-------
DISCLAIMER
This report has been reviewed by the Environmental Research Laboratory,
U.S. Environmental Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the views and policies
of the U.S. Environmental Protection Agency, nor does mention of trade names
or commercial products constitute endorsement or -recommendation for use.
l\
-------
FOREWORD
The protection of our estuarine and coastal areas from damage caused
by toxic organic pollutants requires that regulations restricting the intro-
duction of these compounds into the environment be formulated on a sound
scientific basis. Accurate information describing dose-response relation-
ships for organisms and ecosystems under varying condititons is required.
The Environmental Research Laboratory, Gulf Breeze, contributes to this
information through research programs aimed at determining:
. the effects of toxic organic pollutants on individual species
and communities of organisms .;
. the effects of toxic organics on ecosystems processes and compo-
nents ;
. the significance of chemical carcinogens in the estuarine and
marine environments ;
The concentration and separation of polluting substances from estuarine
water for purpose of indentification and hazard evaluations represent a
significant technical problem. This report describes" research. to develop
simple methods useful in separating trace amounts of organic carcinogens
from marine waters . A unique approach to the separation of non-polar
organic carcinogens from other non-polar organic compounds was explored.
Methods such as these will be useful to the Agency in assessing the
potentical hazard of organic compounds introduced into the estuarine
environments .
Henry F -^Enos
Director
Environmental Research Laboratory
Gulf Breeze, Florida
iii
-------
ABSTRACT
The organic carcinogens benzo(a)pyrene, dieldrin, and N-acetyl-2-amino-
fluorene were recovered on XAD-2 macroreticular resin in yields of 90 percent
or more from distilled water or seawater and in yields of 40 percent or more
from Lake Pontchartrain water containing a high concentration of organic
material. The original solutions contained less than 500 parts per trillion of
carcinogen. These results show that XAD-2 provides an efficient means for
recovering nonpolar organic carcinogens from dilute solutions. More polar
carcinogens such as dimethylnitrosamine were not effectively recovered on XAD-2
columns.
Since XAD-2 binding would not be selective for carcinogens, we investi-
gated methods which might bind carcinogens selectively from a mixture of
organic compounds. We tested the ability of the above carcinogens to bind
to nucleic acid both before and after S9 liver microsomal activation. The
DNA-carcinogen interaction systems included direct binding, equilibrium
dialysis, nuclei binding, and binding to DNA-cellulose. Radiolabeled car-
cinogens were used to quantify the amount bound. Either rat liver nuclei
(0.1 mg DNA) or DNA-cellulose (1 mg DNA) bound is percent of the acetylamino-
fluorene and up to 66 percent of the dieldrin from solutions containing 150
to 280 nmoles of compound. Up to 30 percent of the benzo(a)pyrene from
solutions containing as much as 320 pmoles was bound. Ten-fold or lower
recoveries were found when direct-binding or equilibrium-binding methods
were used. When liver microsomes were used for activation in the binding
systems, less carcinogen was recovered with the DNA fraction. In these
cases, the microsomal protein can decrease the net DNA binding by competing
with the DNA for carcinogen or by converting the carcinogen to products
which do not bind well to DNA.
In the nuclei bindings studies, DNA was isolated from the nuclei by
extraction with sodium dodecyl sulfate and phenol followed by recovery of
the DNA by ethanol precipitation and spooling. The recovered DNA contained
more than 10 percent of the input radioactivity, indicating that a signifi-
cant portion of the input sample may be tightly bound to DNA. The portion
which bound after DNA isolation was not -dissociated by further extractions
with chloroform-isoamyl alcohol or 1 percent sodium dodecyl sulfate solution.
At least some of this material may be covalently associated. Whether or not
DNA-binding is more specific for recovery of organic carcinogens than is XAD-2
remains to be explored.
This report is submitted in fulfillment of Contract No. R805656 by Gulf
South Research Institute under the sponsorship of the U.S. Environmental
Protection Agency. This report covers the period November 1, 1977, to
October 31, 1978, and work was completed December 15, 1978.
IV
-------
CONTENTS
Foreword iii
Abstract .- iv
Figures vi
Tables vi
1. Introduction 1
2. Conclusions 3
3. . Experimental Procedures 4
Background 4
XAD-2 Adsorption Method 5
DNA-Cellulose Adsorption of Carcinogens 11
Adsorption of Carcinogens to Isolated Rat Liver Nuclei.... 17
Direct Binding of Carcinogens to DNA 22
Equilibrium Dialysis Binding of Carcinogens to DNA 23
Comparison of the DNA7binding Methods 26
References.
31
-------
FIGURE
Figure a&e-
1 Binding of Benzo(a)pyrene to DNA - Nuclei Binding Method; BioGel
21
A5 Chromatography
TABLES
Number
1 XAD Adsorption of Carcinogens °
2 DNA-Cellulose Column Chromatography of Organic Carcinogens 14
3 DNA-Cellulose Binding of Organic Carcinogens (Batchwise Incubation
Studies) 15
4 Nuclei Binding of Organic Carcinogens 20
5 Direct Binding of DNA with Carcinogens 24
6 Equilibrium Dialysis Binding of DNA With Carcinogens 27
7 Comparison of DNA Binding Methods 28
VI
-------
SECTION 1
INTRODUCTION
Estuarine waters are subject to pollution from oil spills and influx
of industrial effluents from rivers. The pollutants can be directly toxic
to marine life or can be concentrated by marine organisms which are then
eaten by fish or man and thereby indirectly or directly rendered dangerous
to human health. The concentration and separation of substances from
estuarine water for the purpose of establishing their existence and hazard
represent a significant problem since the substances are diluted by large
quantities of saline water.
The generalized parameters typically used to characterize water (biolog-
ical oxygen demand, total organic carbon, etc.) represent only a cross-
section of the materials that raw and treated industrial waste effluents
may present to estuarine water bodies. The application of these generalized
criteria for assessment of the danger to man and the environment cannot
necessarily protect a given section of the population or environment from
hazardous or toxic compounds.
Novel methods are needed for further isolation and analysis of the
most dangerous substances, since dilution of these substances in the water
1
-------
prevents adequate characterization. In addition, due to the complexity of
the mixture, one method of characterization could interfere with another.
For example, characterization of a mutagen by the Ames Salmonella/microsome
test could be prevented by the presence, in the mixture, of chemicals toxic
to the test bacteria. Also, certain substances could poison the microsomal
system and prevent activation of a premutagen to a mutagen. Techniques
such as gas chromatography/mass spectrometry, while extremely sensitive,
are very costly and do not provide an assessment of the danger of a particular
mixture to animals.
Our goal in the present research has been to investigate simple methods
which might be useful in separating organic carcinogens (polycyclic aromatic
hydrocarbons, in particular) from dilute solutions and in further separating
these from nontoxic, or toxic but noncarcinogenic, substances. For the
initial separation of nonpolar organic substances from water, XAD-2 binding
was used due to its low cost and ability to achieve high recoveries. The
presumed ability of carcinogens to bind to DNA has been investigated as a
means of separating such species from other nonpolar organics. In the
present studies, only radioactively labeled, pure carcinogens have been
tested for XAD-2 and DNA binding. Additional studies are needed in which
actual samples obtained from estuarine waters are tested.
-------
SECTION 2
CONCLUSIONS
The polycyclic hydrocarbons dieldrin, benzo(a)pyrene, and N-acetyl-2-
aminofluorene bind well to XAD-2 resin and can be recovered from very
dilute solutions in impure waters in yields greater than 50 percent.
Dimethylnitrosamine was poorly recovered under these conditions. All of
these compounds were strongly adsorbed to nuclei or DNA-cellulose and less
strongly to DNA in direct binding studies. From 10 to 80 percent of the
compound could be recovered in the bound state. Specificity of the binding
could not be adequately investigated in the time allotted in the grant
study.
-------
SECTION 3
EXPERIMENTAL PROCEDURES
RECOVERY OF ORGANICS FROM DILUTE AQUEOUS SOLUTIONS BY XAD-2 ADSORPTION
Background
Amberlite XAD resins (Rohm and Haas) are polystyrene divinylbenzene
copolymers formed into beads (20-50 mesh) with a macroreticular structure
(1-3). The beads allot* a high flow rate and yet possess a large hydrophobia
2
surface area (>300 m /g) (3). They are, therefore, capable of adsorbing
large quantities of hydrophobic species from water.
There have been many reports describing the recovery of dissolved
organics on XAD resins. Recovery of chlorinated biphenyls (4), chlorinated
pesticides including dieldrin (5), surfactants (3) , polycyclic aromatic
hydrocarbons (1,2), phenols (5), and other classes of nonpolar organic
compounds in yields greater than 80 percent from 50-500 ppb or lower in
solutions in water or seawater have been described. The recovery of adsorbed
.compounds from the resin is simpler than that from activated carbon (2,5)
and there is less chance for chemical conversion of the compounds prior
-------
Co analysis. A general method for recovery of organics from water using
highly purified XAD-2 has been outlined in detail by Junk, et_ al_. (5).
Recently, XAD-2 has been used for recovery of mutagenic substances from the
urine of cigaret smokers, indicating that even very impure fluids can be
utilized (6).
XAD-2 Adsorption Method
The procedure which was used for recovery of organic carcinogens from
seawater followed that of Junk, _et_ al_. (5). The XAD-2 resin obtained from
Rohm and Haas was purified by sequential solvent extractions with methanol,
acetonitrile, and diethylether in a Soxhlet extractor for 8 hours per
solvent. The washed resin was stored in glass-stoppered bottles under
absolute ethanol. The sorbent column was prepared by pouring a slurry of
the resin in ethanol into the column (Bio-Rad glass Econo columns, 0.7 cm
inside diameter x 10 cm were suitable for this purpose); the ethanol was
subsequently replaced by water passed through a charcoal filter and glass
*
distilled.
Sample solutions containing carcinogen were prepared using water from
several sources: distilled water, Lake Pontchartrain water taken at New
Orleans, Louisiana, and seawater collected from the Gulf of Mexico near
Biloxi, Mississippi. The water was spiked with radiolabeled carcinogens at
concentrations ranging from 3 to 800 parts per trillion (pg/liter). The
radio-labeled carcinogens which were used were dieldrin-1,2,3,4,10-C-14
(specific activity, 1.34 mCi/mmole, California Bionuclear Corporation),
dimethylnitrosamine [methyl-C-14] (specific activity, 3.7 mCi/mraole, Califor-
5
-------
nia Bionuclear Corporation), beiizo(a)pyrene-H-3 (ICN, specific activity, 24
Ci/romole), and N-acetyl-2-aminofluorene [9-C-14] (specific activity. 26
mCi/mmole, Schwarz/ Mann). These carcinogens were chosen to provide
examples of a chlorinated hydrocarbon pesticide (dieldrin), a nitrosamine
(dimethylnitrosamine), a polycyclic aromatic hydrocarbon [benzo(a)pyrene]
and an aromatic araine (N-acetyl-2-aminofluorene). All have been shown to
be capable of binding to DNA (7). Dieldrin and benzo(a)pyrene are believed
to be important contaminants of seawater (5,8).. Aromatic amines may con-
taminate seawater from effluents of dye manufacture, and nitroso compounds
are ubiquitous environmental pollutants (8). Therefore, these classes of
carcinogens should be illustrative of the problems which could arise in any
potential scheme for isolating and separating organic carcinogens from
seawater. These carcinogens may all require activation to exert carcino-
genic effects. They have been shown to be mutagenic in in vitro'tests and
to be activated by microsomal oxidizing enzymes to species capable of
binding to DNA. Furthermore, all have been shown to induce tumor formation
in animals, although their carcinogenicity to humans is still unknown.
The dilute solutions were passed through the XAD-2 columns at flow
rates between 3 and 18 ml/min. Solutions made in seawater and lake water
were first passed through a glass fiber filter (Whatman GF/C) to remove
suspended solids. The effluent water from XAD-2 adsorption was discarded.
After the spiked water solution was passed, the column was washed with
several column volumes of distilled water. The adsorbed carcinogen was
then eluted with 4 to 5 column volumes of diethylether. The diethylether
was evaporated in a stream of nitrogen, and the residue placed in a scintil-
lation vial. The sample was counted in toluene scintillant (5 g PPO, 0.13
-------
g POPOP per liter) in a Beckman LS 133 counter. Percent recovery was
calculated from the amount of carcinogen recovered (cpm) and the amount
initially dissolved. These values are given in Table 1. Varying the XAD-2
bed volume, flow rate, and carcinogen concentration had little effect on
recoveries. Recovery was significantly less for dieldrin from lake water
than from distilled water or seawater. Most of this loss was found to be
due to adsorption of labeled dieldrin to particulates in the water which
were removed by filtration. Such loss was not significant for the other
carcinogens tested.
Due to the volatility of dime thy Initrosamine, the ether was not evaporated
after the desorption of the dime thy Initrosamine from the XAD-2; instead,
aliquots were assayed directly in the scintillation fluid. As can be seen
in Table 1, recovery of the dimethylnitrosamine was less than 5 percent of
the input amount. Therefore, binding of such species to XAD-2 could not be
used to recover them efficiently from dilute solutions in water.
The results given in Table 1 confirm and extend those reported by
other workers. Junk, e_t al. (5) have reported a 93 percent recovery on
XAD-2 of dieldrin from a 20 ppt solution in water. Osterroht (2) showed
that aldrin, an isomer of dieldrin, could be recovered in 58 percent yield
from a 10 ppb solution in seawater. The polycyclic aromatic hydrocarbon
phenanthrene was recovered from seawater in almost 100 percent yield from a
10 ppb solution but in lower yields from 100 ppb or 1 ppb solutions (62
percent and 57 percent, respectively). The above recoveries were measured
using gas chromatography and comparison with standards instead of using
radioactive samples as in the presently reported study. In the study by
7
-------
TABLE 1. XAD ADSORPTION OF CARCINOGENS
00
Carcinogen
dieldrin
dieldrin
dieldrin
benzo (a) pyrene
benzo(a)pyrene
AAFft
AAF
dimethyl-
nitrosamine
dimethyl-
nit rosamine
Carcinogen
input
(total cpm)
3600
3600
3600
400000
80000
51000
51000
82500
82500
XAD
Water Carcinogen Solution voiume
volume concentration Water flow rate
(liters) (ppt) source (ml/min) (cm )
1 430 Distilled
1 430 Sea
3.78 114 Lake
Pontchartrain
1 5.2 Sea
0.32 3.15 Lake
Pontchartrain
1 90 Lake
Pontchartrain
1 90 Sea
0.6 833 Distilled
0.9 555 Sea
13 5.03
8 8.05
7 11
7 16.6
3 13.3
3 14.1
3.4 15.1
6 12.1
18 14
Carcinogen
recovered
(total cpm)
2570
2635
1660
296900
48000
40500
51099
4135
1500
Recovery
of
carcinogen
(percent)
71.4
73.2
46.1
74
60
79.4
100
5
2
*N-acetyl-2-aminofluorene
-------
Yamasaki and Ames (6) , recoveries were measured by the Salmonella/micro-
some (9) test. In the Yamasaki and Ames study, benzo(a)pyrene and N-
acetyl-2-aminofluorene were recovered from urine using XAD-2 in yields of
19 percent and 89 percent, respectively, from 0.8 ppm solutions. The
present results show that the N-acetyl-2-aminofluorene is recovered from
seawater and lake water in high yields.
From the present studies and from the work-of others it can be con-
cluded that XAD-2 can be used to recover certain organic carcinogens from
seawater in high yields. A possible exception is the recovery of more
polar carcinogens such as dimethylnitrosamine. Yamasaki and Ames (6)
found that certain metabolic derivatives of carcinogens (for example glucuro-
nides) were not bound well by the nonpolar resin. The effect of environmental
oxidation oh polarity of polycyclic aromatic species and their ultimate
ability to be recovered on XAD-2 has not been explored and would be an
important factor in evaluating the usefulness of XAD-2 adsorption as the
only method for concentrating organic carcinogens from dilute solution in.
seawater. Other possible methods for concentrating organics from marine
waters, such as reverse osmosis, suffer from the fact that the maximum
achievable concentration is 10- to 25-fold due to the increase in salt in
the concentrate and the resulting increase in osmotic pressure. The use of
bioadsorption by marine organisms such as the clam or mussel are potentially
useful since 1000-fold concentrations of certain organics can be achieved (10)
However, only small quantities of the concentrates can be obtained and
these must be extracted from the tissue of the organism, thereby complicating
the recovery process.
-------
SEPARATION OF CARCINOGENS FROM IJONCARCINOCENS BY DNA-BINDING
Current theoretical insight into the mechanism of carcinogenesis
considers the first step in the process to result from the binding to and
alteration of DNA by a carcinogen (7)- It is now considered axiomatic that
carcinogens are mutagens, compounds able to alter DNA thereby causing a
permanent, hereditarily transmitted change in the genetics of the altered
cell. Therefore, carcinogens must bind to DNA by either covalent or non-
covalent attachment. If such binding is fairly specific for carcinogenic
(and mutagenic) species, then at least a partial separation of such species
from nonmutagenic species might be possible. Most carcinogens (often
termed procarcinogens or promutagens) require metabolic activation before
they can chemically alter DNA. The most common activation processes are
oxidations by the microsomal mixed function monooxygenase system, e.g.,
arylhydrocarbon hydroxylase. Such functionalization is necessary before
the carcinogen can bind covalently to DNA base nitrogens. The mechanism of
oxidations have been widely discussed in the literature (11-13) and in
certain cases presumed ultimate carcinogens identified from so-called
proximate species. It is known that many proximate carcinogens can also
bind noncovalently to DNA, although such binding may not result in mutagene-
sis (14).
In the present study, an investigation has been made of different
conditions for interacting DNA with both activated and unactivated samples
of the carcinogens benzo(a)pyrene, dieldrin, N-acetyl-2-aminofluorene, and
dimethylnitrosamine. In each case, both total binding and total covalent
10
-------
binding have been estimated. The question of specificity of the interaction
(carcinogens versus noncarcinogens) has not been answered in this study-
DMA-Cellulose Adsorption of Carcinogens
DNA-cellulose chroraatography has been used for the isolation and
purification of DNA-binding enzymes and proteins. The use of immobilized
DNA has at least two potential advantages over direct interaction of DMA
with substrates in solution. First, a higher ratio of DNA to substrate is
possible during the interaction process if the solution containing sub-
strate is passed slowly onto a column of DNA-cellulose. This should maximize
the amount able to bind. Second, the bound material (noncovalently bound
only) can be selectively desorbed using certain elution solvents according
to the binding strength for the individual component.
DNA-cellulose was prepared according to the procedure described by
Alberts and Herrick (15). A solution of calf thymus or Salmon sperm DNA
(Sigma), at 1-3 rag/ml in 0.01 M Tris-HCl, pH 7.4, 10~3M EDTA (Tris-EDTA)
was transferred to a beaker and Whatman CF-11 cellulose was added until the
paste thickened. The mixture was allowed to sit at room temperature until
dry. The dried mixture was ground to a powder and evaporated by lyophiliza-
tion to remove any remaining water. The dry powder was suspended in Tris-
EDTA for 24 hours at 4°C and washed twice to remove nonbound DNA. The DNA-
cellulose was stored as a frozen slurry in Tris-EDTA. For the experiments
described below DNA-cellulose was prepared with a DNA content of 0.09 to
0.48 mg DNA per packed ml of cellulose. The amount bound was determined by
the uv adsorption (A26Q) of an aliquot heated to 100°C for 20 min in Tris-
EDTA to release the DNA from the cellulose.
11
-------
The DNA-cellulose powder was loaded into Bio-Rad Econocolurans, 0.7 x '
10 cm, in Tris-EDTA. As controls, columns were prepared with cellulose
which had gone through the same preparative steps employed with the DNA-r
cellulose. The columns were washed thoroughly with Tris-EDTA buffer before
adding solutions of the carcinogens.
Since the organic carcinogens chosen for the DNA-binding study all
require activation by microsomal enzymes to exert their mutagenic effect,
samples were added to DNA-cellulose with and without prior activation. The
activation system used was that recommended by Ames et al. (9) for Salmonella/
microsome testing. A liver homogenate was prepared from 200 g male Sprague-
Dawley rats which were injected with 500 ing/kg arochlor 1254 five days
prior to sacrifice. The livers were homogenized in 3 volumes of 0.15M KC1
with a Tekmar Tissuemizer for 1 rain at top speed. The homogenate was
centrifuged at 900 x £ for 10 min at 4°C in an International B60 centrifuge.
The supernatant (S9) was collected and stored at -75°C in 1 ml aliquots.
The activity of the S9 was checked by determining its ability to convert 5 ug
*
benzo(a)pyrene to mutagenic products in the Ames Salmonella/microsome test.
An aliquot of 0.05 ml S9 (0.5 ml S9 mix, see below) gave a 9-fold increase
compared to control in revertant colonies for Salmonella strain TA98.
In the experiments in which the effect of S9 activation was tested,
incubation (1 hour at 37°C) of the DNA-cellulose with carcinogen and" S9 mix
(contains per ml: 0.1 ml S9, 8 umoles MgCl_, 33 umoles KC1, 5 umoles
glucose-6-phosphate, 4 umoles NADP, and 100 umoles sodium phosphate, pH
7.4) was performed in a batchwise manner in a screx/ cap culture tube instead
of a column. Conditions were identical for samples not activated with
microsomal mix, except that S9 was not included.
12
-------
In the column experiments (no S9 activation), the column was further
eluted with 5 column volumes of Tris-EDTA after the sample had passed into
the column bed. This was followed by a linear gradient consisting of 5
column volumes of Tris-EDTA and DMSO. Finally, the column was rinsed with
toluene until no further radioactivity appeared in the wash. The residual
radioactivity on the column, if any, was determined by counting aliquots of
the DNA-cellulose (or cellulose) and the solution obtained by heating the
DNA-cellulose (or cellulose) at 100°C for 30 mio in DMSO:Tris-EDTA (1:1).
After these treatments, less than 1 percent of the input radioactivity was
eluted from the column. A similar procedure was used in the batchwise DNA-
cellulose studies except that washes after the buffer washes were with 100
percent DMSO. In these studies after each wash the cellulose was centrifuged
down and the supernatant withdrawn. Washing with each solvent was continued
until no further radioactivity was eluted.
The results of these studies are given in Table 2 for the column
experiments and Table 3 for the batchwise incubation studies. In the
column studies with dieldrin and benzopyrene (Table 2), most of the carcinogen
(97 percent and 96 percent, respectively) remained bound to DNA-cellulose
after the elution with buffer only, whereas for acetylaminofluorene much
less (16 percent) was initially bound. The binding of dimethylnitrosamine
was not tested. For acetylaminofluorene in the column studies (no S9
activation), following the DMSO and toluene washes, 96 percent of the input
radioactivity was recovered. Approximately 70 percent of the benzopyrene
remained bound to the DNA-cellulose after extensive washing with DMSO and
toluene. As can be seen in Table 2, carcinogen association with cellulose
not containing DNA also occurs. This material, hox«?ever, is more readily
13
-------
TABLE 2. DNA-CELLULOSE COLUMN CHROMATOCRAPHY OF ORGANIC CARCINOGENS
Carcinogen
dieldrin + S9
dieldrin
dieldrin
(control)
bcnzo(a)pyrene
benzo(a)pyrene
(control)
AAF*
AAF (control)
Input
quantity
cpm
(nmoles)
765000
(286)
765000
(286)
765000
(286)
885800'
(0.04)
885800
(6.04)
2.6 x 10
(51)
2.6 x 10
(51)
Flow
rate
(ml/min)
0.06
0.06
0.06
0.03
0.03
6 0.03
6 0.03
1 Amount cluted with different
Cellulose ^V1A solvents (cpm)
n UNA
voiunie
_ present
(cm ) (mg) Trls-EDTA
5.0 1 20622
5.0 1 18530
5.0 0 18410
5.0 0.45 34800
5.0 0 19110
4.3 1.6 2145000
4.3 0 1589000
DMSO
gradient
84000
188352
392880
227192
503102
400300
876000
Toluene
11480
24710
324390
14580
21130
18500
62800
Total
recovered
(percent
of
input)
16
30
96
31
61
96
97
Amount
left on
column
(percent
of
input)
84
70
4
69
39
4
3
*H-acetyl-'2-aminof luorene
-------
TABLE 3. DNA-CELLULOSE BINDING OF ORGANIC CARCINOGENS (BATCHWISE INCUBATION STUDIES)
Carcinogen
dieldrin
dieldrin + S9
benzo (a)pyrene
benzo(a)pyrene + S9
N-acetyl-2-aminof luorene
N-acetyl-2-aminof luorene + S9
Input
cpm
(nmoles)
320000
(120)
320000
(120)
840000
(0.04)
840000
(0.04)
1100000
(22)
1100000
(22)
Total
DNA
present
(mg)
0.1
0.1
0.1
0.1
0.1
0.1
Amount eluted
solvents
Trls-EDTA
167172
297612
161547
608904
882705
998633
with different
(cpm)
Dimethyl
sulfoxide
148330
16890
674281
226924
188118
72190
Estimated
percent
of input
sample
bound to
microsomes
_
41
-
53
-
11
Estimated
percent
of input
bound to
DNA-cellulose
47
5
81
27
18
7
-------
eluted by the organic solvent extractions. The results with dieldrin are
similar to those with benzopyrene; most of the carcinogen is bound initially
to either cellulose or DNA-cellulose. After elution with organic solvents
most of the input radioactivity was recovered from the cellulose column but
only 30 percent was recovered from the DNA-cellulose column. When S9 was
included in the dieldrin sample, slightly more radioactivity was bound
after the organic washes.
Somewhat different results were obtained when DNA-cellulose was allowed
to interact with carcinogen in a batchwise manner. In these studies,
following the DMSO extraction, almost all of the input radioactivity was
not bound to the DNA-cellulose. Less than 1 percent of the radioactivity
remained with the DNA-cellulose. The difference in the amount of material
recovered in the Tris-EDTA washes for microsome-treated samples and samples
not treated with microsomes is considered to represent the amount of material
bound to the microsomes. This material would not be in contact with the
DNA-cellulose (or cellulose) and therefore would be recovered in the superna-
tant. The amount bound to DNA-cellulose is considered to be the percent of
the input radioactivity not eluted with Tris-EDTA, (i.e., the amount extracted
by DMSO). This amount is not the amount bound to DNA, however, since in
control studies most of this radioactivity was also retained by cellulose.
The results of DNA-cellulose binding clearly show that a column'method
is preferred to a batchwise method of interacting sample with the DNA since
a tight binding is achievable by the former method but is not found for the
latter. The chlorinated hydrocarbon, dieldrin, and the polycyclic aromatic,
benzopyrene, are readily adsorbed onto DNA-cellulose and cellulose columns
16
-------
from dilute aqueous solutions. They are noc recovered from the DNA-cellulose
following several organic solvent washes, indicating that they can bind
tightly to the DNA in the DNA-cellulose. In the batchwise studies, almost
100 percent recovery of the bound compounds is possible with a DMSO wash
indicating that they are less tightly adsorbed. N-acetyl-2-aminofluorene
is not bound tightly in either method.
The potential use of DNA-cellulose binding -as a method for separating
carcinogens from noncarcinogens has not been assessed in these experiments
so no clearcut conclusions in this regard can be drawn. However, since
cellulose alone has an affinity for these carcinogens it is unlikely that
specificity of interaction could be achieved by this method.
Adsorption of Carcinogens to Isolated Rat Liver Nuclei
Another method investigated for testing the binding of carcinogens to
DNA was the interaction of arochlor 1254-induced rat liver nuclei with
carcinogens. This method was chosen for two reasons: (1) Nuclei have been
found to possess the metabolic activation enzymes necessary for conversion
of inactive carcinogens to DNA-binding carcinogens (16). (2) The native
state of the DNA in nuclei may allow the greatest possible interaction of
sample with DNA. Since the DNA in nuclei is associated with chromatin
proteins, such interaction may reflect the type of binding which would
exist with DNA in the intact organism.
For this study the method of Bresnick e_t al_. (16) was used. The
incubation system included: sodium phosphate buffer, pH 7.4, 100 pmoles;
17
-------
EOTA, 100 vimoles; glucose-6-phosphate, 18 ymoles; glucosc-6-phosphate
dehydrogenase, 6 units; NADPH, 1 pmole; MgCl2> 3 pmole; rat liver nuclei, 5
mg nuclear protein; and radiolabeled carcinogen in a total volume of 2.0
ml. The mixture was incubated at 37°C for 30 min in tubes covered by
aluminum foil to prevent light-induced reactions.
Nuclei were prepared from livers of rats injected intraperitoneally
with arochlor 1254 5 days prior to sacrifice. The excised livers from the
decapitated rats were washed in isotonic buffered saline, cut into small
pieces with scissors and homogenized in a Potter-Elvehjem apparatus with a
Teflon pestle using 20 up-and-down strokes and a medium speed setting. The
homogenate was centrifuged 2 min at 2000 rpm at 5°C in an International
centrifuge. The crude preparation of nuclei was further purified by layering
the pellet suspended in 3 ml 1.2 M sucrose and 2 ml 0.01 M Tris-HCl, 0.01 M
MgCl2 pH 7.2 over a solution of 20 ml 1.2 M sucrose (17). The pellet was
then taken up in 30 ml of 60 percent sucrose and centrifuged at 40000 x j»_
for 60 min in an International B60 centrifuge. The nuclear pellet was re-
suspended in 1.0 M sucrose, 1 mM calcium acetate,, rehomogenized, and centri-
fuged at 3000 x £ for 5 min. The precipitate was used as the nuclei in the
adsorption experiments described above.
After the incubation with carcinogen the nuclei were centrifuged at
3000 x _g_ and the DNA recovered from the nuclear pellet by the following
procedure: To the pellet was added 2 ml of 0.01 M Tris-HCl pH 8.0, 0.1 M
EDTA, 0.25 M NaCl, 0.5 percent sodium dodecyl sulfate, 50 yg/ml proteinase
K. The sample was incubated at 60°C for 16 hours. After cooling, the
solution was extracted with shaking on a Burrell-Wrist Action Shaker with 2
18
-------
ml of buffer-saturated phenol, 30 rain. The aqueous (upper) phase was
collected by centrifugation and the phenol layer washed with a small amount
of buffer. The combined aqueous layers were reextracted with chloroform:
isoamyl alcohol (24:1) until no protein interface was seen after centrifuga-
tion. The aqueous layer was then treated with 2 volumes of 0°C ethanol
and the DNA spooled. The spooled DNA was redissolved in 0.01 M Tris-HCl pH
-4
7.2; 10 M EDTA and aliquots were taken for scintillation counting.
In certain cases the DNA solutions were chromatographed in 0.01 M
Tris-HCl pH 7.0, 0.1 percent SDS buffer on BioGel A5 columns to confirm
that all of the radioactive label migrated with the void volume DNA peak.
Furthermore, the combined column fractions containing the DNA peak were
extracted with chloroform:isoamyl alcohol to verify that none of the radio-
activity was associated with protein which would be extracted by such
treatment. Results are given in Table 4. A representative profile of
BioGel A5 chromatography is given in Figure 1 for benzo(a)pyrene-nuclei
binding. These results show that as much as 18 percent of the input N-
acetyl-2-aminofluorene can be recovered associated with DNA isolated from
nuclei and that 54 percent of the input benzopyrene can be similarly recovered.
Using similar methods, dieldrin and dimethylnitrosamine can be recovered in
approximately 10 percent yield. The percent of the input radioactivity
bound to the nuclei after the incubation period is given in Table 4. For
benzopyrene and acetylarainofluorene almost 100 percent was initially'bound,
whereas for dieldrin and diraethylnitrosamine approximately 50 percent was
initially bound.
19
-------
TABLE 4. NUCLEI BINDING OF ORGANIC CARCINOGENS
to
o
Carcinogen
dieldrin
benzo(a)pyrene
N-acetyl-2-aminofluorene
dimethylnitrosamine
Input
(cpm)
4 x 105
6.9 x 106
1.6 x 106
1.7 x 106
Input
(nmoles)
150
0.32
31
232
Amount
associated
with nuclei
(cpm)
2.3 x 105
6.7 x 106
1.5 x 106
8.5 x 105
Percent
of input
associated
with nuclei
58
97
94
50
Amount
bound to DNA
spooled from
ethanol
(cpm)
4 x 104
1.9 x 106
3.3 x 105
1.4 x 105
Percent
of input
associated with
DNA
10
27
18
8
-------
l
I
1.6
1.4
1.2
0.8
o 0.6
w
0.4
0.2
L
55000
45000
35000
25000
15000
5000
cu
O
12 16
20 24 28 32
36
Fraction Number
Figure 1. Binding of Benzo(a)?yrene to DNA - Nuclei Binding Method
BioGel A5 Chromatography.
21
-------
From these results it can be concluded that nuclei binding of the
tested carcinogens occurs readily. Inclusion of S9 in the activation
system was found to be unnecessary (results not shown). The present studies
did not evaluate the specificity of nuclei binding with regard to noncarcino-
gens versus carcinogens so no conclusions can be drawn concerning this
question. The advantage in using nuclei binding as a method for recovering
biologically active substances from seawater concentrates is the relative
ease of the method and its closer resemblance ta the type of binding of
such substances to DNA in vivo.
Direct Binding of Carcinogens to DNA
Previous studies by other workers have shown that polycyclic aromatic
hydrocarbons (18) as well as chlorinated hydrocarbons (19) can bind to and
alkylate DNA. In addition, it is well known that most carcinogens can non-
covalently associate with DNA by intercalation or by binding to DNA phos-
phate (14). By such association DNA could potentially select for carcinogenic
substances from those incapable (and presumably noncarcinogenic) of such
interaction.
The simplest method for testing the binding of a substance to DNA is
by direct interaction with DNA in the presence or absence of S9 activating
enzymes. For the studies described below, the method of Gelboin (18) was
used with only slight modification. For testing the binding in the presence
of enzyme activation, rat liver S9 mix was used as described before. The
sample was incubated with S9 mix (0.1 ml) at 37°C for 15 min in the presence
of salmon sperm, highly polymerized, double-stranded DNA (0.125 mg). The
22
-------
total volume of the incubation mixture was 1 ml. After incubation, the DNA
was precipitated by addition of cold ethanol (2 volumes) and redissolved in
0.01 M Tris 10 M EDTA. This solution was extracted with chloroforrarisoamyl
alcohol (24:1) several times to remove protein and non-bound label. Following
the extraction, the DNA was again isolated by ethanol precipitation.
Aliquots of the resolublized DNA were determined for radioactivity.
The results of the direct binding studies are given in Table 5. All
of the tested carcinogens are recovered with the DNA in good yields except
dimethylnitrosamine. With dieldrin an increased yield (56 percent compared
to 12 percent) was obtained when S9 was included but with benzo(a)pyrene a
decreased yield was found (7.6 percent compared to 21 percent). This
effect may be ascribed to activation of benzo(a)pyrene to metabolites
incapable of DMA-binding. No effect of S9 was seen with N-acetyl-2-araino-
fluorene. With dieldrin, because it is present in higher amounts than
benzo(a)pyrene, some of the radioactivity may be associated with residual
protein of the S9 mix not removed by the chloroformrisoamyl alcohol washes.
Presumably, less of the dieldrin is activated to non-binding species than
is the benzopyrene sample but dieldrin may be tightly associated with the
microsomal protein. In general, the recoveries are similar to those found
in the other DNA-binding methods (see comparison Table 7, p. 28).
Equilibrium Dialysis Binding of Carcinogens to DNA
As a final method tested for the recovery of certain organic carcinogens
with DNA, the equilibrium dialysis of carcinogen solutions versus DNA in a
dialysis tube was measured. This method was studied for comparison only
23
-------
TABLE 5. DIRECT BINDING OF DNA WITH CARCINOGENS
Carcinogen
dieldrin
Input
S9 (cpm)
+ 19160
22900
Input
(nmoles)
7.5
8.6
Amount
bound
(cpm)
11173
2700
Percent of
input amount
bound
56
12
benzo(a)pyrene
N-acetyl-2-aminofluorene
dime thyInitro samine
85500
244700
397950
519500
418650
438000
-0.004
0.012
7-4
9.7
54
60
6534
51839
19364
24416
126
85
7.6
21
4.9
4.7
0.03
0.02
24
-------
and has little practical relevance due Co the long time (days) necessary to
achieve equilibrium.
In this procedure, 0.25 to 0.50 uCi of radiolabeled carcinogen is dis-
solved in 120 ml of distilled water and added to a cylinder. A solution of
calf thymus, highly polymerized, double-stranded DNA (6 ml containing 11 mg
DNA) was placed in a dialysis sack (12000 molecular weight cutoff) closed
with Spectrum clips.
The dialysis sack is suspended in the cylinder and the fluid stirred.
Aliquots were taken at intervals until equilibrium was achieved. Controls
with no DNA in the dialysis sack were run to ascertain losses of sample to
the walls and to measure the normal dialysis equilibrium. Any preferential
migration of the carcinogen into the dialysis bag over the control is a
measure of carcinogen binding by the DNA. Correction for the loss of net
dialysis volume due to the DNA solute has not been made. To confirm that
the binding was strong, dialysis of the equilibrated solution against
buffer was performed until no further radiolabel was lost from the dialysis
bag. The remaining amount in solution is that bound to DNA.
The results of the equilibrium dialysis studies are given in Table 6.
In most studies equilibrium was achieved after 4 to 6 days of dialysis.
Equilibrium dialysis studies were performed without prior activation of
carcinogens by S9 microsomes. From a measurement of material balance,
losses of carcinogen to the walls and dialysis tubing were approximately
10-20 percent of the total input radioactivity. Dimethylnitrosamine was
not tested in these studies.
25
-------
In Table 6, the amount bound to DNA after equilibration of the carcinogen
solution outside the dialysis bag with the DNA solution inside the dialysis
bag is given as the cpm/ral (inside) minus the cpm/ml (outside) times the
volume of the DNA solution. The cpm remaining inside the bag after exhaustive
dialysis against buffer (0.01 M Tris pH 7.0, 10~ M EDTA) is also given in
Table 6. Using the known specific activity of the carcinogen, the amount
bound per mg DNA can be calculated. This calculation uses the amount
associated with the DNA after exhaustive dialysis to remove extraneous
counts. The controls were run under identical conditions except that
buffer was added to the dialysis sack instead of DNA solution. The controls
at equilibrium gave identical amounts of radioactivity inside and outside
the bag. In addition, all radioactivity inside was lost upon dialysis
versus buffer.
The results show that most of the benzopyrene which is bound to the
DNA at equilibrium remains bound after dialysis, whereas for dieldrin and
N-acetyl-2-arainofluorene only one third or one fourth, respectively, remains
bound. The percent of the input radioactivity bound to DNA can be calculated.
In the equilibrium binding studies 1 percent of the dieldrin, 35 percent of
the benzo(a)pyrene and 1 percent of the N-acetyl-2-aminofluorene remains
bound to the DNA. Except for benzopyrene, such binding is less than that
for the other DNA interaction systems.
Comparison of the DNA-binding Methods
In Table 7 an attempt has been made to compare the different methods
used to recover carcinogen associated with DNA. The percent of the input
26
-------
TABLE 6. EQUILIBRIUM DIALYSIS BINDING OF DNA WITH CARCINOGENS
ho
Carcinogen
dieldrin
benzo(a)pyrene
N-acetyl-2-amino-
f luorene
Input
(cpra)
198960
214030
752160
Input
(timoles)
75
0.011
14
Amount of
DNA
(mg)
16
11
16
Amount
bound to DNA
at equilibrium
(cpm)
4770
110184
40188
Amount
remaining bound
after dialysis
against buffer
(cpm)
1786
74030
10410
Amount
bound
per mg
DNA
(pmoles)
43
4
17
-------
TABLE 7. COMPARISON OF DNA BINDING METHODS
to
CO
Amount Bound! to DNA
Dieldrin
Method
DNA-cellulose
(column)
DNA-cellulose
(batchwise)
nuclei binding
direct binding
equilibrium
binding
percent
of
input
(cpm)
66
47
10
12
1
pmole
mole DNA
phosphate
63000
187000
19000
2700
14
Benzo(a)pyrene
percent
of
input
(cpm)
30
81
27
21
35
umole
mole DNA
phosphate
9
100
153
7.6
1.3
AAF*
percent
of
input
(cpm)
1
18
18
5
1
pinole
mole DNA
phosphate
110
13000
8000
1100
5.7
Dimethylnitrosamine
percent
of umole
input
(cpm)
NTb
NT
8
0.02
NT
mole DNA
phosphate
NT
NT
24000
31
NT
''•N-acetyl-2-aminofluorene
M
Comparison is made only of samples not treated with S9.
NT « not tested
n
Amount bound to cellulose alone was not subtracted.
-------
amount bound and the amount of carcinogen bound per mole DNA phosphate in
the different binding studies have been compared. In the DNA-cellulose,
batchwise binding studies the amount associated with cellulose has not been
distinguished in Table 7 from the amount possibly bound to DNA. As a means
of carcinogen recovery, nuclei binding and DNA-cellulose binding are to be
preferred to direct or equilibrium DNA binding. However, the binding of
carcinogen to cellulose or protein in these methods would render them far
less specific for the separation of carcinogens-from other related organic
species which would presumably bind equally well. In the DMA-cellulose
studies, specific recovery of the DNA was not performed in the same way
(chloroform extraction, agarose chromatography) as for the other methods.
However, the binding to cellulose and DNA-cellulose is sufficiently great
to prevent recovery by toluene or DMSO extraction.
In the nuclei binding and direct binding studies, carcinogen associated
with the DNA was recovered by methods used to purify DNA from proteins.
Much greater binding was found in the nuclei studies compared to direct-
binding studies, particularly for dimethylnitrosamine. This difference
could be partly due to inadequate purification of the nuclei-derived DNA.
The presence of DNA-associated proteins could cause an increase in the
amount of carcinogen recovered in these studies. From the S9 microsome
activation studies the conclusion can be made that these carcinogens can
bind readily to protein.
The studies which have been described used carcinogens at the mmolar
concentrations associated with the radioactive specific activity as received.
Therefore, the maximum possible binding has not been explored. The effect
29
-------
of concentration on DNA-binding would certainly be valuable for future
discussion of this method as a means of recovery of organic carcinogens
from solutions. Since the different carcinogensj particularly benzo(a)
pyrene, are at different concentrations a comparison of the binding strength
of the different types of carcinogens is not possible. Benzo(a)pyrene was
at such high specific activity that its maximal binding probably was not
achieved. The fact that dieldrin shows the greatest binding may be the
consequence of its much lower specific activity rather than an intrinsically
higher association constant.
30
-------
REFERENCES
1. Riley, J.P-, and D. Taylor. The Analytical Concentration of Traces of
Dissolved Organic Materials from Sea Water with Amberlite XAD-1 Resin.
Anal. Chim. Acta, 46:307-309, 1969.
2. Osterroht, C. Development of a Method for the Extraction and Deter-
mination of Non-Polar, Dissolved Organic Substances in Sea Water. J.
Chrotnatog., 101:289-298, 1974.
3. Kennedy, D.C. Treatment of Effluent from Manufacture of Chlorinated
Pesticides with a Synthetic, Polymeric Adsorbent, Amberlite XAD-4
Environ. Sci. Technol., 7:138-141, 1973.
4. Harvey, G.R., W.G. Steinhauer, and J.M. Teal. Polychlorobiphenyls•in
North Atlantic Ocean Water. Sci., 180:643-644, 1973.
5. Junk, G.A., J.J. Richard, M.D. Grieser, D. Witiak, J.L. Witiak, M.D.
Arguello, R. Vick, H.J. Svec, J.S. Fritz, and G.V. Calder. Use of
Macroreticular Resins in the Analysis of Water for Trace Organic
Contaminants. J. Chromatog., 99:745-762, 1974.
31
-------
o. Yamasaki, E., and B.N. Ames. Concentrations of Mutagens from Urine by
Adsorption with the Nonpolar Resin XAD-2: Cigarette Smokers have
Mutagenic Urine. Proc. Natl. Acad. Sci. U.S.A., 74:3555-3559, 1977.
7. McCann, J., and B.N. Ames. Detection of Carcinogens as Mutagens in
the Salmonella/Microsome Test: Assay of 300 Chemicals: Discussion.
Proc. Natl. Acad. Sci. U.S.A., 73:950-954, 1976.
8. Kraybill, H.F- Global Distribution of Carcinogenic Pollutants in
Water. Ann. N.Y. Acad. Sci., 298:80-89, 1977-
9. 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.
10. Parry, J.M., D.J. Tweats, and M.A.J. Al-Mossawi. Monitoring the
Marine Environment for Mutagens. Nature, 264:538-540, 1976.
11. Burke, M.D., H. Vadi, B. Jernstrom, and S. Orrenius. Metabolism of
Benzo(a)pyrene with Isolated Hepatocytes and the Formation and Degrada-
tion of DNA-binding Derivatives. J. Biol. Chem., 252:6424-6431, 1977.
12. Selkirk, J.K. Divergence of Metabolic Activation Systems for Short-
term Mutagenesis Assays. Nature, 270:604-607, 1977.
13. Sims, P., and P.L. Grover. Epoxides in Polycyclic Aromatic Hydrocarbon
Metabolism and Carcinogenesis. Adv. Cancer Res., 20:165-274, 1974.
32
-------
14. Waring, M.J. Drugs which Affect the Structure and Function of DNA.
Nature, 219:1320-1325, 1968.
15. Alberts, B.M., and G. Herrick. DNA-Cellulose Chromatography. Math.
Enzymology, 21:198-217, 1971.
16. Bresnick, E., T.A. Stoming, J.B. Vaught, D.R. Thakker, and D.M. Jerina.
Nuclear Metabolism of Benzo(a)pyrene and of'(+)-trans-7,8-Dihydroxy-
7,8-Dihydrobenzo(a)pyrene. Arch. Biochem. Biophys., 183:31-37, 1977-
17. Chauveau, J., Y. Moule, and C. Rouiller. Isolation of Pure and Unaltered
Liver Nuclei, Morphology and Biochemical Composition. Exp. Cell Res.,
11:317-321, 1956.
18. Gelboin, H.V. A Microsome-Dependent Binding of Benzo(a)pyrene to DNA.
Cancer Res., 29:1272-1276, 1969.
19. Rocchi, P-, G. Prodi, S. Grilli, and A.M. Ferreri. In Vivo and In
Vitro Binding of Carbon Tetrachloride with Nucleic Acids and Proteins
in Rat and Mouse Liver. Int. J. Cancer, 11:419-425, 1973.
33
-------