
-------
Table 2
Exposure Chamber Atmosphere Concentrations (Study 2)
8 Hours Per Day 7 Days Per Week
Exposure Chamber
Component Diesel Exhaust Clean Air
Carbon dioxide, (C0?)% 0.29 0.040
Carbon Monoxide (CO) ppm 19.72 2.0
Total Hydrocarbons (THC) pprnC 7.84 2.0
Nitric oxide, (NO) ppm 11.23 0.11
Nitrogen dioxide, (N02) ppm 2.65 0.07
Nitrogen oxides (NO + N02) ppm 13.88 0.18
Sulfur dioxide, (S02) ppm 2.06
Particulate
Mass mq/m3 6.39 0.0
1052
-------
•t>;'-' i sf'^if' --^;--
»%i
"
^ ''
\,J%!*. ii
KM* , ». >Jt
Figure 1 SEM photomicrograph of diesel exhaust particu-
late. Arrows point to small particulates.
1053
-------
Figure 2 SEM photomicrograph of diesel participate in-
jected intraperitoneally.
1054
-------
exposed to clean air only; and another group was injected
intraperitoneally with urethan (20 trig/mouse) and placed in
the control air chamber. The animals were exposed for eight
weeks to the diesel exhaust and then sacrificed 30 weeks
after cessation of exposure.
In Study 2, 240 female A/Strong mice weighing 18-20 grams
were randomly divided into two groups and placed ten animals
per cage. One half of each group (60 animals) was given 1 mg
urethan intraperitoneally as previously described. The mice
were exposed for seven and one-half months in chambers to
either raw diesel exhaust or clean control air.
Four additional groups of A/Strong male mice weighing 19-20
grams each were purchased in lots of 200, 220, 240 and 199,
respectively. These animals were randomly divided into two
groups (raw diesel exposure and clean control air exposure)
and placed in their respective exposure chambers. The groups
1, 2, and 3 were exposed until the mice were 36 weeks of age;
group 4 was exposed until the mice were 44 weeks of age.
At the time of sacrifice, animals from all groups were killed
by an IP injection of an overdose of pentobarbital sodium.
The lungs were immediately removed and washed by swirling
them in normal saline to remove any accumulated blood or
debris from the plural surfaces. The lungs were then placed
in Tellyesniczky's fluid (Humason, 1972). Five animals
chosen at random from each group had all organs preserved in
ten percent buffered formalin for a complete histopatho-
logical examination. Any lesions observed in the remaining
animals were also preserved for further histopathological
examination.
After two to five days the lungs were removed from the
Tellyesniczky's fluid and the lobes severed from the primary
bronchus. The presence and number of adenomas in the lungs
were determined by the presence of milky white nodules. A few
typical nodules were selected for histopathological con-
firmation of the pulmonary adenoma. Any questionable lesions
were examined histopathologically for confirmation of the
tumor's presence.
RESULTS
Study 1: Diesel Exhaust Exposed Animals. In general, no
gross changes in condition or appearance were observed in the
mice. During the study several animals died or were found
missing in the different groups. However, the number of
deaths was not significantly different in the exposed and
control groups with the exception of the longest exposure
1055
-------
animals (groups 4) See Table 3. The mean body weights of the
mice exposed to diesel exhaust are shown in Figures 3 through
10. Exposure to diesel exhaust did not affect the growth rate
when compared to the controls.
At necropsy, the lungs from mice exposed to the raw and
irradiated diesel exhaust for eight weeks were grey in
appearance with numerous dense black areas. The animals
exposed for 30 weeks were uniformly black in appearance. On
closer gross inspection of the eight week exposed animals,
it appeared that the black areas outlined the intermediate
and small airways of the lungs. Histopathological ex-
amination revealed that macrophages containing large amounts
of black granular material were accumulated around the
terminal bronchioles. In addition, smaller number of macro-
phages containing the black granular material could be seen
in groups dispersed throughout the lungs (Figure 11). Bron-
chial lymph nodes were easily identifiable on gross ex-
amination because of their color due to the presence of
macrophages containing black particles. A section of a
bronchial lymph node from a mouse exposed to nonirradiated
diesel exhaust is presented in Figure 12. The black deposits
are composed of aggregrations of macrophages containing
granular diesel particulate material. These macrophages
appear to be more concentrated in the medullary spaces
although smaller aggregrations may be seen in the cortical
areas of the lymph node.
No tumors were observed by gross examination in tissues other
than lungs in the diesel exhaust exposed and control air
mice; however, two tumors were observed in the skin of mice
injected with urethan. The incidence of lung tumors in the
mice from each treatment group is given in Table 4. Com-
parison of the tumor incidence among the different treatment
and control groups indicated no significant trend. The A/
Jackson mice did not show an increased incidence of tumors in
the diesel exposed over the controls. The incidence of
tumors in the males was slightly higher when compared to the
females, but was similar to the A/Strong mice. In comparing
the data, there was no significant difference between the
treated and control groups. The incidence of lung tumors in
the control group was nearly identical to that obtained in
other studies demonstrating the stability of the Strain "A"
mice to the induction of lung tumors. In^Strain A/Strong
mice sacrificed 30 weeks after the first injection, Stoner,
et al (1976) reported 31% incidence of lung tumors with an
average number of 0.28 tumors per mouse in the untreated
controls and a 37% incidence of lung tumors with an average
number of 0.42 tumors per mouse in mice injected with 0.85%
NaCl solution. All of the 20 mg urethane injected animals had
1056
-------
OO
01
^—
tO
h-
00
00
01
1 —
o
°_J3
(0
tO
C
OJ
00
E O)
4-> O O JD
tO £ S- E
Q o
•r- 00 E
S =s
1 —
00
IB 00
E S-
•i- o
c CM E
1—
l|
O
o o
^ •— E
1—
00
r— S-
*X3 O
-t-> E
O ^
1 — 1—
O 00
to
1 — E
to T-
00 •!- £Z
S- +-> C 4-
O •!- i— i O
• S-
O 3
O CU
i- 00
-l-> O
C Q.
O X
II II
t 1 1 1 i
C
0
to
>> en
T3 -i-
3 00
•4-> Ol
CO Q
f^.
^^
IT)
CM
CSJ
^^
CTl
0
CM
^
O
CM
0
O
i —
ro
CTi
CJ
[•^
^^
un
CM
CM
^^
CTt
0
CM
o
CM
CM
O
0
IO
CO
UJ
r_
1
ea;
1 —
CM
^^
LO
UD
"^
O
1
o
n
CO
1 —
CM
O
( —
i —
0
o
i —
CM
^^
LT)
tD
^^
O
1 —
O
CM
^
CO
o
1 —
1 —
CO
CTi
UJ
CM
1
c£
O
CM
^^
ID
^j-
i —
^-^
i —
1
!
CM
^
CM
O
CM
i —
O
°
O
CM
^^
ir>
^j-
i —
^ — .
i — -
1
o
00
to
CM
CM
O
CM
( —
LO
O
UJ
oo
1
c£
_o
~ — .
0
CO
^^
1 —
CM
to
en
£
en
en
Ol
0
LO
^^
O
CO
^^
1 —
o
CM
^
^
o
o
CTl
LU
,-j.
i
c^
1057
-------
,-••••'
0 24 6 82 4 6 6 10
| - eipoiurt - (h --- —
M 16 11
post txposure -
Figure 3 Mean body weights of Strain A/Strong mice exposed
to diesel exhaust and controls.
Figure 4 Mean body weights of strain A/Strong mice injected
with diesel particulate and control materials.
1058
-------
0 2 4 6 1 3 5 7 9 11 13 15 17 19 21 23 25 27
I e x po au r t II po si »iposu re— I
WEEKS
Figure 5 Mean body weights of Strain A/Jackson mice exposed
to diesel exhaust and controls.
26-
25- -
24-
23-- 8
22-- o 'if m
21-- Q **
20--
19--
18--
17- -
85
O
Si*
o
D
O CONTROL
• EXPOSED
D CONTROL c URETHANE
• EXPOSEDcURETHANE
J
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
WEEKS EXPOSED
Figure 6 Mean body weights of Strain A/Strong mice exposed
in the urethan plus diesel exhaust promotion
study.
1059
-------
GROWTH CURVE — EXPERIMENT A-1
g
325-
30.0-
275-
250- -
20.0- -
17.5- -
15.0- -
125- •
10.0- •
E E E
C E
E
PLOT OF WT/WEEK SYMBOL IS VALUE OF GROUP
E
C E
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32
WEEK
Figure 7 Mean body weights of Strain A/Strong mice exposed
to diesel exhaust and controls in Study A-1.
32.5-t-
27.5- -
GROWTH CURVE — EXPERIMENT A-2
20.0- -
C C C
EC C E
C E
C E
C E
C E
175-
15.0-
12.5 —
10.0-
I PLOT OF WT/WEEK SYMBOL IS VALUE OF GROUP
•C
E f 0001
E
III II I
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32
WEEK
Figure 8 Mean body weights of Strain A/Strong mice exposed
to diesel exhaust and controls in Study A-2.
1060
-------
GROWTH CURVE - EXPERIMENT A-3
32.5-
30.0-
275-
I
2 22.5-
E E C C C
E C
E C C
E C
17.5-
15.0-
12 5 —
10.0-
c
PLOT OF WT/WEE
E
- C
E
C
1 1 1 1 1
< SYMBOL IS VALUE OF GROUP
p = 093
1
0246
10 12 14 16 18 20 22 24 26 28 30 32
WEEK
Figure 9 Mean body weights of Strain A/Strong mice exposed
to diesel exhaust and controls in Study A-3.
GROWTH CURVE - EXPERIMENT A-4
34-
33-
32-
31-
30-
29-
| 27-
5 26-
25-
24-
23-
22-
21-
20-
19-
E C
E
C
C
PLOT OF WT'WEEK SYMBOL IS VALUE OF GROUP
p =090
0246
-I—I—I—I—r
10 12 14 16 18 20 22 24 26 28 30 32 34
WEEK
Figure 10 Mean body weights of Strain A/Strong mice exposed
to diesel exhaust and controls in Study A-4.
1061
-------
*- * - i
1'* f ' *f J
; ' .*• I • 4 .,. -"^ ';-••
-:f-d'4 <"«--;* -;
*''* '«*'JSi ^CK- *
. *'' Jai'f^ 3IS* ***
*.*™-'^;
**« »X
5'>
Figure 11 High power photomicrograph showing black granular
particulate containing macrophages in alveoli.
1062
-------
Figure 12 Section of bronchial lymph node from Strain Amice
exposed to nonirradiated diesel exhaust. The
heavy black material is composed of aggregrations
of macrophages containing the granular diesel
particulate material.
1063
-------
cu
o
•r-
s:
QJ
ro
QJ
^_
I—
c
ro
c~
1 *
QJ
— i
-Q
c
ro
o
i-
•3- 4-1
c
cu o
,— (_j
_a
cu
0
CL
X
LoJ
1 —
QJ
00
QJ
Q
£-
,:
00
O
E
1—
CD
13
1
O CU
00
• Z5
O 0
-~^
CU oo
CD S-
ro o
s- E
cu rs
> I—
00
-c s-
-l-> O
••- E
3. ^
1—
QJ
0 CD
•r- C
s: =s
— i
i —
ro
4-1
c
1 — 1
"^^.
00
S-
o
>
• 1 —
(
-^
^J
c
QJ
4-J
(O
QJ
S_
1—
CU QJ
roi — ro CL • — oootoo-
IOCMI — •!— rounrotrj-r-
• • • i ' • ^^ ^^ • | ^
OOO>— CD O O O i—
=i =3
s: s:
to co co co r^
• — i — i — i — i —
^^ ^^ ^^ ^^ ^^
LO to LTJ i — r^
"~ '~
en CM CM ro
i — CM OJ CM
^^ ^^ ~°^ ^"^
i — to ro ro co
CM CM
.
0
II
CL
tj_
~j^j
O O O CD O
CM CM CM CM C\J CM
_c to oo co co r^
,^
3
un LO un LO to
CM CM CM CM CTt
— ^ ^-^ ^-^ ^-- •
CJ1 CM CM CO CM
i — CM CM CM
II
00 00
CM QJ QJ
X 00 i — 00 i —
cu ro cu ro
•—£•—£
ro cu ro QJ
s: u- s: LU.
4-> 4-1
00 OO
13 ZS
ro 4-i ro
r" oo c~
x rs x
QJ ro cu
-OX -D
QJ QJ QJ
4-J -4-J
rO "o QJ ^
QJ -I- QJ O -r-
U T3 4-J -X -i- -0 *
•i-roro.— C s: ro i— c
s:s--^oro s-oro
S--OS--C C S- S- -C
CD T- ro 4-1 4-1 O "i" 4-1 4-1
CCS-CCU OOCCO)
OOS-Oi- ^^OOS-
S-Z1-10Z3 U^0r3
4_i ro
oo "~^
0
un
i —
O
n
Q.
<4-
TD
ro
_c
4-1
3
i —
ro
•
LO
II
CM
1064
-------
multiple tumor nodules. In some cases, the nodules had
coalesced into a large tumor mass making it impossible to
determine the number of initial nodules.
Diesel Particulate Injected Mice. Three groups of 30 mice
each were injected intraperitoneally with 235 ug, 117 ug and
.47 ug particulate, respectively, three times weekly. The
highest dose level was chosen to correspond to approximately
the inhalation dose assuming 50% retention. Using the diesel
particulate concentration of 6 mg/m^, a minute volume of
0.024 litter and 50% retention of the particulate, the weekly
inhaled dose would be approximately 725 ug. The highest dose
group received 705 ug/week. There was no evidence of
toxicity in any of the treatment groups as indicated by
general appearance and growth rate. The mean body weights
for the different groups are shown in Figure 4.
The number of tumors in the mice injected with diesel
particulate and the control mice is given in Table 5. There
was no significant difference between the incidence of tumors
in the injected and control mice.
Study 2: The dose of urethane for this group was interpolated
from Depaola (1959) as a dose of urethane below which one
would expect to find an increase in tumors.
No apparent dose related growth effects or lesions other than
lung tumors was apparent. The lung tumor counts are
presented in Table 6. The results are significant using the
Chi Square test p = .00077.
Male and 30 and 38 Week Exposure. The first three groups
designated Al to A3 showed no dose-related growth effects,
gross lesions or microscropic lesions attributed to the
diesel exposure. The results of the tumor counts and
analysis are presented in Table 3. The fourth group was
exposed approximately eight weeks longer and showed a sig-
nificantly shortened survival time in the exposed groups.
The higher mortality occurred in the final phase of the
exposure and was thought to be the result of lessened
resistance to disease.
The randomly chosen animals for histopathological examina-
tion showed no diesel-related lesions which could be at-
tributed to exposure. The lungs, except for questionable
nodules were not routinely examined for histopathological
lesions.
1065
-------
LO
OJ
rd
1—
3
O)
t> •"
•^o
Z? s-
,_; +->
c
SS
^- ^
cn
E s
£3
•<:5
-l~ OJ
^2
o ^
^ o
1 — •"""
^fe
= O-
-~
' — ( —
u_ 1 —
t/1
c- $_
+-> O
3 rs
1—
ai
o a
•i- C
SI 13
CO r—
O -r-
> C
LO
^_
O)
-I-J CO
c: o
CD o
E
OJ +J
S- O
1— 1—
O OO U3 OO
oo oo o oo
• • • •
o o o o
0000
00 OO OO OO
^-^ ' — ^ ^^ *^^
CO O CXI CO
o o o o
oo oo oo oo
o o o o
OO OO OO 00
, —
o
S-
^_)
c
o
o
-a
a>
cn cn cn -4->
E E E 0
O)
U3 OO i — -i-o
• • • c
LO OJ i — •!-
„ = » a
Q. Q. Q. Q.
3 3 13 13
o o o o
S- S- S- i-
C-D CD CS C£3
O)
a.
• 1—
4-1
^~
^
E
0s)
OJ
^x^
CTi
o
oo
cr>
OJ
c:
m
-C
ai
S-
=3
cn
s:
o
OJ
*
CL
~^
o
^_
C3
un
C\J
<
0
CO
oo
^^
r~-
o
oo
00
OJ
'o
^_
4_>
d
o
o
-a
a>
4->
rC
-1 —
3
CT.
o
00
tfi
tf~)
II
CSJ
X
1066
-------
Table 6
Survivors and Number of Tumors in the Control
and Urethan-Diesel Exposed Groups of Animals
Treatment
Survivors
Original
Number
Survivors
No. of
Animals
with Tumors
Number
of
Tumors
Control
Control +
Urethan
Exposed
Exposed +
Urethan
58/60
52/60
56/60
59/60
93.5
83.8
90.3
95.2
4
9
14
22
5
13
18
23
1067
-------
DISCUSSION
Generally, the results of these studies are negative with the
exception of the urethane injected exposed animals (promotion
study). Even though the control animals' tumor rate is below
that which is expected for Strain "A" female mice, the
hypothesis that the test is positive can be tested by
substituting the known tumor value for our Strain A/Strong
control mice. This comparison is shown in Table 7. These
latter control mice are males, whereas the promotion animals
are all females. The literature references report the lower
spontaneous rate occurs in females which would tend to make
the test more significant.
The Strain "A" mouse lung tumor model has been used by several
investigators as a bioassay for assessment of carcinogenicity
of polycyclic hydrocarbons. Shimkin and Stoner (1975)
summarized the findings on ten polycyclic hydrocarbons and
related compounds following a single intravenous injection of
approximately 0.25 mg of aqueous dispersions with sodium
sulfosuccinate as a wetting agent. Some of the polycyclic
hydrocarbons such as 3-methylcholanthrene, dibenz(a)anthra-
cene and benzo(a)pyrene were potent carcinogenic agents while
benz(a)anthracene and 5-methoxy-7-propyl-benz(a)anthracene
demonstrated no activity. In these tests ten to 19 animals
were used for each compound to detect the tumor incidence. In
our experiments no increase in the incidence of lung tumors
was found following as much as nine months of inhalation ex-
posure to diesel exhaust alone, however, the addition of
urethane in low doses did elicit a positive response to the
combination of urethane and diesel exhaust. The role of
urethane as an initiator and/or promotor for lung cancer is
known. In the case of our promotion study, neither the
urethane exposure or diesel exhaust exposure alone produced a
significant increase in lung nodules. The two exposures
together, however, did produce a weakly positive significant
increase in pulmonary adenomas in the Strain "A" mouse.
Diesel particulate was present in the lungs and draining
lymph nodes 26-30 weeks after cessation of exposure. The
duration of the polycyclic hydrocarbons in the lungs after
cessation of exposure cannot be ascertained because the
elution rate of this material off the surface of the diesel
particle under physiological conditions has not been de-
termined. There is little information on the factors
governing release or biotransformation of the polycyclic
hydrocarbons by cells phagocytizing the particulate ma-
terial .
1068
-------
Table 7
Comparison
(Femal
Group
1
Control
2
Control +
Urethan
3
Exposed
4
Exposed +
Urethan
5
Historical
Controls
of Urethan
e) Mice and
Survivors
Initial
58/60
52/60
56/60
59/60
60/602
Promoted Diesel Exposed A/Strong
Historical Controls (Male)
Chi Square Test
Mice with Groups^
Lunq Tumors p value
4 Group 1, 2, 3, 4
p = .00077
9 2, 3, 4, 5
p = .0021
14 2,5 3,5
p - .92 p = .21
22 3,4 2,4
p = .16 p = .019
10
Due to performing several tests on these data significance
should be at the p - .01 level.
This figure represents a portion of historical controls
and is not representative of the total number or of the
percentage of survivors.
1069
-------
The presence of the particulate material in the lymph nodes
indicates the transport of diesel particulate from the
alveoli to the interstitium and the lymphatics. The mech-
anism of transport from the alveoli to the interstitium is
most likely the same as shown by Adamson and Bowden (1978) who
have demonstrated the presence of carbon within sessile
macrophages of the hilar lymph nodes 12 hours following in-
tratracheal instillation of carbon in mice. They found
migration of free carbon particles through the type 1 cells
to the interstitium where they were rapidly phagocytosed by
interstitial macrophages. Also the presence of carbon in
hilar lymph nodes 12 hours before appreciable amounts of
carbon had been taken up by interstitial macrophages, sug-
gests that carbon can reach the lymph nodes in the free
state. It still remains to be determined if the diesel
particles can serve as a mechanism of transport of polycyclic
hydrocarbons to other tissues.
The method used to generate exhaust and inhalation route of
exposure correspond closely to the human potential for
inhalation of diesel emissions. The exposure generation
system used in this study is probably as close to en-
vironmental reality as can currently be designed for studying
such emissions. The concentration of the particulate in the
exposure was several orders of magnitude higher than would
normally occur under severe environmental conditions and the
clearance mechanisms of the lung were obviously able to
remove only part of the diesel particulate as indicated by
the presence of the material in the lungs 26-30 weeks after
cessation of exposure. In vitro tests of extracts of diesel
particulate emissions have indicated microbial mutagenicity.
The significance of an increased incidence of tumors in these
mice underlines the need of animal inhalation exposures when
trying to assess environmental risk to humans.
Reznik-Schuller and Mohr (1977) demonstrated the development
of pulmonary adenomas in Syrian Golden hamsters after in-
tratracheal instillations with automobile exhaust con-
densate. They pointed out that the high tumor rate may, in
part, be due to the failure of the mucociliary defense
mechanism to clear the airways of a large portion of the
condensate. Further, they indicated that the instillation
procedure does not correspond to the human situation of
inhalation and that the instilled material^was administered
as a condensate, whereas most of the inhaled substances are
gaseous, or bound to smaller carrier particles which are
likely to be removed through mucociliary action. An
increased incidence of skin tumors has been found by some
investigators (Kotin et al, 1955) following skin painting of
diesel extracts. This study illustrates the necessity of
1070
-------
continued investigation in order to fully assess the car-
cinogenic potential of atmospheric carcinogens. Although
initially negative by the inhalation route diesel exhaust,
when potentiated with a promoter/initiator, demonstrated a
positive response. There are numerous carcinogens and
cocarcinogens in the environment and the possibility of
interaction is present.
REFERENCES
1. Adamson, I. Y. R. and Bowden, D. H. (1978). Adaptive
response of the pulmonary macrophagic system to carbon.
Lab. Invest. 38:(4), 430-438.
2. Adervont, H. B. and Shimkin, M. B. (1940). Biologic
testing of carcinogens. II. Pulmonary-tumor-in-
duction-technique. J. Natl. Cancer Inst. 1:225-239.
3. Depaola, J. A. (1959). Influenced altered atmospheric
oxygen on urethan-induced pulmonary tumors in mice. J_^
Natl. Cancer Inst., 23:535-540.
4. Humason, G. L. (1972). Animal tissue techniques. 3rd
Edition. W. H. Freeman Co., San Francisco.
5. Kotin, P., Falk, H. L. and Thomas, M. (1955). Aromatic
hydrocarbons III. Presence in the particulate phase of
diesel-engine exhausts and carcinogenicity of exhaust
extracts. AMA Arch. Ind. Health, 11:113-120.
6. Moore, W., Orthoefer, J., Burkart, J., Malanchuk, M.
(1958). Preliminary findings on the deposition and
retention of automotive diesel particulate in rat
lungs. 1978 Annual Meeting, Air Pollution Control
Association Proceedings, Houston, Texas, June 25-29.
7. Reznik-Schuller, H. and Mohr, U. (1977). Pulmonary
tumorigenesis in Syrian Golden hamsters after in-
tratracheal instillations with automobile condensate.
Cancer 40:203-210.
8. Shimkin, M. B. and Stoner, G. D. (1975). Lung tumors in
mice: Application to carcinogenic bioassay. Advan.
Cancer Res. 21:1-58.
9. Stokinger, H. E. (1977). Toxicology and drinking wa-
ter contaminants. Jour. AIM A, July 399-402.
10. Stoner, G. D., Shimkin, M. B., Troxel, M. C., Thompson,
T. L., Terry, L. S. Test for carcinogenicity of
metallic compounds by the pulmonary tumor response in
Strain "A" mice. Cancer Res. 36:1744-1747.
1071
-------
General Discussion
G. SIGNER: During a 36-week period if you inject one
milligram of 3-methyl cholathene per mouse, you get 40 to 50
lung tumors per mouse. With this dose of urethane you usu-
ally get one tumor per milligram and with 20 mgs of urethane
you get about 20 tumors per mouse. If diesel exhaust is
active at all, it is extremely weak in this system. I sug-
gest that you conduct some experiments where you keep the
animals around for a little longer period of time because
some of the weaker agents require a longer time period to
produce a tumorigenic response. This would permit you to
pick up some weak positive responses which you won't get in
a 36-week period.
C. RUDD: Intraperitoneal injections intrigue me. When
you suspend the diesel particles in distilled water, do the
diesel particles stay on the top or are they in suspension?
How do you succeed in injecting it?
J. ORTHOEFER: The diesel particulate at the concen-
tration we had does not stay in suspension. In order to
achieve an accurate injection dose we agitated the par-
ticulate constantly with magnetic stirring bars. We then
pour about 0.5 ml into a syringe and inject each of five
mice with 0.1 ml/mouse. By continuing the stirring motion
there was no problem with keeping it in suspension.
C. RUDD: What size needle did you use?
J. ORTHOEFER: A 25 gauge needle was used for the IP
injection.
C. RUDD: From IP injection, would you suspect that the
diesel particulates travel to the lung or does the organic
matter leach off the particle?
J. ORTHOEFER: The cytopathology of the internal organs
show massive amounts of the carbon particles throughout the
abdominal cavity and lymph nodes. It is possible that some
material may leach off the particle and enter the bloodstream
to be carried to the lungs.
1072
-------
Session VIII
EPIDEMIOLOGY STUDIES INVOLVING HUMAN EXPOSURE
TO DIESEL EMISSIONS
Chairman:
Robert Waller
A Review of the Literature: Human Health Effects Associated
with Exposure to Diesel Fuel Exhaust.
Calabrese, Edward J., Gary S. Moore, Ruth Ann Guisti,
Carol A. Rowan, and Elizabeth N. Schulz.
Trends in Lung Cancer in London in Relation to Exposure
to Diesel Fumes.
Waller, R. E.
Human Data Associated with Diesel Exhaust.
Lachtman, Dennis S.
A Retrospective Cohort Study of Diesel Exhaust Exposure in
Railroad Workers: Study Design and Methodologic Issues.
Schenker, Marc B., M.D., and Frank E. Speizer, M.D.
Characterization of Diesel Exposure Groups.
Hansknecht, Donald F., Richard A. Ziskind, and Michael
B. Rogozen.
An Industrial Hygiene Characterization of Exposures to
Diesel Emissions in an Underground Coal Mine.
Wheeler, Robert W., P.E., Frank J. Hearl, and Michael
McCawley.
1073
-------
A REVIEW OF THE LITERATURE: HUMAN HEALTH EFFECTS
ASSOCIATED WITH EXPOSURE TO DIESEL FUEL EXHAUST
Edward J. Calabrese, Gary S. Moore, Ruth Ann Guisti,
Carol A. Rowan and Elizabeth N. Schulz
Division of Public Health
University of Massachusetts, Amherst, Massachusetts
While the need to assess the potential impact of diesel fuel
exhaust on human populations is now recognized as an impor-
tant task because of the projected increase in the use of
this fuel, the epidemiological data upon which to base an
assessment are quite limited. There have been nine major
epidemiological studies which have attempted to evaluate the
influence of diesel fuel exhaust on human populations. In
all cases, these studies have dealt with occupational ex-
posures to diesel fuel exhaust. This is of concern since a
prime consideration is the influence of diesel exhaust on
the general public as a result of ambient, not occupational
exposures.
These studies have covered a broad range of occupations in
which exposures may occur to diesel fuel exhaust including
bus and railroad mechanics, coal miners, salt, iron, and
potash mine workers. These studies have been conducted in
several different countries - England, Egypt, Sweden, and
the United States by a variety of research teams. Health
outcomes considered include deaths due to lung cancer or
other causes; morbidity due to pneumonoconiosis, bronchitis;
asthma; influenza; and respiratory infection. Other non-
specific indicators of pulmonary function were included
such as reduced pulmonary function test values, and report-
ing of symptoms such as cough, wheezing, shortness of breath,
and phlegm production. All the studies reviewed were con-
ducted between 1956 and 1978.
1074
-------
Lung Cancer and Morbidity and Mortality Due to Other Causes
Three of the studies reviewed considered lung cancer mortal-
ity or mortality from other causes as a health end point
(Raffle (1), Kaplan (2), Waxweiler et al (3)). A major and
unavoidable limitation to these studies is that widespread
dieselization of transportation and mining operations did
not generally occur until the 1950's or even later in some
instances. This resulted in a total industrial exposure
period to diesel exhaust emissions of from only several years
to generally not more than 25 years depending on the study.
Thus, there may not have been enough time to detect a signi-
ficant increase in lung cancer mortality due to extended
latency periods.
Raffle (1) found no excess lung cancer mortality in a study
of diesel exposed London Transport bus mechanics aged 45 to
64 between 1950-1954. Several methodological limitations
must be considered when interpreting the findings of this
study.
There is evidence that smoking habits may have differed
between the highly exposed group and their occupational
controls as diesel exposed garage workers were not permitted
to smoke on the job while some petrol-fueled buses were
still in use (4). No control for the smoking habits of the
workers was attempted. Also, diesel exposure was not well
documented. Commins, et al. (5), reported on the analysis
of air samples from two of the London Transport garages
considered to have relatively poor ventilation. These
analyses conducted during the follow-up period indicate that
the exposure to diesel components in these two garages may
not have differed substantially from the general ambient
levels to which the occupational and general population con-
trol groups were exposed to at this time.
While the lack of consideration of smoking differences, and
the apparent absence of marked differences in respiratory
carcinogen exposure between the diesel exposed workers and
control groups seemed to bias against finding any lung can-
cer differences, the Raffle (1) study design tried to com-
pensate by comparing deaths, retirements and transfers due
to cancer among the mechanics to only deaths from lung can-
cer in the general population, thereby slightly increasing
the number of diesel related lung cancer cases in several
comparison groups. However, this procedure did not result
in any marked change in the annual rate of lung cancer cases.
Kaplan (2) also found no association between the level of
occupational exposure to diesel exhaust and lung cancer
deaths in his study of railroad workers. Again, there was
no control for smoking habits of workers and the individual
exposure to diesel exhaust was not well documented.
1075
-------
Waxweiller, et al. (3), found no significant mortality dif-
ferences between diesel exposed and non exposed potash miners.
Causes of death considered include (1) tuberculosis, (2) ma-
lignant neoplasms, (3) influenza and pneumonia, and (4) "other
respiratory diseases" i.e. bronchitis, pneumonconiosis etc.
The maximum diesel exposure in this study, however, was only
17 and 10 years, respectively, depending on the group of
workers considered and individual exposure, length of employ-
ment and previous occupational exposure are not controlled
for.
Respiratory Morbidity
Limited research has assessed the potential relationship of
exposure to diesel exhaust and specific types of morbidity
including bronchitis, asthma, influenza, and respiratory
infection. These health effects would have great signifi-
cance for high risk segments of the population such as the
very young and the elderly. Unfortunately, the two studies
which attempt to address this issue El Batawi and Nowier (6),
and Jorgensen and Svensson (7) are not very helpful with
respect to evaluating the potential health effects of diesel
exposure because of problems in study design and execution.
El Batawi and Noweir (6) found a higher frequency of upper
respiratory tract disease, bronchitis, asthma, peptic ulcers,
gastritis, and high blood pressure in a survey of Alexandria
diesel bus mechanics than had been reported in studies of
other occupational groups. However, the lack of a compari-
son group and the extremely high proportion of heavy smokers
and night shift workers makes it difficult to assess the
effects of the diesel exposure. Jorgensen and Svensson (7)
examined rates of bronchitis and respiratory infection among
diesel exposed iron miners. While the data seem to indicate
a synergistic effect of smoking and exposure to diesel mine
environments, the exposure to diesel exhaust cannot be sep-
arated from the effect of exposure to the mine environment
as a group of surface mine workers was used as a control.
Respiratory Symptoms and Pulmonary Function
The body of the literature suggests an association of diesel
exposure, reduced pulmonary function and increased reporting
of non specific pulmonary complaints. Reger and Hancock (8),
in a large study of over 700 diesel exposed coal miners and
their matched non diesel exposed coal miner controls, found
significantly more diesel exposed miners who reported symp-
toms of persistent cough, phlegm, and exacerbation of cough
and phlegm. Pulmonary function was generally found to be
poorer among exposed workers, and the disparity of symptoms
and pulmonary function between exposed and nonexposed workers
was found tq increase with the length of employment. How-
ever, decreases in pulmonary function did not seem to be
consistently related to the level of diesel usage in the
1076
-------
diesel mine and some complaints. However, wheezing and
shortness of breath were reported more frequently among the
control miners. Additionally, diesel exposure is compounded
here as in all the other mining studies with exposure to the
mine environment, and the similarity of the diesel and non
diesel mines on factors other than diesel usage is not well
documented. It should be mentioned that the control group
was drawn from a group of miners evaluated nearly a decade
prior to assessment of the diesel exposed workers (9). While
selection of such a control group is not inappropriate, it
does present several methodological questions which may
affect the outcome of the study. For example, while individ-
uals were carefully matched for smoking status, it is poss-
ible that the cigarettes smoked in 1969, at the time of the
aquisition of the control group data, contained on average
fewer filtered brands with more tar and nicotine so as to
have differentially affected the controls. Also, because
the exposed group and controls were evaluated in different
studies, years apart, differences in technique and in coach-
ing effect may have resulted in a lack of uniformity of
data collection between the two groups.
In a subsequent study, Reger, Hankinson, and Merchant (10)
examined a subset of 60 diesel exposed workers from the
matched pair study. These workers were selected to repre-
sent a broad spectrum of mine jobs. Fifty-five non exposed
mine workers were population matched to the exposed workers
and changes in pulmonary function from the beginning to the
end of the work shift were examined. While the absolute
decreases for the five pulmonary function parameters con-
sidered were consistently greater for the exposed miners,
none of the between-group differences were found to be sig-
nificant. Population matching was imperfect and resulted
in an older control group with a longer mean underground
exposure. While age and length of occupation were not found
to be directly associated with ventilatory function, a more
complex relationship could exist.
Gamble, et al. (11), examined the relationship between the
level of exposure to NC>2 and changes in pulmonary function
over the course of the work shift in from five salt mines
in the U.S. of which four used diesel equipment to varying
degrees. Levels of NC>2 were found to be directly related
to the change in pulmonary function for four of five pul-
monary function tests considered. The level of respirable
particles was not found to be associated with pulmonary
function changes. Battigelli, et al. (12), reported no
significant differences in pulmonary function or respiratory
complaints (coughing, phlegm, dyspnea, etc) between diesel
exposed railroad house workers and non exposed yard workers.
Smoking was not well controlled for and controls were found
to have a higher smoking index (packs/day x years smoked)
than did the exposed group. An additional potential source
1077
-------
of bias is the large number of unmatched exposed workers in-
cluded in analysis. No information is given as to the sim-
ilarity of these unmatched exposed workers to their unmatched
counterparts.
The long term significance of decrements in pulmonary function
and of non specific respiratory complaints is not clear.
Gamble, et al. (11), suggest that decrements in pulmonary
function may be early indicators of increased susceptibility
for irreversible lung pathology. A longitudinal follow-up
of these studies would be of great value in elucidating this
question.
In summary, the existing literature indicates:
1. No evidence of excess cancer mortality due to exposure to
diesel fuel exhaust. However, existing studies must be
interpreted with caution in light of methodological limi-
tations specific to each study. In fact, since the three
lung cancer published studies (1)(2)(3) do not contain
sufficient information on diesel exhaust exposure char-
acterization, differential patterns of smoking history,
as well as having latent periods often much shorter than
25 years, it is not possible to adequately assess the
hypothesis that diesel exhaust is an etiologic agent in
respiratory cancer.
2. Insufficient evidence to evaluate the relationship between
exposure to diesel exhaust and other lung disorders, in-
cluding bronchitis and respiratory infection.
3. Some evidence to suggest that exposure to diesel exhaust
may be associated with reduced pulmonary function. The
long term significance of this reduction has not yet been
evaluated.
Clearly, more research is needed to investigate the potential
long term and immediate health effects of exposure to diesel
exhaust in the occupational environment. Additionally,
research must be conducted to investigate the potential
health effects of exposure to diesel exhaust in ambient air
on the general public. Previous studies have dealt with
occupational groups in which workers tend to be self-select-
ing and therefore may not be at comparable risk to adverse
health effects than the general public. Before accurate
risk assessments are derived, those groups likely to be at
high risk to the potential adverse health effects of diesel
exhause must be identified and fully evaluated.
Following is a summary table of the human health effects
studies which illustrates study design, exposures, health
endpoints examined, and results, as well as comments on
the study strengths and weaknesses. In addition, the de-
tailed assessment of these studies, a 70 page document, will
be available through EPA.
1078
-------
I-
co
CO
r-
o
LU
LU
LU
LJJ
I
H
LJJ
I
.ure group may
lave received
iigrnf icantly
0
$ £ E
QJ O W
X *4- 1-
°?S
T3 ^
C 4-1 Ol
5 t E
C. !- 3
330.
U 0 .-
01 D- £>
.C C
4-> S 0>
c i—
s- -o c
fO (/> J-l
i- <— L.
J>
£ °-£
if S .
QJ Lr> 5
DI 1- O
nj QJ i —
L -M i —
QJ a
1 S&
E tj -M
Q.-O 4-
3 •-
i •£ >^
O 4-1
^ 4J ^
i— C i—
U. r- +j .
.^-^0
^_ Q.
1
4-1 0 OJ
"S £ i- r^
for smoking
exposure group
ib smoking res-
ns could reducf
• "c i -C un
W 0 4-1 C
>> ^ i- U
m ..^ ro ro L
f
» 1
QJ ^
E° ..
cancer r
exposure
cT"§i
3 r-
C
^£
fi,
>1-JD
"§ 0
4-1 4-1
s1
c -u
-o
.^^
1 CL
« ° 1
J= >, cn
^ O
s2 S S"
01 •-
ra
>i Ol 1.
- >
ro £-
^g 3
QJ
§
s_
T3
i
"oi •
£ "§
QJ -^
isis for evalu-
;ociation of ex
sorted outcomes
•a
i- Cri L
+-= c ^
O 4J o
i
•o si
sst
.uS-u
SS 0
1 r- "O
^5
cn >^
i 1-
"j?l
£" &
I- »* I- OJ
4-> CM O; .il
0 E
o 4-> a> >j
— "• *; o ^
S
Sr^- CJ-M
§g£fc|
i " u" ? S
o c: QJ ji
-o
?^§
1 Is
i 0 *"
o c
Z u m
•^•^ X
£
t. 4-1
|| |
55U
S °5
c t
cn
>* 1-
3
4-1 I-
s s
c
i- -M C E U
a. ra £ 5 S "§>
I
LJ_
O
DC
' ^
CO
p- 0 5 Dl
1079
-------
o
O
UJ
Q
CO
O
LLJ
LL
U-
LLJ
T.
UJ
I
O
DC
+Jt
§ ^ £
-— U O — TO £! ^ O 4- O <
OJ ^ i I
U f- 4J O> O
o .— oj c c x i—
a; i- o. E
i- 4- e 01
ji w ro o -
XTOt-iiU^tU S-.E
-
~
'a; i- c
I o !
If "£ £
1--— »i — "oJO-
^0 > QJ 0
) - i
in.— a;
) ro _ii ^ "0 .x i
: X D U D D !
) QJ 3 O 4^ 3 (
1080
-------
1 — 1
•o
J
o
0
CO
LU
Q
D
1 —
CO
CO
1—
o
UJ
U-
U_
UJ
X
t-
_J
UJ
X
~l~
u_
0
oc
1
CO
1
1
f
^
c
5
0
i
•7^
O
I*
O >
X
C
D i-
si
o
t^
4-
0
1 II
= «!"
01 oi .
•- s- -c >,
ti en "3
C 4-> "rD
4-1 •- 4-J
Significantly more
exposed miners
reported symptoms
of persistent and
?;
>> ^ 1
C ^ X
S 0 "
3 U r-
sill
i- O
JJIJ^
^- —
o^ Sa"c
"c ^ " c,
o"°Sfc
l^jl
"C <1J Z3 1
3 O -C "O
Q. X QJ
' S °* O
T3| 0) CL
S> U ^ CU
OflJj: i
ClJ Cl U !=
i?
o3 c "5 >,
t
0
°
|| »
3 S- O -M QJ
"c -1+!- ,r— -P 3 X O C C
Q.T3I-OJI---OI 4-J't- XCOOJ
X O> E r- r— 4-> E Cl/IUJ ^1-
^Iplll lsld:sl
!i||l!{!ili|!1J
o^ o "^g- Ol •C3i~>l1^ u
oo.oioj>)J-c:oi-i-oc 01 i__c_i,:
I-JCH-E: i. >, 3 ^3 foOJ o_c 4-j£iu
U >,OiC_^i- CQJOIl-3Cn. C"3 O
ScrtJCEorcC!-OOC,COr-QX
§ ^
^ O OI
£ £5
i- C
i- 0- c
§§5|
<*1
1 >1
T3 ' 3 T3 "O
I!fl§iil-:
COJ E-i? E SS-HJ"1 i^
ui i: (— i-
g -o -a E ro ^ cl-*- *o 8
.— 4-J
O 3
C-C 4J
C j- T3 C C
4-> 3 3 I- OI O
§ QJ S_ V.
c -a L-* £
O "D r— C OJ 3
r- C O 3 > V
& 4- « 4-
1_ S- 1- Oi 0
f r- *0 0 0 C
r— Q. C
35 ^ ° O 4^
o O i/> o ro
IA I- 4-J 4-J OI
1 sl§ IT;
n3 > »- QJ C
.— 1- tu O QJ 4^
_C *
£ °t
^i!
a^«
c? u u
III
o.
OJ p E
0 -E O 3
i. QJ Q. c
1 <" g- ^J *~> g
•a ( ui OJ u oi Q.
OJ^ > L E I
XJO 5 04- 0
1 "O
Si QJ _C ** r— 4-1
i- r- fD J- O 3
cj s: E 01 <_i o
0
15
"1 "5 S^
-J"S^
01 C E «
"SJ^ >
+J |— r- -O
ro 01 4-> i—
£ S 1||
U 3 T3 I- 3
U **~ "3 ^ 0
f- >, O +J D- .
-0 i- u o x -o
QJ 4-1 C T3 L.
j3 fO O E 4- QJ
i— -i- ti O T3
O i- 4-^ ^
C •- QJ O) E
•o m o ^- u o
3 in E -o
for the exposed
workers, none of
the between group
differences was
significant.
^Ji "i
i~ in u * o in r- U
S i- IB 3 Q. O
— "n3 £ *~
SO C -4-1 U
-t> O l^ S- i-
T3 4-1 OI O.
TO C
CL c: OJ 01
3 4- S U • QJ
O O O C -Q
t ^ >, 0
P 4-J 4-J >
QJ 1_ QJ ,
O_ rtJ <— 4- ro
^ f. E rtJ F— "^
£7- 34-1
ro T3 01 a C
r- T3 (fl
-------
0)
=3
O
UJ
Q
D
;§•§ op-t
S o; »«- u •— i/> 014-1 (
T," 5! £ & g. . H- "I4
oiio 01 XT3 OITJI_C
i Q, S
^ ^ "~ £ a) i! ^ "^ 4? "^ Q] S "^ 4-) ^u+j^c
•aroci-oO-ci— CL >^cc CS_'CDC<
OS-QJC:4->>-4-J'OI-4J CO ^4-^C(
14- 01 is "*~ <- o ai ro"Dro >, <»- ^- T3 m s
S^!<
QJ flJ 4) X W 0)
Ctl 4-> L Q; c T3
CO
H
O
LU
u_
LL
LU
= .i
3 E P-4- 3 T3 'o.CEOirtl --UOjii
4JX3 W°CW)-C 1- -O
7s • r. 0+JU--0 c:
^iS
CO
UJ
I
LL
O
>-
CE
^ c"g
L) c: u >,^ ^
3 £J E QJ 00 c:
;! »£•
C ><"-b "o
ill
>,?S I fc
r- _^ C ^
IIS 0?
se?
U Oi S
1082
-------
References
1. Raffle, P. 1957. The health of the worker. British
Journal of Industrial Medicine, 14:73-77.
2. Kaplan, I. 1959. Relationship of moving gases to car-
cinoma of the lung in railroad workers. Journal of
the American Medical Association, 171:2039-2043.
3. Waxweiler, R., J. Wagoner, and V. Archer. 1973. Mortal'-
ity of potash workers. Journal of Occupational Medicine,
15(6):486-489.
4. Mage, D. 1978. Visit to investigate the background of
the epidemiological study on the health of diesel bus
workers. (Received from EPA, 26 West St. Clair St.,
Cincinnati, Ohio.)
5. Commins, B., R. Waller, and P. Lawther. 1957. Air pol-
lution in diesel bus garages. British Journal of
Industrial Medicine, 14:232-239.
6. El Batawi, M. and M. Noweir. 1966. Health problems re-
sulting from prolonged exposure to air pollution in
diesel bus garages. Industrial Health, 4:1-10.
7. Jorgensen, H. and A. Svensson. 1973. Studies on pul-
monary function and respiratory tract symptoms of
workers in an iron ore mine where diesel trucks are
used underground. Journal of Occupational Medicine,
12(9):348-354.
8. Reger, R. and J. Hancock. Coal miners exposed to diesel
exhaust emissions. Appalachian Laboratory for Occupa-
tions! Safety and Health. National Institute for
Occupational Health and Safety. Morgantown, West
Virginia, 26505.
9. Morgan, W., C. Keith, D. Burgess, G. Jacobsen, R. O'Brien,
E. Pendergrass, R. Reger, and E. Shaub. 1973. The
prevalence of coal workers pneumoconiosis in U.S. coal
miners. Archives of Environmental Health, 27:221-226.
10. Reger, R., J. Hankinson, and J. Merchant. Ventilatory
function changes over a work shift for coal miners
exposed to diesel emissions. Appalachian Laboratory
for Occupational Safety and Health. National Institute
for Occupational Health and Safety. Morgantown, West
Virginia, 26505. In: Proceedings of the First NIOSH
Symposium, Cincinnati, Ohio, 1978.
11. Gamble, J., W. Jones, J. Hudak, and J. Merchant. Acute
changes in pulmonary function in salt mines. (Received
from EPA, 26 West St. Clair St., Cincinnati, Ohio,
June, 1979).
12. Battigelli, M., R. Mannella, and T. Halch. 1964. Envi-'
ronmental and clinical investigation of workmen exposed
to diesel exhaust in railroad engine houses. Industri-
al Medicine, 3:121-124.
1083
-------
General Discussion
J. VOSTAL: You have mentioned that you would like to
differentiate between occupational exposures and nonoccu-
pational exposures, and you have stated that it could be
influenced by a self-selection process by the workers.
Since the most important question we are addressing con-
cerns the lung, could you be more specific as to how
anyone could be preselected to lung cancer susceptibil-
ity.
E. CALABRESE: I am not sure that the most important
question is that of respiratory cancer. From my own
prospective, respiratory cancer is only one factor to
consider along with a whole host of respiratory dis-
orders, in terms of the effects of diesel exhaust on
individuals, including bronchitis, etc. At the present
time I think these should enjoy an equally important
concern. With regard to a self-selected group for detec-
tion of respiratory cancer, that would be a difficult, if
not impossible, situation. Regarding occupational epi-
demiological studies, the concept of the healthy worker
effect results from either self-selection or selection by
the occupational physician at employment. In the latter
case individuals looking for a job might be turned away
because they may be carrying an identifiable type of
disease process. Preselection may be very effective for
respiratory disorders or some other type of illness but
usually it is more apparent with regard to death by cir-
culatory disorders. The healthy worker effect is not
usually as large with respect to cancer, but can be a
very important factor in regard to noncancer illness.
1084
-------
TRENDS IN LUNG CANCER IN LONDON
IN RELATION TO EXPOSURE TO DIESEL FUMES
R. E. Waller
MRC Toxicology Unit
Clinical Section
St. Bartholomews Medical College
London
ABSTRACT
Mortality from lung cancer among males in England and Wales
is now beginning to decline, and the urban excess that has
persisted for many years is also becoming smaller. There is
nothing in these . 'onal trends to suggest that pollution
from diesel vehicle.,, which has been increasing in recent
decades, would be involved. Some preliminary results are
presented from a long-term study of the incidence of lung
cancer among London Transport staff, including men who work
in bus garages where there are high concentrations of smoke
from diesel buses. Over the 25 year period from 1950-74,
the numbers of cases reported in each of the several job
categories have been below those expected on the basis of
greater London rates. The absence of data on smoking habits
makes it difficult to determine whether the small differ-
ences in rates between job categories are of any importance,
but the standardized mortality ratios are all well within
the range of those found in national studies of lung cancer
mortality and smoking in relation to occupation.
For many years there has been concern about the possible
influence of exposure to fumes from motor vehicles on the
incidence of lung cancer. This arose partly as a result of
the demonstration of carcinogenic comounds such as benzo(a)-
pyrene in urban air (Waller, 1952) and specifically in
1085
-------
emissions from petrol and diesel engines (Kotin et al.,
1954, 1955). In the case of diesel engines, emissions of
such compounds are applicable only under poor operating
conditions, when there is much smoke, and it is clear that
at least prior to the implementation of the Clean Air Act,
coal fires were more important sources of benzo(a)pyrene in
British towns than motor vehicles (Lawther and Waller,
1976). With the gradual elimination of coal smoke in London
(and in most other major cities) over the past 25 years,
concentrations of benzo(a)pyrene in the air have fallen
dramatically, as illustrated in Table 1.
It is difficult to determine from an examination of trends
in lung cancer mortality in the general population whether
any component of air pollution may be involved in the
aetiology. Figure 1 shows the age-specific death rates for
lung cancer among males in England and Wales arranged
according to date of birth. Rates at each age increased
steeply in relation to date of birth up to the beginning of
the century, but men born since about 1911 have been exper-
iencing slightly lower death rates than their predecessors.
While the substantial fall in smoke and benzo(a)pyrene
concentrations in the air might have had some oearing on
the recent decline in death rates, it seems unlikely that
exposures to any component of air pollution could account
for the massive increase that occurred earlier. Certainly
there is nothing in these trends to implicate exposure to
diesel fumes in particular, for the rise in death rates
began before these could have had any impact and the decline
started when concentrations would have been increasing
rather than decreasing.
It is of course clear from the many epidemiological studies
that have now been done (e.g., Doll and Peto, 1976) that
smoking is the dominant factor in the aetiology of lung
cancer, and the trends displayed in Figure 1 can be accoun-
ted for largely by the switch from pipe to cigarette smoking
amongmen earlier this century, with the downward turn in the
curves reflecting effects of reductions in smoking in more
recent years coupled with a change towards filter cigarettes
with a lower tar yield.
A better impression of the possible effect of air pollution
can be obtained by examining trends in urban and rural areas
separately. Table 2 shows part of a large table that has
been prepared, giving sex-age specific death rates for lung
cancer in Greater London and in rural areas of the country.
The main feature is that while death rates in London have
always been higher than those in rural areas, the gap
between them is closing gradually. These trends also may be
dominated by changes in smoking habits as the latter become
1086
-------
LUNG CANCER
AGE-SPECIFIC DEATH RATES
KXXDi
1
ENGLAND and WALES
MALES
Ifc
60-
55-
81 91 I9O
YEAR OF BIRTH
21 31
Figure 1. Trends in age-specific death rates for lung
cancer (males) in England and Wales, arranged
in relation to year of birth.
1087
-------
Table 1
CONCENTRATIONS OF BENZO(A)PYRENE IN AIR AT SITES IN
CENTRAL LONDON, 1949-73. BASED ON 24 HOUR SAMPLES
AGGREGATED FOR YEARLY PERIODS
Period
1949-51
1953-56
1957-64
1972-73
Sampling site
County Hall
St. Bartholomew's Hospital
County Hall
St. Bartholomew's Hospital
Medical College
Benzo(a)pyrene
g/1000 m3
46
17
14
4
Table 2
LUNG CANCER: AGE SPECIFIC DEATH RATES. MALES AGED 60-64.
GREATER
LONDON AND RURAL
DISTRICTS OF ENGLAND
AND WALES
Year of
Death
1951-55
1956-60
1961-65
1966-70
1973
Year of
Birth
1891
1896
1901
1906
1911
Deaths per
Greater
London
350
420
450
405
370
100,000
Rural
Districts
170
240
270
295
290
These rates are approximate, being interpolated from tab-
ulations in broader age-ranges. Figures for the single
year 1973 are given as an indication of those in the most
recent quinquennium, and these are subject to revision.
1088
-------
more uniform between town and country, but the downward
trend in London could be related to the decline in pollution
there. Clearly there is no indication from these trends
that pollution from diesel engines, that is liable to have
increased rather than decreasedin recent decades, has had
any perceptible effect on lung cancer mortality in the
general population.
To consider possible effects of pollution arising speci-
fically from diesel engines it is necessary to turn to
occupational groups with enhanced exposures. Men who work
in garages when diesel-engined bases are parked and serviced
present one opportunity, and to follow that up, Raffle
(1957) examined lung cancer incidence among several cate-
gories of London Transport employees during the five-year
period 1950-54. It was shown then that the incidence among
the diesel bus garage workers was no higher than that
expected among the general population. The duration of
exposure of the men involved to diesel fumes might however
have been quite short up to that time, since diesel buses
were only introduced in the 1930's, gradually replacing
petrol-engined vehicles until, in the 1950's they finally
replaced trarns and trolley-buses also. It is however now
possible to consider the situation over a much longer
period, since records of lung cancer cases arising during
service among men in five categories (bus drivers, conduc-
tors, engineers in bus garages, engineers in the central
works, and motormen or guards on the Underground) have been
maintained continuously since 1950. Some preliminary
results from this simple incidence-rate study for the 25
year period up to 1974 are presented below, and the analysis
will be completed when data up to 1979 have been assembled,
to make 30 years in all.
The study is limited to men in the age-range 45-64, and the
numbers employed year by year in the five job categories
considered are shown in Figure 2. In the early years there
were some 20,000 men in all, but reductions in the number of
vehicles in service, and changes in operating procedures,
have led to a gradual reduction in the numbers employed in
connection with buses. The change in riumers has been
accompanied by a changed in age-structure, and it has been
essential therefore to define the population at risk in
narrow (5 year) age ranges ready for trie subsequent calcu-
lation of expected numbers of lung cancer cases.
The lung cancer cases comprised those recorded on death
while still a member of staff within one of the designated
job categories, plus transfers to alternative work within
London Transport or ill-healtn retirement following diag-
nosis of the disease. In this way all cases arising during
1089
-------
LONDON TRANSPORT
MALES 45-64
STAFF NUMBERS
I95O -74
8OOO -
6OOO
u.
u.
fe
O 4OOO
o
Z
2OOO
TOTAL MAN-YEARS
!25Yr. PERIOD) 42O.7OO
ENGINEERS (WORKS)
J I
1955
I960
1965
I97O
1975
Figure 2. Numbers of staff in five job categories in London
Transport, 1950-74.
1090
-------
service could be related strictly to the population at risk
in the appropriate category. Ascertainment was believed to
be complete, within the limitations of normal diagnosisand
death certification, but some care was needed to ensure
that cases involving transfer to alternative work were not
counted a second time on death.
Ideally, incidence rates in the general population built up
in a similar manner from first diagnosis or death records
would be required for comparison purposes. Such data are
available from cancer registries for recent years, but in a
study extending back to 1950 it was considered more satis-
factory to use death records only. In a disease with such a
poor survival rate as lung cancer, the difference between
these two bases is not substantial, and the question of any
bias introduced by limiting general population comparison
data to death records only is discussed further below.
Detailed tabulations of death rates in the general popula-
tion had been prepared as part of a wider study of trends in
lung cancer mortality in urban and ruraly areas of England
and Wales, illustrated in Figure 1 and Table 2 above. Most
of the Lorid Transport staff lived within the Greater London
area, and expected deaths were therefore calculated by
applying Greater London rates to the population at risk
within each job category. The data were assembled in 5 year
periods and 5 year age ranges, finally aggregating results
to obtain the total expected deaths in the age range 45-64
for the period 1950-74.
There were in all 667 cases of lung cancer reported over the
25 year period. The majority (81%) were notified on death,
and the proportion of other cases (transfers to alternative
work, or ill-health retirement) was greatest for bus drivers,
possibly as a reflection of the greater need to remove them
from their existing job on initial diagnosis of the disease.
The- numbers of cases occurring within each of the job
categories are shown in Table 3, along with the numbers that
would be expected on the basis of Greater London death
rates. In each category the observed numbers are below
those expected, and this may in part be a reflection of the
"healthy worker" effect that has been seen in other studies
of occupational groups subject to some degree of medical
screening on entering employment. This is not however
likely to be an important factor in the case of a disease
such as lung cancer, that does not manifest itself at an
early stage.
While the highest mortality ratio is that among bus garage
workers, it does not differ significantly from that among
rnotormen and guards on the Underground, who do not have
1091
-------
Table 3
LUNG CANCER CASES AMONG LONDON TRANSPORT STAFF
IN RELATION TO THE NUMBER EXPECTED ON THE BASIS OF
GREATER LONDON DEATH-RATES
1950-74, MALES AGED 45-64 ONLY
Job
Category
Bus Drivers
Bus Conductors
Engineers,
Garages
Engineers,
Central Works
Motormen and
Guards
Total
Man-Years
at Risk
175,909
93,095
86,054
30,031
35,610
420,699
Expected
Deaths
346.8
174.5
197.1
63.2
67.7
849.2
Observed
Cases
259
130
177
42
59
667
Mortality
Ratio, %
75
75
90
66
87
79
1092
-------
any special exposure to diesel fumes. The major problem in
interpreting the differences in mortality ratios between the
several job categories is however that the smoking habits of
the men are not known. The study is based only on existing
staff records, and there has not been any opportunity to
collect personal information from the men.
The results can perhaps bets be put in perspective by
linking them with the data on lung cancer and smoking
published recently by the Office of Population Censuses and
Surveys (1978). It has been shown that in England and Wales
as a whole, lung cancer death rates within broad occupa-
tional categories are correlated with the proportion of
(current) smokers 'in those categories. The findings are
reproduced in Figure 3, identifying the point representing
transport workers in the general population, and then
superimposing the data from the present London Transport
study. For this purpose the lung cancer cases have been
expressed as percentages of the numbers expected i nthe
population of England and Wales rather than Greater London,
to allow direct comparison with the national data in terms
of SMR's (Standardized Mortality Ratios). Clearly there are
limitations to this approach, including the different time
spans involved in the London Transport and the national
data, but sine there has been no substantial change through
the years in the ratio of observed to expected cases in the
London Transport series, then the comparisons should be
valid. The conclusion to be drawn is that the London
Transport data would be consistent with the national figures
if the proportion of smokers among garage workesr and the
motormen and guards was similar to that in the general
population, and below average in the other groups. None of
the London Transport lung cancer rates is as high as that
among transport workers in general. The inclusion of a
small proportion of cases at diagnosis rather than at death
in the London Transport series should not make any appre-
ciable difference to the SMR's, although if some of those
men lived on for several years (perhaps taking them beyond
age 64) then the tendency would be to overestimate the
London Transport lung cancer rates in relation to those
of the general population.
A limited amount of environmental sampling has also been
done in connection with this study. At the time of the
original work by Raffle (1957), observations of smoke
and the associated polycyclic hydrocarbons were made at two
representative bus garages in London (Commins _et _al_., 1957).
The conclusion then was that while concentrations of smoke
were substantially enhanced in the garages, the amounts of
polycyclic hydrocarbons present were in general dominated
by outside sources (principally coal fires), and it was
necessary to select warm days in the summer to detect the
effect of emissions from the buses.
1093
-------
LUNG CANCER MORTALITY & SMOKING BY OCCUPATION
200 -
CM
1^
I
O
r-
o>
=. \50\-
OC.
10
O
z
U
O
IOO
5O
LONDON
TRANSPORT
SMR's
ENGINEERS,GARAGES >
-MOTORMEN/GUARDS
CONDUCTORS
DRIVERS
. CENTRAL WORKS
• • A TRANSPORT
WORKERS
ENGLAND
_L
_L
O 50 IOO ISO
PROPORTIONAL SMOKING RATIO (1972)
Figure 3. Lung cancer mortality and smoking by occupation,
England and Wales, males 1970-72.
1094
-------
In view of the changes that have taken place during the past
20 years, both in terms of the control of emissions from
coal fires and of the types of buses in services, further
sampling has been done recently int he same two garages as
before. Concentrations of smoke and polycyclic hydrocarbons
have again been determined, and Figure 4 illustrates find-
ings from just one section of the results, for benzo(a)-
pyrene at Dalston Garage. This shows results for a single
day, divided up into periods of several hours at a time,
corresponding with different phases of activity in the
garage. The lower part of the diagram relates to the recent
measurements, and corresponding data for 1957 are included
above for comparison. It can be seen that the background
concentrations of benzo(a)pyrene in the outside air are now
very much smaller than they were, and the contributions from
the buses (represented by differences between the unshaded
and shaded columns) are also mainly smaller. The present
concentrations are clearly very small in comparison with
those that used to exist in the general air of London
(particularly during the winter months) some 20 or more
years ago, and the indications are that the overall exposure
of garage workers to benzo(a)pyrene during their working
lives would not differ much from those of the general
population. There is no doubt though that their exposure to
smoke would be greater than that of the general population,
since concentrations have at all times been higher in the
garages than outside. It may be that there are other
constitutents of the smoke with potential carcinogenic
activity, but the epiderniological data reported above o not
indicate any increased risk of lung cancer that may be
attributed in any way to the diesel fumes.
Despite these negative findings, arrangements are in hand to
continue and extend the present investigation, and for some
groups of workers it will be possible to carry out a cohort
study, in which the men will be followed up to death irres-
pective of whether they remain in employment with London
Transport.
1095
-------
3O -
25
20
- 15
Z 10
o
IS|
I
INSIDE GARAGE
OUTSIDE (ROOF)
JUNE 1957
JULY 1979
13-18
p p
18-23 23-1
P
1-5
Q
5-9
P SITE
9-13 TIME
Figure 4. Concentration of benzo(a)pyrene in Dalston
garage: 1979 compared with 1957. Site P was
at the main entrance to the garage and Site Q
at one end, close to parked buses.
1096
-------
REFERENCES
Commins, B. T., Waller, R. E. and Lawther, P. J. (1957).
Air pollution in diesel bus garages. British Journal
of Industrial Medicine, U_, 232-239.
Doll, R. and Peto, R. (1976). Mortality in relation to
smoking: 20 years' observations on male British
doctors. British Journal of Medicine, _2, 1525-1537.
Kotin, P., Fall, H. L., and Thomas, M. (1954). Aromatic
hydrocarbons II. Presence in the particulate phase of
gasoline engine exhausts and carcinogenicity of exhaust
extracts. AMA Archives of Industrial Hygiene and
Occupational Medicine, _9> 164-177.
Kotin, P., Falk, H. L., and Thomas, M. (1955). Aromatic
hydrocarbons III. Presence in the particulate phase of
diesel-engine exhausts and the carcinogenicity of
exhaust extracts. AMA Archives of Industrial Hygiene
and Occupational Medicine, U_, 113-120.
Lawther, P. J. and Waller, R. E. (1976). Coal fires,
industrial emissions and motor vehicles as sources of
environmental carcinogens. INSERM Symposium Series,
50, 27-40.
Office of Population Censuses and Surveys (1978). Decennial
Supplement, 1970-72. Occupational mortality. Series
DS No. 1. H. M. Stationery Office, London.
Raffle, P. A. B. (1957). The health of the worker.
British Journal of Industrial Medicine, 14, 73-80.
Waller, R. E. (1952). The benzpyrene content of town air.
British Journal of Cancer, 6, 8-21.
1097
-------
General Discussion
L. JOHNSON: At the beginning of your presentation you
showed a dramatic decrease in black smoke in London and
attributed the remainder to diesels. Do you have any
idea of the proportion contributed by light-duty, heavy-
duty,- and stationary diesels?
R. Waller: No, much of the pollution in the early
years could not have been contributed by the diesel, but
it is only in the most recent times, since about the
early '60's, that diesel pollution has become signifi-
cant.
A. Kolber: I noticed an absence of morbidity data in
your studies. I have tried to do some work of a similar
nature and also find a great lack of morbidity informa-
tion in this country. I thought that morbidity informa-
tion was available in England though.
R. Waller: It is, yes. The study I presented, how-
ever, was just related to mortality. It is not quite
clear whether we can really rely'on this paper. We have
social security records too and these have been used.
One has seen differences in occupational groups in sick-
ness, but it is by no means certain that these really
reflect true health effects.
J. HANCOCK: Were these mortality causes all from
death certificates?
R. WALLER: All those presented here were obtained
from other sources. The death certificates are slightly
better than the rest. Some do not reflect the death
certificate but those carrying authorization have their
own card and when a diagnosis such as mine is made, or
anything that is related to it, they would show it. I
would say that the diagnoses are as good as, and possibly
better than, the general population at large.
F. SPEIZER: I appreciate why you are continuing your
study for another five years. There might also be other
ways of analyzing the data. This is a question that
epidemiologists run up against all the time, estimating
what the appropriate expected numbers are for a popula-
tion such as this. It is not totally clear to me that
the general mortality for the greater London area is the
appropriate expected value. If you try to use one of the
other working groups to estimate the expected values, it
looks to me as though you might get relative risks for
the engineers that might approach 1.5, 1.6.
1098
-------
R. WALLER: We have taken the whole and tried to look
at the data of that population and then looked at respective
values for all of the engineers and they are higher. You
will perhaps note that another factor is time and the
underground workers. As you can see, these are the least
likely to be exposed to pollution, and looking at the
trends over the years, we also tried to decide whether
any of the groups have been increasing more rapidly than
the general population or decreasing more rapidly. The
underground men seemed to have come up slightly, more
rapidly, than any of the others. It brings in all kinds
of other problems when you ask what are the comparison
groups. There is another problem and that is the ethnic
origins. They actually went out in the West Indies to
recruit workers and brought them back. This I think will
also pull their weight down because these other ethnic
groups have lower rates of lung cancer. We will then
have to consider how different from the general population
these are.
F. SPEIZER: That raises a further question in regard
to when some of these people started smoking and how
smoking affects some of these groups as well.
R. WALLER: I didn't go into detail. In fact, we have
not done sufficient tracer studies. I think the ethnic
groups have started smoking later than the others, and I
think some groups increased more rapidly because they
have caught up.
D. HOFFMANN: I understood it is fairly difficult, in
retrospect, to get the smoking data on these people.
Isn't it possible that there are some nonsmokers in this
country? When I collected data there was never a time
that more than 60 percent of the male population were
cigarette smokers. When you have your 600 lung cancer
cases, and I guess in England when you didn't smoke you
were somewhat unusual, isn't there a way in retrospect to
see how many nonsmokers were diseased?
R. WALLER: In the cases, or in the whole population?
D. HOFFMANN: No, in your 600 cases?
R. WALLER: Yes, it is possible. We have gone back to
the records. There are some notes in the history, so we
may make some point on this.
1099
-------
HUMAN DATA ASSOCIATED WITH
DIESEL EXHAUST
Dennis S. Lachtman
Director, Health Sciences
Envirotech Corporation
Menlo Park, California
ABSTRACT
A review of the literature concerning data rela-
tive to human exposures from diesel engines is
presented. Epidemiologic evaluations among
workers exposed to diesel exhaust are discussed.
Morbidity and mortality data are critically
reviewed. The outline of a mortality and morbid-
ity study to be completed in 1981 among construc-
tion workers exposed to diesel exhaust is given.
INTRODUCTION
Government and labor sources have speculated that
diesel exhaust exposures may have adverse health
consequences. It is widely recognized that the
number of components within diesel exhaust are
extremely numerous. The diverse emission stream
of diesel exhaust is similar in complexity to that
which we experience every day in the form of gaso-
line exhaust and urban smog.
The diversity of substances within the exhaust
stream from diesel engines, in combination with
the lack of comprehensive animal and human data
on the subject, make estimates of health effects
extremely difficult. Yet, circumstances dictate
that at certain periods evaluations and decisions
become necessary. At this point in time, the
following discussion will attempt to place a
perspective on the health implications of diesel
exhaust exposures.
1100
-------
HUMAN STUDIES
Although a number of studies have attempted to
evaluate the health effects of diesel exhaust on
animals and humans, none of these efforts have
looked at ambient exposure levels of diesel
exhaust as would be experienced by the general
population. The available human response data
from diesel exhaust exposures have either been
collected from occupational groups or from volun-
teers exposed to simulated occupational exposures.
Those constituents found in diesel exhaust that
are of primary health concern are oxides of
nitrogen, particulates and sulphur oxides. Since
control of sulphur oxide emissions is possible
by use of sulfur-free fuels, it might be consid-
ered that the agents of primary concern are
oxides of nitrogen and particulate matter. For
the sake of simplicity, particulate matter can
be categorized in terms of carcinogenic and non-
carcinogenic response. A great deal of informa-
tion has been focused on benzo(a)purene (BaP) as
an occupational and environmental carcinogen.
While it must be recognized that use of BaP as an
index for all polynuclear aromatic hydrocarbons
(PNA) leaves much to be desired, it remains as
one of the better indexes for PNAs.
Earlier reports (1, 2) have estimated that PNA
levels in many occupational settings, as indexed
by BaP, are roughly similar to those levels
found in ambient urban environments. If BaP can
be used as an index of the carcinogenic hazard of
diesel exhaust, then the following discussion of
occupational environments has relevance to ambient
diesel exhaust exposures.
The following will discuss health data from epidem-
iologic studies among diesel bus workers, diesel
railroad workers, underground coal miners exposed
to diesel exhaust and underground non-coal miners
working with diesel equipment.
The general categories for human studies involving
exposures to diesel exhaust emissions can be con-
veniently organized into two distinct categories.
1101
-------
The first category would involve those studies
that look at lung-function testing and evaluate
short-term symptoms. Data from medical question-
naires is used to determine subjective symptoms
of respiratory health status. The second category
involves mortality studies looking at death
certificates among occupational cohorts. These
studies are more useful in terms of evaluating
long-term diseases, including cancer.
SYMPTOMATOLOGY AND LUNG FUNCTION TESTS
In 1964, Battigelli et al. (3) reported on cohorts
of railroad workers with respect to occupational
exposure to diesel exhaust emissions. In this
study, there were 210 workers in the exposed
cohort having an average of 9.6 years of exposure
with a 154 worker control group. Results failed
to show a significant difference in pulmonary
function performances between the two groups.
There were observed differences which were signi-
ficant with respect to respiratory complaints
associated with smoking habits.
In one of the only investigations of its type,
Battigelli (4) exposed volunteers to diesel
exhaust emissions for short periods of time and
recorded pulmonary resistance. The exposure con-
centrations of the controlled diesel exhaust
exposures -were comparable to realistic values
found in railroad shops. The results found no
significant difference in pulmonary resistance
among the exposed and non-exposed volunteers.
A study by El Batawi and Noweir (5) in 1966
reported results of a questionnaire survey of
Egyptian workers in diesel bus garages. The
reported findings were based on subjective com-
plaints in the exposed population. Larger than
expected prevalence rates of upper respiratory
tract disease, chronic bronchitis, asthma, peptic
ulcers, gastritis and high blood pressure were
observed. The investigators reported that many
of the workers were heavy smokers. It was
suggested that the observed symptoms could be
easily attributed to excessive tobacco consumption
habits.
1102
-------
Jorgensen and Svensson (6) reported results of an
investigation of pulmonary function by categories
of smoking, age, and surface work versus under-
ground employment in Swedish iron ore mines.
Basic findings were that smoking and working
underground were synergistic with respect to
frequency of chronic productive bronchitis.
Symptoms appeared early among underground miners
who smoked; however, a falling off of symptom
frequency was observed with increasing length
of underground employment. Non-smokers under-
ground exhibited a frequency of chronic bron-
chitis that was similar to smokers employed on
the surface. The diminishing frequency of
respiratory tract infections in these underground
miners is inconsistent with a decrement in
immunologic competence. While the results of the
study demonstrated some age-related differences
with respect to pulmonary changes, the authors
stated that no significant differences were
observed that could be related to employment above
or below ground.
In 1969, Yamazaki et al. (7) reported on a study
evaluating the effects of diesel exhaust gases
on railway workers. Lung function equipment
included a spirometer and a peak flowmeter. A
total of 475 workers were observed. The control
group consisted of 67 individuals not receiving
diesel exhaust exposures during their work
routines. Questionnaires were used that included
data on the type of work site, age, job desig-
nation, air pollution, residence, allergic
factors, heart disease history, subjective
respiratory symptoms, tobacco habits, height and
weight. No reductions of pulmonary function were
found in excess of 10 percent for the five
respiratory function variables tested. Inspection
and repair sheds showed results that were greater
in magnitude in comparison to workers assigned to
tunnel areas. Work sites had greater observed
effects on pulmonary function results than smok-
ing. Data analyses utilized multi-variate
statistical techniques relying on quantification
theory. It was concluded that no significant
effects were observed as demonstrated by the
failure to obtain a greater than 10 percent reduc-
tion of pulmonary function.
1103
-------
In 1978, Attfield (8) reported on a respiratory
health survey taken in 21 metal and non-metal mines.
This study used a combination of lung function
tests and questionnaires based on those of the
Medical Research Council of Great Britain. A
total of 4,924 men were evaluated. The purpose
of this effort was to update a former study in
silica mines performed by the Public Health-
Service in 1958 through 1961 (9) and correlate
health effects to measures of diesel use and
silica exposure levels. The analysis evaluated
only white male miners. Several parameters of
exposure were evaluated. The variables included:
1. number of diesel machines in use;
2. years of underground exposure; and
3. exposure to NC>2 and aldehydes.
Symptoms of persistent cough, persistent phlegm
and shortness of breath did not demonstrate any
association between the amount of diesel units
in use and adverse health effects.
No consistent associations between three indices
of lung function (FVC, FEV, and FEF5Q) and diesel
use were found. The general picture presented by
the analysis is one of conflicting results.
There was an observed increase in lung function
with increasing length of diesel exposures.
While it is plausible that diesel exhaust was not
harmful in these mines, it was reported that
other possible reasons could account for these
negative findings. These negative findings may
have occurred for a number of reasons including
the observed wide variation in the length of
underground exposure, the potential for the indices
of pulmonary function to have been inappropriate,
and the possibility that the differences between
high and low exposure categories were not signifi-
cant.
Gamble, et al. (10) reported on a study that eval-
uated workers at five salt mines. Two of these
mines used large amounts of diesel equipment while
two mines used lesser amounts. One mine did not
utilize diesel equipment. The study was designed
to determine if diesel particulates plus nitrogen
dioxide exposures were associated with pulmonary
function changes over a work shift. A total of
1104
-------
246 miners were tested. Maximum forced expiratory
volume in one second (FEV) and forced vital
capacity (FVC) values were used to obtain peak
flow and flows at 257., 507= and 7570 (FEF25, FEF5Q,
FEF75) of FVC. The analysis of these data
demonstrated a decrement in pulmonary function
associated with nitrogen dioxide exposures. No
decrement in pulmonary function was observed with
respect to diesel particulates.
In 1979, Reger and Hancock (11) reported on a
study that evaluated miners working in coal mines
using diesel-powered machinery. This study,
which used matched pair analysis, evaluated pul-
monary function tests and questionnaires. Data
from miners exposed to diesel exhaust were matched
to miners working in coal mines not using diesel
machinery. The data for the miners in the five
coal mines using diesel equipment were collected
in 1978. The data for those miners working in
coal mines not using diesel machinery were derived
from the (1973-1974) second round of the National
Coalworkers Pneumoconiosis Study. The response
rate for the exposed group was 92 percent. The
response rate was not reported for the control
group. Selection from the control group to con-
stitute the "matched pair" for each exposed miner
was done to correlate the variables of geo-
graphic area, smoking status, weight, height and
years underground. In all, 722 pairs were
evaluated. Incidence of cough and phlegm symp-
toms were greater for the miners exposed to
diesel exhaust than their non-exposed counter-
parts. The miners exposed to diesel exhaust,
however, had significantly fewer complaints of
moderate to severe dyspnea and wheezing compared
to their non-exposed counterparts. Results of
pulmonary function tests were mixed.
The mean difference in pulmonary function vari-
ables including FEV, FVC, FEFrQ and FEF-,5 indicate
on the average that the diesel-exposed group had
statistically lower pulmonary function than
controls. Conversely, the diesel-exposed miners
had improved pulmonary function that was statisti-
cally significant (unpublished report - 12) in
terms of peak flow in comparison to controls.
1105
-------
Some interesting associations were observed when
the pulmonary function data was segregated by
individual mines. Among those miners evaluated
in Mine 4, improved lung function differences for
FVC and FEV were observed in comparison to their
matched controls. On the average, the men of
Mine 4 were reported to have longer exposures
to diesel emissions in comparison to the other
diesel mines evaluated. Differences in pulmonary
function for Mine 3 were markedly greater than
for any of the other mines. Of the five pulmon-
ary function variables tested, decrements in
pulmonary function among the miners exposed to
diesel exhaust were found to be statistically
significant for three variables in Mines 1 and 3,
and a corresponding decrement for two variables
was observed in Mine 2. Of the five pulmonary
function variables evaluated for the matched
pair differences in Mines 4 and 5, one was found
to be statistically significant (Peak Flow,
Mine 5) in terms of improved pulmonary function
(unpublished report - 13) among the diesel-
exposed mines and another variable (FEF-,,- -
Mine 5) was statistically significant for a
decrement in pulmonary function. The eight
remaining differences among the matched pairs in
Mines 4 and 5 were statistically insignificant.
The lung function and symptomatology data lack
sufficient consistency to allow any conclusion
regarding the health effects of diesel exhaust.
MORTALITY STUDIES
In 1957, Kaplan reported on a mortality study on
railroad workers for the years 1953-58 inclusive
(14). The workers were divided into separate
groups: operating personnel exposed to exhausts
from diesel and steam engines (e.g. , trainmen)
and non-exposed workers (e.g., office personnel).
The study comprised 235,110 person-years. A
total 6,506 deaths in the study population were
reported; only 154 were attributed to lung cancer,
although 192 would have been expected on the basis
of age-specific rates. There was no increase in
lung cancer mortality in any of the groups studied.
Lung cancer among London transport workers servic-
ing bus equipment was reported by Raffle in 1956
(15). This study found no excess of lung cancer
1106
-------
'"among any group of London transport staff such
as would have been expected if diesel exhaust
were a serious contributory factor in producing
an excess of lung cancer in urban dwellers".
In England a recent follow-up study has been
reported through the auspices of EPA which accounts
for 25 years of exposures by diesel transport
workers in London bus garages (16). The lung
cancer rate for the garage workers was not statis-
tically different from that of any other group of
employees, such as underground train drivers and
guards. Additionally, the lung cancer rate was
less than that found among the general population.
An epidemiologic investigation performed by NIOSH
evaluating the mortality experience of potash
miners and millers was authored by Waxweiler,
Wagoner and Archer in 1973 (17). Although the
study was principally concerned with causes of
mortality (especially lung cancer) related to
the question of individual constitutional pre-
disposition, two of the eight mines investigated
had long histories of diesel engine usage for
underground transportation purposes. One mine
had used diesels for approximately 24 years and
the other for approximately 16 years at the time
of the study.
The investigators subdivided the underground
cohort according to which men had worked in
either of the two "dieselized" mines and when.
On the basis of the mortality experience of the
total cohort of underground miners, 46.7 deaths
would have been expected in the dieselized
population of 6,733 person-years when, in
actuality, only 31 deaths occurred. This NIOSH
study concluded that " ... no major cause of
death exceeded expectations among men who had
worked in the diesel-using mines." In addition,
the study concluded:
"... No causes of death were signifi-
cantly different between miners who
worked in dieselized mines and those
who worked in other mines. It may be
noteworthy that the 'other respiratory
disease' category, which was high
among underground workers, was not
different between diesel and non-diesel
workers."
1107
-------
The investigators stated that this latter find-
ing may be related to the "insufficient elapsed
time since the start of diesel usage for chronic
or long latent-period diseases, such as emphysema
or lung cancer, to be manifested as excess
deaths in the relatively small exposed group."
However, it should be emphasized that diesels
had been in use in the two mines for approximately
16 and 24 years, respectively. This study repres-
ents the only mortality investigation published
in the scientific literature that was conducted
in a dieselized mine in the U.S.
HEALTH RESEARCH STATUS
To date, results from health studies on diesel
exhaust exposures have been inconclusive. The
human health literature has not shown any
evidence that exposures from diesel engines can
be associated with adverse chronic health effects
such as cancer or other long-term illnesses.
Future studies involving diesel exhaust with
longer exposure periods to account for the twenty-
to-thirty year latency period expected for
chronic diseases such as cancer would help
alleviate some of the present uncertainties in
the assessment of the health hazards from diesel
exhaust emissions.
It appears that two types of epidemiologic
studies are needed to provide an improved health
hazard assessment of diesel exhaust exposures
among humans. The first type of study would be a
cross/sectional morbidity study among a popula-
tion exposed to diesel exhaust. This form of
survey could be based upon an examination of an
occupational population. Properly devised
sampling techniques evaluating the factors of
age, race, geographical location, job location,
and other variables effecting exposures would be
necessary to assure both representativeness of the
sample and delineation of distinct exposure
categories. Morbidity studies have limitations
in terms of extrapolating results to chronic
illness. In the case of diesel exhaust, such
studies would help evaluate the acute effects on
respiratory and other organ systems effected or
associated with diesel exhaust exposures.
1108
-------
Chronic health hazards including cancer endpoints
can be evaluated by use of mortality data. Al-
though a prospective epidemiologic approach would
follow an exposed cohort for thirty years and
could offer a more ideal study design, time con-
straints limit such approaches.
HEAVY EQUIPMENT OPERATORS
Currently in progress is an epidemiologic study
designed to evaluate mortality trends and cancer
incidence among an occupational cohort exposed
to diesel exhaust. This study, which is scheduled
for completion by 1982 (personal communication -
18), will evaluate the medical records from a
Northern California union of heavy equipment
operators. Mortality patterns in the selected
cohort will be compared to those expected among
other males of the same age in California or
other appropriate areas within the United States.
The hypothesis of the mortality study is, "Do
men exposed to diesel exhaust exhibit different
mortality patterns than similar workers without
exposure to diesel exhaust?"
The cohort will be selected from those individuals
having one year of work history between a period
of January 1, 1958, through December 31, 1978.
The eventual cohort is estimated to comprise some
25,000 to 40,000 individuals and will consist of
up to 500,000 man years of working experience.
This cohort will also be evaluated in terms of
cancer incidence. This procedure will utilize
the California Tumor Registry (CTR), which con-
tains the incidence of all tumors reported within
five San Francisco Bay Area Counties contributing
to the CTR. This phase of the study will use
membership computer tapes and correlate the data
with members residing in the five Bay Area
Counties who were alive and had a working year of
membership by the end of 1969. Individuals
joining this union after 1969 will be included
in the cohort following one year of their date of
initiation. This cohort will be matched with
those cancer patients in the five Bay Area
Counties contained in the CTR on an annual basis
beginning with 1969. It should be possible to
identify each local union member with one year or
more membership and evaluate the incidence of
tumors. This cancer incidence study will have an
1109
-------
advantage over the mortality study because it
will pick up the risk of developing tumors as it
includes those exposed workers with tumors that
are living.
An important aspect of these efforts will be the
environmental sampling component. The environ-
mental sampling is designed to evaluate levels of
diesel exhaust exposures that characterize the
study populations in terms of work classifica-
tions. The sampling procedures will evaluate
numerous job sites where construction workers are
involved with heavy equipment using diesel-
powered machinery.
Another feature of this epidemiologic study will
enable a data collection network to form a worker
health registry. This health registry can be
updated and used for future prospective epidemi-
ologic evaluations of working cohorts with longer
exposure periods to diesel "exhaust.
SUMMARY
The preceding has discussed the available human
diesel health literature with respect to both
short-term and chronic indicators of health.
Studies looking at lung function data and respira-
tory complaints from questionnaires were evaluated.
No conclusive evidence suggesting an association
between diesel exhaust exposures and adverse
health effects have been established. Mortality
studies evaluating cancer incidence among occupa-
tional cohorts have also failed to associate
diesel exhaust exposures with long-term diseases,
including cancer.
More studies involving humans evaluating short-
term and long-term health effects are urgently
needed. A review of a study among construction
workers exposed to emissions from heavy equip-
ment powered by diesel engines was discussed.
While the entire question of human health effects
from diesel exhaust cannot be settled on the
basis of this particular study, the large cohort
should help clarify the question of whether or
not diesel exhaust exposures adversely affect
workers in terms of long-term illness.
mo
-------
REFERENCES
1. Lassiter, D. V., andT. H. Milby. 1978.
Health effects of diesel exhaust emissions,
Berkely, California. Environmental Health
Associates, Inc.
2. Lachtman, D. S. 1978. Health factors
associated with the use of diesel equipment
in underground mines. In: Proceedings,
Mini Symposium - Diesels in underground coal
mines, SME-AIME Meeting, Lake Buena Vista,
FL, September, 1978.
3. Battigelli, M. C., R. J. Marmella and T. F.
Hart. 1964. Environmental and clinical
investigation of workmen exposed to diesel
exhaust in railroad engine houses. Indiana
Medicine and Surgery, 33:121-124.
4. Battigelli, M. C. 1965. Effects of diesel
exhaust. Archives of Environmental Health,
10:165-167.
5. El Batawi, M. A., and M. H. Noweir. 1966.
Health problems resulting from prolonged
exposure to air pollution in diesel bus
garages. Industrial Health.
6. Jorgensen, H., and A. Svensson. 1970.
Studies on pulmonary function and respira-
tory tract symptoms of workers in an iron
ore mine where diesel trucks are used under-
ground. Journal of Occupational Medicine,
September"; 1970.
7. Yamazaki, K., T. Mogi, Y. Nishimoto and
T. Komazawa. 1969. The effects of diesel
exhaust gas on the body. Report No. 2. An
analysis of pulmonary function tests.
Railway Labor Science, No. 23, pp. 1-11.
8. Attfield, M. D. 1978. The effect of expos-
ure to silica and diesel exhaust in under-
ground metal and non-metal mines. In: ACGIH
Proceedings, Denver, Colorado, NoveinEer 6-7,
1978.
mi
-------
9. Public Health Service. 1963. Silicosis
in the metal mining industry. Pub. 1076
USGPO.
10. Gamble, J., W. Jones, J. Hudak and
J. Marchent. 1978. Acute changes in
pulmonary function in salt mines. In:
ACGIH Proceedings, Denver, Colorado,
November 6-7, 1978.
11. Reger, R. and J. Hancock. 1979. Coal
miners exposed to diesel exhaust emissions.
In: Proceedings, Health implications of
new energy technologies, Park City, Utah,
April, 1979.
12. Environmental Health Associates. 1979.
Report to the American Mining Congress:
Examination and interpretation of the draft
NIOSH report and data tape sets: On res-
piratory health: Coal miners exposed to
diesel exhaust emissions, Reger and Hancock,
September 21.
13. Environmental Health Associates. 1979.
Report to the American Mining Congress:
Examination and interpretation of the draft
NIOSH report and data tape sets: On res-
piratory health: Coal miners exposed to
diesel exhaust emissions, Reger and Hancock,
November 16.
14. Kaplan, I. 1959. Relationship of noxious
gases to carcinoma of the lung in railroad
workers. J.A.M.A., 171:2039.
15. Raffle, P. A. B. 1957. The health of the
worker. Brit. J_._ Indust. Med. , 14:73-90.
16. Bruce, R. M., and D. T. Mage. 1978.
Report on trip to England and Denmark.
United States Environmental Protection Agency
report, August 2-14, 1978, pp. 19-23.
17. Waxweiler, R. J., J. K. Wagoner and W. C.
Archer. 1973. Mortality of potash workers.
JOM, 15:406-489.
18. Milby, T. 1979. Personal communication,
Berkeley, California, November, 1979.
1112
-------
General Discussion
SPEAKER: On pack years of cigarette consumption, they
are matched qualitatively exactly as to the smoking stat-
us. We thought it was unnecessary on the basis of pack
years because as you know they were also matched fairly
well on age. Later on we have done as you have done with
the data also looking at pack years and as you might
expect, they are roughly equivalent overall even though
the matched pairs vary between them in pack years. Sec-
ondly, on the study of acute effects the changes in lung
function over a work shift indicated that the controls we
used were older than the exposed people. This was all we
had at the time and this did worry us to some extent.
However, we found that the age of a person seemed to have
almost no importance whatsoever relative to the pulmonary
performance of the individual over an eight hour work
shift. Hence, I think the age effect in terms of an
eight-hour shift change is somewhat cancelled out. Smok-
ing certainly had an effect. No one denies for a moment
that aging has an effect on pulmonary functions. Cer-
tainly it does.
D. LACHTMAN: That is a good point. I had thought
that to be true from the data we got in the Hatfield
Study. It is a fairly large study and I think it was in
that study that we actually found some increasing re-
sults, with a decreasing time of the people - that is
easy to explain. In terms of people smoking, you are
right. You matched qualitatively very well, so we played
with your first data batch.
S. KAPLAN: On the Environmental Health Associate
study on cancer incidence, do you have any information as
to what proportion of your teamsters migrate outside the
five-county bay areas that is covered by the cancer regis-
try? I think this is something that if it is going to be
carried out will have to be addressed in your analysis.
D. LACHTMAN: That can be looked at. In fact, the
numbers that must be followed include some of the team-
sters who work in that union but don't live in the bay
area. These will have to be eliminated and it is going
to decrease the numbers considerably. I think there is
some analysis being done in that regard but I don't think
it is complete.
S. KAPLAN: I am also talking about people who lived
there at the time thay qualified and then subsequently
moved away, who will not be picked up.
1113
-------
A RETROSPECTIVE COHORT STUDY OF DIESEL EXHAUST EXPOSURE Ih
RAILROAD WORKERS: STUDY DESIGN AND METHODOLOGIC ISSUES*
Marc B, Schenker, M,D.** and Frank E. Speizer, M,D.
Channing Laboratory, Depts. of Medicine,
Peter Bent Brigham Hospital and Harvard Medical School
Boston, MA 02 I I 5
ABSTRACT
Despite the experimental evidence for a carcinogenic ef-
fect of diesel exhaust, only a few epidemiologic studies
have evaluated this question. All of the existing studies
have major study design or methodologic weaknesses that
may account for the absence of a consistently observed ef-
fect. Some of these issues which are discussed in this
paper include source of subjects, required sample size,
cohort selection and confounding. Study design issues are
related to a proposed epidemiologic study to evaluate the
possible carcinogenic effect of diesel exhaust in U.S.
railroad workers. Data for the study come from the U.S.
Railroad Retirement Board. Components of the study in-
clude: I) a retrospective cohort analysis of approximately
80,000 male railroad workers, 2) a case-control study of
300 incident lung cancer cases and matched controls in
railroad workers, and 3) actual environmental monitoring
of worker exposure to diesel exhaust.
These approaches will allow for quantitative assessment
of both level and duration of diesel exhaust exposure and
consideration of the major confounding factor (cigarette
smoking), thus minimizing the major drawbacks of al-r pre-
vious studies.
* Supported in part by Feasibility Study Grant from the
Harvard University - Massachusetts Institute of Technology
Health Sciences Center, funded by the National Institute
of Environmental Health Sciences (ES02I09).
** Supported by National Research Service Award from the
National Institute of Environmental Health Sciences
(ES05I57)
1114
-------
An epidemiologic study of diesel exhaust as a potential car-
cinogen presents many unique methodologic problems in addi-
tion to the more general difficulties of studying a possible
occupational or environmental carcinogen. I would like to
discuss some of these study design problems and relate them
to previous epidemiologic investigations of diesel exhaust
effects. Finally, I will describe a study we are under-
taking of this question on U.S. railroad workers and relate
some of these issues to that study.
Study Design
A few large scale cross-sectional cancer surveys have re-
corded occupation data and determined cancer rates, particu-
larly for respiratory cancers, in occupations with potential
diesel exhaust exposure. These studies have found increased
Standardized Mortality Ratios (SMR's)t in diverse occupa-
tions with possible diesel exposure including construction
workers, railroad brake and switchmen, and truck drivers
(2,3). Such large descriptive surveys can only be considered
hypothesis generating since positive findings may be due to
other exposures, confounders or chance alone. In addition,
occupation data are limited and potentially biased, usually
being based on only the last known occupation. A final
weakness in these studies is that quantitative and qualita-
tive exposure to diesel exhaust or any other substance is
only inferred.
Analytical or hypothesis testing studies are basically of
two types, case-control and cohort. A case-control approach
has been used at RoswelI Park Memorial Institute to look for
trends in cancer incidence by occupation (4). They compare
occupation in patients with cancer to the occupation of
patients who are found not to have cancer. This approach
can control for confounding factors such as smoking habits,
but results relating to diesel exposure have been limited by
small numbers of cases in individual occupations, crude
occupational histories and no data on actual exposures. In
these studies no significantly increased relative risks
(RR's) in lung cancer for railroad engineers or firemen or
for truck or bus drivers were found (4,5).
Case-control and cohort studies have also had reduced statis-
tical power in studying possible causal effects of diesel
exhaust and cancer because of the absence of a "marker"
The SMR i^ the number of deaths, either total or cause-
specific, in a^given occupational group expressed as a
percentage of the number of deaths that would have been
expected in that occupational group if the age-and-sex
specific rates in the general population had obtained (1).
1115
-------
tumor, as is the case for mesothelioma from asbestos or
angiosarcoma from vinyl chloride exposure. While such a
"marker" tumor does not exist for diesel exposure, the use
of more specific histologic classification of tumors in
epidemiologic studies may increase the sensitivity and
specificity of these investigations. An excess of oat
cell carcinomas of the lung has been noted in populations
exposed to known occupational carcinogens including asbestos,
chromium, uranium and chloromethyl methyl ether (6-8). A
case-control study of oat cell cancers of the lung from the
Massachusetts Tumor Registry found an excess of transporta-
tion equipment operatives compared to controls with central
nervous system tumors (9). While this result is not speci-
fic for diesel exhaust, it suggests that pathologic assess-
ment may be a potential tool for greater power in future
case-control and cohort studies of diesel exhaust exposed
populations.
Most of the existing studies have used a cohort design and
the remainder of this paper will focus on that epidemiolo-
gic approach.
Source of Subjects
A major weakness of existing studies that relate to diesel
exhaust exposure is the absence of exposure data for dif-
ferent occupations. Extensive use of diesel engines is
found in heavy duty trucks, bus and construction, railroad
and maintenance vehicles and in some types of mining.
Within these industries maintenance and repair personnel
may have higher exposures than drivers of diesel vehicles
(10), but there are scant data on actual exposure levels.
For example, exposures of truck drivers or railroad engineers
must be measured before these occupational categories are
appropriately considered exposed to diesel exhaust. A
Finnish study of railroad conditions found particulate
levels over five times as high in diesel roundhouses as
in diesel locomotive cabs (10). This type of information
will increase the sensitivity of future epidemiologic
studies but has not been utilized in the design of studies
done up until this time.
A related issue is what component of diesel exhaust should
be considered in evaluating exposure dose. Experimental
evidence suggests that the particulate fractions, which con-
tain the polycyclic aromatic hydrocarbons (PAH), are the
principle carcinogenic components of diesel exhaust (11-14).
However, particulate fractions may vary in their PAH content,
and other exhaust gases such as N02 and S02 may modify
the carcinogen effect (15,16). Data are only beginning to
be generated on comparability of diesel exhaust from
1116
-------
different sources.
Number of Subjects
Considering only cancer of the respiratory tract one can
make estimates of the number of cases required to see an
effect due to diesel exhaust exposure. These estimates
use an approximation of equal potency of equal amounts of
benzo(a)pyrene (B(a)P) in products of coal combustion and
in diesel exhaust. Using estimates developed in the
British Gas Works studies, the expected relative risk may
be in the range of 1.5 to 2.5 (17). Existing measurements
of diesel exhaust exposure suggest levels will not exceed
the mean concentration of 3yg/m3 B(a)P to which the gas
workers were exposed, however, further environmental mon-
itoring studies are necessary to determine actual exposures
(10,18). There are no data to suggest that exposures will
equal those of coke-oven topside workers, where RR's as
high as 10 for lung cancer were seen (19).
The rate of lung cancer for an unexposed population can be
determined from cancer mortality statistics (20). It is
then a straightforward calculation to determine the required
cases of lung cancer to observe a statistically significant
result for different RR's (21). This is done in Table 1
for a probability of lung cancer in an unexposed population
of 0.0259*, one sided test.
TABLE 1
REQUIRED CASES OF LUNG CANCER IN A COHORT STUDY TO OBSERVE
RELATIVE RISK
RR
N = Cases of Lung Cancer
1.5
2.0
3.0
5.0
2313
687
165
80
a=.05 6=0.2
4850
1440
346
168
a=.01
0=0.1
This probability was calculated for the white male cohort
described in the study at the end of this paper. It is
based on age and sex specific respiratory cancer rates for
Connecticut, 1962-64 (20).
1117
-------
The existing studies relating to occupational diesel ex-
haust exposure and cancer are listed in Table 2 with the
number of cancers observed.
Author
Kaplan, 1959 (22)
Hueper, 1955 (23)
Raffle, 1957 (24)
TABLE 2
Population
RaiI road Workers
Ra iI road Workers
London Transport
Workers
Hannunkasi, 1978 (25) Railroad Workers
Luepker, 1978 (26)
Teamsters
Number of Cases
154, lung cancer
133, lung cancer
96, lung cancer
47, all cancers
34, lung cancers
Even the largest study, which observed 154 cases of lung
cancer, may not have had enough statistical power to ex-
clude a false-negative result (8 type error).
Obviously, numerous factors may alter these size require-
ments. Errors in exposure classification or differences in
rates of smoking between exposed and unexposed populations
will increase the number of cases required. A more general
carcinogenic effect or an effect on non-malignant causes of
death (e.g. respiratory disease) may decrease the size of
the required cohort.
Cohort Selection
One important factor in selecting a cohort is adequacy of
the observation interval for the latency period of lung
cancer. Studies of other populations exposed to PAH's
suggest an incubation (latency) period of 10 to 25 years (27).
Investigations of possible carcinogenesis from diesel ex-
haust must allow for a similar interval from the beginning
of potential exposure to the end of the follow-up.
A related consideration is the time period of use of diesel
engines. Diesel locomotives were first used in quantity
by the railroads in the 1940's and only by the mid 1950's
did they represent even the majority of locomotives in ser-
vice (Fig 1). A similar sequence of use was seen in buses
and trucks. Prior to use of diesel locomotive, most rail-
road engines were steam units. These engines potentially
exposed the workers to the product of coal combustion and
to asbestos from the thick asbestos insulation around the
boilers. Exposure of subjects in a cohort of railroad
workers to these engines represents a potentially significant
confounder in diesel exhaust exposure studies.
1118
-------
DIESEL LOCOMOTIVES IN SERVICE
UNITED STATES 1930-1977
lOOn
1940
1950 I960
YEAR
1970
1980
Figure I. Diesel locomotives as a percent of all Class I
locomotives in the United States. Docked line indicates
50% diesel locomotives in service, corresponding to the
year 1952.
Thus, an optimal cohort for study must have adequate dura-
tion of exposure and interval of observation to allow for
a 10-25 year latency period, but in the case of the rail-
roads, workers cannot have had potential exposure much be-
fore the widespread use of diesels in the 1950's. Un-
fortunately, most of the studies of railroad workers and
lung cancer deaths were done in cohorts exposed in the
1950's and earlier (Fig 2). They therefore cannot be con-
sidered to reflect effects of diesel exhaust.
Confound!ng
A final consideration in evaluating existing or planned
epidemiologic studies of diesel exhaust exposure is the
effect of potential confounders. A confounder is a factor
related to the exposure being studied and also a risk
indicator of the disease (28). The major potential con-
founder in diesel exhaust studies, particularly those
assessing lung cancer or respiratory disease as an outcome,
is cigarette smoking. Two considerations are important,
1) Is there a difference in smoking rates of the diesel ex-
posed and the reference population, and 2) if a difference
in rates exists, what is the direction and magnitude of any
effect that might be seen.
1119
-------
COHORT STUDIES OF CANCER
IN RAILROAD WORKERS
3100
at
in
5
"EXPOSURE" FOLLOW-UP
133 LUNG CANCERS
iHH 154 LUNG CANCERS
47 CANCERS
ALL TYPES
1930
1940
1950 I960
YEAR
1970
1980
Figure 2. Three cohort studies of cancer in railroad
workers. Percent diesel locomotives in service is plot-
ted on the same time scale.
No cohort study on diesel exhaust effects contains direct
information on smoking habits of the observed population.
An estimate of present smoking in railroad workers can
be made from a prospective study of cardiovascular diseases
in U.S. Railroad employees that was begun in 1957-1959 (29).
Comparing rates of current smokers in that study with data on
U.S. males in 1955 obtained by the National Clearinghouse
for Smoking and Health (30) indicates only small differences
(Table 3). Differences in smoking rates of this magnitude
can be shown to have only a weak confounding effect (31).
The use of comparison populations within the industry being
studied is another method to further reduce this possible
confounding effect, particularly if one controls for covari-
ates such as socio-economic status.
Proposed Study
This discussion has focused on reasons why the existing
epidemiologic studies may have not been able to show an
effect if one exists or may not have been specific for
effects of diesel exhaust exposure. I would like now to
briefly describe an epidemiologic study we are undertaking
to evaluate this question.
The data base for the study comes from the files maintained
by the U.S. Railroad Retirement Board (RRB) on all U.S.
1120
-------
Railroad workers. These computerized files contain demo-
graphic and job category information on all U.S. railroad
employees in active service and on retirees with at least
10 years of service in the railroads. All male railroad
workers between 50 and 64 years of age with 10 to 19 years
of railroad service in 1964 will be potential subjects.
From this pool a cohort of approximately 80,000 subjects
will be selected (Fig 3). Selection will be based primarily
on job category and specific railroad company. This will
allow inclusion of workers in the highest diesel exhaust
exposed job categories and from the companies with the
longest history of diesel engine use. The cohort will also
contain an appropriate control of workers in non-exposure
job categories.
TABLE 3
Comparisons of Current
Smokers in Railroad Employees
and United States Population,
Male; Age 45-54
Current Smokers
U.S., 1955 56.9
Railroad Employees, 1957-1959
Clerks 57.5
Switchmen 62.2
Executive 54.7
The cohort will be followed through 1978 and all deaths in
active service and after retirement will be ascertained and
coded for cause from death certificates. Based on U.S.
male cancer mortality rates, one can estimate that 2,483
cases of lung cancer will occur. This should allow detection
of a RR of at least 1.5 for lung cancer between exposed and
non-exposed populations.
Another proposed component of the investigation is a case-
control study of approximately 300 incident cases of lung
cancer reported to the RRB for 1976-1978. Each case will
be age and sex matched with a case of non-respiratory can-
cer and another case of non-malignant non-accidental cause
of death. Next of kin of all cases will be contacted for
questionnaire information on potential confounders such
as cigarette smoking and residence (urban vs. rural). Hos-
pital records will also be requested on lung cancer cases
and histo logic material reviewed by a consultant pathologist.
1121
-------
DIESEL LOCOMOTIVES IN SERVICE AND
PROPOSED RETROSPECTIVE COHORT STUDY
OF U.S. RAILROAD WORKERS
"EXPOSURE" FOLLOW-UP
IDENTIFY
COHORT
1930
1940
1950 I960
YEAR
1970
1980
Figure 3. Proposed retrospective cohort study described
in the text. Minimum potential exposure is 10 years.
Percent diesel locomotives in service is plotted on the
same time scale.
A third and very important part of the study will be actual
environmental and personal monitoring of levels of diesel
exhaust exposure for different job categories and work
areas. This will be done in different parts of the country
and over two seasons. This information will be integrated
into the cohort selection and epidemiologic data analysis.
We hope this combined approach will provide useful evidence
about the possible occupational carcinogenic effect of
diesel exhaust and also relate to increased diesel exhaust
in the general environment.
The retrospective cohort component will evaluate a unique
population of workers that is both large enough and has an
adequate duration of potential exposure and follow-up to
detect a relative risk for lung cancer of 1.5. This
component will be supplemented by case-control study of
incident cases of lung cancer. By this method we will
be able to directly control for potential confounding factors
and to evaluate the occurrence of type specific cases of
1122
-------
lung cancer. The use of environmental and personal expo-
sure monitorfng data in selecting the cohort for study
and evaluating the data will greatly increase the efficiency
and validity of the results.
As mentioned, this study is just beginning. We anticipate
it will take 3 years to collect the appropriate information
and complete the analyses. Thus, it will be an expensive
study both in terms of time for people and money. However,
since automobiles and light duty trucks are predicted to be
25% diesel by 1985, this study may be the last opportunity
we have to measure the effect of one source of diesel ex-
haust before it becomes part of the general environmental
pollution, in Which case added complexity may make it
impossible to select out the effects that might be present.
1123
-------
REFERENCES
1. MacMahon, B., T.F. Pugh. 1970. Epidemiology
Principles and Methods. Little Brown and Co.,
Boston.
2. Guralnick, L.G. 1963. Mortality by occupation
and cause of death among men 20-64 years of
age in the United States, 1950. U.S. Dept.
of HEW, National Vital Statistics Division,
Vol. 53.
3. Menck, H.R., and B.E. Henderson. 1976. Occu-
pational differences in rates of lung cancer.
J_. Occup. Med. 18: 797-801.
4. Decoufle, P., K. Stanislawczyk, L. Houten, et
al. 1977. A retrospective survey of cancer in
relation to occupation. DHEW, (NIOSH) Pub.
No. 77-178.
5. Viadana, E., I.D.J. Bross, L. Houten. 1976.
Cancer experience of men exposed to inhalation
of chemicals or to combustion products. J^
Occup. Med. 18:787-792.
6. Hueper, W.C. 1966. Occupational and Environ-
mental Cancers of the Respiratory Tract. New
York: Springer-Verlag.
7. Churg, J., M. Kannerstein. 1970. Occupational
exposure and its relation to type of lung can-
cer, in Morphology of Experimental Respiratory
Carcinogenesis, U.S. AEC Division of Technical
Information, Symposium Series, Dec.
8. Figueroa, W.G., R. Raskowski, W. Weiss. 1973.
Lung cancer in chloromethyl ether workers.
H. Engl. J_. Med. 288:1096-1097.
9. Wegman, D.H., J.M. Peters. 1978. Oat cell
cancer in selected occupations. vK Occup.
Med. 20:793-796.
10. Heino, M., R. Ketola, P. Makela, et al. 1978.
Work conditions and health of locomotive
engineers. I. Noise, vibration, thermal
climate, diesel exhaust constituents, ergo-
nomics. Scand. J_. Work Environ. Health 4:
Suppl. 3, 3-14.
1124
-------
11. Schenker, M.B. Diesel exhaust - an occupa-
tional carcinogen? J_. 0ccuj3. Med. (In Press).
12. Santodonata, J., D. Basu, P. Howard. 1978.
Health effects associated with diesel exhaust
emissions. Literature review and evaluation.
U.S. Environmental Agency. No. 600/1-78-063.
November.
13. IARC Monographs on the Evaluation of Carci-
nogenic Risk. 1973. Certain polycyclic
aromatic hydrocarbons and hetereocyclic
compounds. IARC, Lyon.
14. National Research Council. 1972. Particu-
late polycyclic organic matter. National
Academy of Sciences. Washington, D.C.
15. Pitts, J.N., K. Van Cauwenberghe, D. Grosjean,
J. Schmid, D. Fitz, W. Belser, G. Knudson, P.
Hunds. 1978. Chemical Transformations of
Polycyclic Aromatic Hydrocarbons in Real and
Simulated Atmospheres and their Biological
Implications. Presented at the Third Inter-
national Symposium on Polynuclear Aromatic
Hydrocarbons, Batelle - Columbus Laboratories,
Columbus, Ohio, Oct. 25-27, 1978.
16. Laskin, S.M. Juschner, R.T. Drew. 1970.
Studies In pulmonary carcinogenesis. In:
Hanna, M.G., J.R. Gilbert, (Eds.) Inhalation
Carcinogenesis, U.S. AEC Symposia Series.
#18, p. 321.
17. Doll, R. 1972. The causes of death among gas
workers with special reference to cancer of
the lung. J5R_. J_. Ind. Med. , 9:180-185.
18. Batigelli, M.C., R.J. Mannella, T.F. Hatch.
1964. Environmental and clinical investiga-
tion of workmen exposed to diesel exhaust in
railroad engine houses. Ind. Med. Surg.,
33:121-124.
19. Christian, H.A. 1962. Cancer of the lung in
employees of a public utility. A fifteen year
study. J_. Occup. Med. , 4:133-139.
20. Doll, R., C. Muir, J. Waterhouse, (Eds.).
1970. Cancer Incidence in Five Continents.
UICC Springer-Verlog: New York.
1125
-------
21. Snedecor, G.W., W.G. Cochran. 1967. Statis-
tical Methods. Iowa State University Press,
6th Edition.
22. Kaplan, I. 1959. Relationship of noxious
gases to carcinoma of the lung in railroad
workers. J..A.M..A. 171:2039-2043.
23. Hueper, W.C. 1955. A quest into the envi-
ronmental causes of carcinoma of the lung.
Public Health Monograph No. 36. U.S. Sept.
HEW, PHS.
24. Raffle, P.A.B. 1957. The health of the
worker. B_r. J_. Ind. Med. 14:73-80.
25. Hannunkasi, E. Jarvinen, T. Partonen. 1978.
Work conditions and health of locomotive
engineers. II. Questionnaire study, mor-
tality and disability. Scand. J_. Work.
Environ. Health. 4:Suppl 3, 15-28.
26. Luepker, R.V., M.C. Smith. 1978. Mortality
in unionized truck drivers. J. Occup. Med.
20:677-682.
27. Cole, P., M.B. Goldman. 1975. Occupation,
in Persons at High Risk of Cancer. Fraumeni,
J.F., (Ed.). New York. Academic Press.
28. Miettinen, O.S. 1974. Confounding and
effect-modification. Am. J_. E p i d e m i o 1 .
100:350-353.
29. Taylor, H.L., H. Blackburn, J. Brugh, R.W.
Parlin, T. Puchner. 1967. Railroad
Employees in the United States. Acta. Med.
Scand. (Suppl.). 460:73.
30. Appendix. 1979. Cigarette Smoking in the
United States, 1950-1978 in Smoking and
Health, A Report of the Surgeon General.
U.S. Dept. HEW, DHEW. Public. No (PHS)
79-50066, U.S. Govt. Printing Office,
Washington, D.C.
31. Axelson, 0. 1978. Aspects in confounding
in occupational health epidermiology. Scand.
J. Work. Environ, and Health. 4:85-89.
1126
-------
CHARACTERIZATION OF DIESEL EXPOSURE GROUPS
Donald F. Hansknecht
Richard A. Ziski rid
Michael B. Rogozen
Science Applications, Inc.
1801 Avenue of the Stars, #1205
Los Angeles, CA 90067
ABSTRACT
Current diesel epidemiological studies focus upon occupa-
tional exposures because public exposure to diesel emissions
has until now been minimal. With the pronounced market
penetration of diesel passenger automobiles predicted for
the 1980s public exposure will be markedly increased. Care
must be exercised for the utilization of occupational
cohorts in order to derive the greatest benefit from their
study. Each group will have its own characteristics of
exposure which may vary across specific job classifications,
worksites, and levels of activity to note but a few para-
meters of concern. Additionally all groups are subjected
to accompany toxic agents which may act as confounding
factors. Issues discussed in this paper include establish-
ing criteria useful to contrast occupational groups to
determine their suitability to test hypotheses on health
effects and examining indicators of exposure and measurement
methods.
EXPOSURE GROUPS
Current diesel epidemiological studies focus upon occupa-
tional exposures because public exposure to diesel emissions
has until now been minimal. The groups that have been
1127
-------
identified as possibly having above average exposures,
include railroad workers, truck drivers, miners, heavy duty
equipment workers, diesel mechanics, and trash collectors.
Within each of the occupations, the exposure environment may
vary according to job classification, worksite, and level of
activity. Furthermore, each of these groups is exposed to
other airborne agents, which may act as confounding factors.
In order to assess their applicability for utilization as an
epidemiological cohort a great number of considerations need
to be made including but not limited to:
o The statistical requirements of the study protocol
o The characteristics of the exposure including the
period, level and the relationship across job or
equipment classification
o Confounding factors such as other substances and
working conditions.
Of course this merely touches on points needed for consider-
ation of performing such studies such as population mobility,
availability of records, etc.
In order to provide proper exposure data for an epiderniolo-
gical study, two questions must be addressed:
1. What are reliable indicators of methods of measure-
ment of exposure to diesel emissions in the working
environment?
2. Do the levels of exposure differ sufficiently among
occupational groups to enable the confirmation of
hypothesis on health effects?
Diesel exposure measurements reported in the literature and
obtained in field studies by SAI are summarized for compari-
son in Table 1. Although the wide variety of measurement
techniques makes it difficult to compare values, it is
evident that a measurement program must be carefully tail-
ored to the exposure environment under study. Both personal
and area monitoring for gases and particulates are neces-
sary. Because of their importance in health effects,
respirable particulates need to be collected and analyzed
in such a manner as to separate and identify components of
diesel origin. Techniques for these analyses will be
described.
In attempts to determine whether there are significant
differences in the extent to which individuals in different
cohorts are exposed, several outcomes are possible:
1123
-------
o Exposures are essentially the same for all putative
exposure groups
o Exposure differences are statistically significant
but meaningless from a health standpoint
o Exposure differences are both statistically and
medically significant
o Exposure differences become meaningful if exposure
group categories are redefined.
A study design based upon a randomization or non-parametric
test is expected to be useful in determining which of these
outcomes applies.
Table 1 summarizes a group of studies which measured work-
place concentrations of several diesel emission species. It
is difficult to make meaningful comparisons with these data
and/or to extract answers to either of the above posed
questions. Difficulties are many and but to mention a few
include:
o There are a wide variety of measured species,
average times, conditions and techniques utilized.
o Little personal sampling has been done, rather area
sampling is principally available.
o Significant changes in occupational equipment and
conditions have occurred over the extended periods
of interest.
o Determination of exposure across job categories,
although necessary, is usually unavailable.
o Similarly the separation of exposure level from
background level is essential but not usually
conducted rigorously.
Looking at the table allows some gross intercomparison among
occupations. Our first hand experience centers around buses
and diesel tractors, earthmoving equipment, and truck repair
sites. In all our studies attempts were made to determine
and separate out the background and self contamination
contributions. Several limited pieces of data not included
in the table include other gases measured inside and outside
of diesel tractors:
1129
-------
oo
LU
Q
r—
O
r—
Q;
h~
LU
O
•^
CD
O
^
UJ
CQ
s:
<:
UJ
O
1
a.
0
3
— i
UJ
oo
UJ
Q
U.
O
>-
^
y
OO
^
U"
^3^
I —
CO
u
u
ai
0)
4-
O)
to
ra
^
o
•r—
-J-)
S- CO
ro IE
Q- -^
ai
0) E
XI
T3
5.
CO
QJ
C£
—
O Q.
O d
CM
0 E
z: a.
a.
E
O Q.
^ a.
i. CD
O C
4— T-
CO ra
••- S-
CO OJ
m ec
a>
o
fO
o.
I.
o
3
4-
o
a.
1—
J TO
_o r***
r-^. 3}
CT> —1
1 — 1
" +-*
O) fO
•4-* QJ
3 -4— >
OO> oo
Lf>
*
< — 1
I
dT> "O
• Q
0 ^
LO LO
V UD
LO
XI O
LO
CO
1
LO CO
r-H LO
eee> ID
ai
c
C T-
o s:
i % r_
O ro
3 O
1- O
CO TJ
C C
o 3
U 0
s_
'aj s-
d CU
C "O
3 C
1— rD
^
+j
11 (J
-o
c r~^.
o en
LO ^H
C.
r".
0
LTt
•
cj-
r-l
CO
-O
^-1
CO
CM
O
UP
LO
CM
s^-
•
T-H
S-
o
!_ !-
-C -C
1 1
T-H CM
O)
C
S-
0)
Q-
Q.
0
0
-o
c
s-
cn
a)
c
, —
1 >
a)
r —
CO -O
c en
•i— i-H
E
O
0
0)
LO
CM
t— t
t
OJ
CO
o
^
§
Q
l_
r~
CO
OJ
Ol
ra
S-
(3
CO
3
CO
CO
r —
en
i — t
A
O
a.
"^
*o
C-J
•
o
1
en
0
o
LO
i — i
i
i — i
o
o
1 — 1
o
o
o
CM
O
CM
O
0
to
i.
J=
o
CM
a.
o
-C
t_
fO
Q.
CU
^
T3
ro
O
• r~
fa
a;
o
rj1
•a
iT3
Ul
i_
u
-a co
c: >--
^o o^
LO i— i
r^.
i — i
! 1
•
ro
i-^.
oo
o
I — 1
•
ro
CM
t/i
I—
-C eC
CO?
,^
ra
o
+j ru
a; co
i. ra
0 (_>
a)
"£
i— O
••- 3
fO
o:
•g
S
-a
c
c^
i- en
QJ <-H
^
fO
r?
<3-
i
LO
i — 1
,_£.
«
O
i
o
1 — 1
o
en
un
i
i— t
ID
OJ
C
£
_
ra
o
o
•o
c
o
S-
CD
s_
CU
c
p^
en
— 4
^
., —
t/1
., —
^
~^~-
o
CO
1
o
CO
1
T— 1
0
0
CM
1
r-H
O
^
to
xy
U
3
t—
rv
CO
tv
., —
Q
en en
r^ r— *
en er>
— f ^H
c c
aj QJ
N N
P °
O O
^ C£
^O r— 1
• t
O O
1 1
r-H cn
o o
0 0
^~
1 1— t
1 — 1 v
Lf) C\J
• (
O O
1 1
c\j ro
0 0
0 0
uo
o
LO •
• < — 1
CO 1
i tr>
i-H O
o o
c c
., — ., —
E E
CM CM
•M
C
U
!_ 3
ra S-
UJ 1—
,
O
s_
o
n-
co
3
r—
ro
>
Ol
c
-o
o
a.
CO
OJ
J_
o
u
'1
o
J-.
OJ
S-
fl)
o
c
ro
_c
"^
CO
CO
(U •
r— CO
r- Ol
ra
"
2i i-^
OJ ro
S
CO 0)
C
o -u
•1- l-
+-> ro
ra 5 •
• i- ai to
C | ' | ' 4->
S C 00 ro
O CD T3
cue
.* C -f- 0
CO C
3 U XI
CU II
II CM-H
O -i- Q
=>-^ <^>-z.
ro JO U "O
•
O
II
CM
O
•z.
0
CM ro
O .C
z J-J
1 CT1
C
• "x E
ai o 3
4-> ^ CO
ro = to
i — ra
3 -a
O CU >,
•r- | * _Q
+-> l-
S- O T3
ro O. OJ
Q. 01 +J
a: ro
'ra II -i-
O O CO
1 — ^1 1 1 1
OJ f- cn
1130
-------
TABLE 2. EXPOSURE INDICATORS
Supporting Features
Reliably measured
Sufficiently stable
Interferences known
Well differentiated from
background
Personal sampler available
Detracting Features
NO
and
N02
CO
Exposure well below TLV
(NO)
Relationship between
NO-N02 varies markedly
Stable
Few interferences
Personal sampler available
Not excessively emitted
Poor differntiation from
urban background
Exposure well below TLV
S02
Personal sampler available
Less data available
Fuel content varies
Conversion to other
species
Alde-
hydes
Known to be toxic
Thought to be useful indi-
cators of diesel exposure-
over background levels
Field measurements
present difficulties
Cost competitive detec-
tion methods relatively
insensitive
Parti- Useful indicator
culates (respirable)
Personal sampler available
Analysis permits differen-
tiation from background
Area sampling is costly
and useful only for
longer period averages
Gas
o Sulfur dioxide
o Gaseous hydrocarbons
(methane equivalent)
o Zone
Result of Summary
o Approximately ambient levels
o Factor of 4 to 6 times ambient
(20-27 ppm)
o Fraction of ambient
The review of the literature does not allow definitive
guidelines to be formulated concerning the most useful
exposure indicators. Summarizing the supporting and de-
tracting features of the compounds usually measured we
have the results in Table 2.
Special mention should be made concerning the role of
particulate measurements. Clearly they rate a high priority
in supporting an epidemiological study by quantifying
exposure to potentially toxic emissions. In designing a
recommended characterization of the particulate environment
we have included the following:
1131
-------
Total and respirable ( <2.5 pm) participate frac-
tions of both area concentration (upwind and down-
wind) and personal exposure should be taken
Diesel emission measurements - analyses for organic
and elemental carbon
Upwind dust - analyses to determine elemental
composition
Relative proportions of total mass, organic and
elemental carbon, and carbonates should characterize
diesel contribution and allow differentiation. The
interrelationship between two or more elements such
as aluminum and calcium could provide additional
verification.
SAMPLING APPROACH
In order to evaluate the relationship of exposure within the
occupational group being considered, i.e., the second
question posed, comments are needed concerning structuring a
sampling approach while considering the problem of large
numbers of job classifications and/or equipment types being
used. Thus one wishes to know how meaningful such divisions
are and determine how to stratify our cohort into exposure
groups not solely based upon the number of years of employ-
ment. Stated in another way we want to take steps to
explore which of the four outcomes results:
o Are exposures essentially the same for all subgroups?
o Are exposure differences statistically significant
but meaningless from a health impact viewpoint?
(low exposure case)
o Are exposure differences both statistically and
medically significant?
o Are exposure differences meaningful if postulated
exposure group categories are redefined?
It is to differentiate among these possible outcomes that
a sampling aporoach is proposed and illustrated by example.
The sampling approach is summarized and the hypothesis to
be tested are as follows.
o Objective: Evaluate the exposure differences among
job classifications or diesel equipment types
1132
-------
o Randomized (non parametric) sampling approach -
i.e., no particular distribution of exposures is
assumed rather a "reference distribution" based
upon actual outcomes is used.
o Assume A and B the exposure groups of interest with
m_ and n randomly selected test vehicles respectively
and means X^ and Xg
o Hypothesis to be tested: Are A and B useful and
valid exposure groupings?
To proceed with the test compare the difference in the means
of the initially distributed groups with all alternative
unique arrangements. This process is summarized as follows:
o To test hypothesis compare T/\ - Tg with distribution
of all possible alternative rearrangements
o Number of unique arrangements is (m + n)!/m!n!
o The fraction of arrangements exceeding T/\ - Xg is a
measure of the significance level of the hypothesis
o Sensitivity depends on sample size (greater number
of combinations the lower the minimum level of
significance inferred by the test).
For the case where we sample equally from all candidate
groupings of occupations (i.e., m = n) the benefits of
increasing the sample size are to reduce the probability
that the hypothesis is unjustified. This is quantitatively
shown as follows:
o For equal size groups N = (2n)!/n!n!
Sample # Possible Minimum Level of
Size (n) Combinations (N) Significance (P)
2 6 0.17
3 20 0.05
4 70 0.014
5 252 0.004
6 924 0.0011
7 3432 0.00029
o Thus if goal is P < 0.01, a minimum of 5
samples for each group are necessary.
Obviously for uneven sample sizes across candidate groupings
there is less efficiency.
1133
-------
TABLE 3. EXAMPLE OF NONPARAMETRIC TEST: CARBON MONOXIDE
~IN SCHOOL BUSES (ALL CONCENTRATIONS IN PPM)
Group A Group B
Vehicle No.
1
2
3
4
5
6
7
8
9
CO Concentration
9.3
7.9
25.3
67.9
15.5
13.0
8.0
46.1
38.1
Vehicle
1
2
No. CO Concentration
15.5
7.0
XA = 25.678 XB = 11.250
o IA - XB = 14.4 and N = 55
o 12 combinations of Y/\ - "Xg groupings exceeded 14.4
o Probability that the difference of 14.4 was due to chance
was 0.22. Too high to justify the hypothesis
o If 9 measurements had been made in Group B, N = 18!/9!9!
= 884 x 55.
To illustrate consider two populations of school buses,
i.e., different manufacturers. Other factors such as route,
mileage, depot, maintenance are equal. Are the populations
significantly different in their interior carbon monoxide
concentration - hence driver exposure. For this case only
two vehicles were available of group B whereas nine were
tested in group A. Results are summarized in Table 3.
In this case the probability was 22 percent that the differ-
ence in the mean values could have been the result of
chance. As noted if 7 additional measurements could have
been taken for group B the number of combinations used to
provide comparison with the grouping mean values would have
increased by a factor of 884 making the test of the hypo-
thesis much more powerful and sensitive. We conclude the
nonparametric test approach offers a useful tool in most
cases.
1134
-------
REFERENCES
Apol, A., 1973. Health hazard evaluation/toxiclty deter-
mination, Union Pacific Ralroad, Pocatello. Idaho
(NTIS: PC 229 161).(Cited in Sanders and Peay,
1978).
Commins, B. T., R. E. Waller and P. J. Lawther, 1957. "Air
pollution in diesel bus garages", Brit. J. Ind. Med.,
Vol. 14, pp. 232-239.
Johnson, J., D. Carlson et al., 1976. The development and
application of advanced mine air monitoring techniques
to mines using diesel-powered equipment. Prepared by
Michicagn Technological University for Mining Enforce-
ment Safety Administration.
Lawter, J. R. and D. A. Kendall, 1977. Effects of diesel
engine emission on coal mine air quality. Prepared
by A. D. Little, Inc., for U.S. Bureau of Mines, USBM
Open-File Report 46-78 (NTIS: PB 282 377).
Sanders, M. S. and J. M. Peay, 1978. Health and safety
implications of diesel locomotive emissions, U.S. Navy
Personnel Research and Development Center, San Diego,
California, NPRDC TR 78-17.
Stewart, D. B., P. Mogan and E. D. Dainty, 1978. "Diesel
emissions and mine ventilation", Canadian Mining and
Metallurgical Bulletin (March), pp. 1-8.
Sutter, E., 1975. "Diesel engines in tunnel construction:
measurement of their fumes in air", Staub. Reinhalt.,
No. 11 (in German).
SAI-1 Ziskind, R., T. Carlin, et al., 1977. Toxic gases in
heavy-duty diesel truck cabs. Prepared by Science
Applications, Inc., for U.S. Department of Transporta-
tion, Federal Highway Administration, FHWA-RD-77-139.
SAI-2 Unpublished Measurements taken in 1979.
1135
-------
AN INDUSTRIAL HYGIENE CHARACTERIZATION
OF EXPOSURES TO DIESEL EMISSIONS
IN AN UNDERGROUND COAL MINE
Robert W. Wheeler, P.E.
Frank J. Hearl
Michael McCawley
National Institute for Occupational Safety and Health
944 Chestnut Ridge Road
Morgantown, West Virginia 26505
ABSTRACT
Problems exist in the United States' effort to achieve
energy self sufficiency. Increasing coal production to meet
President Carter's energy self sufficiency is a prime
problem for the rest of the century and beyond. The use of
diesels in underground coal mines has been suggested as a
mining method to aid in this needed, increased production.
Many questions exist about the effects on humans in such
environments. NIOSH Divison of Respiratory Disease Studies
has undertaken a research effort to characterize the envi-
ronments of existing diesel coal mines. The results of one
of the studies will be presented. Preliminary assessments
of carbon monoxide, nitrogen dioxide, C}-C5 aldehydes
and organic acids, aliphatic hydrocarbons, sulfates, total
and respirable dust, and polycyclic aromatic hydrocarbons
(PAH) are presented. Nitrogen dioxide and total aldehydes
are suggested as possible species to quantify diesel expo-
sure.
1136
-------
INTRODUCTION
Diesel powered mining equipment is extensively used today in
underground metal and non-metal mines. The U.S. Bureau of
Mines estimates there are more than 4,400 diesel units being
used in metal and non-metal mines as of 1978 (1). The first
diesel vehicle was introduced in a Western coal mine in
1946. Since that time, 33 coal mines have decided to use
diesel powered equipment in their mines (2).
During 1976, NIOSH began a five-year effort to study the
health implications of diesel use in coal mines. This
effort includes inhalation toxicology studies, morbidity and
mortality cohort studies, research industrial hygiene
characterizations, and control technology assessments.
The objective of the industrial hygiene portion of this
study was to characterize occupational exposures in die-
selized coal mines. The benefits of the characterization
are two-fold. First, meaningful toxicity testing requires
knowledge of realistic exposure levels from the occupational
setting. As a minimum, such levels must be known to allow
valid extrapolations for dose-response relationships. Second,
documenting existing levels in diesel coal mines establishes
a data base for future epidemiological studies, and evalua-
tion of control efforts.
METHOD
Certain chemcial species known to be present in diesel
exhaust were selected for study (3,4). These were parti-
culates, nitrogen dioxide, carbon monoxide, carbon dioxide,
cyclohexane extractable hydrocarbons, aldehydes, aliphatic
hydrocarbons, organic acids, and sulfates. The collection
and analytical methods are summarized in Table 1. The
sampling plan was to place a sampler on each major piece of
equipment within the breathing zone of the operator. Area
samples were also taken in the intake air, the return air,
at the haulage station, and at the feeder-breaker as located
on the map in Figure 1. Samples were positioned so as not
to interfere with mining operations. The samplers were put
in place before mining began and collected after mining was
finished to obtain full shift samples. Gas detector tubes
for CO, C02, S02, and N02 were collected on a time-
available basis.
1137
-------
Table 1. SAMPLING AND ANALYTICAL METHODS
Exposure
Method
Analysis
CO
N02
C02, S02
CI-GS Aldehydes
Aliphatic Hydro-
carbons
Total and Respir-
able Dust
Cyclohexane Sol-
uble Fractions
Organic Acids
Combined Carbox-
ylic Acids
Detector Tubes
Ecolyzer
Passive Dosimeters
Ecoloyzer
Detector Tubes
Impinger
Charcoal Tubes
Low Flow Pumps
DM 800 Filters
WO/W Cyclone
MSA Model G Pump
2.0 LPM
Silver Membrane
Filter B
Filter WO/W Cyclone
MSA Model G Pump
15 ml Impinger
NaOH 2 hrs.
MSA Model G pump
1 Liter/Min.
Colorometric (5)
Gas Chromatograph (6)
Gas Chromatograph (7)
Gravimetric
Sonication (8)
Gravimetric
Ion Chromatograph (9)
1138
-------
Figure 1. Typical mine section.
WORKING FACE
CONTINUOUS MINER
MACHINE
ROOF BOLT
MACHINE
HAULAGE
ROUTE
CONVEYER
BELT LINE
FEEDER BREAKER
KEY:
COAL
PILLAR
*
6_
CINDER BLOCK
PERMANENT WALL
TEMPORARY
VINYL CURTAIN
1139
-------
BACKGROUND
A coal mine is a dynamic system in which workers move from
place to place throughout the grid of tunnels. Machinery is
mounted on wheels or treads and is also mobile during a
shift. Since the objective of mining is the removal of
coal, the actual shape of the workplace also changes as the
mine advances into uncut coal. Ventilating air is directed
through the mine tunnels using cinder block walls, vinyl
curtains, and auxiliary forced air systems, Figure 2. Using
these techniques, the air flow pattern is changed to provide
clean air to the advancing work area. The ventilation
pattern is illustrated by the block diagram shown as Figure
3. Most of the mined-out tunnels in the studied mine were
20 ft. wide and 6-14 ft. high.
Coal was removed from the working face using a Jeffery
electric powered continuous miner machine. Two Wagner
Teletrams, model MTT-F20-18(S), were used to haul coal from
the continuous miner to the feederbreaker. These teletrams
were powered by Cat D330 diesel engines equipped with water
exhaust gas conditioners. These vehicles are approved under
Schedule 31, for use in gassy mines. At the feederbreaker,
the teletrams dump the coal where it is partially crushed,
loaded onto a conveyor belt, and transported to the surface.
Once the continuous miner has advanced to the last roof
support, it is withdrawn from that area and moved to the
next working face, where the mining process continues. An
electric powered roof bolter machine is then brought to the
freshly mined face. Roof supports are installed to prepare
for additional cuts in that area.
RESULTS
As seen in Table 2, production from the mine varied widely.
Day-to-day variability in the sampler, therefore, is to be
expected. Mining operations frequently encounter equipment
problems requiring maintenance and causing delays. This
problem is reflected in the various down times; when no coal
was being mined. Ventilation rates also varied throughout
the tunnel network and from shift to shift. This too is
responsible for some of the sample variability. Table 2
shows the ventilation rate for the section return to illus-
trate this point.
Table 3 presents the results of total and respirable dust
measurements. The total dust measurements are divided into
area and breathing zone samples. Both kinds of samples show
1140
-------
Figure 2. Ventilation pattern across the working face.
Figure 3. Ventilation network.
1141
-------
Table 2. PRODUCTION AND VENTILATION
Date Shift
5/15/78 1
5/15/78 2
5/16/78 1
5/16/78 2
5/17/78 1
5/17/78 2
5/18/78 1
5/18/78 2
5/19/78 1
Location
Intake Air
Haulage Area
Feeder Area
Return Air
Breathing Zone
Samples:
Continuous
Miner Op.
Roof Bolt
Operator
Teletram
Operator
Tons
Mined
840
340
600
840
924
984
456
1284
864
Table
No. of
Samples
6
5
6
6
7
3
6
Down Time Section Return
Minutes Ventilation x 1000 CFM
200 112.0
125
250 81.3
95
95 87.7
60
335 86.7
20
80 74.9
3. TOTAL DUST
mg/irr
Geometric
Geometric Mean Standard Deviation
0.49 3.08
1.33 2.17
0.96 1.54
1.64 4.24
2.15 3.63
1.13 2.36
0.91 3.54
1142
-------
the wide variability, reflected in the geometric standard
deviation, that would be predicted from the production
ventilation variability. The highest area level was ob-
tained from the return air, which is basically the exhaust
ventilation for that portion of the mine. The highest
breathing zone sample was obtained for the continuous miner
operator, which is considered a category that normally
experiences higher dust levels, whether the mine is diesel
or not. Respirable dust data are shown in Table 4. The
conclusions concerning total dust are the same for respir-
able dust, though on the average, levels are slightly lower
except for return air.
Nitrogen dioxide data in Table 5 are 8-hour time-weighted
averages (TWA). The NIOSH recommended standard for nitrogen
dioxide exposure contains a fifteen minute ceiling value of
1 ppm. While the observed eight-hour averages were well
below this value, there is no way to determine from these
measurements if the limit was exceeded during any fifteen
minute period. However, these results are important in
evaluating the diesel contribution to mine contamination
since combustion is the only nitrogen dioxide source in this
mi ne.
Data on cyclohexane extractable material shown in Table 6
were collected on the total and respirable dust samples.
Although the means are low, there are excursions above the
1979 TLV of 0.2 mg/rrr shown in the range.
The aldehyde data in Table 7 show relatively low levels on
the average; however, these may serve as an indicator of the
diesel exposure.
The data in Table 8 relating other measurements show the
relatively low levels of other species. The measurement of
total hydrocarbons, however, was affected by the high
moisture content of the mine atmosphere. Future studies
should therefore avoid this methodology.
DISCUSSION
Due to the variability of mining operations, a single day's
sample may not adequately reflect average concentrations for
research purposes. Intercomparison of samples, however,
from different work stations such as the feeder area or
haulage area should be done for same day samples, to elimi-
nate the large day-to-day variation.
Particulate levels are probably dependant on coal dust more
than diesel emissions because:
1H3
-------
Table 4. RESPIRABLE DUST
Location
Intake Air
Haulage Area
Feeder Area
Return Air
Breathing Zone
Samples:
Continuous
Miner Op.
Roof Bolt
Operator
Teletram
Operator
Location
Intake Air
Haulage Area
Feeder Area
Return Air
Breathing Zone
Samples:
Continuous
Miner Op.
Roof Bolt
Operator
Teletram
Operator
No. of
Samples
5
6
5
6
6
3
7
Table 5.
No. of
Samples
41
17
14
25
16
10
15
Geometric
Geometric Mean Standard Deviation
0.42 2.85
0.78 1.99
0.34 2.02
1.76 2.07
1.68 2.47
0.91 1.97
0.58 1.68
NITROGEN DIOXIDE
ppm
Geometric Geometric
Mean Standard Deviation Range
0.08 3.66 .01-. 30
0.25 1.68 .09-. 62
0.34 1.52 .17-. 68
0.26 1.56 .14-. 67
0.24 1.66 .11-. 46
0.03 2.80 .01-. 17
0.21 1.52 .08-. 45
1144
-------
Table 6. CYCLOHEXANE EXTRACTABLE FRACTION
mg/nr
Location
Intake Air
Haulage Area
Feeder Area
Return Air
Breathing Zone
Samples:
Continuous
Miner Op.
Roof Bolt
Operator
Teletram
Operator
Location
Intake Air
Haulage Area
Feeder Area
Return Air
Breathing Zone
Samples:
Continuous
Miner Op.
Roof Bolt
Operator
Teletram
Operator
No. of
Samples
12
9
19
18
6
2
7
Table 7.
No. of
Samples
8
7
15
16
5
2
3
Geometric
Mean
0.03
0.08
0.08
0.09
0.07
0.07
0.04
Geometric
Standard Coefficient
Deviation of Variation
1.97 82%
1.98 76%
2.13 74%
2.63 67%
2.59 70%
1.10 96%
2.93 72%
TOTAL ALDEHYDES
ppm
Geometric
Mean
2
31
32
20
13
0
11
Geometric
Standard Deviation
4.45
1.65
3.23
1.71
1.24
1.00
1.56
Range
0-68
16-55
0-144
8-46
0-18
0
7-17
1145
-------
Table 8. OTHER RESULTS AND MEASUREMENTS
Carbon Dioxide .04-.08%
Carbon Monoxide 0-2 PPM
Carboxylic Acids 2-6 ppb
Sulfates < 50 ug/m3
Hydrocarbons Charcoal tubes affected by
high relative humidity
Relative Humidity 87-95%
Temperature 54° F
Pressure 28.97-29.36 in Hg.
o The cyclohexane soluble .fraction is relatively
constant regardless of the dust levels as seen by
the coefficient of variation which does not change
greatly; and
o The observed dust levels are highest near the
sources of coal dust generation.
More indicative of the diesel emissions than particulate
levels are N0£ levels, which in this mine are due to
combustion. However in other mines these levels may be
due to blasting, thus interfering with diesel exposure
monitoring. A better candidate, therefore, is total alde-
hydes which show the expected pattern for exposure. High
levels are found around the tram and along the haulage areas
and low levels are observed for the roof bolter who normally
works in the fresh intake air.
SUMMARY
Using available equipment, we have begun to characterize the
diesel coal mine environment. We have established ranges of
contaminants that may be expected. We have identified
compounds which may be used to quantify worker's diesel
exposure. Finally, we intend to pursue an analysis of the
correlation between production, ventilation, and observed
concentrations in the dieselized coal mines.
1146
-------
REFERENCES
1. Stefanko, R., R. Ramani, and G. Kenzy. Evaluation of
Diesel Equipment Deployment in Underground Coal Mines,
Volume 1. Validation Experiments for Models of Diesel
Exhaust Contamination of Mine Atmoshperes. Pennsylvania
State University, University Park, PA NTIS Publication
PB 288 716, (1977).
2. U.S. Department of Labor, Mine Safety and Health Admin-
istration. Private Communication, DCT, (1979).
3. NIOSH Workshop on the Health Effects of Diesel in
Underground Mines. Unpublished Proceedings, (1977).
4. Stewart, D. B., J. P. Morgan and E. D. Painty. Diesel
Emissions and Mine Ventilation CANMET Mining Research
Laboratories Report MRP/MRL 77-59, (1977).
5. Palmes, E. D., A. F. Gunnison, J. D. Mattio, and C.
Tomzyk. Personal Sampler for Nitrogen Dioxide. Amer.
Ind. Hyg. Assoc. J. 37:570-577, (1976).
6. NIOSH Analytical Method P & CAM 235 NIOSH Manual of
Analytical Methods, 2nd Edition, pg. 235, (1977).
7. NIOSH Analytical Method P & CAM 127 NIOSH Manual of
Analytical Methods, 2nd Edition, pg. 127, (1977).
8. Bel inky, B. Analytical Method Devised for This Study,
(1978).
9. NIOSH Analytical Method P & CAM 127 NIOSH Manual of
Analytical Methods, 2nd Edition, pg. 127, (1977).
1147
-------
PANEL DISCUSSION
ON THE
HEALTH RISK ASSESSMENT OF DIESEL EMISSIONS
Dr. Roy Albert, Chairman
Cancer Assessment
Office of Research and
Development - USEPA
Washington, DC
Dr. Allan Moghissi
Environmental Scientist
Office of Research and
Development - USEPA
Washington, DC
Dr. William Balgord
Environmental Resources
Technology, Inc.
Brookfield, CT
Dr. Jaroslav J. Vostal
General Motors Corporation
Research Lab.
Biomedical Science Department
Warren, MI
Dr. Albert:
We'll move right into the panel on health risk assessment of
diesel emissions. Let me introduce the panel. Next to me
is Dr. William Balgord who is from Environmental Resources
Technology Incorporated. Next to him is Dr. Alan Moghissi
who is from the Office of Research and Development at EPA.
Furthest down the line is Dr. J. Vostal from the General
Motors Corporation Research Laboratory.
The ground rules for the panel are that each speaker will
take five or ten minutes to make a brief statement; we will
have a few minutes of discussion among the panel members
themselves and then open the panel to questions and abuse
from the floor. I'll start off, Dr. Balgord will be next,
then Dr. Moghissi, and Dr. Vostal.
I will speak from the standpoint as Chairman of the CAG or
Carcinogen Assessment Group of EPA. This is a group that
1148
-------
was called into being by the guidelines for risk assessment
that were adopted by the Environmental Protection Agency in
1976, which has provided guidance for the EPA approach to
assessment of carcinogens since then. This group has been
involved in doing the risk assessments for most of the
agency since 1976.
I wanted to very briefly present a point of view that is
embodied in the interim guidelines for risk assessment on
how health risks from carcinogens are evaluated and then go
on to talk about the specific issues as they are currently
perceived with respect to diesel particulate emissions.
I might say that the guidelines for the assessment of cancer
risks that have been adopted by the EPA are consistent with
those developed by the Interagency Regulatory Liaison
Group; the IRLG is made up of the principal regulatory
agencies dealing with regulation of carcinogens. Within the
last few months a document put out by the Federal Regulatory
Council dealing with a national cancer policy also supports
the EPA approach to risk assessment.
The approach to the assessment of carcinogenic risks is no
different from the assessment of any other risks. It
involvew two questions, How likely is the risk to occur and
second, what's the impact if it does occur? Namely, How
likely is the agent in question to be a carcinogen and
second, If it is assumed to be a carcinogen, how much cancer
is it likely to produce?
The assessment always tries to give these two answers. The
first is a qualitative evaluation and the second is a
quantitative evaluation. The qualitative evaluation is
based on the weight of evidence, the nature of evidence, its
scope and quality. A judgement is made about the weight of
evidence recognizing that the best evidence consists of
repeated epidemic!ogic studies that show the same effects
under different circumstances backed up by animal studies
which identify the causal agent. At the other extreme we
can have a marginal response in one sex, strain, or species.
Evidence is regarded as suggestive that is based on short-
term in vivo and in vitro tests, such as mutagenesis or cell
transformation and initiation studies in the mouse skin. I
don't believe that any regulatory action has been taken on
the basis of suggestive data.
So the qualitative judgement is based on the weight of
evidence approach. It clearly accepts animal tests as a
surrogate for human response as well as the appropriateness
of testing at high dose levels.
1149
-------
The second aspect of risk assessment involves the quantita-
tive estimate of risk. It is clearly understood that at the
present state of our knowledge, any quantitative assessment
of risk at low levels of exposure can't be anything more
than a crude ball park estimate.
Quantitative assessment requires exposure data and some
means of extrapolating from high doses where observations
are made, either in human epidemiologic studies or animal
studies, to the low doses associated with population expo-
sure.
In essence, the EPA, supported by the IRLG, has adopted the
linear nonthreshold extrapolation model as the model of
choice. This is the extrapolation model that has the most
scientific support although it is generally recognized that
the foundation for this or any other extrapolation model is
quite thin.
The linear nonthreshold model also provides a reasonably
conservative approach to estimating risks at low exposure
levels. For a regulatory agency as a guardian of public
health, this is an appropriate approach.
I think it is clear that the linear nonthreshold dose-
response extrapolation model is the dominant concept in
quantitative risk assessment; it involves the notion that
there is no such thing as a safe dose and produces cal-
culable numbers of cancer cases even at extremely low levels
of exposure where large populations are involved. Indeed
this concept effectively swamps out consideration of any
other health risks not associated with the acceptance of
a similar extrapolation model. In short, with the linear
nonthreshold dose-response, if one protects against car-
cinogens, one can be pretty sure that one is going to be
protecting against almost everything else.
It is clear that in the quantitative estimation of risks it
is far better to use epidemiologic evidence than animal data
because of the uncertainties in the extrapolation from
animals to humans in addition to the uncertainties of
extrapolating from high levels to low levels.
The assessment of diesel health risks is clearly in an
evolutionary state. I think one thing that has come out of
this meeting is the recentness of most of our information.
At the present time we have to deal with information that
has a lot of uncertainty. The impetus as you well know for
concern about diesel emissions comes from positive mutageni-
city tests plus the knowledge that when one combusts any
organic material, carcinogens are going to be produced.
1150
-------
The initial risk assessment for diesel emissions had a
strong element of urgency associated with it. By early
1980 very large investments were going to be committed by
automobile companies, and EPA had to give some signal as
to whether it regarded the health hazard in diesels as
important enough to be taken into account in the planning
by the automobile industry.
We were faced with the need for coming up with some sort of
risk assessment strategy that could produce a defensible
evaluation in a matter of months. There certainly were no
animal studies available that could form the basis for an
assessment.
The approach that was used was to take into account the
available epidemiological evidence of lung cancer in res-
ponse to exposure to combustion products since combustion
products are more or less analogous to diesel exhaust
particulates. This, in effect, involves lung cancer in
coke oven workers, lung cancer in cigarette smokers, and
lung cancer in roofers.
The point was to use these data to obtain the linear extra-
polation dose-response slope for the purpose of estimat-
ing the magnitude of a cancer risk, given the exposure
estimates for diesel particulates and taking into consi-
deration the relative potencies of diesel particulates
with respect to coke oven emissions, cigarette smoke tar
and roofing tar.
You heard a presentation of the research results that have
been prdouced in the last few months. I must say that I
have to compliment the people involved. In think an extra-
ordinarily effective and very high quality effort is being
made.
I think it is fair to say that the data that has just been
assembled for presentation has not really been digested. It
is too early to say much about the resultss which are incom-
plete in any event. I can say that we are certainly looking
forward to a close scrutiny of the material that is being
developed as the basis for going ahead with the carcinogenic
risk estimates. That is essentially what I have to say, and
I pass to you, Dr. Balgord.
Dr. William Balgord:
E&RT has been involved in the diesel health issue for about
the last two and a half years, dating back to our participa-
tion in the DOT fuel economy proceedings that established
1151
-------
the 1981-85 fuel economy standards for light-duty vehicles.
We have had a sustaining interest in the health implications
of diesel emissions and share your impression that a great
deal of information has been generated at this point. At
the time we first looked into this, there wasn't very much
available on the subject. But I'd like to address what I
see right now as some barriers to completing the tasks
necessary for arriving at a quantitative risk assessment
that could be generally supported. Some of these items
are rather specific - so please bear with me.
I am concerned with a general lack of representativeness
regarding the way many of the (animal exposure) tests
are being conducted with respect to the engine duty cycle,
that is, the way the engine is operated and how that bears
to the way vehicles are operated in the real world. In
reading several papers, I understand that some of the
engines are being operated at constant speed during the
sampling periods and, correct me if I'm wrong, in no case
was consideration given to collecting samples during
cold-start engine operation. Duty cycles vary considerably
from lab to lab as do sample collection methods. As a
suggestion of this difference, I note that at one labora-
tory the concentration of N0£ in the (animal) exposure
chambers is on the order of 5 or 6 percent of the total
NOX as compared with concentrations that are fairly
typically 15 to 20 percent in other labs. If N02 is a
critical ingredient, along with particulate, I really
feel we ought to get this together so that the results
from different laboratories can be compared when we are
through. I'm impressed that there have been only two engine
types and vehicle types tested extensively. Admittedly when
the work began, there were not that many types of diesel
engines available in passenger vehicles to choose from.
But this is changing. Now Volvo has one. Peugot and
the Japanese are building diesels and so is a domestic
manufacturer.
I think it would be important for interchange of samples
to take place, particularly between General Motors (which
has undertaken a very ambitious program) and EPA. In this
way, we could have a feeling that the samples used during
the course of the experiments had the element of compara-
bility. It strikes me that all the particulate sampling is
done with very little time delay from the moment the exhaust
leaves the engine (or tailpipe in vehicle tests) to the
point where the animals are exposed (or samples taken for
cell studies).
I believe General Motors says this is on the order of 15
seconds. Again correct me if I'm wrong, but I don't see
this as materially different from the situation in the EPA
1152
-------
laboratory. Now this is all well and good from the stand-
point of applicability to streetside exposures - a person
walking near vehicular traffic along a street in downtown
Los Angeles or in Manhattan - this may represent this
situation. But I'm not convinced that emissions that have
resided in the atmosphere for 15 minutes, a half hour, or
several hours relate to the juvenile exhaust products
pertaining to the diesel exposure studies described at
this meeting. There is really too much opportunity for
subsequent chemical transformation. This aspect needs
to be considered more fully.
The engines as operated during the tests are very well
attended to. We need to take account of the fact that the
general population of vehicles, historically, has not had
that kind of careful attention. Realistically we can't
expect a car owner to bring in his diesel every month or
so to have the injectors changed. The cost would be pro-
hibitive.
A related issue is what we can expect in terms of changes
in diesel fuel over the next four to six years. Currently
there is quite a latitude in the ASTM specification for
commercial diesel fuels. Typically, fuels do not vary
much from supplier to supplier. But there is latitude
within the specification for considerable deterioration
in the quality of the fuel. And what with the uncertainties
of the world petroleum supply, prospects are that we may
have to turn to syn-fuels for at least part of the diesel's
requirements. The experimental results we are now generat-
ing with very tight specification fuel - low sulfur and
whatever else - may not give us an adequate picture of what
these same diesel vehicles are going to emit in the near
future.
I mentioned a concern with atmospheric transformations of
PAH. This ties in directly to a matter of concern many of
us have with the so-called artifacts of sampling. No doubt
artifacts sometimes occur and have to be guarded against to
avoid jumping to unwarranted conclusions. However, I would
suggest that PAHs as a class undergo a series of chemical
transformations in the atmosphere which eventually convert
them to innocuous compounds. The concern is with how they
get there. PAH, the hydrocarbon itself- as we have heard
many times during these meetings - requires activation
before it can express its inherent mutagenicity. On expo-
sure to the atmosphere, the PAH may be altered in such a
way that human populations will be exposed to a series of
intermediates, some of which require no prior activation.
We need to progress in selecting and developing dispersion
models which would apply to diesel emissions, and for that
1153
-------
matter, to related sources. How are these to be dealt
with and how are we to bring them into a suitable compari-
son with the contributions from diesels in cogeneration
applications which are just now coming on the scene? From
what I can tell, this is a subject which has not yet been
addressed.
Concerning estimates of risk to human populations, we
need models which adequately take into account all segments
of the exposed population according to various medical
histories, susceptibilities to disease, and potential
susceptibility to carcinogens. Unless this is done, there
is no way to justify picking out an individual as typical
of the population and attempting to make a simplistic
assessment of human risk applied to X million like indi-
viduals.
One particular item came up a number of times relating to
the relative potency of particulate derived from gasoline or
diesel engines. We need to lay out ground rules on how
these apples and oranges should be compared. From what I
understand of EPA's experiments with the gasoline Mustang,
the carburetor was maladjusted. I don't see that as repre-
sentative of a typical gasoline vehicle in today's economy.
With gasoline prices having reached current levels, the
average driver can ill afford a sticking choke. He'd have
it fixed.
I would also encourage EPA to begin epidemiological studies
addressing various populations occupationally exposed to
diesel emissions.
Dr. Albert:
Thank you Dr. Balgord. Dr. Moghissi.
Dr. Moghissi:
I shall discuss the importance of fuel and fuel additives
rather than the engine as sources of pollutants.
Let me define what I believe risk assessment ought to be.
Risk assessment, or more precisely, environmental health
risk assessment should consider the release of a toxicant,
its transformation in various environmental media, the
direct and indirect exposure to humans and finally, the
health affects as a result of the exposure. As far as the
impact of diesel is concerned, one would want to consider
1154
-------
the components of the exhausts as well as those of the
fuel -- individually analyzed and identified, their trans-
formation in the atmosphere, and their health effects as a
result of incorporation.
Obviously that is not very realistic. Recently I made a
cost estimate, which showed the cost of such an investiga-
tion to be unacceptably high.
What I am going to say is the result of an effort to write
regulations for fuel and fuel additives, and what I will
describe are potential options or potential methods for
regulating fuel and fuel additives, which have an important
impact on regulating the diesel.
Starting with 1975 EPA has registered fuel and fuel addi-
tives. In 1977 the Congress -- not very pleased with the
efforts of EPA -- added a section to the Clean Air Act
requiring EPA to develop procedures and protocols for
testing fuel and fuel additives for their health and
environmental effects.
I have prepared a manuscript on this subject, which will
be printed in the proceedings of this symposium. The
essential features of the manuscript will be discussed
here.
The fuel and fuel additives should not adversely affect
the emission control system. The importance of this
fact is obvious. One would also want to have the fuel
and fuel additives analyzed particularly for elements
other than C, H and 0 because presumably, unless some
elements get stuck in the engine, they will appear in
the exhaust. One would expect to find nitrogen in the
exhaust in elemental form, as nitrogenoxides or maybe
some nitrogen-containing materials such as hydrocarbons.
Therefore the analysis of the original material is im-
portant.
In evaluating combustion products one has various alterna-
tives:
Alternative number 1 is to disregard the combustion pro-
ducts. A number of people in this audience would appre-
ciate this alternative. The idea being that we are already
regulating the so-called regulated pollutants and if an
engine does not exceed the prescribed quantity of hydro-
carbons, why bother? In addition, since we know the organic
content of the fuel, why bother with the analysis of combus-
tion products? The disadvantage of this alternative is
its lack of detection of a fuel that generates harmful
materials.
1155
-------
The second alternative is the exact opposite, namely a
complete chemical analysis and testing of all combustion
products. This option would include transformation and
health effects testing on every compound. This alternative
is not very realistic because, although ideal, it is very
expensive.
The alternative three would require that one could collect
the exhausts and somehow try to fractionate them much
like the method presented by Dr. Huisingh. For example,
the exhaust mixtures are taken and fractionated into
basic, neutral and acidic fractions; subsequently these
fractions are evaluated for their health effects. The
trouble with this alternative is that if a compound is
readily decomposed in the atmosphere, it is getting the
same credit as a compound that would not be readily de-
composed. In fact, unless one analyzes the materials,
it would be very difficult to conduct a risk analysis
because one would not be able to look at the transforma-
tion and various other mechanisms that are required in
the risk analysis.
The fourth alternative would set up a threshold for compo-
nents of the exhaust. For example, all components falling
below a certain percent (1, 0.5, 0.1, etc.) would not be
considered for testing with the option that, if for whatever
reason, one finds a material that is of significance,
although it falls below that threshold, one would still
consider it for testing.
Although there may be other alternatives, a closer look
indicates that they are mostly variations to these basic
four alternatives.
Once one establishes the candidates for testing, they
can be subjected to the tests established by various
offices of the EPA. The Office of Toxic Substances in
EPA has developed certain protocols for the health effects
testing. Some of these protocols are proposed, others
are in various stages of preparation. After completion
of these tests, one starts to make a risk analysis by
establishing the quantity that is released, the air con-
centration, atmospheric chemical reactions, and then
one selects an area where the people might be exposed
to these materials such as gas stations and establishes
potential acute effects. In the case of carcinogenic
effects, because our present perception indicates that it
has no threshold and is linear with dose, it doesn't make
any difference if one exposes a smaller number of people
to relatively large concentrations, or a larger population
to smaller concentrations. Statistically speaking, one
comes up more or less with the same number of effect.
1156
-------
The prerequisite is that exposure occurs at the linear
and dose rate independent region. . Once one does all
of these, one comes up with a risk assessment for can-
cer.
Of course there are a number of other diseases aside
from cancer. However, the mechanism for those effects
would be basically the same, provided one knows at what
point the threshold is and provided one can make certain
assumptions relative to the shape of the dose effect
curve.
As of this date the discussion centers on the triggering
point of these tests. Some of these tests are rather
expensive and should not be needlessly mandated. In
addition, the question is, which one of these options
would one want to take? Because some of the points re-
quire a resolution, I'm sure we would be very glad to
hear from you. Thank you.
Dr. Albert:
Thank you Dr. Moghissi. Dr. Vostal
Dr. Vostal:
First of all, I have to admit that all of us who have
participated in the meeting during the last two days
have been quite impressed by the large amount of research
effort involved in the study of potential health effects
of diesel emissions. Particularly, we appreciate how
much work has been done on the in vitro biological acti-
vity of diesel particles by the research team of the U.S.
Environmental Protection Agency. In my comments, I would
like to concentrate on several items which are of impor-
tance if the objectives of all the research programs on
the potential health effects of diesel emissions are to
be achieved.
The U.S. EPA Mobile Sources Research Program published
on January 24, 1979 formulated the governmental research
objectives and indicated that
"Recent test results and the usual carcinogenic proper-
ties of soot from incompletely burned organic material
have raised the question of whether particles emitted
1157
-------
from diesel engines are carcinogenic. From this, two
further questions arise, to wit:
1) What is the increased risk to public health if a
significant fraction of new cars have diesel
engines?
2) And how can the emissions from diesel cars be made
less hazardous?"
Research programs aiming to answer the questions were
outlined and presented to the scientific, industrial and
other interested communities for their information and to
receive their comments and concerns in a public Scientific
Review Meeting at Arlington, Virginia, on December 12-13,
1978. We have had the possibility to comment on the
appropriateness of the program during this meeting and we
feel that some of our comments represent a serious criticism
of the U.S. EPA research approach.
Our rationale for the risk assessment of potential health
effects of diesel emissions differ from the EPA approach in
at least three major research objectives:
First, we feel that mutagenicity assays or other short-term
laboratory tests alone cannot serve as a sole basis for
health effects assessment; second, that experimental ani-
mal inhalation models must be based on field-relevant
conditions; and third, that long-term human exposure studies
are the most significant factor in risk assessment.
In this meeting, we have heard the preliminary data from
studies conducted by the U.S. EPA, as well as by the indus-
try. We have learned that organic solvent extracts of
diesel particulates have been extensively tested for muta-
genic effects in a wide array of short-term laboratory tests
with unequivocal results; we have listened to the first
reports on the observed lack of significant health effects
in the long-term animal inhalation experiments; and finally,
we have been told in the last evening session, that the
organic solvent extracts of particulates obtained from one
type of diesel engine produced benign skin tumors when
applied on the skin of a sensitive strain of mice in large
doses, whereas extract from other engines displayed negative
or only marginally positive results.
On the other hand, we have also heard presentations that
bring the interpretation of the obtained laboratory data
into a completely different perspective. Thus e.g., Dr.
Siak presented an approach which questions if it is scien-
tifically appropriate to use an organic solvent to extract
1158
-------
the hydrocarbons absorbed on the core of the diesel par-
ticle. The strength of such solvent to solubilize the
organic matter is many times stronger than anything found
in the human body and, therefore, cannot be used to sim-
ulate mechanisms by which the biologically active compounds
can be released in the organism for interaction with the
sensitive cells of the respiratory system. It would seem
more appropriate if - in addition to the dichloromethane
extract - extracts obtained by real biological fluids
were used in the skin painting testing, particularly if
the results of these tests are to be used in the risk
assessment process.
Similarly, I feel that the question of the maximum toler-
able dose used widely in many experiments with the inten-
tion of obtaining a positive biological response in the
shortest possible time needs to be reexamined. In my
presentation yesterday, I showed that the calculated pul-
monary dose of the diesel particulate deposits under the
highly exaggerated exposure conditions exceeds the real-
life situations that may be expected in the year 2000 in
large urban communities, by four orders of magnitude. As
a result, the exaggerated local accumulation of milligram
amounts of particles may produce entirely different biolo-
gical responses than those expected when only microgram
amounts of diesel particulates are in prolonged contact
with the sensitive respiratory cells.
Finally, data presented last night on the comparative
studies of relative mutagenic and carcinogenic potencies
of cigarette smoke condensate, coke oven emissions, roof-
ing tar and organic solvent diesel particulate extract
seems to indicate that the expectations that such a com-
parative process can be used in the risk assessment of
potential health effects of diesel particulates have
been premature. At least, it will not be an easy role
for the governmental research teams to use the compara-
tive data for extrapolation of well-documented epidemi-
ology studies on carcinogenicity of cigarette smoking
into a risk assessment process predicting the potential
carcinogenic effects of diesel particulate exposure, if,
contrary to all epidemic!ogical experience, the relative
potency of cigarette smoke condensate is completely nega-
tive in skin painting carcinogenicity tests and the ben-
zo[a]pyrene content is lower than 1 nanogram per mg of
extract. Similarly surprising is the finding that we
cannot obtain a sample of the organic solvent diesel par-
ticulate extract that could be declared as a representa-
tive sample of diesel emissions. Large differences in
biological activities of soot from different types of
engines ranging from completely negative to clearly posi-
tive test results indicate that either our tests or our
1159
-------
sampling procedures in the form we have in our hands today
are not satisfactorily representative to provide a mean-
ingful conclusion on the risk involved in the wider use
of diesel engines on our roads. Considering the wide
variations in the chemical composition of the product
obtained from using identical fuel in different types
of engines, as was indicated by the unbelievably large
range of benzo[a] pyrene concentrations (from 2 to 1200
nanograms per mg extract in different engine samples in
Dr. Huising's presentation), we have to admit that the
present way of testing the diesel emissions needs to be
improved before we can make any conclusions on their po-
tential harm to human health. Similarly, the proposed
comparative methods of short-term test extrapolation
should not be accepted unless samples which are compared
for relative potencies are compatible with the level
and quality of emissions under which the epidemiological
data were obtained in the exposed populations 15-30 years
ago.
In this respect, I would like to return to the third
point of the research objectives presented in the intro-
duction of my comments and express my concern that the
U.S. EPA approach has unjustifiably ignored a set of epi-
demiology data which can be important in the risk assess-
ment process. I think that we have learned a lot about
the potential health effects of diesel emissions from
this afternoon's presentation by Dr. Waller from the
Medical Research Council in England. Personally, I feel
that in spite of the stated shortcoming, the London Trans-
port Study remains the most meaningful and significant
information for the risk assessment of potential health
effects of diesel particulate that we have in our hands
today, particularly since we cannot expect the results
of other proposed epidemiology studies such as the inves-
tigation of the railroad workers described by Dr. Schenker
today, sooner than about three years from now.
I don't want to repeat Dr. Waller's presentation, but
I would like to emphasize the simple fact that despite
the several thousand diesel buses and more than 10,000
fully dieselized taxicabs that have been used in the
London city area during the last decades, the general
incidence of lung cancer in the London population has
been steadily decreasing since the 1960's, a fact which
fully concurs with the conclusion of the British Medical
Research Council study on garage workers exposed to
high concentrations of diesel particulates. Unless we
can have other significant evidence which would contra-
dict this data, I feel we cannot abstain from its use
in the risk assessment process, particularly when we
realize that we have to make the important decision
1160
-------
on the increased use of diesel engines today. The scienti-
fic community has been repeatedly assured that the automo-
tive industry does not have any intention of increasing
diesel engine use on the highways if the public health would
be endangered from diesel particulate- emissions, and I would
like to conclude by a direct quote of Mr. E. M. Estes,
President of the General Motors Corporation, published in
the Environmental Health letter on November 15, 1979:
"Despite all the research, it is unlikely that by the
end of 1979 we can say flatly and without any qualifi-
cation that the diesel possess absolutely no threat to
human health, regardless of how remote -- if we can
ever say that about anything. I do hope that we will
soon be able to say that there is no unreasonable
hazard -- that the benefits of using diesels in cars
and light trucks outweigh any possible unsubstantiated
risks."
As you see, the automotive industry depends fully on the
wisdom and realistic views of the scientific community in
making decisions important for the benefits of the society
and all of us today. Thank you.
Dr. Albert:
Thank you Dr. Vostal. I think in view of the lateness of
the hour and since the panel has already had its say, maybe
we ought to open it up to the floor now. Does anybody, --
Question: Mr. Fitiansky (?): I'm Steve Fitiansky from the
Environmental Science and Technology Magazine in the New
York Times. First I'd like to congratulate Dr. Moghissi on
his caution by quoting price figures. He's quite right.
You never can tell when a member of the press is going to
be located in the audience.
Dr. Vostal has presented some of his interpretations of the
various results presented here and I was wondering if some
of the other members of the panel could comment on whether
they believe any consensus has emerged.
We've seen a lot of results which appear to be inconsistent.
I wonder if you would comment on first whether there is any
concensus on that mutagenicity, second carcinogenicity, and
third noncarcinogenic respiratory affects.
The second question is whether you believe that mutagenicity
itself is a matter of concern regardless of carcinogenicity.
Answer: Dr. Albert: Well I can start this off. I think
one of the points that was raised by Dr. Vostal had to do
1161
-------
with the importance of epidemiology. Certainly nothing
can be said except to support the position. For example,
if the continued evaluation of the London garage workers
is negative, or even if it isn't and some good exposure
estimates are generated, which don't exist at the moment,
one can use the data to put an upper limit on the magni-
tude of the risks. Negative data on a few thousand people
is not the last word when one is dealing with exposures
that involve millions.
In the past, negative epidemiology data on carcinogenic
responses has played a very important role in risk assess-
ment in terms of putting a lid on any interpretations
associated with animal data and essentially putting an
upper limit to the response.
So far as concensus about carcinogenicity and mutage-
nicity is concerned, I don't think there has been an
opportunity for people to consider the data. Most of
it has just been presented here at the meeting and most
of it I'm sure is brand new to virtually everybody in
the audience. I would say there hasn't been time to con-
sider the matter in light of whether or not there is a
concensus.
Does anybody have anything to say?
Answer: Dr. Vostal: There was one important observa-
tion which came out from the discussions during this
meeting. A broad spectrum of different tests was used
in the U.S. Environmental Protection Agency studies to
investigate the mutagenic effects of diesel particulate
extracts. Surprisingly, we have seen that widely dif-
ferent test results were reported from testing of iden-
tical samples. Positive mutagenic effects found in one
test contrasted with entirely negative results in other
assays. Consequently, a tested sample could be declared
mutagenic or nonmutagenic, depending on the assay used.
The inconsistency of the observed mutagenic effect was
not easily explained, even by the original investigators
who developed the procedures and conducted the tests.
I feel that it is necessary to further look into the
underlying mechanisms of the postulated mutagenic effects
in various biological assays before we correlate the
results of short-term laboratory mutagenicity tests with
the possibility of a meaningful prediction of the poten-
tial carcinogenic hazards of chemical exposures in human
populations.
Answer: Dr. Albert: I think on that score it is worth
pointing out that the material that was presented here
really represents a pioneering effort to apply a battery
1162
-------
of short-term tests coupled with whole animal studies to
the evaluation of carcinogenic risks.
I think it really is a new effort and undoubtedly has a long
way to go, but I think even on inspection of the data, it
has obviously got a lot to contribute.
Answer: Dr. Moghissi: Well, I think Dr. Albert said it as
well as anybody could say it. It's just a brand new field
and some of the papers presented here are so new, that one
has to sit down and digest them. I think the bottom line is
the question if diesel engine exhaust is carcinogenic?
That's really what the people are asking.
You don't have to be very smart or do many experiments to
find out that there are carcinogenic materials within the
exhaust of the diesel engine as they are in many other
combustion products. The relevant question is not if the
diesel engine causes cancer in population, rather how
many cancers in addition to the naturally occurring or
spontaneously occurring cases.
As I see it, the job of the science is to give at least a
semi-quantitative value as to the number of those added
cancers. After that the scientists are no smarter than
anyone else to say if the society is willing to accept those
added numbers of cancer cases. The decision of the risk
at that point is no longer a scientific but a societal
one that our society must decide by appropriate societal
representation.
I don't think there is any question that there carcinogenic
materials in diesel exhaust. You analyze it and you find
that there are several papers and there are many other fully
nuclear automatic materials, hydrocarbons and heterocyclic
(?) materials in there, and you can imagine that it is.
Question: Speaker: IIT Research Institute. I wonder if
Dr. Vostal realizes that he weakens the scientific basis of
the number of his arguments by his constant insistence upon
doing the requirements for doing low dose and animal testing
for carcinogenics and that it has been well accepted by
every organization of regulatory nature - by the National
Cancer Institute's Bio (?) Program, by FTA (?) - and hope-
fully by EPA, that you cannot get anything but a negative
answer if you do low dose testing in 50 animals. If
you hope negative answers, that is the ideal way to get it,
but otherwise you have to give massive doses.
Answer: Dr. Vostal: This is a very important point. How-
ever, a basic question still remains open if a short expo-
sure to a high dose produces effects identical with the
1163
-------
long-term breathing of extremely low concentrations. In
the testing for potential carcinogenic effects, the U.S.
Environmental Protection Agency proposed and used the
linear, nonthreshold model as the most conservative approach
to protect the innocent, and the discussion of the validity
of this approach would require a longer discussion. How-
ever, the same concept should not be applied for the poten-
tial local effects of inhaled air pollutants on the respira-
tory system where the effects of high concentrations cannot
always be found after low doses in spite of extremely
prolonged exposures. This was the main reason why our
long-term diesel health effects program incorporated three
different levels of exposure instead of only one maximum
tolerated dose used by the U.S. Environmental Protection
Agency group.
Question: Dr. Kerd: I'm Dr. Kerd with the United Mine
Workers. Dr. Vostal, I gathered from point three of where
you said the information should come from, if that you are
advocating continuing long-term human exposure studies.
Would you clarify that?
Answer: Dr. Vostal: Originally, we have had plans to
study the respiratory effects of the diesel particulate
exposure in short-term human volunteer studies using well-
controlled low-level exposure chambers, but the fact that
diesel particulates were suspected as a human carcinogen
prevented the realization of the project. However, other
studies could be done on population groups who had been
exposed to various diesel concentrations in their occupation
during the past. The London Transport study, which I have
mentioned briefly is only one of them. If we can identify
any singular source of data which would help us to analyze
potential health effects of prolonged diesel exposures, we
are prepared to use the retrospective approach as effec-
tively as possible.
Question: Dr. Kerd: I'm quite in agreement with that,
however without the word retrospective in there, it becomes
almost, if you read it at first glance, as though you are
advocating a continued exposure in order to get the infor-
mation necessary to permit the use of diesels which I think
you ought to correct on your slide.
The other thing is the fact that I find that your statement
by the President of this rather large corporation to which
you referred, is saying that we ought not to let these
things delay the use of diesels above ground. I find it
very comparable to the position that co-operators are now
taking and advocating the use of diesels underground in coal
mines. They are not willing to see what the results are by
the tests that are being made at the present time. They are
1164
-------
so anxious to rush these instruments of evil underground,
that it is almost appalling, and there is tremendous
resistance building up against this use of humans as
guinea pigs.
Question: Dr. Weisenberger: Dr. Weisenbeger from Cummins
(?) Engine Company. I would like to ask the panel to
reflect on one sentence which we see- in both the IRLG
documents and the CAG documents and that is the statement
that the linear-threshold has the most scientific basis.
Answer: Dr. Albert: I indicated what I thought was the
available evidence to support it, or maybe I didn't. One
of the major lines of supporting argument is the close
correlation between mutagenesis and carcinogenesis. Both
appear to reflect attack by agents on DNA, the cell's
genetic material, with an expression in terms of either
mutation or neoplastic transformation or even cell death.
There is very strong evidence from mutagenesis, particularly
in single cell organisms where one can study very large
numbers of such organisms, that there is a linear non-
threshold dose response pattern. This emerged first with
radiation and then with chemical carcinogens.
The second point is that there is emerging evidence from
analysis of the two stage carcinogenic process in the
mouse skin. Namely, that the first stage has a linear
nonthreshold dose-response pattern. Those who heard Dr.
Slaga's presentation must have been struck with the char-
acter of the dose response curves that he showed for skin
irritation. These were very consistent with a linear
nonthreshold dose response; indeed these observations
have been made in other laboratories too, my own for
one.
Third, there is a limited body of epidemiologic evidence
which supports the possibility of the linear nonthreshold
dose response in the sense that the observed evidence
is consistent with it. This first emerged in the analy-
sis of leukemia cases from atom bomb survivors; there
are a few other instances where one can interpret the
data as consistent with the linear nonthreshold response
pattern.
Question: Mr. Fenelli: Benson Fenelli from the University
of Cincinnati. From this meeting that impression has been
put on carcinogenics and mutageneics of diesel exhaust. I
think other after affects have been neglected. General
1165
-------
toxicological studies have been few - effects status
have not been satisfactorily done or they have not even
been attempted. Could you comment, other than the car-
cinogenic and mutagenic study, why don't we center our
attention to more health effects in general?.
Answer: Dr. Vostal: We have seen during the discus-
sions in this meeting that attention has been also paid
to other potential health effects of diesel emissions,
rather than concentrating only on mutagenesis and car-
cinogenesis. Thus, our laboratory presented several
papers yesterday that discussed possible changes in
lung function, morphology and biochemistry. We have
studied the effects on enzymatic composition of tissue
cells and blood plasma and we have other ongoing studies
centered on potential effects in immunological response.
As you may remember, so far we have not found any effects
after nearly twelve months of exposure that would be of
serious concern or specific for diesel particulates.
Question: Speaker: I have two comments. The first relates
to the comments you offered Dr. Albert on the linear non-
threshold model which certainly would agree with most of
what you said but I think one comment was perhaps subject
to misinterpretation. You said that the use of the model
swamped out any other risk projected using alternative
models and thus it would protect against anything else.
In essence, to paraphrase what you said, I think that
you offered that as related to that specific agent.
Clearly we have I think in this specific situation an
option in which we are not looking just at control of
the specific agent, but we are forced to make compari-
sons to other risks, risks that may not be adequately
described with the linear nonthreshold model, and so I
think the situation becomes very much more complex. It
falls really on the heels of the last gentleman's comment
as we talk about effects that may be produced by other
agents.
I think that is necessary to keep this in mind as we look
at the use of the linear nonthreshold model, especially
in comparative assessments.
The second is that it seems to me that one of the les-
sons that comes through the last three days is something
that is perhaps related to the study of other materials.
We've heard data indicating very clearly the presence
of potent mutagens within diesel exhaust extracts; how-
ever, the majority of the data at hand today, if not
negatives, are at least indicating the extent to which
diesel exhaust particles are relatively weak carcino-
gens.
1166
-------
I think this is important as we look to the issue of other
materials that are introduced into our environment and
that are in our environment now, and may be introduced in
the future. As most of us I think are aware, we have to
exercise a great deal of caution in the interpretation
of short-term screening tests in terms of mutagenicity.
They in themselves cannot be the basis for taking final
action on a product whether it should or should not be
present. I think this certainly concurs with your com-
ments and several of the comments of other panel members.
Answer: Dr. Albert: I would agree. This relates to
the question whether mutagenicity is enough to raise a
concern about regulatory action. I think the essential
ingredient that is missing from mutagenicity data derived
from short-term tests is the issue of whether or not there
is penetration of the agent in question to the germ cells.
There really hasn't yet been a coherent approach to the
regulation of mutagens as with carcinogens. Guidelines
for assessment of mutagens in the Environmental Protection
Agency are being attempted at the present time.
Question: Mr. Culver: Allen Culver from Research Triangle
Institute. I have been amazed in the past six or seven
years by the amount of cooperation that has occurred between
analytical chemists and biologists in investigating problems
such as the diesel particulate problem, and I also respect
very much Dr. Vostal's comments; and in order to help
resolve some difficulties in that area, the ball can be
thrown a little now to the chemists.
We know what compounds are heavy-duty carcinogens and
we know what they are; we have a little bit of an idea
as to how much is required to produce a tumor. What we
don't have a good idea of at the moment is a good quanti-
fication of the presence of these compounds in extracts
from diesel particulates; however, the extracts may be
made.
I think that the chemist serves a very important func-
tion while the biological development lags and is being
developed; the chemist can serve a role in trying to
quantify the important carcinogens present in the diesel
emissions under various conditions. I think that may
now be a mandate in this field of research.
Answer: Dr. Moghissi: We discussed that subject among some
of our colleagues, and you will find differences of views in
this area. I happen to agree that the chemistry has to go
hand in hand with the biology; you had better know what you
are feeding to your animals or letting your animals inhale.
1167
-------
There are, however, some people who believe that you can
simply take a mixture and do biological experiments. There
is some justification to this approach, and in certain
well specified cases one comes up with satisfactory answers.
I happen to believe that in the long range there is nothing
better than systematic separation and biological evaluation
of each compound.
Question: Dr. Albert: Are there any other comments?
Question: Mr. Friedman: Bob Friedman, Bureau of Mines.
Underground we are trying to ease the problems of chemist,
biologist, and so forth doing analysis, by going to control
measures; there is some hope of controls underground. We
are beginning to see the light at the end of the tunnel.
Now I know that GM is, for example, and probably other
companies as well, experimented with control devices such
as particulate removal on exhaust. How well they are
going with that right now? Are you willing to say?
Answer: Dr. Vostal: The automotive industry studies
intensively all possible devices for particulate removal
or exhaust modification. However, the question is entirely
technical, and I do not feel that I am the proper person
to answer it.
Question: Dr. Albert: Anybody know how well...
Answer: Dr. Klimisch: Dick Klimisch, GM Research Labor-
atories.. I am not a spokesman for GM on this subject,
but I will answer the question since no one else is around
now who can do it better. There are certain problems
in applying the underground control systems to the above-
ground situation. Obviously we are working on traps
and things like that, but we can't use these gigantic
water scrubbers underground or we will lose the fuel
economy advantages that moved us toward diesels in the
first place. The biggest problem currently is the in-
teraction between nitrogen oxides and particulates, that
is, trying to get both of those emissions down at the
same time. We are making progress but we are not yet
there.
Question: Dr. Albert: We'll take one more.
Question: Mr. Peters: Bill Peters, MIT. I'd just like
to thank Dr. Balgard for drawing our attention to the need
to consider the operating conditions of the engines and
also the fuel type; the latter could be very important in
defining the optimum utilization for various synthetic
1168
-------
fuels, fuels based not only on coal, but also on shale and
perhaps even biomass, and so I just appreciate your mention-
ing that.
Question: Dr. Albert: This will be the last.
Question: Mr. Louflin: Dennis Louflin. There is one thing
that I am concerned about. What is the possibility of
inaccurate interpretations of risk? I think there are some
dangers that people don't really discuss. For example, in
the regulatory council statement and the IRLG statement,
there is some indication there that the screening is set up
to overestimate risks. I think one has to be mindful (?)
that one ought to protect the work force but only to the
degree that one does not do that to an extreme, in which
case you tend to overstate risks. I think this could be
very significant.
For example, if control technology is adapted for diesels
in the coming years because of the risks compared to auto-
motive, and if a hasty decision is made, then it seems as
though the public at large tends to suffer.
I think in the IRLG screen they don't really address what
the consequences are of overstating the risks, and I think
it might be significant because I noticed in the IRLG screen
that there is no mechanism for correcting a false identifi-
cation of harmful agents such as carcinogens. I think that
in this carcinogens issue that it is actually in comparing
things to other indexes like coke-oven (?), it's not going
to happen by helping the diesel issue that it isn't taken
lightly because perhaps substitute agents can be more
harmful. Dr. Albert you might comment on the IRLG screen.
I think to some degree it does overestimate. Have they
looked at the risks of overestimating?
Answer: Dr. Albert: Well, I think that there hasn't been
any major regulatory action taken on the basis of short-term
tests. Conformatory studies in animals for carcinogenicity
are required. But it seems to me that you have a point.
I think there is a tendency in the risk assessment business
to err on the conservative side, and there obviously has
to be a limit to that. This is a matter of balance and
judgment given the present state of our knowledge about
these matters.
1169
-------
-------
APPENDIX
REGULATORY OPTIONS FOR THE DEVELOPMENT OF
HEALTH AND ENVIRONMENTAL TESTING OF
DIESEL FUELS AND FUEL ADDITIVIES
A. Alan Moghissi and H. Matthew Bills
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC 20460
It is well known that mobile sources have a significant
impact on air quality. Traditionally, the U.S. legislation
has been directed at the automobile engine to reduce the
quantity of air pollutants. However, in recent years, the
significance of fuel and fuel additives in the reduction of
air pollution has been recognized. The Clean Air Act and
its recent amendments have made provisions for registration
and testing of fuel and fuel additives. The U.S. Environ-
mental Protection Agency (EPA) has an ongoing registration
program since 1975 and has registered several thousand
materials. The 1977 amendment to the Clean Air Act required
the development of regulations for testing of fuel and fuel
additives including deadlines to develop protocols for
environmental and health tests and criteria for testing.
The discussion of the legislative mandate of the EPA is
beyond the scope of this paper. This paper limits its
scope to options for these tests and the framework for the
development of appropriate regulations.
Health and environmental tests are advantageously divided
into two groups. One group would identify compounds or
mixtures of compounds to be tested, and the other group
would specify which tests are to be conducted on those
compounds or mixtures of compounds identified in the first
A-l
-------
group of tests. As a matter of convenience, the compounds
identified in the first group of tests, henceforth, will
be referred to as candidate test materials (CTM).
The first group of tests, as mentioned previously, may
lead to the identification of CTMs through various mech-
anisms; however, these tests also may be conducted to
assure compliance with present regulations such as emis-
sion of the so-called regulated pollutants.
CHEMICAL ANALYSIS AND FUEL PROPERTIES
It is readily understood that the knowledge of elemental
composition of fuel and fuel additives is useful in eval-
uating their environmental impact. In this case, one
is not concerned with C, H, or 0; rather the heteroatoms
such as N, S, and P and trace metals are of concern.
Because most elements are converted to their highest oxi-
dation stage during the combustion process, exhaust will
consist of C02, H20, N2, NOX, S02/S03, P205 and oxides
of various metals. The availability of the elemental
composition of the fuel would therefore permit an esti-
mate of environmental release of various pollutants.
Additionally, various analytical techniques such as atomic
absorption spectroscopy permit rapid and comparatively
inexpensive analysis of any trace elements down to 0.01
percent.
In addition to elemental composition, the availability of
chemical composition of fuel and fuel additives is impor-
tant. For example, the evaporative emissions are signifi-
cantly impacted by the composition of the fuel mixture.
Also, the presence of polynuclear aromatic compounds in
significant quantities would be undesirable; therefore, a
minimum knowledge of chemical composition is both desirable
and necessary.
Among fuel properties of concern are distillation curve and
cetane number (octane number for gasoline) because they
impact the proper operation of the engine and thus impact
the environmental releases of pollutants.
EMISSION TESTING
Emission testing, as proposed or promulgated in EPA regula-
tions, refers to testing automobiles for compliance with
regulated pollutants, namely CO, NOX, and total hydrocar-
bons for gasoline engines and particulate emissions for
A-2
-------
diesel engines. Emission testing does not lead to identi-
fication of a specific pollutant; rather, it assures that
the total emissions of an automobile is not adversely
affected by the introduction of new fuel, a new additive,
or a new combination of known fuel and fuel additives.
Because the evaporative emissions can be greatly influenced
by the fuel and fuel additives, evidence should be presented
that a new fuel or a new additive will not cause excessive
evaporative emission from an automobile which has met
the standards. Evaporative emissions can be sometimes
predicted on the basis of certain p'roperties of the fuel.
Several oil companies have developed theories to predict
evaporative emissions «and have experimentally tested these
theories. Because of the progress in this field, it is
sometimes possible to consider predictive models as sub-
stitutes for testing for evaporative emissions provided
they have been confirmed by at least a limited degree of
experimental data.
COMBUSTION PRODUCT TESTING
As mentioned previously, the EPA has proposed or promul-
gated limits of emissions of carbon monoxide, nitrogen
oxides, hydrocarbons (HC), and particulates from diesel
engines; however, health and environmental impact of auto-
mobile emissions depend upon the composition of HC (and
particles from diesel). For example, it is readily con-
ceivable that the HC fractions resulting from two different
fuels may have very different health effects, although
both emit the same quantity of HC per vehicle mile. The
purpose of combustion product testing should be to identify
those materials which will be considered for testing for
their environmental and health effects. To facilitate
the discussion of these alternatives, a new term in the
combustion products is defined. This fraction, emission
products other than common oxides (EPOTCO), consists of
the entire exhaust with the exception of oxides of carbon,
nitrogen and sulfur; water; nitrogen; oxygen and noble
gases.
Several alternatives may be considered in the identification
of CTM.
Alternative 1 - Disregard Combustion Products
It is well-known that the collection and identification of
combustion products are difficult and not standardized.
A-3
-------
Combustion products in EPOTCO are numerous and their produc-
tion depends upon so many engine parameters, such as engine
temperature and air-to-fuel ratio, that great difficulty
would be experienced in developing a standard procedure for
the identification of EPOTCO. Because the total quantity of
pollutants of concern is already limited, EPOTCO would not
be analyzed or identified and would not be considered in the
health and environmental testing program.
The advantage of this alternative is its simplicity and
ease of regulatory process. Its major disadvantage is a
disregard for potentially harmful materials in the combus-
tion products.
Alternative 2 - Complete Analysis
All compounds in EPOTCO are analyzed to the lowest level of
presently available analytical techniques. In so doing, one
would choose specific engine cycles which are representative
of the current year models of the automobiles. Each one of
the identified compounds would be subjected to environmental
and health effects testing.
The advantage of this alternative is the comprehensive
analysis of the exhaust and the possibility of considering
every single identified compound for health and environmen-
tal testing; however, the required extreme care in control-
ling engine parameters, sample collection, and the unusually
high sophistication in the analysis make such a system
unrealistic and expensive.
Alternative 3 - Collection of the Exhaust, But No Analysis
Toxicity tests will be conducted with the entire exhaust
or fractions of the exhaust such as gaseous and other
fractions. For example, particulates from a diesel engine
would be collected and used for toxicity testing, either
directly or after fractional separation.
Also, this alternative suffers from major disadvantages.
The presence of one compound with sufficient acute toxic
properties woul-d make it difficult and occasionally impos-
sible to observe potential chronic effects. In addition,
materials with widely different environmental persistence
would be given equal ranking in their toxic effects. For
example, if a compound is readily decomposed in an aqueous
medium in the environment within a short period of time, it
would be given equal consideration with a compound which is
A-4
-------
persistent for a long time. Other disadvantages of this
alternative include the extreme difficulty in collecting a
"realistic" sample. Compounds which, because of dilution in
the atmosphere, would not react with each other may react
under conditions of toxicity testing and generate new
materials having significantly different toxicity properties
than those originally present.
Alternative 4 - Partial Analysis
Only those compounds which are above a certain percentage
of the total EPOTCO fraction will 'be identified. This
alternative is a more simple and less expensive version
of Alternative 1. Although changes in engine parameters
may cause differences in the composition of the EPOTCO
fraction, it is anticipated that many compounds will appear
above the threshold level regardless of the engine para-
meters, provided these parameters are kept within realistic
range of conditions. For example, all compounds appearing
in concentrations of more than 0.1 percent in the EPOTCO
fraction would be tested.
The advantage of this alternative is that it covers a major
part of EPOTCO fraction without the requirement of a compre-
hensive analysis. The disadvantage of this alternative is
the possibility that highly toxic materials appearing in
concentrations below the threshold level will not be identi-
fied and thus will not be considered for further testing.
The choice of the threshold level is significant in that it
establishes the number of materials for further testing.
For the gasoline engine, a level of 1 percent by weight in
the EPOTCO fraction would cover about 20 compounds, a 0.5%
by weight, about 35 and a level of 0.1 percent by weight,
about 70 compounds (1-12). The number of compounds for
diesel are unknown at this time although their number should
not be significantly different from that of the gasoline
engine. Regardless of threshold chosen, there may be
compounds below the threshold which could be of potential
hazard and should be identified and tested.
CANDIDATE TEST MATERIALS (CTM)
Once various components of the fuel, fuel additive, and
combustion products are identified, they become potential
candidates for further testing. These CTMs could be a
simple compound or, depending upon the selection process,
mixtures of various compounds. Obviously, one has to
A-5
-------
establish, at least on a semiquantitative basis, potential
exposure of the public before a CTM goes through a series
of tests usually associated with the evaluation of a com-
pound.
BIOLOGICAL TESTS
The Office of Pesticides and Toxic Substances (OPTS) of EPA
has proposed various rules for biological testing of pesti-
cides and toxic chemicals (13). Many of these tests are
generally applicable tests for toxicity evaluation of a
chemical. For example, acute inhalation toxicity tests and
chronic carcinogenicity tests can be conducted with pesti-
cides as well as with components of fuels. However, as a
general rule, the route of exposure to fuel and fuel addi-
tives and their combustion products is usually dominated by
inhalation and thus this route should predominate. Also,
because of potential skin exposure as a result of spillage,
this route should be considered. A convenient way to
categorize biological and environmental tests is as follows:
Acute Toxicity Testing
These tests are designed to evaluate potential hazard of
fuels and fuel additives on a localized and concentrated
basis. For example, potential toxic effects of fuels and
fuel additives to the customers of self-service gas stations
will be evaluated using the results of these tests. Acute
toxicity tests are well standardized and have been used for
registration of pesticides and are being proposed for
premanufacturers testing of toxic substances. On the basis
of potential exposure routes, the following tests could be
considered:
Acute Inhalation Toxicity
Subchronic Inhalation Toxicity
Acute Dermal Toxicity
Subchronic Dermal Toxicity
Eye Irritation
Chronic Toxicity Testing
These tests are designed to evaluate the long-term poten-
tial health impact of fuel and fuel additives and their
combustion products. Protocols for these tests have been
A-6
-------
described in proposed guidelines for registration of
pesticides and are being developed for toxic substances.
Examples of these tests are:
Mutagenicity
Teratogenicity
Oncogenicity
Other Chronic Effects
Environmental Testing
The purpose of environmental testing is to establish
the indirect exposure of population to fuels and fuel
additives and the resultant products of their combustion.
Examples of these tests are:
Soil Degradation
Atmospheric Stability
Plant Uptake Studies
Because the exposure to these matrials predominates
in the urban areas, plants grown in the urban environ-
ment should be considered for testing.
RISK ANALYSIS
The biological and environmental tests are conducted
to avoid unnecessary and unwarranted human health risks.
Therefore, at least a semiquantitative analysis should
be conducted to evaluate potential effects of a fuel
or fuel additive. Methods for risk analysis are being
developed by various offices of EPA and by several other
U.S. agencies and other organizations. The EPA, along
with several other agencies, has published its cancer
policy (14) and quantitative techniques for radioacti-
vity have been in operation for many years (15). It
is fully recognized that a quantitative risk analysis
for many toxicants may be possible only after extensive
testing, or may be unfeasible because of economic reasons.
This problem is, however, common to all toxic chemicals
and is being addressed by EPA and a number of other agen-
cies. Once a policy decision is made as to the approach,
it is applicable to all toxicants including fuel and fuel
additives.
A-7
-------
described in proposed guidelines for registration of
pesticides and are being developed for toxic substances.
Examples of these tests are:
Mutagenicity
Teratogenicity
Oncogenicity
Other Chronic Effects
Environmental Testing
The purpose of environmental testing is to establish
the indirect exposure of population to fuels and fuel
additives and the resultant products of their combustion.
Examples of these tests are:
Soil Degradation
Atmospheric Stability
Plant Uptake Studies
Because the exposure to these materials predominates
in the urban areas, plants grown in the urban environ-
ment should be considered for testing.
RISK ANALYSIS
The biological and environmental tests are conducted
to avoid unnecessary and unwarranted human health risks.
Therefore, at least a semiquantitative analysis should
be conducted to evaluate potential effects of a fuel
or fuel additive. Methods for risk analysis are being
developed by various offices of EPA and by several other
U.S. agencies and other organizations. The EPA, along
with several other agencies, has published its cancer
policy (14) and quantitative techniques for radioacti-
vity have been in operation for many years (15). It
is fully recognized that a quantitative risk analysis
for many toxicants may be possible only after extensive
testing, or may be unfeasible because of economic reasons.
This problem is, however, common to all toxic chemicals
and is being addressed by EPA and a number of other agen-
cies. Once a policy decision is made as to the approach,
it is applicable to all toxicants including fuel and fuel
additives.
A-8
-------
CONCLUSIONS
There are a variety of options for regulating fuels and fuel
additives. The selection of an option should be based upon
sound economic considerations and on potential risk to the
population. In many cases, one can readily make appropriate
decisions solely on the basis of a chemical analysis. In
other cases, toxicity data may be available and the cost of
regulating a fuel or an additive may consists of a litera-
ture search and application of sound scientific models to
predict potential hazards of fuels, fuel additives, and
their combustion products. In other cases, it may be neces-
sary to conduct a series of biological tests. Fortunately,
protocols for biological tests are either developed or are
being developed by various offices of EPA. The regulations
should spell out methods for exempting a fuel or an additive
for certain tests after exposure is negligible.
REFERENCES
1. Black, F. and L. High, Automotive Hydrocarbon Emission
Patterns in the Measurement of Nonmethane Hydrocarbon
Emission Rates, Paper No. 770144 Society of Automotive
Engineers, Warrendale, Pennsylvania, March 1977.
2. Papa, L. J., Gas Chromatography - Measuring Exhaust
Hydrocarbons Down to Parts Per Billion, Paper No. 670494
On: Vehicle Emissions - Part III, PT-14, Society of
Automotive Engineers, New York, New York, 1971 pp.
43-65.
3. Hum, R. W., J. R. Allsup and F. Cox, Effect of Gasoline
Additives on Gaseous Emissions, Part I EPA 650/2-75-014,
U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina, 1974, pp. 76.
4. Seizinger, D. E., B. Dimitriades, Oxygenates in Automo-
tive Exhaust, Effect of an Oxidation, RI 7837, Bureau
of Mines, Bartlesville, Oklahoma, 1973, pp. 21.
5. Pepelko, W. E., J. G. Orthoefer, and Y. Y. Yang, Effects
of 90 Days Exposure to Catalytically Treated Automobile
Exhaust in Rats, Environ Research, 19(8): 91-101, 1979.
6. Gabele, P. A., J. N. Braddock, F. M. Black, F. D. Stump,
and R. B. Zweidinger, Characteristics of Exhaust Emis-
sions from a Dual Catalyst Equipped Vehicle, EPA 600/2-
77-068, United States Environmental Protection Agency,
Research Triangle Park, North Carolina, April 1977, pp.
29.
A-9
-------
7. Van Notta, 0., and R. D. McMillian, A Study of Emis-
sions from 1965-1975 Light-Duty Vehicles in Archeium,
California and St. Louis, Missouri, EPA 460/3-76-003,
U.S. Environmental Protection Agency, Ann Arbor,
Michigan, April 1976, pp. 443.
8. Bradock, J. N., Gaseous, Particulate, and Sulfur-
Related Emissions from Non-Catalyst and Catalyst
Equiped Vehicles, EPA 600/2-77-237, U.S. Environ-
mental Protection Agency, Research Triangle Park,
North Carolina, December 1977, pp. 59.
9. Oberdorfer, D. E., The Determination of Aldehydes in
Automobile Exhaust Gas, In Vehicle Emissions Part III,
PT-14, Society of Automobile Engineers, New York, New
YOrk, 1971, pp. 32-42.
10. Harkins, J. H. and S. W. Nicksie, Ammonia in Auto
Exhaust, Environmental Science and Technology, 1(9):
751-752.
11. Dietyman, H. E., J. R. Smith, M. Parnes, and E. R.
Francis, Analytical Procedures for Characterizing
Unregulated Pollutant Emission from Motor Vehicles,
EPA 600/2-79-017, February 1979, p. 495.
12. Stump, F., Oxygenated Compounds in Automobile Exhaust
Gas, Chromatographic Procedure, MSERB-ESRL-EPA,
Research Triangle Park, North Carolina.
13. U.S. EPA Federal Register. Vol. 43, 29, 696, 1978;
Vol. 43, 37, 336, 1978; and Vol. 44, 2242, 1979.
14. U.S. Regulatory Council Regulation of Chemical Carcin-
ogens, U.S. Regulatory Council, Washington, DC.
15. Moghissi, A. A., R. E. Marland, F. J. Longel, and K. F.
Eckerman, Methodology for Environmental Human Exposure
and Health Risk Assessment in Dynamics, Exposure and
Hazard Assessment of Toxic Chemicals, R. Hague, ed.,
Ann Arbor Science, Ann Arbor, Michigan, 1979.
A-10 ftUS GOVERNMENT PRINTING OFFICE 1980-757-064/0191
-------