DNA adducts in target and nontarget tissues of 3,2'-dimethyl-4-aminobiphenyl in rats.

3,2'-Dimethyl-4-aminobiphenyl (DMAB) is a potent carcinogenic aromatic amine which demonstrates multiorgan tropism in rats. Using polyclonal antibodies against DMAB-DNA adducts, an immunohistochemical procedure as well as an ELISA were applied to investigate the relationship between DMAB-DNA adduct formation and tumorigenicity. Dose-related nuclear staining was observed 24 hr after application of the carcinogen but specificity in terms of sites of tumor development was lacking. No observable decrease in staining intensity was evident in most organs by 168 hr after administration of DMAB. Specific DNA lesions which could be responsible for carcinogenesis were not detected by the 32P-postlabeling method. The tumorigenic response of the ventral prostate in five strains of rats was roughly paralleled by DMAB-DNA adduct levels generated in the tissue. Strong enhancement of bladder tumor development by combined administration of the antioxidants, butylated hydroxyanisole, or butylated hydroxytoluene, with DMAB, was well correlated with an increase in DNA adducts. Our findings so far suggest that DNA adduct formation itself does not determine the carcinogenic organotropism of DMAB. Other factors (including cell proliferation and promotion by exogenous agents) may play important additional roles. For individual target organs or tissues, however, there seems to be a correlation between adduct levels and carcinogenic potential. ImagesFigure 1. AFigure 1. BFigure 1. CFigure 1. DFigure 2. AFigure 2. BFigure 2. CFigure 2. DFigure 3. AFigure 3. BFigure 3. CFigure 3. DFigure 3. EFigure 3. F


Introduction
The process of chemically induced carcinogenesis generally is divided conceptually into initiation and promotion stages. In the initiation stage, reactive carcinogen and/or metabolite(s) covalently bind to DNA in target tissues to form DNA adducts. These modifications lead to development of DNA damage, resulting in the production of initiated cells. These cells are then susceptible to promotional stimuli, which leads to tumor formation. The initiation event is believed to be an irreversible process and therefore a key point in chemical carcinogenesis.
3,2'-Dimethyl-4-aminobiphenyl (DMAB) is a potent multiorgan carcinogenic aromatic amine that targets the colon, Zymbal gland, preputial gland, uri-nary bladder, prostate, pancreas, and subcutis (1)(2)(3)(4)(5). Tissue specificity is also exhibited within the prostate complex (4)(5)(6). The seminal vesicles are also targeted and develop atypical hyperplasias. Clarification of the relationships between the carcinogenicity of DMAB and DNA damage in specific organs or tissues is an important goal in understanding mechanism(s) of chemical carcinogenesis. Specific antibodies have recently been developed against such carcinogen-DNA adducts and used for the sensitive detection of adduct formation in tissues (7)(8)(9)). An immunohistochemical approach using such specific antibodies makes possible the precise localization of adduct formation at the cellular level (10,11).
We have developed polyclonal antibodies specific for DMAB-DNA adducts. In this article, we describe the immunohistochemically visualized distribution of DNA adducts, and present a quantitative comparison of adduct levels in relation to tumor response.
Male F344 rats (Charles River, Japan), 9 weeks old, received DMAB (>98% purity) dissolved in 0.5 ml of corn oil by sc injection at various doses and 24 hr later all animals were killed and autopsied. The influence of orchiectomy and chemical castration with estrogen on adduct formation was also studied. Organs or tissues were fixed in cold acetone for 1 week and processed routinely for preparation of paraffin sections. Adducts in paraffin sections were visualized immunohistochemically as described previously (13).
Tables 1 and 2 summarize data for tissue or organ specificity of carcinogenic action and dose-related staining intensities in rats given sc injections of 200, 50, and 10 mg DMAB/kg bw. Dose-dependent staining was observed with only very weak intensity at a level of 10 mg. At a dose of 200 mg/kg, almost all organs examined demonstrated positive staining, at least to a certain degree, regardless of their DMAB-carcinogenic target status ( Figure 1). In the male accessory sex organs, strong staining was observed in the ventral and lateral prostate lobes; the seminal vesicles exhibited prominent nuclear staining, lighter staining being evident in the dorsal and anterior (coagulating gland) lobes ( Figure 2). A positive reaction was found for both the acinar and Environmental Health Perspectives ductal epithelial cells of the prostate. Target sites of DMAB-carcinogenicity are the ventral prostate and seminal vesicles and other target organs, including the urinary bladder, colon, preputial glands, Zymbal glands, sebaceous glands, and pancreas. All exhibited positive staining.
However, nontarget organs such as the kidney, lung, brain, and salivary glands also exhibited positive staining. The positive sites in the liver were the hepatocytes and sinusoidal endothelial cells distributed uniformly throughout the lobules and those in the kidney were the distal and collecting tubule cells. Positive binding in the salivary glands was located in the acinar cells of the submandibular gland and the duct epithe-lial cells of the sublingual glands. A weak positive staining was also found in primary spermatocytes of the testis and accessory glands of the skin. In the brain, glia cells and choliod plexus were positive, while neurocytes were negative. Heart muscle proved negative. Investigation of DMAB-DNA adduct persistence continued until   (4,5,7,16,23). Cln parentheses are staining intensities of interstitial cells of the crypts. dTumors could be induced by administration of DMAB and then testosterone propionate. @Glia cell/ependymal cell. 168 hr after the administration of DMAB revealed no clear changes in staining intensity in organs other than the urinary bladder, colon, ileum, and duct epithelium of the preputial glands (Table 2). Orchiectomy caused severe atrophy of all parts of the accessory sex organs associated with a marked to moderate decrease in staining intensity. Mild to moderate atrophy induced by chemical castration with ethinyl estradiol was accompanied by moderate to slight decrease in staining intensity.

Strain Differences in Prostatic Carcinogenesis and Adduct Formation
Tumorigenic response to DMAB in the prostates of F344, ACI, Lewis, CD, and Wistar strain rats was examined in relation to adduct formation (14). Rats received DMAB sc at a dose of 50 mg/kg bw 10 times, once every other week. The experiment was terminated at week 60. The final survival rates were 44, 28, 46, 27, and 0% in F344, ACI, Lewis, CD, and Wistar rats, respectively. All Wistar strain rats died by week 50. Carcinomas of the ventral prostate were found in 46, 50, 5, 0, and 0% of F344, ACI, Lewis, CD, and Wistar strain animals, respectively ( Table 3).
The amounts of DMAB adducts in DNA samples isolated from the prostate and liver of the five strains of rats 24 hr after administration of a single sc injection of DMAB 200 mg/kg bw were quantitatively analyzed using ELISA. The results were calculated as numbers of adducts per 106 nucleotides (Table 3). Ventral prostate values for the 5 strains ranged from 2.75 adducts per 106 for F344 and Wistar animals, to 1.00 adduct per 106 nucleotides in Lewis rats. The dorso-lateral lobe levels were about 45 to 70% of those in the ven-  (Table 4).  (15) aMeans ± SE of four determinations. bSignificantly different from the corresponding control value at p < 0.001. CSignificantly different from the corresponding control value at p < 0.01. dSignificantly different from the corresponding control value at p < 0.05. Modification of DMAB-multiorgan carcinogenesis by simultaneous treatment with BHA or BHT was studied using male F344 rats (15). Four-week old animals were given DMAB (sc injection of 50 mg/kg bw once a week for 10 weeks) during administration of BHA (2.0% in diet for 11 weeks) or BHT (1.0% in diet for 11 weeks). The experiments were terminated at week 55, at which time the incidences of urinary papillomas and carcinomas in rats given DMAB plus BHA or BHT were more than 70%, whereas no bladder tumors were evident in animals given DMAB alone ( Table 5). The development of preneoplastic lesions in the liver and pancreas, however, was reduced approximately one half by BHA and BHT. No modification by BHIA or BHT was evident for these tumor inductions in other organs. The urinary bladders of rats given DMAB with BHA or BHT had two or three times the levels of DMAB-DNA adducts in rats given DMAB alone (Table 5). In contrast, the amounts of the adducts in the livers of rats given DMAB plus BHA or BHT were significantly lower than those given DMAB alone. Suppression of adduct formation in the liver by BHT was stronger than by BHA. Those in the colon were also decreased by the concomitant treatment with BHA or BHT.
Volume 102, Supplement 6, October 1994 Immunohistochemically, the nuclear staining intensity of the urinary bladders of rats given DMAB plus BHA or BHT was clearly stronger than that of rats given DMAB alone. In the livers, however, the intensities were weaker in the antioxidant groups. BHT suppressed the formation of the adducts particularly in the centrilobular areas, whereas BHA exerted diffuse effects. Demonstmtion ofSpeific DNALesions by the P-Postlab& Method DMAB-DNA adducts were analyzed by the 1-butanol extraction method (17). DNA (10 pg) was digested to 3'-mononucleotides at 370C for 3.5 hr with micrococcal nuclease and spleen phosphodiesterase as described previously (18). DMAB-DNA adducts were then extracted with water-saturated 1-butanol from 8-jg samples of DNA digests according to the method of Gupta (17), and then 32P-labeled by T4 polynucleotide kinase with 2.0 pM[y-32P]-ATP (7000 Ci/mmoles) at 37°C for 1 hr. The labeled solutions were applied to a PEI-cellulose layer. Adducts were located by screen-enhanced autoradiography at -80°C for 2 days. Amount of total nucleotides were determined as described previously (17) and adduct levels were calculated from radioactivities of the spots for each adduct divided by the radioactivity for the total nucleotides (17). Figure 3 illustrates autoradiograms of DNA adducts in the prostate complex and liver of rats exposed to 150 mg/kg of DMAB 24 hr before sacrifice. The levels AB were 37.1, 14.6, 6.8, 15.3, 6.7, and 39.7 adducts per 109 nucleotides in the ventral, lateral, dorsal, anterior lobes, the seminal vesicles, and the liver, respectively. Four spots usually were observed in both the prostate complex and the liver. The ratios of adduct levels of no. 2 and no. 3 spots to total adducts were lower in the prostate complex (about 10%) than the liver (about 30%). There were no clear differences in the ratio of each spot among prostate lobes, but no. 5 spot could only be clearly detected in the ventral prostate and not in other organs.

Discussion
Immunohistochemical staining of DNAcarcinogen adducts revealed that DMAB metabolites bind to DNA in a variety of organs. This approach has the advantage of allowing localization at the cellular rather than the tissue or organ levels. However, observed distribution of adducts did not obviously correspond to the reported and observed sites of tumor development in DMAB-exposed rats. While all target organs of DMAB examined, including the colon, small intestine, ventral prostate, seminal vesicles, preputial gland, urinary bladder, sebaceous gland, pancreas, and liver were positive, nontarget organs such as the kidney, salivary gland, testis, and anterior, lateral, and dorsal prostate lobes also demonstrated considerable formation of DNA adducts. Similar anomalous findings in nontarget tissues have been reported earlier (19). Quantitation of .. D C X pr t00: 0 : adduct levels also did not reveal any direct correlation with DMAB-organotropism (12). In addition, persistence of DMAB-DNA adducts up to 168 hr after the single carcinogen treatment in the present experiment did not allow an explanation of organotropism on the basis of differential repair.
DNA modification through covalent binding of carcinogen(s) is thought to be an essential step for the activity of genotoxic carcinogens but not in itself enough to initiate carcinogenesis. In addition to possible formation of DNA adduct types that are nonrelevant to carcinogenesis, a high efficiency of DNA repair and therefore inadequate fixation of heritable DNA lesions also have been suggested as complicating factors. The chemical natures of the main DNA-DMAB products were identified by Flammang et al. (20) as N-(deoxyguanosin-8-yll-DMAB (60-70%), 5-(deoxyguanosin-N -yl)-DMAB (2-3%) and N-(deoxyadenosin-8-yl)-DMAB (1-3%). At present, antibodies specific for individual isolated DNA adducts are not available, and detection of the forms primarily responsible for carcinogenesis awaits the development of such analytical tools.
Many carcinogenic compounds that have no effect under normal conditions induce liver carcinomas when applied during the period of regenerative cell proliferation after partial hepatectomy (22). DMAB was found to induce rat pancreatic tumors when given at doses necrogenic for the pancreas (13) and cell proliferation at the time of carcinogen action is generally accepted to play an important role in the initiation of carcinogenesis, presumably via fixation of heritable DNA lesions. Accordingly, one might expect development of liver tumors in rats given DMAB only in association with partial hepatectomy, since this carcinogen is not particularly toxic in this organ. Since carcinogenesis is a multistage process, a second complicating factor that should be recognized is that promotion may be required for the development of visible tumors in organs with Environmental Health Perspectives * . . :: DMAB-DNA adduct formation that are normally considered to be nontarget tissues. A good example for this is apparent from experiment 3. Under normal conditions, DMAB induces prostate tumors only in the ventral lobe. However, long-term treatment with TP after DMAB injection results in tumor development also in the lateral and anterior prostate and seminal vesicles.
There is a possibility that the adduct patterns in different tissues might vary with single and multiple dosing and indeed our studies with DMAB have shown that carcinogenic response is related to the dosing schedule (5). The multiorganotropicity of DMAB carcinogenicity is also dependent on the strain of rat investigated. With respect to prostate tumorigenesis, F344 rats showed a high susceptibility not shared by other strains. N-OH-DMAB, a proximate form of DMAB, can be further metabolized by O-acetylation to yield a carcinogenic form that is capable of reacting with nucleic acids (20,23). However, the prostate tumorigenic response could not be correlated with the activity of this enzyme in the ventral lobe (24), suggesting that prostate-susceptibility to DMAB carcinogenicity is related to both activation within the target organ and disposition and excretion patterns of the chemical and/or its metabolites in the whole body. The observed differences in the numbers of DMAB-DNA adducts in the ventral prostate among the five strains of rats, on the other hand, did demonstrate a good correlation between these lesions and the tumorigenic response, except in the Wistar rat case, which had a poor survival period. The very weak staining in atrophied prostate tissue resulting from orchiectomy might account for the decreased tumor response in the prostate under such conditions (21). Loss of activating enzyme(s) in the epithelial cells in this case may be one of the reasons why DMAB-DNA adduct formation is decreased.
The striking finding that a simultaneous administration of DMAB with BHA or BHT strongly enhanced urinary bladder tumor induction while inhibiting the appearance of preneoplastic lesions in the liver and pancreas offered direct support for an adduct role. Thus, quantitative assay of carcinogen-DNA adduct formation clearly demonstrated that the amounts of DMAB-DNA in the urinary bladder, liver, and pancreas correlated well with the tumorigenic response; both BHA and BHT increased levels of DMAB-DNA adducts in the urinary bladder, but decreased those in the liver and pancreas. A discrepancy was observed only for the colon where a decrease in the amount of DMAB-DNA adducts was found but no change in tumor incidence occurred. The reason might be that although the drop in DNA adducts in the experimental groups was statistically significant, it was not low enough to result in decreased tumor induction. In fact, no apparent difference in nuclear staining intensity was present in the immunohistochemically stained colon samples.
The mechanisms by which BHA and BHT modulate DMAB carcinogenesis seem to be the same as those proposed in the case of modulation of 2-AAF carcinogenesis by BHT (25), i.e., changes in carcinogen metabolism resulting in increased or decreased ultimate carcinogenic species, because BHA and BHT are inducers of drug metabolizing enzymes such as cytochrome P450s, epoxide hydroxylase, glucuronyltransferase, and glutathione Stransferase (26). Increased excretion of glucuronic acid conjugates (27) of N-OH-DMAB resulting from an alteration of liver metabolism by BHA and BHT could have played an important role in enhancing urinary bladder carcinogenesis and inhibiting hepatic carcinogenesis in the present experiment. Another factor that could have acted in the enhancement of urinary bladder carcinogenesis is increased DNA synthesis in the urinary bladder epithelium induced by BHA and BHT (28,29). Therefore, synergistic effects of increased reactive metabolite(s) of DMAB in the urine and increased DNA synthesis in the urothelium are medicated as the mechanism behind strong enhancement of urinary bladder carcinogenesis.
In conclusion, the data regarding initial formation of DNA adducts by DMAB and their persistence suggest that DNA lesions themselves are not sufficient to explain tumor induction. Involvement of other factors, including cell proliferation and promotion, require further investigation. The findings on staining differences and BHAor BHT-modulation, however, do indicate that there is a correlation between adduct levels and carcinogenic potential for individual target organs or tissues.