Monoclonal antibody to aflatoxin Bl-modified DNA detected by enzyme

Proc. Natl Acad. Sci. USA Vol. 78, No. 7, pp. 4124-4127, July 1981 Biochemistry

Monoclonal antibody to aflatoxin Bl-modified DNA detected by enzyme immunoassay (hybridoma/carcinogen-nucleic acid interactions/ultrasensitive enzyme radioimmunoassay)

AAGE HAUGEN*, JOHN D. GROOPMANtt, IH-CHANG Hsu*, GLENN R. GOODRICH*, GERALD N. WOGANt, AND CURTIS C. HARRIS* *Human Tissue Studies Section, Laboratory of Experimental Pathology, Division of Cancer Cause and Prevention, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20205; and tLaboratory of Toxicology, Department of Nutrition and Food Science, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139

Contributed by Gerald N. Wogan, March 26, 1981

ABSTRACT Monoclonal antibodies were obtained after fusion of mouse P3 x 63 myeloma cells with spleen cells isolated from BALB/c mice that had been immunized with aflatoxin B1-adducted DNA complexed with methylated bovine serum albumin. Selected hybridomas were found to produce monoclonal antibodies specific for aflatoxin B1-modified DNA containing both the 2,3dihydro-2-(N7-guanyl)-3-hydroxyaflatoxin B1 and the putative 2,3-

system (6, 7). The compound is frequently detected as a contaminant of grains and peanuts, especially under conditions of storage and transport where inadequate facilities permit fungal growth. The problem is especially acute in developing countries. AFB1 forms covalently linked adducts with guanine in DNA after oxidative metabolism to a highly reactive 2,3-exo-epoxide,

dihydro-2-(N5-formyl-2',5',6'-triamino-4'oxo-N5-pyrimidyl)-3-hy-

droxyaflatoxin B1, suggesting that these DNA adducts share a common antigenic determinant. The monoclonal antibody was not reactive towards the free aflatoxin B1-guanine adducts in solution, seven other aflatoxin derivatives, or benzo[a]pyrene-adducted DNA. A noncompetitive ultrasensitive enzyme radioimmunoassay could measure 15 fmol of aflatoxin B1-DNA adducts in 10 ng of DNA and was at least 100-fold more sensitive than the standard enzyme-linked immunosorbent assay. Competitive enzyme-linked immunosorbent assay with these monoclonal antibodies reliably quantitated aflatoxin B1 adducted in vivo to rat liver DNA at adduct levels of one aflatoxin B1 residue per 250,000 nucleotides. The competitive ultrasensitive enzyme radioimmunoassay was determined to be at least 6-fold more sensitive than the competitive enzyme-linked immunosorbent assay in analysis of aflatoxin B1-adducted DNA. Therefore, enzyme immunoassay using monoclonal antibodies will be useful analytical tools for studying both the molecular interactions of aflatoxin B1 with DNA and the occurrence of aflatoxin B1-DNA adducts in biological specimens from people exposed to this environmental carcinogen. Many chemical carcinogens produce electrophilic derivatives that react with sites in DNA to form covalently linked nucleic acid adducts (1). These adducts may produce conformational changes in the DNA double helix and cause mutation by miscoding during transcription. Such alterations in DNA structure may be requisite events in potentiating some forms of toxicity and initiating carcinogenesis. Recent development of highly sensitive enzyme immunoassays-e.g., enzyme-linked immunosorbent assay (ELISA) and ultrasensitive enzyme radioimmunoassay (USERIA) (2-4)-represent important advances for specific detection of DNA modification by carcinogens at low levels. Furthermore, the availability of monoclonal antibodies (5) of defined specificity and homogeneous binding characteristics could greatly increase the sensitivity of these analyses. Aflatoxin B1 (AFB1) is produced by some species of Aspergillusflavus and A. parasiticus. This mycotoxin is acutely toxic and highly heptocarcinogenic in many animal species and is mutagenic for a variety of eukaryotic and prokaryotic cells that either contain or are supplemented with a metabolic activation

the major adduct being 2,3-dihydro-2-(N7-guanyl)-3-hydroxyaflatoxin B1 (AFB,-N7-Gua) (8-10). This adduct has been identified in a wide variety of systems such as rat liver in vivo (11), cultured human bronchus and colon (12), and Salmonella typhimurium cultures incubated with a rat liver microsomal activation system (13). AFB1-N7-Gua in DNA contains a positive charge on the imidazole ring of the guanine moiety and scission of this ring would produce the putative 2,3-dihydro-2-(N5formyl-2', 5', 6'-triamino-4'-oxo'N -pyrimidyl)-3-hydroxy-aflatoxin B1 (AF-FAPyr) (8). This putative derivative has been found to be a persistent adduct in DNA of rat liver in vivo (14, 15) and in human lung cells in culture (16) after acute or chronic administration of AFB1. We report here the development of a monoclonal antibody directed towards AFBI-modified DNA and preliminary data from enzyme immunoassays with this antibody. This development will facilitate further the detection and quantification of biochemical lesions produced in DNA by the carcinogen and, therefore, may contribute to further understanding of the importance of DNA modification in the carcinogenic process. Further, adaptation and application of the technique also could make possible the detection and quantification of AFB1 adducts in DNA of tissues of people exposed to the carcinogen and, thus, contribute materially to an objective assessment of its possible role in the causation of human disease.

MATERIALS AND METHODS Cell Line and Media. The myeloma cell line used for fusion, P3 x 63 Ag8, was obtained from J. D. Minna (National Cancer Institute). The cell line was grown in suspension in Dulbecco's modified Eagle's medium (DME medium) with 20% (vol/vol) fetal calf serum and 20 ,ug of8-azaguanine (Sigma) per ml. Cloning medium consisted of DME medium with 20% fetal calf Abbreviations: AFB1, aflatoxin B1; AFB,-N7-Gua, 2,3-dihydro-2-(N7guanyl)-3-hydroxyaflatoxin B1; AF-FAPyr, 2,3-dihydro-2-(N5-formyl2', 5',6'-triamino-4'-oxo-N5-pyrimidyl)-3-hydroxyaflatoxin Bj; B[a]P, benzo[a]pyrene; DME medium, Dulbecco's modified Eagle's medium; ELISA, enzyme-linked immunosorbent assay; HAT medium, hypoxanthine/aminopterin/thymidine medium; USERIA, ultrasensitive enzyme radioimmunoassay. t Present address: Human Tissue Studies Section, Laboratory of Experimental Pathology, Division of Cancer Cause and Prevention, National Cancer Institute, National Institutes of Health, Bethesda, MD 20205.

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U. S. C. §1734 solely to indicate this fact. 4124

Biochemistry: Haugen et al. serum supplemented with 10% (vol/vol) NCTC-135 medium (GIBCO) containing 3 mM glutamine and nonessential amino acids concentrated 4-fold. For selection against parental myeloma cells, 16 uM hypoxanthine/0.5 1LM aminopterin/100 tM thymidine (HAT medium; Sigma) was added to the cloning

medium. Preparation of in Vitro AFBI Metabolite-Modified DNA. Calf thymus DNA (Sigma, type I) was modified by 3H-labeled AFB1 (3H-AFB1; specific acitvity, 5 Ci/mmol; 1 Ci = 3.7 X 1010 becquerels) from Moravek Biochemicals (City of Industry, CA) either by using phenobarbital-induced rat liver microsomes as described (9) or by using a modified chemical procedure with m-chloroperoxybenzoic acid (10). The amount of 3H-AFB1 metabolite bound to DNA was quantified by liquid scintillation and by high-pressure liquid chromatography as described by Croy et al. (11, 15). The DNA hydrolysates were quantitatively analyzed for specific adducts with a model 204 liquid chromatograph equipped with a model 440 detector (254 and 365 nm) and a #-Bondapak C18 column (all from Waters Associates) eluted at 1.0 ml/min with 18% (vol/vol) ETOH/10 mM KOAC, pH 5.0, at ambient temperature. DNA was determined by the method of Burton (17) as modified by Giles and Myers (18). Preparation of in Vivo AFB1-Modified DNA. Male CDF Fischer rats (Charles River Breeding Laboratories) weighing 100-150 g were given intraperitoneal injections of AFB1 dissolved in 50 p.1 of glass-distilled dimethyl sulfoxide (Burdick and Jackson Laboratories, Muskegon, MI). Animals were killed 2 hr after injection of 1, 0.1, or 0.01 mg of AFB1 per kg of body weight, and liver DNA was isolated as described (11, 15, 19). Immunization. AFB1-modified DNA was dissolved in phosphate-buffered saline (125 ,ug/ml). Methylated bovine serum albumin (Sigma; 1% in water) was then added with mixing until a final methylated albumin/DNA weight ratio of 1:1 was attained; insoluble complexes formed on mixing. The mixture was then emulsified by mixing with an equal volume of complete Freund's adjuvant (Miles). Female BALB/c mice, 3 months old, were given two intraperitoneal injections 3 wk apart, each consisting of 50 ,ug of methylated albumin-AFB,-DNA complex in 0.3 ml. Blood samples were removed from the tail 3 wk after the second injection, and the serum was assayed by ELISA for antibody activity. The mice were given a final booster immunization of 50 ,ug of methylated albumin-AFB,-DNA complex in 0.1 ml of saline into the tail vein 2 wk later. 'Cell Fusion and Cloning. Three days after the booster injection, animals were bled and spleens were aseptically removed and mechanically dissociated by passage through a sterile steel mesh. Cells were fuised by incubating 108 spleen cells with 107 myeloma cells in the presence of 50% polyethylene glycol 1000 (Baker Chemical, Brick Town, NJ) and were stirred gently for 1 min. Over the next 5 min, 10 ml of DME medium was added with gentle shaking ofthe tubes. The cells were pelleted by centrifugation and resuspended in hypoxanthine/thymidine medium at a density of 2-4 X 106 cells per ml and incubated overnight at 37°C in a 75-cm2 flat flask. The next day the culture was diluted to 1 X 106 cells per ml with HAT medium and distributed in 0. 1-ml aliquots into 200 wells of Costar multiwell plates. To each well was added 1 drop of HAT medium on day 7 and then on each succeeding second day. Isolated colonies were visible in 1-3 wk, at which time the medium was tested by ELISA for anti-AFB,-DNA adduct activity. Cells from positive wells were cloned by a limiting dilution procedure (20) in 96-well plates with mouse thymocyte feeder cells (107 cells per ml). Ten days later the wells were examined for single clones that were tested for specific antibody activity. Hybridomas were grown as ascites tumor cells in BALB/c mice and pretreated by intraperitoneal injection with 0.5 ml of

Proc. Natl. Acad. Sci. USA 78 (1981)

4125

pristane (Pfaltz and Bauer, Stamford, CT) 4-7 days prior to injection of 1-10 X 10' hybridoma cells. Screening of Hybridomas. To test monoclonal antibodies produced by the hybridoma cultures, plastic microtiter U-bottomed plates (Dynatech Laboratories, Alexandria, VA) were coated by drying in each well 10 ng (in 60 A.l of phosphate-buffered saline) of DNA either with or without AFB1 modification (1 adduct per 20 nucleotides). Media from hybridoma cultures were diluted 1:5 to 1:100, and 100 p.l was added to each microwell. This antigen-antibody reaction was enzymatically amplified by goat anti-mouse alkaline phosphatase conjugate (Kirkegaard and Perry Laboratories, Gaithersburg, MD); the enzyme substrate, p-nitrophenylphosphate (Sigma) was hydrolyzed to the yellow product, p-nitrophenol. The details of the ELISA procedure have been described by Hsu et al. (3, 4). Enzyme Immunoassay. Microtiter U-bottom plates were coated with DNA either with or without AFB1 modification by drying 10 ng of DNA in 60 1.l of saline per well for 16-24 hr at 370C. In a noncompetitive assay, the 10 ng of DNA that adsorbed to the solid phase contained 1500, 500, 150, 45, 15, or 0 fmol of AFB1-modified DNA having equimolar amounts of AFB,-N7-Gua and AF-FAPyr. The wells were washed with saline/Tween and incubated with 0.2 ml (per well) of saline/ Tween containing 2% fetal calfserum at 370C for 1 hr to prevent nonspecific binding of antibody. The antigen-antibody reactions were then carried out by adding to the adhered DNA in each well 0.1 ml ofascites diluted 1:20,000 in saline/Tween and incubating at 37°C for 90 min. In the competitive assay, diluted ascites fluid (1:10,000) was first mixed with an equal volume of competitor and then immediately applied to the well. In the com'petitive experiments, calfthymus DNA was used as a control competitor for the monoclonal antibodies. We determined that 50% inhibition was obtained with 10 ,ug of calf thymus DNA per well. Therefore, any background contributed by calf thymus DNA was subtracted from the data obtained with the AFB1-adducted DNA in order to calculate the inhibition curves found in Fig. 1. After the antibody was added to the wells, the plates were further incubated for 90 min at 37°C with 0.1 ml of affinity-purified goat antimouse IgG conjugated to alkaline phosphatase per well (Kirkegaard and Perry Lab.; 1:250 dilution for ELISA and 1:1000 dilution for USERIA). The alkaline phosphatase activity that adhered to the plates was measured by adding 0.1 ml of substrates per well, either 1 mg of p-nitrophenylphosphate per ml for ELISA or 106 cpm in 100 pmol of [3H]adenosine-5'-monophosphate (New England Nuclear) for USERIA. The assays were done in triplicate. Details of the USERIA procedure have been described (2,4). Determination of Antibody Specificity. The plates were coated with 10 ng of DNA per well, either with or without AFB1 metabolite modification (1 per 20 bases). Competitive inhibition of antibody reacting with antigen bound to the solid phase by free adducts and aflatoxin-related compounds in the reaction mixture was carried out by incubation of 0.1 ml of saline/ Tween containing antibody (ascites 1:20,000 dilution) and various amounts of either adducts or aflatoxin-related compounds per well. The plates were then washed and incubated with 0.1 ml of anti-mouse IgG alkaline phosphatase conjugate (1:250 dilution) per well. Finally, 0.1 ml of p-nitrophenylphosphate solution (1 mg/ml) was added to each well and the amount of p-nitrophenol produced was estimated by A405 nm after a 3-hr incubation. Isotype Determination. AFB1-modified DNA and calf thymus DNA (10 ng per well) were added to each microwell. After incubation with 50 ul of ascites fluid, goat antiserum specific for mouse IgGj, IgG2, or IgA (Meloy Laboratories, Springfield,

Proc. Natl. Acad., Sci. USA 78 (1981)

Biochemistry: Haugen et al.

4126

A

-4 50

0 ._

._-

B

._-

100

_

-4

Table 1. Ig isotype of monoclonal antibodies* IgM Sample IgG1 IgG2 IgA 1-A-2 0.02 0.99 0 0 1-B-2 0.01 0.72 0 0 0 NMS 0 0 0

Isotype analysis of the Igclass ofhybridoma antibodies. Ascites from hybridomas 1-A-2 and 1-B-2 at dilution 1:20,000 were tested. The ELISA procedure was modified by using goat antisera specific for mouse IgG1, IgG2, or IgA (1:5000 dilution) or rabbit anti-mouse IgMk from BALB/c myeloma (1:5000 dilution) in the second incubation before adding rabbit anti-goat IgG-alkaline phosphatase conjugate (1:5000 dilution) or goat anti-rabbit F(ab')2 antiserum-alkaline phosphatase conjugate (1:5000 dilution), respectively. Normal mouse serum (NMS) was used as control. * Absorbance at 405 nm.

0

50 0

I11111111 111I1I1I O.01

01

11H I1111111

1.0

10.0

1

11

100.0

AFB1, pmol FIG. 1. (A) Comparison of competitive inhibition curves obtained with monoclonal antibody 1-B-2. Increasing concentrations of various compounds were mixed with diluted 1-B-2 in USERIA (1:100,000 dilution (m) and in ELISA (1:20,000 dilution) (e). (B) ELISA data. o, Data points obtained with in vivo-adducted AFB,-DNA; *, data points obtained with in vitro-modified AFB1-DNA at a level of one AFB1 residue per 20 nucleotides.

VA) was added. Thereafter, rabbit anti-goat IgG-alkaline phosphatase conjugate (Miles-.Yeda, Israel) was added. For IgM determination, rabbit anti-mouse IgMk from BALB/c myeloma (gift from M. Pawlita, National Institutes of Health) and goat anti-rabbit F(ab')2 antiserum conjugated to alkaline phosphatase (Northeast Biomedical Laboratories, South Windham, MA) was used. Finally, p-nitrophenylphosphate was added for 1 hr at 370C, and A4015 was determined. The assays were done in duplicate.

antibodies 1-A-2 and 1-B-2 both of which were found to be of subtype IgG1. A comparison of the sensitivity of using these antibodies in noncompetitive ELISA and USERIA for the determination of AFB1-modified DNA is found in Table 2. USERIA was up to 100-fold more sensitive than ELISA and easily detected 15 fmol of AFB1 adducts in a total of 10 ng of DNA. This detection level is similar to previous results with DNA-modified by other carcinogens, acetylaminofluorene and benzo[a]pyrene (B[a]P) (3, 4). Fig. 1A compares the sensitivity of competitive USERIA and ELISA assays for quantitating AFB1-adducted DNA. The modified DNA used was produced by the in vitro technique and contained one adduct per 20 nucleotides. High-pressure liquid chromatographic analysis of this material showed that- 95% of the bound AFB1 was present in the form ofAFBI-N7-Gua or AFFAPyr. Under the conditions used, USERIA was found to be 6- to 7-fold more sensitive than ELISA in detecting AFB1-DNA adducts with a 50% inhibition of 1.0 pmol. In a second experiment, AFB,-adducted DNA was isolated from livers of rats injected with 1.0 and 0.1 mg of AFB1 per kg of body weight and killed 2 hr later. The competitive ELISA inhibition curves obtained with this in vivo-modified DNA are found in Fig. 1B, which contains data compiled from in vivo DNA samples ad-

nm

RESULTS BALB/c mice were injected with AFB1 metabolite-modified DNA complexed with methylated bovine serum albumin as an immunogenic carrier, in complete Freund's adjuvant. Sera from these animals were tested 6 wk after initial injection and screened for antibodies specific for AFB,-modified DNA, as described above. The spleen cells from one of five positive mice were fused with mouse myeloma cells (P3 x 63 Ag8), and 2-3 wk later hybridomas were present in about 200 microtiter wells. Eight hybridomas were found to secrete antibodies reactive with AFB1-modified DNA* and, to a lesser degree, calf thymus DNA. These hybridomas, were cloned by a limiting dilution procedure and were retested for specific antibody production. Large quantities of monoclonal antibody were obtained by harvesting ascites fluid and sera from BALB/c mice injected with hybridoma cells. The properties of two monoclonal antibodies produced by this protocol, referred to as "1-A-2" and "1-B-2," are described in this report. 1-A-2 and 1-B-2 were initially characterized for immunoglobulin type and subgroup because hybridomas are known to produce antibodies from the various classes of immunoglobulins (21). Table 1 contains the results of isotype determination of

Table 2. Detection of AFB1-DNA adducts in the ELISA and

USERIA* Monoclonal antibody 1-A-2

AFB1-DNA adducts, fmoi/10 ng of DNA 1500 500 150 45 15

1-B-2

ELISA 130 ± 17t 95 ± 8t 55 ± lit 35 it 24

2

Calf thymus DNA

26 ± 4

1500 500 150 45 15

142 ± 7t 98 5t 47± 4t

26

3

16± 6

USERIA 81 ± 12t 761± 2t 63 ± 3t 48 ± 8t 43 ± 6t 30 ± 2

83 ± 16t 75 ± 15t 60± 4t 59 ± 13t 49 ± 5t 39 ± 2

24 1 CalfthymusDNA * Antigen, AFB1-modified DNA, was a 1:1 mixture of AFB,-N7-Guaand AF-FAPyr-modified DNA in a total of 10 ng of DNA coated per well. Primary monoclonal antibody was diluted 1:20,000 and alkaline phosphatase-anti-mouse IgG conjugate was diluted 1:250 for ELISA and 1:1000 for USERIA. ELISA values are expressed as absorbance (x 10-2) at 405 nm. USERIA values are expressed as dpm x 10-3. t Student's t test, P < 0.05; value compared to calf thymus DNA

Proc. NatL Acad. Sci. USA 78 (1981)

Biochemis": Haugen et al. ducted at levels of one adduct per 30,000 and 250,000 nucleotides. Further experiments determined if these monoclonal antibodies were specific for the major AFBI-DNA-base adducts (AFB,-N7-gua and AF-FAPyr) and other oxidative metabolites of AFB1. These monoclonal antibodies apparently recognize only AFB1 bound to DNA because the following concentrations of compounds resulted in less than 2% inhibition: 11,000 pmol of AFB1, 530 pmol of aflatoxin B2, 32,000 pmol of aflatoxin P1, 5100 pmol of aflatoxin M1, 10,000 pmol ofaflatoxin B2, 260 pmol ofaflatoxin G1, 2000 pmol of 2,3-dihydro-2,3-dihydroxyaflatoxin Bl, 2100 pmol of AFBl-N7-Gua, and 2100 pmol of AF-FAPyr. Finally, we also determined that 5.5 pmol of B[a]P metabolite bound to DNA at a level of 1.24 adducts per 100 nucleotides was not recognized by either 1-A-2 or 1-B-2 antibody. These data support the conclusion that these monoclonal antibodies are specific for an altered conformation in DNA resulting from the covalent binding of AFB1. DISCUSSION The monoclonal antibodies described here are specific for AFBl-modified DNA containing either or both AFBl-N7-Gua and AF-FAPyr covalent adducts. These antibodies are able to recognize differences in confarmational changes induced by the covalent binding of AFB1 to DNA, but they do not react with free adducts (AFB,-N7-Gua, AF-FAPyr) in solution.The simplest explanation of these findings is that the antigenic determinants detected by these monoclonal antibodies are fixed in the same configuration in DNA whether the adduct form is AFBl-N7-Gua or AF-FAPyr. This result is different from data obtained with rabbit polyclonal antiserum against products of B[a]P metabolites covalently bound to DNA. This antiserum was found to recognize both B[a]P-modified DNA and the isolated B[a}P-DNA adducts (22), although the level of interaction with the free B[a]P-DNA adducts was a factor of eight less than that with the covalently bound DNA. By modifying radioimmunoassay and ELISA, Harris and coworkers recently reported the development of an USERIA (2). A noncompetitive USERIA is 10- to 1000-fold more sensitive than radioimmunoassay or ELISA for the quantitation of rotavirus, cholera toxin, acetylaminofluorene-modified DNA, and B[a]P-modified DNA (2-4, 22, 23). We report here that noncompetitive USERIA can detect 15 fmol ofAFB1 adducts bound to 10 ng of DNA. This represents a level of modification of one AFB1 residue per 2200 nucleotides. However, competitive enzyme immunoassays are at least 100-fold more sensitive than noncompetitive procedures. By using a competitive ELISA, we readily detected AFB, adducts formed in vivo in rat liver DNA at levels of one AFB1 residue per 250,000 nucleotides. This demonstrates that the monoclonal antibodies are capable of recognizing AFB,-DNA adducts at levels of biological significance. Further refinement of the competitive USERIA technology should permit analysis of samples containing even lower levels of modification.

4127

At the present time, radioactivity is the most sensitive procedurefor determining low levels ofcarcinogen-modified DNA, but this is obviously useful only under experimental laboratory conditions. Highly sensitive enzyme immunoassays do not suffer from this limitation and should be useful probes for studying DNA repair, carcinogen metabolism, and other pharmacokinetic parameters of DNA modification. Finally, these monoclonal antibodies and enzyme immunoassays can be used to determine carcinogen adducts in biological specimens collected from people exposed to environmental carcinogens such as AFB1. The authors wish to express their appreciation for the technical and secretarial assistance of I. A. McClendon and S. Dorfman, respectively. 1. Miller, J. A. (1970) Cancer Res. 30, 559-576. 2. Harris, C. C., Yolken, R. H., Krokan, H. & Hsu, I. C. (1979) Proc. Natl. Acad. Sci. USA 76, 5336-5339. 3. Hsu, I. C., Poirier, M. C., Yuspa, S. H., Yolken, R. H. & Harris, C. C. (1980) Carcinogenesis 1, 455-458. 4. Hsu, I. C., Yolken, R. H. & Harris, C. C. (1981) Methods En-

zymol., in press. 5. Kohler, G. & Milstein, C. (1975) Nature (London) 256, 495-497. 6. Busby, W. F. & Wogan, G. N. (1979) in Food-Borne Infections and Intoxications, eds. Reimann, H. P. & Bryan, F. L. (Academic, New York), 2nd Ed., pp. 519-610. 7. Wogan, G. N. (1973) in Methods in Cancer Research, ed. Busch, H. (Academic, New York), Vol. 7, pp. 309-344. 8. Lin, J. K., Miller, J. A. & Miller, E. C. (1977) Cancer Res. 37, 4430-4438.

9. Essigmann, J. M., Croy, R. G., Nadzan, A. M., Busby, W. F., Reinhold, V. N., Buchi, G. & Wogan, G. N. (1977) Proc. Nati. Acad. Sci. USA 74, 1870-1874. 10. Martin, C. N. & Garner, R. C. (1977) Nature (London) 267, 863865. 11. Croy, R. G., Essigmann, J. M., Reinhold, V. N. & Wogan, G. N. (1978) Proc. Natl. Acad. Sci. USA 75, 1745-1749. 12. Autrup, H., Essigmann, J. M., Croy, R. G., Trump, B. F., Wogan, G. N. & Harris, C. C. (1979) Cancer Res. 39, 694-698. 13. Stark, A. A., Essigmann, J. M., Demain, A. L., Skopek, T. R. & Wogan, G. N. (1979) Proc. Natl. Acad. Sci. USA 76, 13431347. 14. Wogan, G. N., Croy, R. G., Essigmann, J. M., Groopman, J. D., Thilly, W. G., Skopek, T. R. & Liber, H. L. (1979) in Environmental Carcinogenesis, eds. Emmelot, P. & Kriek, E. (Elsevier/ North Holland, Amsterdam), pp. 97-121. 15. Croy, R. G. & Wogan, G. N. (1981) Cancer Res. 41, 197-203. 16. Wang, T. V. & Cerutti, P. A. (1979) Cancer Res. 39, 5165-5170. 17. Burton, K. (1956) Biochem. J. 62, 315-323. 18. Giles, K. W. & Myers, A. (1965) Nature (London) 206, 93. 19. Groopman, J. D., Busby, W. F. & Wogan, G. N. (1980) Cancer Res. 40, 4343-4351. 20. KcKearn, T. J. (1980) in Monoclonal Antibodies, eds. Kennett, R. H., McKearn, T. J. & Bechtol, K. B. (Plenum, New York), p. 374. 21. Rajewsky, K., von Hesberg, G., Lemke, H. & Hammerling, G. J. (1978) Ann. Immunol. 129, 389400. 22. Poirier, M. C., Santello, R., Weinstein, I. B., Grunberger, D. & Yuspa, S. H. (1980) Cancer Res. 40, 412-416. 23. Hsu, I.-C., Poirier, M. C., Yuspa, S. H., Grunberger, D., Weinstein, I. B., Yolken, R. H. & Harris, C. C. (1981) Cancer Res. 41, 1091-1095.