United States
Environmental Protection
Agency
Environmental Research
Laboratory
Duluth MN 55804
Research and Development
EPA-600/S3-83-076  Nov. 1983
Project  Summary
Detection  of  Mutagenic/
Carcinogenic Alteration
in  Fish
David E. Hinton, James E. Klaunig, Michael  M. Lipsky, Rhona M. Jack, and
Benjamin F. Trump
  The feasibility of using fish as bioassay
organisms to detect mutagenic/carcino-
genic substances in the aquatic environ-
ment was tested in species not common-
ly employed in chemical carcinogenesis
bioassay. Microsomal fractions from
livers of channel catfish (Ictalurus
punctatus). fathead minnow (Pimephales
promelas). bluegill sunfish (Lepomis
macrochirus), brown bullhead (Ictalurus
nebulosus). and mummichogs (Fundulus
heteroclitus) were used. Data obtained
from these species were compared to
those obtained in rainbow trout (Salmo
gairdneri) - a more commonly employed
fish  - and mammalian species. The
presence and relative amount/activities
of microsomal protein, cytochromes P-
450  and bs, NADPH cytochrome C
reductase, aminopyrine demethylase,
and aryl hydrocarbon hydroxylase were
determined. The effects, both morpholog-
ic and biochemical, in exposure to the
known MFOS inducing agents, PCBs,
benzo(a)pyrene, and 3-methylcholanth-
rene  were studied.  Exposure  caused
induction of enzymes and proliferation
of endoplasmic reticulum  membranes
of hepatocytes.  High pressure liquid
chromatographic analysis of benzo(a)py-
rene  (BP) metabolism in  catfish and
trout were performed. Both species
produced the following metabolites of
BP: 9,10-diol BP; 4,5-diol BP; 7,8-diol
BP. 9-OH BP; 3-OH BP; and quinones.
Catfish postmitochondrial supernatant
converted BP and 2-acetylaminofluorene
(AAF) into mutagenic intermediates in a
microbial mutagen system. Catfish liver
cells  were isolated and maintained with
high  viability (98%) for 10 days. When
incubated with  H-BP, these cells
showed preferential accumulation of
label over nuclei. Subsequent liquid
scintillation analysis of cell fractions
obtained by cesium chloride centrifuga-
tion revealed radioactivity in DNA
fractions. Foci of hepatocellular altera-
tion, hyperptastic areas and bile ductular
hyperplasia  were  seen  in channel
catfish  chronically exposed to  the
chemical carcinogen AAF. These data
indicate the  suitability of conducting
further  studies on this ubiquitous
species  designed  to determine dose-
response characteristics to various
chemical carcinogens. On the basis of
microsomal metabolism and cellular
response, it appears feasible to use fish
tissue to test for mutagenic/carcinogenic
compounds in the aquatic environment
and to develop bioassay methodology
for testing possible carcinogenic proper-
ties of new chemical formulations prior
to their  introduction into the aquatic
environment.
  This Project Summary was developed
by  EPA's Environmental Research
Laboratory, Duluth, MN. to announce
key findings of the research project that
is fully documented in a separate report
of the same  title (see Project Report
ordering information at back).


Introduction
  A majority of human cancers are due to
chemical carcinogens in the environ-
ment. Since the oceans ultimately  be-
come the reservoir for every pollutant en-
tering the biosphere, the importance of
the aquatic environment in considera-
tions of carcinogenic effects of chemical
pollutants cannot be overly emphasized.

-------
Correlated Biochemical and
Morphologic Studies of Effects
of Xenobiotics on Fish Liver
  To characterize the response of certain
microsomal components in channel
catfish, we administered a commercial
PCB mixture (Aroclor  1254*) as seven
daily intraperitoneal  injections of 50
mg/kg b.w.  Via correlated biochemical
studies on hepatic microsomal mixed
function oxidative system (MFCS)enzymes
and ultrastructural studies of hepatocytes
we analyzed acute and subacute responses.
A  similar approach  was  used with
rainbow  trout  exposed to 3-MC  (50
mg/kg i.p.  x 4  days). Our results on
hepatic MFOS components of the channel
catfish and rainbow trout  and their
response to Aroclor  1254 and 3-MC
exposure  are summarized in the full re-
port.
  Acute exposure (7 days) resulted in  a
moderate increase in  some MFOS com-
ponents.  Cytochrome P-450 increased
approximately 33% in channel catfish.
Specific activity of NADPH-cyt-c-reductase
was increased to 1.6 times control
values. The greatest change was observed
in amount  of cytochrome b5, which
increased 2.5-fold.  Aminopyrine deme-
thylase activity,  microsomal protein and
liver to body weight ratios did not change
after acute  exposure. In contrast, rat
MFOS components were markedly increased
after an  identical concentration and
duration of exposure.
  Subacute (21  days) of PCB treatment
resulted  in a large increase in catfish
MFOS components  with  respect to
controls and acutely exposed fish.
Amounts of  cytochromes  P-450 and bs
increased three and two times  over
controls  respectively. NADPH-cyt-c-
reductase was elevated to a level similar
to that seen after  the acute exposure.
Aminopyrine demethylase, unchanged
by acute treatment,  was increased three-
fold over controls after subacute treatment.
Again, microsomal  protein and liver to
body weight ratios were unchanged.
Thus, the response of the catfish MFOS to
acute PCB  treatment is a  moderate
induction, much less than is seen in the
rat.  Subacute exposure resulted in  a
greater degree of induction.
  When compared to control morphology,
changes  were also seen in hepatocyte
ultrastructure of channel catfish after
acute PCB  exposure.  Dilated cisternae
and  meandering tracts of  endoplasmic
reticulum (ER)  were seen. Minimal
•Mention of trade names or commercial products
 does not constitute endorsement or recommendation
 for use.
increases in agranular ER profiles were
noted. These appeared as focal, discrete
aggregations.  Increased lipid droplets
also characterized hepatocytes of acutely
exposed channel catfish.
  Subacute exposure of catfish resulted
in alterations in agranular ER not seen in
acutely exposed  fish liver.  Tubular and
vesicular  profiles, nearly absent  in
control catfish hepatocytes, were increased
in number and often accumulated in large
aggregates. In addition, parallel stacks of
agranular ER in continuity with granular
ER were observed. Often these patterns
contributed to large whorls of ER membranes
in affected cells.
  Biochemical levels of induction follow-
ing  acute  and subacute Aroclor  1254
exposure in the  catfish correlated well
with  ER content and morphology. Higher
levels of MFOS induction were accompanied
by apparent increases in agranular ER
content and variations of ER morphology.
  Acute 3-MC exposure caused significant
induction of rainbow trout  MFOS com-
ponents. Cytochrome P-450 increased to
over three times  control amounts.
NADPH-cyt-c-reductase activity was
elevated two times over controls.  As in
the catfish, neither  microsomal protein
nor  liver to body weight ratios of trout
were altered by 3-MC treatment.
  Morphologically, acute MFOS induction
correlated with  apparent increases in
trout liver endoplasmic reticulum.  Ultra-
structural  alterations caused  by  acute
PCB and 3-MC exposure involved granular
ER content  and morphology, although
focal increases in agranular membranes
were noted. The major response after 21
days of exposure to Aroclor 1254 was
extensive proliferation of ER.

ARYL Hydrocarbon
(Benzo-a-Pyrene) Hydroxylase
(AHH) in  Channel Catfish Liver
  Polycyclic  aromatic  hydrocarbons
(PAH), an important class of environmen-
tal pollutants  which cause tumors  in
experimental animals including fish and
have been implicated in human carcino-
genesis, are activated  by the  NADPH-
dependent  MFOS.  That portion of the
MFOS responsible for this metabolic
activation is the AHH system. AHH from
livers of control fish and from livers of fish
exposed to 3-MC or to Aroclor 1254 was
characterized as to pH, temperature, and
protein concentration optima. Linearity of
reaction over time was established.
  With  respect to temperature, an
optimum of 30°C was found for control
catfish liver microsomes. This corresponds
to the temperature  optima  reported for
rainbow trout,  Finnish lake trout,  and
various marine species. A shift to a higher
temperature range (30°C - 40°C)followed
treatment with 3-MC or Aroclor 1254 in
the catfish. The induced system seems to
have  quite a broad temperature range
with no sharp drop at higher temperatures
such as that found in the trout.
  The effect  of  in  vitro addition  of
benzoflavone (BF) on AHH was studied in
control, Aroclor 1254 and 3-MC-pretreated
fish. Inhibition of activity was seen in 3-
MC-pretreated samples at  all ranges of
BF  tested (Figure  1).  Inhibition ranged
from 12% at .001 mM BF to 76% at 1 mM
BF. PCB treatment left AHH less sensitive
to BF inhibition at most concentrations
than  did 3-MC treatment.  In control
samples, .001  mM BF had no effect on
AHH  levels; however, at concentrations
of .01 mM, up to 200% enhancement of
activity was seen. Enhancement decreased
with increasing BF concentrations, until
at 1 mM BF, 22% inhibition was seen. No
enchancement was seen in samples from
either treated group at all concentrations
tested.
  Studies  thus far have demonstrated
that two forms of AHH can be distinguished,
one in control fish liver microsomes and
the other in microsomes from 3-MC  or
PCB-pretreated  fish.  The enzymes are
distinguished by the differential effect of
BF on AHH. Both 3-MC and Aroclor 1254
treatment in vivo cause the behavior of
the AHH enzyme to change. Characteris-
tically, the induced enzyme is similar to
the control enzyme with respect to pH
optimum but  differs in temperature
optimum and BF sensitivity (at the same
concentration).
  Benzo-a-pyrene, administered as a
single i.p. injection of 100 mg/kg b.w. in
corn oil, induced its own in  vitro metabo-
lism within 48 hr. in livers of three of four
channel catfish  (Figure 2-A). Differences
in response of exposed individuals was
apparent with AHH values  ranging from
120 to 350% of control values. Taken as a
group, the  mean value for exposed fish
was 1.90± 1.41 compared to 0.50 ±0.04
for controls (p<.025).
  When BP was administered as six daily
i.p. doses  of 25  mg/kg (Figure 2-B),
induction of AHH was seen and individual
variation among fish was  reduced. The
mean AHH value in exposed  fish  was
2.67 ± 0.97. This  is,  on the average, a
513% increase  over the control mean of
0.52 ±0.01. These values were significant-
ly different (p<0.01). When BPwas given
as a single gastric intubation, induction of
AHH  occurred  in  80% of  treated  fish
(Figure  2-C).  Interindividual variation
was apparent. The mean for exposed fish

-------
                                            O Microsomes from control liver
                                            • Microsomes from 3-MC-treated livers
                                            A Microsomes from PCB-pretreated livers
    200 -
    /50 -
c
o
CJ
    700
             .0007
            .007       .07         .7

                 mM 7,8 Benzoflavone
Figure  1.
Effect of 7,8 benzoflavone concentration on in vitro A HH activity. Control equals the
same sample  without BF.  Data  illustrated are from a single representative
experiment.  Triplicate assays were performed with variation between replicates
less than 10%.
a 5-°
o
1 40
AHH
WBP/20 m
\> Co
b b
o
nmole 3-
b

0


4 r+f
& n
rf
1

T
1



©


3
f3
rf


rt


•4-


-i-



©

V
5 T
+
•h


*





Figure 2.   Effects of BP on hepatic AHH activity in channel catfish. Slashed bars represent
           the average of control fish ±S.D. between animals. Numbers over the bars represent
           the number of fish in the experiment. Clear bars represent individual treated fish ±
           S.D. for triplicate assays on each animal. Treatment: (A) one i.p. injection of 100 mg <
           BP/kg b. w.. sacrificed after 48 hours; (B) six daily injections of 25 mg BP/kg b. w..
           sacrificed after 24 hrs; (C) one gastric intubation of 100 mg BP/kg b. w., sacrificed
           after 14 days.
(1.73 ± 0.95) was 208% of control values
and was significantly different (p <.025).

3-MC Induction of Channel
Catfish AHH
  The effects of five daily i.p. injections of
20 mg  3-MC/kg  body weight  upon
channel catfish liver  AHH are shown
(Figure 3). All fish exposed inthis manner
exhibited induction with the individual
response ranging from 10 to 25 times
control values. The experimental mean of
4.20  ±1.73 was statistically  different
from the control mean of 0.27 + 0.12 (p
<.01).
  When constant dose (25 mg 3-MC/kg
b.w.) was maintained for 4 days and fish
were then killed daily up to one week
after exposure, it was possible to deter-
mine the duration of effect. These data
are presented  in the full report.
  AHH  was increased dramatically by
one day after the cessation of exposure.
At day 4 following cessation of exposure,
the AHH activity was 24 times the control
mean, and by day 7 after treatment, AHH
was  19-fold  the control mean.  The
deviation between individual fish decreased
markedly as the time of exposure length-
ened. The highest  values  for  NADPH-
cytochrome-c-reductase and cytochrome
P-450 were seen at day 4. At this time,
the mean cytochrome  P-450 estimation
was 2.2 times over the control value. By
day 7,  cytochrome P-450 values  were
only slightly higher than controls.
  Microsomal  protein  concentration in
catfish appeared to increase  in the fish
killed 4 and 7 days after treatment ended;
however, no clear cut correlation between
induction and protein concentration was
seen. The mechanism for  induction in
fish  has not  yet been explained. New
enzyme synthesis or enzyme modification
is one possibility; however, changes in
the membrane composition or conforma-
tion should also be considered.

Metabolism of Benzo-a-Pyrene
by Microsomal Fraction of Fish
Liver
  Metabolites formed by the reaction of
fish liver microsomal fractions with 80-
100 nmoles of 3H-BP (specific  activity
150-200 fiC\/umo\e were  analyzed by
high pressure liquid chromatography
(HPLC). Each ml of assay volume included
0.1  M phosphate buffer, pH 7.0-7.2, 3.0
mM MgCI, O.1  mM EDTA, 0.4 mM NADP,
10 mM  glucose-6-phosphate, 15-20/uCi
BP, and 1.5 Units of glucose-6-phosphate
dehydrogenase as the NADPH generating
system. BF (0.1 mM) was added in 20/ul of
acetone  in some experiments.  Protein

-------
      6.0
  .c
   r
?!
  •S
   o
   I
     4.0
     2.0
Figure 3.
Effect of 3-MC on hepatic AHH activity in catfish. Five fish were injected on five
consecutive days with 20 mg 3-MC/kg b. w.. and sacrificed after 24 hrs. The slashed
bar represents the mean of three control fish ± 1 S.D. between animals. Clear bars
represent individual treated fish ± 1 S.D. for triplicate assays.
concentration was maintained at 0.1 mg
per assay. After incubation in a shaking
water bath at 30°C for 20 minutes, the
reaction was stopped by the addition of 1
ml  acetone. The  contents of 10  assay
tubes were combined, and metabolites
were extracted three times with ethyl
acetate. Radioactivity remaining in the
aqueous phase was less than 0.1 % of the
total. The  ethyl acetate extracts were
pooled, flash evaporated, and the  meta-
bolites were resuspended in 0.1  ml of
glass distilled methanol.
  Metabolite separation was performed
using a Varian  high pressure  liquid
                             chromatograph with a 25 cm Whatman
                             Partisil column (PXS10/250DS) of inside
                             diameter, 4.6mm.  Column temperature
                             was  ambient. For  analysis, 10-20//I of
                             the methanolic extract was injected into
                             the HPLC.
                               A methanol-water gradient was used
                             to elute metabolites.  Two  solvents
                             designated 'a' and 'b' were used. Solvent
                             'a' was  30:70 methanol:  water, and
                             solvent  'b' was 70:30 methanol water.
                             Initially, concentration of 'b' in 'a' was
                             25%. This  was increased  linearly at
                             3%/min and at a flow rate of 1.2 ml/min
                             to 100% 'b'. This reverse phase chroma-
tography, proceeding from a more to less
polar solvent, elutes polar components
first. Approximately 200 fractions  were
collected (0.4 min/fraction) in order to
elute all metabolites and unmetabolized
parent BP  which, due  to  its nonpolar
nature, is eluted last.
  All samples were co-chromatographed
with  14C-labeled BP metabolites  from
rat liver microsomes  as  an internal
standard or with pure individual metabo-
lites to  validate identification of peaks
(absorbance monitored at 254 nm). Since
the  rat  liver  BP metabolite  pattern is
well characterized,  it  was used  as a
biological standard. Peaks were also
identified by retention times established
with pure metabolite standards.
  Radioactivity of  eluted fractions was
determined by liquid scintillation counting.
The  3H counts were computed to obtain
specific  activity for each metabolite, and
expressed as pmoles BP metabolized/min/
mg protein. The 14C counts were  used
only for  identification purposes. The sum
of counts from fractions collected prior to
thefinal BPpeakrepresentedtotal BPme-
tabolized. Individual metabolites were ex-
pressed as a percent of this total. In this
manner, quantitative differences in acti-
vity  of various metabolites were  com-
pared in control and treated fish to estab-
lish effect of in vivo exposure of  xenobi-
otics upon in vitro BP metabolism.
  Table  1 provides a summary of results
in terms of percent metabolism of BP into
the various metabolites. Three different
control  and  3-MC-pretreated  catfish
were used to obtain these data. The 3-MC-
pretreated catfish  showed  an apparent
increase in activity toward the formation
of all metabolites. Control and 3-MC val-
ues for 9,10-diol BP and 7,8-diol BP were
significantly  different by the  T test
(p<0.05). Of the total radioactivity added
per assay, 3-MC-pretreated microsomes
metabolized 12-28% (3.75-4.8 fjC\) and
control microsomes metabolized 14-22%
(3.0-5.6 A»Ci). The PCB-pretreated catfish
also show an increase in 9,10-diol BPand
7,8-diol  BP, although total metabolism of
BP was only 3% (0.43  /uCi).  3-MC-pre-
treated  trout also showed  increases in
9,10-diol BP and 7,8-diol BP over  control.
Control trout metabolized only 3% of the
labeled  BP (0.64 //Ci), while  3-MC-pre-
treated trout metabolized 13% (2.78//CJ).
  The in vitro addition of 0.1 nM BF to
microsomes  from untreated  catfish
caused an increase inall metabolites over
the straight control preparations, plus a
28.8%  metabolism of  BP (97.9 //Ci)
corresponding  to  the enhancement of
AHH discussed earlier. The presence of in
vitro BF in  3-MC-pretreated microsomal
                                    4

-------
 Table 1.    Summary of BP Metabolite Formation as % of Metabolism
Metabolite
Catfish
                                                                    Trout
                       Control
                                 +BF
    3-MC
                                                    +BF   PCB   Control  3-MC
BP-9,10-diol
BP - 4,5-diol
BP - 7.8-diol
9-OH - BP
3- OH - BP
quinones
total BP
metabolized, %
AHH
nmoles/20 min/mg
protein
1.97±15
3.77 ±.93
7.47±1.27
14.5±2.3
13.17±3.17
12.5 ±1.32

14-22


0.6
3.9
14.0

20.0
33.0
22.0

28.8


1.0
3.77±1.07
6.03±4.03
1.25±1.09
19.2 ±2.85
13.6 ±4.07
16.4 ±3.38

12-28


6.08
4.0
1.9
5.4
5.7
12.0


10


0.7
9.5
2.6
12.6
19.6
5.5
13.8

3


2.71
1.7
1.8
3.3
13.0
8.5
45.0

3


0.54
5.9
1.0
14.0
10.6
12.6
22.0

13


7.37
Data from three individual fish are given for control and 3-MC-pretreated catfish (50 mg/kg b. w. x
4 days). The concentration of BF when added was 0.1 nM. Data for trout represent "pooled"
microsomal fractions from  4 control and 3-MC treated fish (50 mg/kg b. w.).
preparations markedly inhibited formation
of most  metabolites and AHH activity.
Total  metabolism was only 10% (2.76
yuCi). A great deal of variation was seen in
BP metabolism among fish; however,
treated fish generally metabolized more
BP. Some 9,10-diol BP and 7,8-diol  BP
formation may be at the expense of 4,5-
diol BP, which decreased  in 3-MC-
pretreated trout, PCB-pretreated catfish,
and one 3-MC-pretreated  catfish. This
study  has demonstrated  that in vivo
exposure of catfish and trout to xenobiotics
affects the in vitro metabolism of BP.
  BF reduced metabolism of BP significant-
ly in samples from 3-MC-pretreated fish
and enhanced metabolism in control
samples.  This correlates well with
characterization in which AHH was
enhanced in control samples by addition
of 0.1  mM  BF.  AHH,  however, was
inhibited in microsomal fractions from 3-
MC-pretreated fish  after  the in  vitro
addition  of  0.1  mM BF. Since  the
formation of every metabolite is affected,
the inhibition  must  take  place at the
oxidase level.

Microbial Mutagen Assays
  These followed standard  methods
using  Salmonella typhimurium tester
stains  TA-1535, TA-1537, TA-98, and
TA-100.  In order  to incubate NADPH
generating system, S-9 and test compound
directly on  petri  dishes with bacterial
tester strains, an incubation temperature
of 37°C was used. In previous experiments
with catfish liver AHH, 70% of the activity
at temperature optimum (30°C) remained
when assay temperature was maintained
at 37°C.  Following  incubation,  the
number of  revertant colonies was estab-
lished by direct counting.  For a biological
control, S-9 from livers of rats previously
exposed  to Aroclor 1254 was used in
companion trials with fish S-9.
     The results of the microbial mutagen
   assay using S-9 from fish with the four
   tester strains of Salmonella typhimurium
   and AAF, BP and MNNG are given (Table
   2). AAF and BP, when incubated with S-9
   and cofactors of the MFCS, were positive
   in all  four tester strains. MNNG showed a
   variable effect (i.e., positive in TA-100
   and TA-1535 but negative in TA-98 and
   minimal in TA-1537).
     The effect of fish protein (S-9 concentra-
   tion upon the number of revertant col-
   onies observed  when concentration  of
   test compound is kept constant (20 fjg/
   plate) is shown (Figure 4).
     When AAF and  BP  were used,  in-
   creases in S-9 concentration were asso-
   ciated with increases in numbers  of re-
   vertant colonies. This response was near-
   ly linear between S-9 protein concentra-
   tions  of 0.17 and 0.35 mg/plate but flat-
   tened at higher protein concentrations.
   By  contrast, changes in fish S-9 protein
   did not affect the number of MNNG-in-
                                duced revertants.  This  compound does
                                not require microsomal  activation to act
                                as a mutagen in the Ames assay.
                                  The microbial mutagen assay results
                                show that postmitochondrial supernatant
                                (S-9) from channel catfish liver converted
                                AAF and  BP  into mutation-causing
                                metabolites.

                                Culture of Isolated
                                Fish Hepatocytes
                                  The liver is the major site of experimen-
                                tally-induced teleost tumors, and as such,
                                would provide an ideal subject for in vitro
                                assay. The liver also possesses an active
                                and inducible MFOS and has been shown
                                to be sensitive to a broad spectrum of
                                procarcinogens.  The epithelial nature of
                                the liver also qualifies it as a model for
                                epithelial carcinogensis in general.  For
                                these reasons, the culture of isolated liver
                                cells was attempted.
                                  Optimal methods for isolation of catfish
                                liver cells were  developed. Liver pieces
                                were trypsinized  at 25°C for variable
                                periods of time. The results of this study
                                are  shown  in  Table 3. Eight hours
                                trypsinization gave the greatest  yield of
                                viable liver cells as established via trypan
                                blue staining and direct counting with a
                                hemocytometer. The yield of viable cells
                                was 3.25 x 106 per gram  body weight (2.8
                                x106/g liver). An average viability (trypan
                                blue exclusion) of 98% was obtained.
                                  Following  isolation, liver cells were
                                plated on to 90 mm2 culture dishes and
                                allowed to stabilize for 24 hrs. Represen-
                                tative plates were sampled for viability by
                                phase contrast  microscopy  and trypan
                                blue staining after 5 and  10 days. Cells
                                were also sampled for electron microscopy.
                                  The ultrastructural  characteristics  of
                                the isolated liver cells were indistinguish-
  Table 2.
Induction of Mutations Using Salmonella typhimurium Strains and S-9 Fraction from
Catfish Liver
Tester Strain

TA-98


TA-100


TA-1535


TA-1537

Compound*
AAF
BP
MNNG
AAF
BP
MNNG
AAF
BP
MNNG
AAF
BP
MNNG
Revertants/ ug CMPD/mg Protein*,
72.7±21.5
155.1 ±85. 8
No response
110.7±36.1
112.0±68.0
43.2+8.3
86.0±38.0
68.5±22.3
56.6+6.8
41.8±14.8
32.9±13.3
1.3+0.4
  *Compounds: AAF (2-acetylaminofluorene); BP (Benzo(aipyrene); MNNG
   (N-methyl-N'-nitro-N-nitrosoguanidine).
  fFive concentrations of S-9 protein (0.17 - 2.51 mg/plate) used per each of six concentrations
   (5-30 ug/'plate) test compound.
   Values represent the number of revertants per jjg of compound tested per mg protein ± S. D. sub-
   sequent to the subtraction of the number of spontaneous revertants (background level).

-------
  AAF = 2-acetylaminofluorene 20 fig/plate;
  BP = benzo(a)pyrene 20 fig/plate;
  MNNG =
    N-methyl-N'-nitro-N-nitrosoguanidine
    20 fjg/plate.
                                         Table 3.    Effect of Duration of Trypsinization on Catfish Liver Cell Yield
s
5  8
•S>  6
o
I  2
s
tL
Q:
        MNNG x-
Figure 4.
            0.17      0.34     0.84
             Protein Concentration
           Effect offish S-9 protein concen-
            tration  upon  the  number of
           revertant colonies (TA-1535).
able from those of the intact liver. After 5
and 10  days of culture,  increases  in
glycogen content were noted. Cells
during these periods began to aggregate
andjunctional complexes were frequently
noted.
  These  cultures were used to determine
uptake and intracellular localization  of
3H-BP. When cells were exposed to 1.25
fjg  3H-BP (32 //Ci/ml of medium) for 24
hrs. and autoradiograms prepared, grains
were concentrated over nuclei (Table 4).
Net nuclear grains (nuclear number less
cytoplasm and background) averaged 10
(Table 4). Radioactivity  of DNA isolated
from liver cells following 24hrs incubation
with 3H-BP averaged 490 dpm//yg.
  One limitation of studies restricted  to
subcellular fractions is the occurrence of
artifacts  inherent in preparation  which
may damage key enzymes and/or disrupt
compartmentation of others.  The  use  of
intact cells provides a system more akin
to that of the whole organ or tissue from
which the cell type of interest has been
taken. To date, the results  have been
most encouraging. High yields of viable
cells have been routinely obtained. Initial
experiments employing autoradiographic
localization  techniques  after  incubation
of cultures with 3H-BP showed preferen-
tial localization over liver cell  nuclei. This
finding, coupled with recovery of radioac-
tivity in DMA fractions suggests interaction
of BP with cellular components. Additional
studies are needed to determine whether
binding  to cellular macromolecules has
occurred and,  if so, to elucidate the
Duration
hrs
2
4
8
12
Total Number
of Cells
Isolated
x10e
36.8
98.3
243.5
7.3
%
Viable
96
99
98
75
Total Number
of Cells
Per g bw
x/06
0.43
1.31
3.25
0.08
Total Number
of Viable
Cells Per g bw
x/06
0.47
1.30
3.18
0.06
                                         Temperature of trypsinization was 25°C.
                                         Table 4.
                                                   Localization and Quantification of 3H-BP in Subcellular Components of Cultured
                                                   Primary Fish Liver Cells
                                         Experiment Number
                               Autoradiographic Analysis*
                              (grain counts over equal areas)
           Liquid Scintillation f
             (dpm/fjg DNA)
                                                           Nuclei
                                  Cytoplasm
   Intercellular
(space background)
1
2
3
33.4+14.7
30.8±11.8
37.8±12.8
18.4±6.3
19.5±7.1
15.2±5.4
6.2+1.7
6.4+1.2
5.8±1.4
480
380
630
*1xW6 cells exposed to 1.25 3H-BP (32 ud/ml of medium).
 Grain counts = mean ± S.D. of 100 nuclei, 100 randomly-selected cytoplasmic and 100
 randomly-selected intercellular spaces of equivalent area/experiment.
\5x1Oe cell as above.
 Each experiment was performed on  a  separate  cellular isolate  from  individual channel
 catfish.
                                         nature of such binding.  Since the
                                         metabolism of BP varies between species
                                         and between organs in the same species
                                         as well as with  the type of  preparation
                                         used (i.e., whole  cells, microsomal
                                         fractions,  or  nuclear  fractions), further
                                         studies on metabolism, mutagenesis and
                                         in vitro carcinogenesis using  intact
                                         epithelial cells from fish are  needed.

                                         Morphologic  Findings  in
                                         Chronic Carcinogen
                                         Exposures
                                           Chronic dietary exposure  of  channel
                                         catfish to AAF  or FBPA  for up  to  14
                                         months produced no grossly observable
                                         tumors. Since major  emphasis on the
                                         morphologic alterations following carcin-
                                         ogen exposure  was   placed on  liver,
                                         extensive analysis of this  organ was
                                         performed. Control channel  catfish liver
                                         of this study closely  resembled earlier
                                         published descriptions. Exocrine pancreas
                                         was  found in adventitia of portal veins.
                                         Cords  of cells existing as a  dual-plated
                                         muralium were  seen.  Some regions of
                                         hepatocyte cytoplasm were opaque while
                                         other areas were nearly transparent in
                                         H&E stained  preparations. The  latter
                                         corresponded to  regions where glycogen
                                         was  present. At later time periods (12 and
                                         14  months),  regions of control liver
                                         showed alterations   in  architectural
                                         pattern.  In  these areas  hepatocyte
                                         margins were indistinct and areas were
                                         interpreted as  necrotic. In addition,
                                         control  liver  at 14  months showed
                                         occasional foci of round cell accumulation.
                                         These were interpreted as inflammatory
                                         foci and were comprised  primarily of
                                         mononuclear cells.  The above were
                                         located in portal and  in some instances
                                         midzonal regions of hepatic lobules. High
                                         resolution light  microscopic analysis
                                         (HRLM)  (toluidine blue-stained sections
                                         of Epon-embedded material) revealed
                                         cytologic properties of control hepatocytes.
                                         In these, dark  staining material was
                                         arranged  as a  perinuclear cuff with
                                         extensions to cell periphery. The hepato-
                                         cytes  contained  one nucleus and were
                                         cuboidal to pyramidal  in shape.
                                           After  9 months of  exposure to AAF,
                                         focal  sites  of  necrosis with vascular
                                         congestion were encountered in eight of
                                         nine fish. Fat vacuoles were present in
                                         centrolobularand midzonal regions of the
                                         hepatic lobules. In two-thirds  of the
                                         animals,  peribiliary fibrosis was seen. A
                                         common finding in all  fish chronically
                                         exposed to  carcinogens was the biliary
                                         epithelial  response. This included cyto-
                                         plasmic vacuolization and pyknosis of
                                         nuclei in bile ductular and ductal epithe-
                                         lium.  In  addition, diffuse inflammation
                                         throughout the hepatic lobule was noted
                                         in one of  nine  fish after 9  months of
                                         exposure.
                                           After 12 months of exposure to AAF,
                                         one of 12 livers showed a  hyperplastic
                                         focus  in  which  hepatocytes contained.

-------
increased eosinophilia and basophilia,
and existed as trabeculae some 8-10cells
in thickness. Mitotic figures were seen in
some hepatocytes of the foci. Evidence of
a chronic necrotizing process was seen in
7 of the 12 animals studied. Necrotic cells
were seen  throughout entire  regions of
the liver. In 3 of the 12 animals, necrotic
areas  were associated with  localized
vascular congestion. Two-thirds of  the
animals at this time showed evidence of
peribiliary fibrosis and 7 of 12 animals
showed diffuse inflammatory foci through-
out liver lobes. Fatty change was encoun-
tered in a  single  animal at  this time.
Nonspecific tubular epithelial changes
were seen in some kidneys.  All other
organs were nonremarkable. Morphologic
alteration occurred in livers of all 27 fish
surviving 14 months  exposure to AAF.
Microscopic  findings in other organs
were nonremarkable. In general, hepatic
architecture was maintained  in all fish.
Foci and areas of cellular alterations were
encountered in  14 of 27 fish. These
lesions were typically comprised of cells
which showed increased acidophilia and
basophilia.  They were larger than those
seen at 12 months. In  addition,  these
clumps of cells showed numerous mitotic
figures. No tumors were encountered in
any  of the fish studied at this time.
Hepatocyte nuclei  in foci of necrotic
hepatocytes was  observed. Two of  the
five fish showed small foci of inflammatory
cells in liver parenchyma.  Cytologic
features of individual  hepatocytes were
identical to that described in control fish.
Four fish were killed after 2 months of
exposure to FBPA and were processed for
HRLM. In these individuals, normal liver
architectural pattern was seen. In two of
the four animals, small areas were
observed  in which  cells showed a
disorganized pattern  of dark staining
material and it was difficult to distinguish
between individual cells in  these areas.
These foci were interpreted as necrotic.
No  fat  was  observed.  Bile  ductular
proliferation was not seen. Peribiliary
fibrosis occurred  in  two  of  the four
animals studied. All other features were
identical to controls. After 3 months of
exposure to  the  compound,  the only
alterations observed were small sites of
necrosis similar to areas described above.
Cytologic features of individual hepatocytes
were identical to controls. After 6 months
of exposure  to FBPA, a  normal liver
architectural pattern was seen. In one of
three treated fish, dark staining regions
within cytoplasm were irregularlyarranged;
however, the  response was  focal.  No
necrosis, fat, bile  duct proliferation,
 •eribiliary fibrosis or inflammation was
observed. Cytology of individual hepato-
cytes revealed dispersal of dark staining
material peripherally with central portion
of cells containing light staining material.
This pattern of cytoplasmic rearrangement
is similar to that seen in hepatocytes in
which smooth endoplasmic reticulum
has proliferated. Livers from six treated
and six control fish were  studied by
routine light microscopy after 9 months of
exposure  to FBPA. In general,  liver
architecture was well-maintained. How-
ever, in two of the six fish, areas were
seen  in  which  sinusoids  were  not
apparent.  Cytoplasm of  cells in these
regions revealed a hyalinized appearance.
Necrosis was not found at this time. Fat
was  seen as  clear vacuoles  within
hepatocytes in one of the six animals. In
addition, the same  liver showed diffuse
small foci of inflammatory cells. After 12
months of exposure to FBPA, five of six
livers observed showed normal architec-
tural pattern.  In the other, the outer
margins of the liver revealed numerous
bulges and alternating constrictions
resulting  in a  "scalloped" appearance.
Necrosis was seen in two of the six livers
examined at this time. Sites of necrosis
and pyknosis were  common. In three of
the six livers, areas were seen in which
cytoplasm of  hepatocytes  revealed  in-
creased opacity. In addition, these areas
were sites of vascular  congestion. Fat
was observed in two of the six animals as
clear  vacuoles within  hepatocytes of
centrolobular and midzonal regions. Bile
duct  proliferation, not seen in animals
exposed for shorter periods  of  time to
FBPA, now appeared in four of the six
animals.  Bile stasis was indicated by
expanded, perfectly rounded lumina of
ductules  and  ducts and thin lining
epithelial cells suggested increase intra-
luminal pressure. Five of the six animals
studied showed inflammation particularly
pericholangitis. Numerous mitotic figures
were seen in bile ductular epithelium. In
livers of three fish, bile  ductal epithelial
hyperplasia was apparent. One of these
had progressed to papillarly projections
within the lumen. This configuration with
numerous mitotic figures is consistent
with a diagnosis of cholangioma.
  When compared to controls maintained
for identical periods  of time,  livers of
channel catfish exposed to FBPA for 14
months revealed changes. Both  control
and treated livers showed necrotic areas
although these were by far greater in
treated  fish. Areas  of necrosis in  both
groups were associated with fibrosis.
Granulomas were seen in the liver of one
control fish but not in treatedfish. Hepatic
architecture was similar in both groups.
However, cytoplasm of hepatocytes from
treated fish contained  more stamable
area—often completely  filling the cell.
Controls apparently contained abundant
glycogen which, by H&E, did not stain,
making these cells nearly  transparent.
Cytologic features of  hepatocytes from
the  14-month FBPA group included
abundant acidophilic inclusions which
were frequently as large  as nuclei. In
basophilic  regions of  cytoplasm, a fine
vacuolization was seen. Nuclei, generally
one  per cell, were lucent, rounded and
had  a  single  prominent nucleolus.
Treated livers revealed oval cell prolifera-
tion  in  15  of 32  fish  surviving the  14-
month  exposure. Peribiliary fibrosis,
percholangitis and melanomacrophage
centers were common. In 8 of 32 livers,
areas  were seen in  which cytoplasm
appeared hyalinized and gave a general
basophilic hue with H&E. Nuclei in these
regions were more opaque and generally
uniform. However, some nuclei were
enlarged,  oval  in shape and indented.
Cells in these  regions existed  as con-
tinuous sheets and it was difficult to
visualize sinusoids. In one fish, the above
described  region was  seen  as a nodule
which compressed adjacent hepatocytes.
This  led  to a  diagnosis  of minimal
deviation hepatocellular carcinoma. Mito-
tic figures were seen in the above.
  In  the interpretation of liver lesions,
previous  reports  in various teleost
species were reviewed. The response of
the channel catfish liver to AAF and FBPA
was  encouraging. However,  the latency
period was long. In light of the amount of
water required to maintain large numbers
of this size fish and the problem of
disposal of large amounts of contaminated
water,  use  of smaller species  such as
small aquarium fishes may be advisable.
Further study of  carcinogenesis in  the
catfish could be  extended to egg, embryo-
larval exposures in  which concentrated
dosages of  carcinogen are followed by
rinsing  and subsequent culture under
routine conditions. These procedures
may  constitute  a  feasible alternative to
the above-noted problems. Egg,  embryo-
larval exposures need to be extended to
other species in an effort to determine an
ubiquitous, sensitive, indicator organism
for aquatic carcinogen bioassay.
  In  light  of the  projected  increasing
national reliance  on coal-based energy
production, regional  contributions of
environmental PAH  from high point
source  emission  associated with coal
combustion, coking  and  conversion
processes require surveillance of environ-
mental  quality. The  response of the
channel catfish  to BP and 3-MC coupled

-------
   with its ubiquitous occurrence in streams
   and lakes gives added importance to
   continued  studies  with this species
   including:  varying routes of  exposure,
   bioaccumulation, acute and chronic
   morphologic  alterations  and possible
   tumorigenesis.  Such  would help to
   determine the suitability of this species to
   serve as a sentinel organism for aquatic-
   borne PAH in fresh water habitats.

   Conclusions
     Based upon the data obtained from this
   study, the  following conclusions have
   been made.
   • All fish species studied were found to
      contain a hepatic mixed-function
      oxidative system (MFOS) qualitatively
      similar  to that of  other teleosts and
      rodents.
   • Enzyme activity of the fish MFOS was
      increased by exposure to  polychlorin-
      ated biphenyls and polycyclic aromatic
      hydrocarbons, benzo(a)pyrene (BP)
      and 3-methylcholanthrene  (3-MC).
   • Increased enzyme activity correlated
      with proliferation of endoplasmic
      reticulum of hepatocytes.
   • Channel  catfish  and  rainbow trout
      liver microsomal fractions metabolized
      BP into diols, hydroxy derivatives, and
      quinones.
   • Pretreatment of catfish and trout with
      3-MC increased formation of 7,8- and
      9,10-diols of BP. Since diol formation
      proceeds via epoxide  formation,
      evidence indicates fish MFOS activated
      BP to a reactive carcinogen.
   • Catfish MFOS formed  mutagenic
      intermediates of  BP and  AAF in a
      microbial mutagen system.
   • Isolated primary hepatocytes of chan-
      nel  catfish can be maintained with
      high viability for  10  days and  permit
      detailed analysis  of cellular events in
      chemical carcinogenesis as weft" as
      direct interspecies comparison.
          The presence of foci  and areas of
          hepatocellular alteration, nodular
          lesions and bile ductular hyperplasia
          suggest neoplastic  responses to
          chronic carcinogen exposure in chan-
          nel catfish  liver.
          It is feasible to use fish to test for the
          presence of mutagenic/carcinogenic
          substances in the aquatic environment
          and to determine whether new chem-
          ical formulations, proposed for wide-
          spread usage, would have mutagenic/
          carcinogenic potential  in aquatic
          species.
          The use of fish species with relatively
          large  body size  coupled with  the
          chronic nature of laboratory exposures
          designed to demonstrate carcinogeni-
          city of compounds results in sizeable
          quantities of contaminated water
which require decontamination prior
to discharge. The volume of contamin-
ated water can be diminished by use
of closed system exposures  but
remains a problem to be considered in
projects of this nature.
The establishment of regional centers
of excellence to  provide a  safe
environment for testing of potentially
carcinogenic  substances  under con-
trolled conditions is strongly recom-
mended. Such would provide opportun-
ity for collaboration between govern-
ment, industrial and university person-
nel to establish necessary prerequisites
for aquatic carcinogenesis bioassay.

Defined nutritional  requirements of
these species used in long term assay
are needed.
          D. E, Hinton is with West Virginia University. Morgantown, WV 26506; J. E.
            Klaunig is with the Medical College of Ohio. Toledo, OH 43699; M. M. Lipsky. R.
            M. Jack, and B. F.  Trump are  with the University  of Maryland School of
            Medicine. Baltimore, MD 21201.
          Gary E. Glass is the EPA Project Officer (see below).
          The complete report, entitled "Detection of Mutagenic/Carcinogenic Alteration
            in Fish." (Order No. PB 83-253 559; Cost: $ 13.00, subject to change) will be
            available only from:
                  National Technical Information Service
                  5285 Port Royal Road
                  Springfield, VA 22161
                  Telephone: 703-487-4650
          The EPA Project Officer can be contacted at:
                  Environmental Research Laboratory
                  U.S. Environmental Protection Agency
                  6201 Congdon Blvd.
                  Duluth, MN. 55804
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300

-------