Click for next page ( 274


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 273
CIRCULATORY SYSTEM Juchau et al .76~77 summarized a body of literature bearing on the hypothesis that PAHs may play an important role in the pathogenesis of arteriosclerotic lesions. The validity of this hypothesis apart, these investigators clearly demonstrated that aortic tissues from a number of species, including man, have detectable, albeit low, monooxygenase activities using BaP and 7,12-DMBA as substrates. Enzyme activities were comparable with those characterizing mouse skin. Cytochrome P-450 could be detected in primate aortas, and epoxide hydratase activity for BaP 4,5-oxide was identified in homogenates of the arterial walls of chickens and rabbits. The characteristics of the aortic monooxygenase for BaP resembled those of the enzyme system found in other tissues. It could be markedly induced, for example, by 3-MC, polychlorinated biphenyls (PCBs), and 5,6-benzoflavone; and, surprisingly, aortic homogenates produced higher than expected quantities (by as much as a factor of 28) of alkali-extractable metabolites when hematin was added to the reaction mixtures. Interestingly, hematin has been shown in other studies to degrade, in vitro, components of the monooxygenase system.lll The primary BaP metabolites formed in rabbit aortic homogenates were the 3-OH and 9-OH derivatives, phenolic compounds known to be cytotoxic. The authors cited unpublished data to show that the aortic metabolites of BaP form covalent bonds with such macro- molecules as calf-thymus DNA. Treatment of chickens with the inducer 3-MC markedly increased the amount of the PAM-DNA adducts, whereas addition of 7,8-benzoflavone in vitro inhibited binding. Aortic enzymes also have been shown to catalyze the formation of mutagenic metabolites from 7,12-DMBA. Thus, both cytotoxic and mutagenic metabolites of PAHs can be generated in vascular tissues. The possible relation of the formation of these compounds to the initial vascular injury that may presage the local development of an atherosclerotic plaque is of considerable interest. The interaction of benz~aJanthracene and BaP with crystalline human serym albumin in solution has been studied fluorimetrically by Ma et al. 10 Equilibrium studies indicated that both PAHs hind to the pro- tein to the same extent. Evidence of energy transfer from the trypto- phan residue of the protein (increase in the weak B region--395-420 nm--fluorescence of the PAHs) permitted an assessment of the mean dis- tance between the tryptophan and the bound ligand, thus identifying two different binding sites in the same general area. The authors sug- gested that structural differences among hydrocarbons, which may greatly affect their orientations on the protein molecule, influence mainly the selection of the binding site, rather than the binding equilibrium. In vivo BaP associates very little with serum albumin in the presence of lipoproteins. The kinetics of BaP transfer between human plasma lipoproteins have been examined by Smith and Doodyl63 with high-density lipoproteins (HOL), low-density lipoproteins (LDL), and 6-ll

OCR for page 273
very-low-density lipoproteins (VLDL) prepared from fresh unfrozen human plasma by ultracentrifugal flotation. BaP-lipoprotein interactions were analyzed fluorimetrically, and kinetic measurements were determined by stopped-flow techniques. The half-times of BaP transfer between HDLs, between LDLs, and between VLDLs were 40, 180, and 390 ms, respectively. The transfer of these PAHs among lipoproteins of the same density class was about one-twentieth that of pyrene under the same conditions. The rate of BaP transfer between lipoproteins also decreased with increasing size of lipoprotein; at equilibrium in vitro, VLDLs contain about 10 times more of the BaP than LDLs, and LDLs contain 20-50 times more than HDLs. The distribution between plasma and erythrocytes is different for 7,12-DMBA, BaP, benzanthracene, and anthracene, the mass of the PAH being associated with red cells (50, 70, 93, and 100t, respectively). Plasma lipid concentrations and the dynamics of lipid and lipoprotein metabolism clearly may have an impact on PAH distribution in blood and into specific tissues. For example, transfer of BaP is quite rapid, compared with the half-time for either hydrolysis of chylomicron triglyceride (about 2-5 min in humans) or clearance of the most abundant lipoproteins from the circulation (3-5 d in humans). The data of Smith and Doodyl63 concerning the role of plasma lipoprotins in the transport of PAHs corroborated and extended the findings of other invle~t~ai2ors wh6O examined the interaction of PAHs and plasma proteins. , , ,57,1 2 The specific process of BaP uptake from human LDLs into cultured human cells was examined by Remsen and Shireman.149 The cell lines used were WI-38, a human embryonic lung-fibroblast line, and GM 1915, a skin-fibroblast line derived from a patient with homozygous familial hypercholesterolemia; the former cells are LDL-receptor-positive, and the latter LDL-receptor-negative. Thus, in these studies, it was possible to explore the role of LDL receptors in the cellular uptake of PAHs that enter the bloodstream transported by chylomicrons and plasma lipoproteins. The results indicated that cellular uptake of the tritiated PAH by both cell lines from delipidated or serum-free medium varied linearly with concentration, whereas incorporation of PAH bound to LDLs was much less and, at higher lipoprotein concentrations, varied nonlinearly. The prese2ce of the PAH in the LDL preparation did not affect the binding of 1 5I-labe]ed lipoprotein to receptor-positive cells. The study provided several findings of special importance relative to the biologic impact of PAHs--or at least BaP as a model compound--on tissues in vivo. Clearly, although LDLs carry substantial amounts of PAH, the presence of LDL receptors on cel Is is not necessary for tissue uptake. The fact that PAH bound to LDL was incorporated into cells more slowly than PAH in a delipidated serum or serum-free medium raises questions about the biologic significance of experimental models in which increased incorporation of BaP from particles into lipid vesicles has been demonstrated. The data rom these experiments also indicate that cells that may be directly exposed to a PAH (i.e., tracheobronchial, intestinal, and cutaneous cells) before the compound reaches the bloodstream may accumulate PAH in much higher concentrations than cells exposed to the PAH bound to lipoproteins, inasmuch as the latter significantly slowed as well as limited the 6-12 . -

OCR for page 273
cellular uptake of BaP. Finally, the report indicated that BaP previously incorporated into WI-38 cells could be substantially removed (by 55-797) in a 120-min posttreatment study period by 10% delipida ted serum or LDL-containing medium. This finding implies a potential for considerable PAN redistribution and a requirement for a not insignificant period for progression of the hydrocarbon from the plasma membrane to the endoplasmic reticulum, where metabolism takes place. The ability of human monocytes to oxidize BaP and the induction of this enzyme activity by bt~z9agth:gcene have been demonstrated by several investigators. 6, , ,1 Lake and colleagues96 re-examined this problem with the goal of developing a practical assay for measuring whole-cell metabolism of BaP under highly standardized conditions, eliminating--among other problems--the need for a large volume of blood (50 ml) in the fluorometric assay developed earlier for AHH activitity in this cell type. By measuring whole-cell generation if water-soluble BaP metabolites over a 3-d culture period, using H-labeled substrate and closely controlling other character- istics, they provided a useful alternative cell system to that using mitogen-stimulated lymphocytes for characterizing BaP oxidation activity in humans. Because of the advantage gained by the much greater inducibility of AHH activity (-up to 40-fold) in cultured monocytes, compared with mitogen-stimulated 1 ~hocytes (about 5-fold), the monocyte system was used by Okuda et al. to study the contribution of genetic factors to the control of individual variation in AHH inducibility. Ten sets of monozygotic tissues were assayed two to four times and 17 sets of dizygotic tissues one to three times for basal and induced monocyte AHH activity. The results indicated that 55-70% of the individual variation in AHH inducibility of monocytes was genetically determined. Variation in AHH inducibility within subjects in repeat assays was wide and approached the magnitude of the variation between subjects. Thus, a single AHH assay is an imprecise biochemical characterization of a subject. Alternatively expressed, the method then available (late 1977) made it impractical to characterize a population with genetically distinct differences in AHH inducibility. The large intrasubject variation in AHH inducibility of monocytes also indicated that, in addition to the clear genetic influences on this process, unknown environmental or technical factors expressed themselves in the test procedure. An abundant literature exists related to the monooxygenase activity of lymphocytes; the inducibility of this activity by mitogens, which have the property of stimulating lymphocyte transformation, during which a number of metabolic activities are concurrently greatly increased; and the use of mitogen-stimulated lymphocytes to study the genetic control of AHH in man and its relation to the occurrence of Some human cancers--notably those of the l~lg,2 Kouri and colleagues have1 revi2ewed key aspects of this subject; ' McLemore et al. 19-12 have also provided a detailed analysis of the genetics of AHH and its purported relation to human cancer. Only a brief summary of these findings can be included here. 6-13

OCR for page 273
The identification of AHH activity in lymphocytes in 197229,192 and its increase during lymphocyte blastogenesis led quickly to clinical studies, the earliest being that of Kellermann et al. ,87 in which this induced enzyme activity was measured in cultured lymphocytes of normal controls, non-lung-tumor controls, and lung-cancer patients. In a preceding study in the same year, this grouped had examined the genetic variation in AHH activity in lymphocytes of 353 normal subjects and had categorized the population into three groups--low, inter- mediate, and high responders with respect to AHH inducibility; the population frequencies were about 50t, /~0X, and 10%, respectively. The conclusion was reached that the enzyme activity was controlled by two alleles at a single gene locus and that the high and low responders were homozygous and the intermediate group ~7eterozygous for those alleles. In the initial lung-cancer study, there was a virtual absence of cases in the low-inducibility population, and all but two cases were in the intermediate- and high-inducibility categories. All the lung-cancer cases were in heavy smokers; of the 50 subjects, 48 had an average consumption of two packs of cigarettes per day. When the two control groups (normal subjects and a non-lung-cancer tumor group) and the lung-cancer group were compared for risk of lung cancer, those with intermediate and high inducibility (48 of the 50 lung-cancer cases) had risks for lung cancer 16 and 36 times, respectively, the risk in the low-inducibility group. This study prompted considerable controversy over the next few years, during which the findings of Kellermann and associates were cast in doubt. 85 strong correlation (r = 0. 923) was also found by Kellermann et al. between the plasma elimination rate of antipyrine and the rate of BaP metabolism in human lymphocytes from a "carefully selected homogeneous" population, compared with the much lower correlation (r = 0.425) found in a "heterogeneous" population. The authors interpreted their findings as supporting the existence of common oxidative systems or common genetic control of the systems for antipyrine and BaP oxidation. Atlas et al. confirmed that plasma antipyrine half-life is correlated to some extent with AHH inducibility (r = 0.84), although no intrasubject correlations were found between AHH inducibility and the oxidation of other drug substrates, such as phenylbutezone and bishydroxycoumarin. Most importantly, this group, 7 while affirming a significant heritable determinant of AHH inducibility in human lymphocytes, failed to confirm the monogenic model and trimodal distri- bution of AHH indu86bility in the general population, proposed by Kellermann _ al.; rather, the population distributions for AHH inducibility (and for plasma antipyrine half-life) were consistent with polygenic control of both traits in man. In other studies in which the relation of AHH indu-cibility t~ th~ occurrence of lung cancer was re-examined by Paigen et al., 331 4 low AHH activity was found in half the tumor patients studied, in contrast with the earlier findings of Kellermann et al.,87 and no characteristic alterations in this enzyme activity were found in the progeny of these patients. A con- siderable number of technical problems related to the lymphocyte-AHH assay may confound the results obtained in studies of this enzyme 6-14 . . .

OCR for page 273
activity and its relation to human cancer, as noted by Kouri et al.91 However, recent methodologic advances made by this group, particularly the use of cryopreserved lymphocytes and close control of a number of assay variables, have added an important degree of precision to the assay. Chrysene, one of several PAR derivatives (benzanthracene is another), has been shown by Snodgrass et al.164 to induce AHH activity in cultured human lymphocytes (from normal subjects) with BaP as substrate. The individual variation in the monooxygenase activity observed with other inducers was also seen with chrysene. The comparative metabolism of BaP in human lymphocytes and human liver microsomes has been studied by Selkirk et al.,l6u who examined the nature of the metabolites formed by each cellular system. The patterns of metabolizes formed in both cell systems had characteristics quite similar to each other, with some exceptions--for example, among the derivatives formed in a 30-min incubation, all three dihydrodiols produced by liver were absent in the lymphocyte incubation mixture. In a~24-h incubation of lymphocytes, however, all three dihydrodiols formed by liver microsomes were also formed by the blood cells, and new metabolite peaks were observed, presumably reflecting more extensive biotransformation of already formed metabolizes in the reaction mixture. The authors concluded that, although the ratios of some metabolites may differ and although lymphocytes form several more derivatives than does liver, many identical metabolizes are produced in these two human cell types. Schonwald et al.l58 studied the effect of BaP on sister chromatic exchange in mitogen-stimulated lymphocytes of 11 normal subjects and 18 patients with lung cancer. Patients and controls differed neither with respect to the spontaneous rate of sister chromatic exchange nor in their responses to the hydrocarbon, although it did double the number of exchanges in both population groups. Barfknecht et al.l5 studied the ability of dichloromethane extracts of automobile diesel soot at high concentrations (100 mg/m3) to induce trifluorothymidine-resistant mutants in human lymphocytes incubated in the presence of rat-liver postmitochondrial supernatant. A significant induction of such mutants was observed. Anthracene, phenanthrene, and their alkylated derivatives accounted for one-fourth of the observed biologic activity. Among eight related compounds, there was general agreement between responses in lymphoblasts and in bacterial test systems. Phenanthrene was an exception, in that it was positive in the human-lymphoblast test system, but negative in bacteria at a concentration 60 times higher. The data in this report indicate that methyl substitution at some sites of anthracene and phenanthrene greatly increases their mutagenicity in both S. typhimurium and human lymphoblasts. A similar effect for chrysene has been observed. Methylations at the 1 and 3 positions of phenanthrene and the 2 and 9 positions of anthracene result in PAHs that are particularly mutagenic 6-15 At;

OCR for page 273
in the human and bacterial test systems used. Methylations at other positions had the capability of eliminating the mutagenic activity of the PAH derivative. No correlation between the results of the mutagenesis studies with the soot-derived PAHs and the reported capacity of the compounds studied to elicit neoplastic or carcinogenic responses in test animals could be made. REPRODUCTION The title of this section refers collectively to studies related to the ability of some genital tissues (including the placenta) to metabolize or otherwise respond biochemically to Part. There is an abundant and detailed literature on transplacental and peri- natall8 carcinogenesis. These and related topics in reproduction were reviewed in a 1981 special issue of the Journal of Environmental Pathology_and Toxicology and are not summarized here. It is perhaps appropriate, however, to refer to the report by Sir Percival Pott in 1775, 42 in which there was first described an increased incidence of scrotal cancer in chimney sweeps exposed to soot and to note that almost 150 yr elapsed before Yamagiwa and Ichikawal4 demonstrated that the repetitive application of crude coal tar to the rabbit ear produced skin cancer and that the identification of specific carcinogenic coal-tar constituents, such as BaP, required the passage of additional decades.20~43' 8 Over this period, the question of why only scrotal cancers, and not other genital cancers or even other cancers in general, were found in-excess in chimney sweeps appears to - have remained unanswered. Grover _ al.66 investigated the metabolism--including the Specific identification of biotransformation products--of three H-labeled PAHs by nonneoplastic human mammary epithelial-cell aggregates maintained in culture. The lobuloalveolar units from which these aggregates are derived are thought to be the site of origin of many human mammary carcinomas; two of the PAHs studied, 7,12-DMBA and BaP, are known to be relatively potent mammary carcinogens in rats, whereas benz~aJanthracene is not a mammary carcinogen in rats. Tissues from eight patients were studied. The extent of metabolism of the PAHs is summarized in Table 6-4. There was considerable individual variation in PAH metabolism among the subjects studied, but the formation of water-soluble metabolites by the tissue samples accounted, in each instance, for a major portion of the total of each PAH metabolized. The extent of binding of each PAH to cellular DNA and proteins also varied considerably. Interestingly, the extent to which H-labeled metabolites of benz~aJanthracene--a noncarcinogen for mammary tissue in the rat--were bound seemed, from the limited data obtained, to be consistently lower than the binding displayed by the other two PAHs. The results of chromatographic characterization of PAM-DNA adducts formed suggested that, with Hap, the hydrocarbon was activated by the cultured cells through the formation of anti-BaP 7,8-diol-9, 10-oxide, a bay-region diol-epoxide that appears to be responsible for most of the nucleic acid adducts formed in several 6-16

OCR for page 273
other biologic systems. The situation was less clear with 7,12-DMBA, although a portion of the adducts formed with this PAH cochromatographed with adducts present in a DNA hydrolysate that had been treated with anti-7,12-DMBA 3,4-diol-1,2-oxide--a derivative that is also classified as ~ bay-region diol-epoxide. The authors interpreted their data with caution, considering all the factors known to bear on the development of mammary cancer; but the possibility of partial causal relationships among the PAHs, their metabolic transformations, and tumor stimulation is implicit in this work. Stampfer and colleagues168 did similar studies with BaP and cultured mammary epithelial cells and fibroblasts. They showed that the breast epithelial cells were 50-100 times more sensitive (growth inhibition) to BaP than the fibroblasts; that the epithelial cells formed adducts as early as 6 h after addition of the PAH to the cultures; and that the adducts between the 7R anti stereoisomer of BaP diol-epoxide and deoxyguanosine predominated at all times and, with two minor adducts that were consistently present, persisted in the epithelial cells for at least 72 h in a BaP-free medium. No adducts were detected in fibroblasts until 96 h after exposure to the PAH, at which time the type and extent of adduct formation were similar to those observed with epithelial cells. As with the report of Grover et al.,66 caution concerning the direct relation of these findings to the role of PAHs in mammary carcinogenesis is necessary. On this matter, Stampfer and co-workers168 stated, however, that "chemical carcinogens, particularly BaP, should not be minimized as possible factors in the initiation of breast cancer." Mass et al.113 studied 26 specimens of normal human endometrium _ _ ~ to determine the patterns of metabolism of [~H]BaP in short-term explant cultures. Three of the tissue samples were from postmenopausal women; of the remaining 23, it was possible to approximate the stage of the menstrual cycle at which the tissue was removed during surgery. Eight of the latter subjects were smokers. In summary, it was clear that normal human endometrium could enzymatically convert BaP to a wide variety of oxygenated derivatives that cochromatographed with dihydrodiols, quinones, and monohydroxy products of the PAH; sulfation was also identified. HPLC analysis of metabolites revealed marked individual variation in metabolite formation among the subjects studied; smoking did not account for this difference, but some evidence of hormonal influences on the patterns of PAH metabolism was adduced. In a study by Dorman et al.,52 BaP binding to DNA in human endometrial tissue was studied in samples obtained from 41 subjects and, again, a striking (70-fold) range in the observed specific activities of carcinogen binding to UNA was identified (see Figure 6-1~. Tissues obtained late in the proliferative phase or early in the secretory phase of the menstrual cycle had the highest mean specific activity of PAM-DNA binding (Table 6-5~. Binding was significantly reduced when tissue specimens from low-estrogen periods of the menstrual cycle were studied. The reason for this apparent association between estrogen content (actually, the estimated phase of the cycle) 6-17 ~.. . ..

OCR for page 273
and PAM-DNA binding is obscure, but clearly merits further study. Such study would have to deal with the important confounding factor of the broad range of individual variation in binding, which may mask systematic but small changes that can occur during a menstrual cycle, but which cannot now be detected. Namkung and Juchaul28 studied the oxidative biotransformation of BaP in preparations of human placental microsomes with HPLC. The investigations revealed that the use of substrate concentrations high enough to ensure zero-order reaction kinetics markedly inhibited the formation of dihydrodiols in the reaction mixtures. The relative quantities of dihydrodiols generated increased with decreasing substrate concentrations between 200 and 2.7 ~M. Addition of manganese or ferric ions to reaction mixtures altered the ratios of generated phenols to dihydrodiols. Identical results were obtained with C- and 3H-labeled BaP as substrate. The data suggested that considerable amounts of 7,8-dibydroxy-7,8-dibydro-BaP, a proximate mutagen-carcinogen, may be generated in vivo by placental tissues of women who smoke. The formation of PAH metabolite-nucleoside adducts when human tumor placental microsomes were incubated with i3H]BaP and salmon sperm DNA has been studied by Pelkonen and Saarni.L 9 There were significant differences between the PAH metabolite patterns and the nucleoside- metabolite complexes formed, compared with rat liver, for example. Specifically, in the human placenta microsomes, the absence of the nucleoside complex of 9-'nydroxy-4,5-oxide implied the inability of this tissue to form 4,5-oxides of BaP. Indirect evidence of epoxide hydratase activity in placental tissue was obtained. The extent of PAM-DNA binding in this tissue correlated significantly with both 7,8-diol metabolite formation and fluorometrically determined AHH activity. The question of whether the 7,8-dioi-9,10-epoxide of BaP is formed by the human placenta in vivo could not be answered unequi- vocally, but the authors' inferential conclusion is that it is probably formed in the human host. The interplay of possible genetic influences and clearly established regulatory influences of environmental factors on Herman placental AHH has been incisively discussed by the same group.l38 Cigarette-smoking has been shown by Conney and associates42~189~190 to be one of the most potent and consistent inducers of human placental AHH activity yet identified. In the initial report of the group, 189 the enzymatic hydroxylation of BaP could not be detected in nonsmokers in homogenates of placentas frozen immediately after birth and studied within 48 h. In contrast, the enzyme activity was present in all 11 placentas from women who smoked during gestation, although enzyme activity in this small group did not correlate with the number of cigarettes smoked. BaP administration to pregnant rats also was shown to induce AHH activity in the placenta. The effect was related to PAH dose. This study constituted the first demonstration that compounds in cigarette smoke could induce a carcinogen-metabolizing enzyme in human tissues. These studies were 6-18

OCR for page 273
extendedl90 to related enzymatic reactions in human placentas and to other types of pyrenes as probes for AHH-inducing activity in rat placenta (see Table 6-6~. Extremely active inducers included chrysene, 1,2-benzanthracene, pyrene, 3,4-benzofluorene, and a number of related compounds. 90 The wide variability in the induction of AHH activity in human placentas is exemplified by the data in Table 6-7--a range in activity of the enzyme in smokers approaching 1,000-fold (a nearly 2,000-fold range if smokers are compared with nonsmokers). The basis for this extreme range of responses to a chemical exposure (15-20 cigarettes/d for each subject) is not known. However, data presented by Harris et al.70a suggests that pulmonary alveolar macrophages can metabolize BaP to proximate and ultimate mutagens released into extra- cellular space. LUNG - The respiratory tract comprises an extremely disparate and complex set of tissues containing some 40 different cell types.166 As Devereux et al.47 have noted, whereas pulmonary cytochrome P-450 and the metabolism of xenobiotics have been studied with various preparations of lung tissue (microsomes, isolated perfused lung, cells obtained by pulmonary Savage, direct instillation of xenobiotics in various portions of the respiratory tract, etc.), little is known about the localization of the cytochrome P-450 monooxygenase components in the pulmonary system. This section deals exclusively with the metabolic properties of human respiratory tissues with respect to PAH metabolism, but the lack-of information just cited needs to be kept in mind. There are facets of the investigation of Devereux et al.47 in rabbits that probably bear significantly on problems of human pulmonary tissue biotransfo'=ations that depend on cytochrome P-450; these aspects include the observation that the alveolar macrophage that accumulates PAH has little or no measurable cytochrome P-450 or monooxygenase activity58~71~148 and that there is selective cellular distribution of cytochrome P-450 species. The ability of human bronchial epithelial cells to bind and presumably to activate such PAHs as 7,12-DMBA, 3-MC, BaP, and dibenz~ahianthracene was described by Harris and colleagues in 1974.7 Four tissue samples were studied (one control and three lung cancer) in explant cultures, and radiolabeled PAHs were used; radioactivity from all four compounds tested was found in both cytoplasm and nuclei and in all tissue samples studied (see Table 6-8~. The number of tissues examined precluded comparisons between normal and tumorous lung PAH metabolism adduct identification were carried out, ~ activity from the labeled PAHs was found tightly by CsC1 gradient. A more detailed study by this tissues obtained from an additional four patients pulmonary malignancy. and no studies of PAM-DNA althou~h, as noted, radio- bound to DNA isolated groupl95 used , three of whom had Explants of human bronchi also metabolized BaP and released deriv,iives that are mitogenic in the Chinese hamster V-79 cell line. The 7,8-diol of BaP was approximately 5 times more potent as a promutagen than the parent PAH; binding of the diol to DNA was 5-20 6-19

OCR for page 273
times greater than that found with BaP. When 13 samples of bronchial cells were studied with cloned Chinese hamster V-79-4A cells, a positive correlation between ONA-PAH binding (in the cultured bronchial cells) and induction of Or (ouabain-resistant) mutants was found, but no correlation between this mutation frequency and AHH activity was identified. This may be attributable, as the authors noted, to the difficulty in correlating AHH activity with the consequences of the multistep pathway of metabolic activation for BaP. The individual variation in mutation frequency was 9-fold, and the variation in binding of PAH to UNA 5-fold. This important investigation pointed the way toward study of the metabolic activation of chemical carcinogens into promutagens and mutagens directly in differentiated epithelial cells derived from human tissues; and the human tissue-mediated mutagen assay opened the possibility of testing the hypothesis that people differ in mutagenic and oncogenic susceptibility to environmental chemicals, depending on individual capacity to activate and deactivate chemical procarcinogens. Autrup et al.12 compared the metabolism of BaP by cultured tracheobronchial tissues from humans and four other species (mice, hamsters, rats, and cows). They provided evidence that the metabolism of BaP is qualitatively similar in tracheobronchial tissues from humans and from animal species in which PAHs have been shown experimentally to be carcinogenic. A similar study limited to a comparison of human lung microsomal fractions and rat microsomes was carried out by Prough et al.144 The results indicated that human microsomes form a higher percentage of dihydrodiol products from BaP than do rat microsomes. The wide variation of PAH metabolize profiles formed by the 15 samples of human lung studied may be due in part to differences in clinical diagnosis when the samples were obtained. Bronchial tissues cultured in a chemically defined medium were exposed to radiolabeled BaP or its metabolizes, and their binding to TUNA was measured. Radiolabeled metabolizes were prepared by incubating the parent PAH with rat liver microsomes and then purifying and identifying with silica gel and HPLC. The binding data showed that (-~-trans-7,8-diol bound to bronchial mucosal DNA to a considerably greater degree (S- to 23-fold) than did BaP; binding was also much greater (25- to 80-fold) than with the (-~-trans-9,10-diol. The trans-7,8-diol constituted 3-6% of the total identified metabolizes when human bronchi were exposed to BaP. Diol-epoxides were formed from (-~-trans-7,8-diol in two of the bronchial explants, and strong evidence was provided that the major tumor bronchial mucosal DNA-binding BaP metabolize is in fact derived from (-~-trans-7,8-diol.195 The specific adducts formed between DNA and the metabolic intermediates of BaP were not isolated, but the author concluded that the predominant bound metabolite is a single enantiomer of diol-epoxide I derived as indicated above. In an extension of their earlier work, Harris and colleagues69 examined the metabolism of BaP and 7,12-DMBA in explants of human bronchus and made a metabolic comparison with human pancreatic duct explants. As in the prior study, both normal and malignant human bronchi (37 subjects) metabolized BaP actively and in generally similar 6-20 hi . .

OCR for page 273
fashion, except for a higher percentage of organic-solvent-extractable metabolites formed by bronchi from noncancer patients. In addition, prior exposure of the bronchial explants to benz~aJanthracene altered the qualitative features of the metabolite profile of BaP, as analyzed by HPLC. Benz~aJanthracene specifically increased the binding of BaP to cellular DNA and the activity of AHH. Among a group of 28 of the patients' tissues studied, 7,12-DMBA was bound to DNA more often (26 of 28) than BaP. In the comparison with pancreatic duct explants, 7,12-DMBA-ONA binding was consistently lower in the latter tissue than in the bronchial explants. Cohen et al.34 showed, with cultured human bronchial epithelium, that BaP was converted promptly to metabolites that cochromatographed with 9,10-dihydro-9,10-dihydroxy-BaP and 7,8-dihydro-7,8-dihydroxy- BaP. Similar results were obtained with human lung cultures, except that a major metabolite, benzota~pyrene-3-yi hydrogen sulfate, was identified. The biologic activity of this sulfate ester of 3-hydroxy- BaP is of interest, because, owing to its physicochemical properties, it could be extremely persistent in man. - Covalent adducts between DNA and BaP in treated cultured explants of peripheral human lung tissue and in the continuous human alveolar tumor cell line were identified by Shinohara and Cerutti.161 From the chromatographic analysis of digests of ONA extracted from these tissues, it was concluded that both the lung specimens and the human alveolar tumor (A549) cells metabolized BaP to diastereomeric 7,8-dihydroxy-9,10-epoxytetrahydro-BaP intermediates that mostly reacted with the exocyclic amino groups of deoxyguanosine to form N'-~10-t 76 , 8a , 9a- and 96-trihydroxy-7,8,9,10-tetrahydro- benzotaipyrene~yl~deoxyguanosine (dGua-BaP I and II). Although comparable amounts of dGua-BaP I and II were formed in A549 cells, dGua-BaP I was the predominant adduct in the DNA of lung specimens from six different donors. The wide range of metabolic capacities for PAHs exhibited by other buman tissues studied also extends to lung tissue, as shown by Cohen et _ .35 They observed a 44-fold variation in the ability of short-term organ cultures of peripheral lung tissues from human cancer patients to metabolize BaP to organic-solvent-soluble derivatives. The total amounts metabolized ranged from 1: to 96.2t in a 24-h culture period. The authors concluded that, although caution must be exercised in measuring metabolic activities of human tissues derived from diseased patients, the use of short-term organ explant cultures mimics the in vivo metabolic disposition of PAH better than the use of lymphocyte AHH activity would. A solution to the practical problem of obtaining lung tissue from large populations to study the validity of this conclusion is not apparent. Kahng et al.78 concluded from a study of 11 immediately autopsied subjects that bronchial tissue exposed to benz~aJanthracene produced induction responses of AHH that correlated with induced AHH activity in monocytes from the same subjects. A reconfirmation of the wide range 6-21 ~; ~... .

OCR for page 273
194. Yamagiwa, K., and K. Ichikawa 195. 196. 197. . Experimental study of the pathogenesis of carcinoma. J. Cancer Res. 3:1-29, 1918. Yang, S. K., H. V. Gelboin, B. F. Trump, H. Autrup, and C. C. Harris. Metabolic activation of benzo~a~pyrene and binding to ONA in cultured human bronchus. Cancer Res. 37:1210-1215, 1977. Yoshida, D., H. Nishigata, and T. Matsumoto. Pyrolytic yields of 2-amino-9H-pyredot2,3-b~indole and 3-amino-1-methyl-5H pyridot4,3-b] indole as mutagens from proteins. Agr. Biol. Chem. 43:1769-1770, 1979. Yoshinaga, T., S. Sassa, and A. Kappas. A comparative heme degradation by NADPH-cytochrome c reductase alone study of _ and by the complete heme oxygenase system: Distinctive aspects of heme degradation by NADPH-cytochrome c reductase. J. Biol. Chem. 257:7794-7802, 1982. 198. Yoshinaga, T., S. Sassa, and A. Kappas. Purification and properties of bovine spleen heme oxygenate. Amino acid composi- tion and sites of action of inhibitors of heme oxidation. J. Biol. Chem. 257:7778-7785, 1982. 199. Yoshinaga, T., S. Sassa, and A. Kappas. The occurrence of molecular interactions among NADPH-cytochrome c reductase, heme oxygenate, and biliverdin reductase in heme degradation. J. Bio1. Chem. 257:7786-7793, 1982. 200. Yoshinaga, T., S. Sassa, and A. Kappas. The oxidative degrada- tion of heme c by the microsomal heme oxygenase system. J. Biol. Chem. 257:7803-7807, 1982. - lung in railroad workers. J.~.M.A. 171:2039-2043, 1959. 201. Zedeck, M. S. Polycyclic aromatic hydrocarbons : A review. J. Environ. Pathol. Toxicol. 3: 537-567, 1980. 6-76

OCR for page 273
7 SOME FACTORS THAT AFFECT SUSCEPTIBILITY OF HUMANS TO POLYCYCLIC AROMATIC HYDROCARBONS The interaction of chemical pollutants, including the PAHs, and mammalian cells can result in a variety of problems, including toxicity, mutagenesis, carcinogenesis, and teratogenesis. This interaction of chemicals with somatic cells probably results in such end points as cancer, and the interaction of chemicals with germ cells probably results in a variety of hereditary disorders. Many genetic disorders result in a predilection to the development of cancer. The cancer burden in the male population in the United States, although speculative, is distributed approximately as follows: 40% from tobacco-smoking, 10-20% from all diet-related causes, 5t from occupational exposures, it from single-gene inheritance, and 35% from other causes, which may include unknown genetic predisposition and environmental effects.59 The birth-defects burden in the United States is distributed approximately as follows: 5-10% from known teratogens, such as viruses, chemicals, and radiation; 25% from genetic anomalies; and 60-65% from unknown mixtures of genetic predisposi- tion and environmental effects.59 Although monogenic disorders (includ- ing dominants), X-linked recessive disorders, and chromosomal abnor- malities account for only about 5% of the human disease burden, the impact of heterozygous recessively inherited abnormalities similar to the mono- genic disorders is very ill-identified, but could outweigh all other contributions.115 The heterozygous recessively inherited disorders may be the major reason why cancer incidences are not uniformly distributed.95397 In fact, of the millions of people exposed to such environmental chemicals as diethylstilbestrol, estrogen oral contraceptives, vinyl chloride, and cigarette smoke, only a very small proportion develop or express the cancer thought to be associated with these exposures. It is likely that genetic variability within the human population accounts in part for the distribution pattern. As depicted in Figure 7-1, cancer sensitivity can be viewed as a function of inborn susceptibility. Where this inborn or genetic susceptibility is low, cancer expression is low. Where this susceptibility is high (e.g., in single-gene defects), cancer expression is high. The major question is whether the combination of chemical exposure and genetic susceptibility can change significantly the numbers of persons who develop cancer. PAHs are ubiquitous chemicals capable of producing a broad spectrum of biologic responses. Some can cause cancer in a variety of tissues, including lung, liver, kidney, colon, skin, and bladder. In humans, epidemiologic evidence has demonstrated that the incidences of cancers of stomach, nasal cavity and sinuses, lung, and to a lesser extent rectum, testis, skin (e.g., melanoma), brain, liver, pancreas, and hemopoietic 7-1

OCR for page 273
tissue i.e., leukemia) are correlated with area" captaining high con- centrations of industrial pollutants. 10 ,46' 102,1 2, For many persons, the amount of these agents in the environment may be the rate- deterTnining factor for cancer susceptibility. Thus, the primary need would be to identify and measure the amount o f exposure to the environ- mental pollutants. The advent of a variety of in vitro and in vivo bioassays promises the development of me thods for identi Eying chemicals that are potential carcinogens. In animal-model systems, susceptibility to chemically induced cancers is usually dose-related. However, route, duration, and frequency of administration and such genetic factors as species, sex, and strain all tend to modify the relationship. In humans, mixtures containing PAHs can certainly cause cancer, hut inadequacies in the information on age and trauma but especially on duration, frequency, and intensity of exposure and on the size and characteristics of the exposed population make quantitative estimation of dose-response relationships and the concept of thresholds difficult to interpret. EFFECT OF GENETIC D IFFERENCE The hypothetical stages in carcinogenesis are depicted in Figure 7-2. PAHs probably can show biologic effects at any of these stages. Thus, answers are needed to the following questions: Which stages can PAHs modify in humans? Are there naturally occurring variations in the expression of some of these steps in humans? Can a genetic basis be identified for the regulation of these naturally occurring differences? If so, can the differences result from the action of a single gene system? Can a relationship be shown between the express ion in the gene locus and PAH-mediated effects? PAH-induced effects in humans could depend on exposure, uptake, and distribution of the chemicals; their metabolic activation and inactiva- tion; DNA-repair capacity; "promoters"; and the extent of immunocompe- tence. Each of these is discussed below. UPTAKE AND DISTRIBUTION OF PAHs IN TISSUES The distribution of PAHs in tissues or cells depends on the route of exposure. According to the results of Rees et al.,1 the distribution of benzota~pyrene (BaP) in tissue other than at the site of absorption (i.e., intestine) depends on two phases: accumulation of the BaP on the tissue and passive diffusion through the tissue. These two phases underlie these authors' views about the apparent exponential nature of the accumulation of BaP as a function of dose. The exponential increase could be very important, but it must be pointed out that humans are rarely exposed to BaP at concentrations greater than 200 AM (i.e., 50 legal) under "normal" circumstances. Concentrations of a variety of PAHs (e.g., pyrene, anthracene, and BaP) in human tissues average about 1,100 parts per trillion (ppt) in fat tissue and 380 ppt in liver.ll9 BaP can vary 7-2

OCR for page 273
from 0.3 No 15,000 parts per billion (ppb) in bronchial-carcinoma tissue.18 Most of the subjects in the study of Tomingas et al.184 were cigarette-smokers, but no obvious correlation between BaP concentration and extent of smoking was seen. The PAHs observed in addition to BaP included fluoranthene, benzotbifluoranthene, and perylene. Reasons for differences in tissue distribution are not known, but, inasmuch as most of these chemicals are inducers and substrates for microsomal enzymes, tissue variation in cytosol and nuclear receptors could be important. In rodents, the induction of the microsomal mono- oxygenase system by some PAHs depends on the presence of particular cytosol receptor proteins.56314231 3 These receptor proteins are not evenly distributed in all tissues, but are highest in thymus and lung, lower in liver and kidney, lower yet in testes, brain, and skeletal muscle, and not detectable in pancreas, adrenal, or prostate.l5 Most importantly, receptor proteins are found in high concentrations in strains of animals or cultured cells in which PAHs induce the enzyme aryl hydrocarbon hydroxylase (AHH) and are nondetectable in those in which AHH is nonrespons ive . 56 , 142 This correlation also extends to humans, in whom the concentration of a BaP-binding plasma component is correlated with the capacity of lymphocytes to be induced for AHH activity in culture.104 A cytoplasmic receptor for BaP, which did not cross-react with 7,12-dimethy31benz~ajanthracene, has also been reported for human cells in culture. 1 The presence of some of these receptors is under specific genetic control in animal models,l22~143, so uptake and distribution, at least in particular persons, could be under a form of genetic control. METABOLISM PAHs are metabolized in a variety of ways, with the microsomal mono- oxygenases (e.g., AHH) probably most important. Steady-state activities of these enzymes vary Animals and are linked to susceptibility to some PAH-mediated cancers.7 ~ 1 In humans, the data are much less clear. Table 7-1 summarizes the studies that suggest a correlation between high AHH inducibility (and usually high induced-AHH activity) and cancer susceptibility, and Table 7-2 summarizes the studies that suggest the converse. Reasons for the contradictory results probably lie in methodologic variations, such as the use of different cell types, different assays, and different assay conditions. The most easily accessible and therefore commonly used Lyman tissue is She peripheral blood lymphocytes. Nutri- tional state,1 drug intake,8 age,35 and disease state74 influence the capacity of the lymphocytes to respond to mitogen. These influences have not been assessed in determining their relationship to the AHH activity observed in cultured lymphocytes. Variations in AHH activ- ity in lymphocytes have been observed to occur seasonally in some geo- hi 1 ations 128$129,154 but whether they result from in vivo or in vitro factors is not known. 7-3 ~ . it. . ..

OCR for page 273
A variety of in vitro conditions are known to influence AHH activity. The initial concentration of lymphocytes affects the time course and amount of control and induced AHH activity.7 The type and lot of serum Supplement infuence the control and induced AHH activity.5 ,7 In fact, some lots of feta)-galf serum are capable of causing mitogen activation of lymphocytes. O The numbers of cultured T cells may affect the AHH activity observed.67 In studies using cultured human tissue, two important aspects are the question of the variable degree of AHH activity in different cell types202 and the question of large variations between and within individuals in both AHH activity and microsome-mediated BaP-DNA adduct formation. ' 3 Blood monocytes and pulmonary alveolar macrophages are examples of other human cell types whose AHH activity is Correlated with that in lymphocytes or cultured human tissue, but there are problems of accessibility with each of these cell types. If the cell samples are cultured and assayed on the same days, the v5r~2t~30 seems to be ac3eptable.5379 Culturing lympho- cytes ~ ~ or monocytesl2 from fraternal or identical twins at the same time has shown that AHH activity is under a degree of genetic control, and the numbers of genes in question are probably small. Thus, the genetic component most likely results in a unimodal frequency distribution that is skewed in the populations of individuals toward those with higher AHH activity,7831ll rather than the trimodal distribution originally reported.72 To circumvent many of the in vitro problems, the use of cryopre- served tissue may be an alternative, in that lymphocytes can be cryo- preserved before mitogen activation and still have ted capacity to be mitogen-activated and then assayed for AHH activity. The relative AHH activities among the lymphocyte samples from different individuals are similar, whether the assays are conducted on freshly cultured lymphocytes or after cryopreservation.82 Cryopreservation allows the culture and assay at the same time of cells from different organisms collected in diverse geographic locations and over extended periods. The use of cryopreserved lymphocytes 3 control of some basic culture variables--such as initial lymphocyte concentration (1.0 x 106 cells/ml) and lot and type of serum supplement (e.g., human AB serum)--and assaying AHH activity at two times to ensure detection of peak activity can yield the data presented in Figure 7-3. Data were taken on a group of 51 per- sons who were on hospital diets for at least 2 d before phlebotomy, who were not on any medication, and who were eventually followed for complete clinical diagnosis. Viability of cells was measured by assay for the NADH-dependent cytochrome by reductase (using cytochrome c as a sub- strate) activity (Cyt c). Carcinogen-metabo~i~gg activity is presented in terms of units of AHH per unit of Cyt c.7 ~ The degree of mitogen activation was also measured. Data analyses showed that: ~ Cryopreserved lymphocytes from over 957 of the normal and cancer patients were mitogen-activated. 7-4

OCR for page 273
Lymphocytes from lung-cancer patients were mitogen-activated as efficiently as lymphocytes from noncancer patients (actually better; = 0.001). O The 14 highest AHH activities were found in patients with lung cancer, with the mean in the 21 lung-cancer patients (0.89 unit AHH/unit Cyt _) being significantly higher than that in the 30 non-lung-cancer patients (0.47 unit AHH/unit Cyt c). The higher AHH activities were not directly related to higher degrees of blastogenesis and were not related to cigarette-smoking history, tumor type, tumor location, or family history of cancer. Whether high AHH activity is the cause or the result of lung cancer cannot yet be answered. In animal-model systems, some PAHs cause tumors of the lymphoreticu- lar system, and a genetic association for this activity at the Ah locus has been suggested.29~1l3 Although this is only presumptive, there may be a similar relationship in human leukemia patients who were recently shown to express lower AHH activity (as in animal-model systems);ll in other studies, the first-degree relatives of leukemia patients expressed normal AHH activity.90 The results of these studies are interesting and certainly need to be confirmed and extended. The extended studies should be multifaceted; that is, they should simultaneously measure more than one enzymatic end point. Perhaps an appropriate group of assays would include an assay for AHH, as described in the literature; an assay for all B3aP ~abolites via HPLC; an assay for particulate P-450s via immunoassays;1 1, and an assay for mRNA expression of the P-450 genes with cloned UNA fragments containing the P-450 genes.116 Human tissues should be used where possible. There is probably a degree of genetic control of AHH activity in the human population, and this enzyme may play a role in determining susceptibility to PAH-mediated cancer and other diseases. DNA BINDING, DAMAGE, AND REPAIR Many PAHs are converted by the microsomal monooxygenases to forms that bind covalently to a variety of cellular macromolecules, including nucleic acids (see Chapter 5 and Phillips and Simsl40~. Evidence of the importance of DNA binding is exemplified by the observation that varia- tions in DNA-repair capacity seem to play a major role in determining the toxic, mutagenic, and3 t :00s f49ming ac t ivit ie s o f many chemical care i 0 In animal-model systems, the amount of PAH metabolism is determined by the activity of the microsomal monooxygenases, and variations in these enzymes result in concomitant changes in the binding of chemicals to DNA.112 In cultured human tissue, hydrocarbon-DNA binding also occurs as the result of microsomal monooxygenase-mediated metabolism,6~55 and variations in metabolic activity are asso5:iated with concomitant varia- tions in binding of hydrocarbons to DNA. The major DNA adduct often results from the interaction of specific metabolites of PAHs (diol- epoxides) and the N7 of deoxyguanosine.l24,l35 Other products a 7-5 ~; ~` ..

OCR for page 273
found, including interactions with the N4 deoxyadenosine,177 the back- bone phospt~tl,70f ONA, 1 and the exocyclic amino group of deoxy- adenosine. ~ The latter may be important, because its formation from various PAH-like chemicals closely parallels their carcinogenic potencies on mouse skin.27 No apprec~,b{~ syggificity of binding with respect to base sequence is apparent, ~ 8, but binding may be influenced by chromatic structure, with a greater extent of binding associated wi th internuc leosomal regions.6 ,76 ' A potentially important anomaly is that, although in vitro metabolism of BaP to forms that bind to DNA parallels the AHH activity of the micro- somal preparations and the Genetic background of mice used to generate these microsomal samples,13 the in viva results from strains of mice that differ widely in ASH activity so that there is very little strain variation in BaP-DNA binding. ~ Probably more crucial to carcinogenicity is the geometry of the binding in relation to later excision repair by endonucleases.48 The binding of different residues and different chemical groups within residues dramatically affects excisibility. These chemical-DNA adducts are either repaired, not repaired, or misrepaired (see Figures 7-2 and 7-4). The fate of these adducts determines whether a cell remains normal, mutates, or dies. Repair capacity can be separated into two major types--excision repair and postreplication repair.l49 Excision repair is the in situ removal and replacement of chemically modified ONA so that the original DNA sequence is re-established. For a variety of reasons, excision- repair systems usually do not remove all the modified bases; so the ONA very often replicates, even though some unexcised damage may be present. This replicated UNA usually has gaps in the newly synthesized strand opposite the DNA adduct. The gaps are fillip in by postreplication repair--also termed "recombination repair." O Figure 7-3 depicts how these two processes of repair contribute to the cells' survival of the damaging effects of chemicals like PAHs. A combination of both methods is involved in the repair of hydrocarbon-bound DNA.1OO A large number of both constitutive and inducible enzymes are involved in this ONA-repair process.47 The exact role of these enzymes is not known, but it seems that rather small changes in any of the enzyme activities can have great effects on the repair process and eventual bio- logic expression of the DNA adducts. Moreover, it has been recently shown in prokaryotes that the DNA adduct itself is not likely to be mutagenic, but rather that the mutagenic event is induced by the action of the DNA-repair enzymes themselves.] Natural variations in DNA-repair capacity occur in humans. These variations are exemplified by the existence of genetic diseases that are associated with defects in DNA repair. Table 7-3 presents a list of such diseases, their modes of inheritance, the specific tumors associated with them, and their proposed DNA-repair defect. These genetic diseases are 7-6 Am .. .. ..

OCR for page 273
associated with a high incidence of malignancy, compared with the incidence in the general population, and often a specific malignancy is involved (see Table 7-3~. The incidences for persons who are genetically homozygous for xeroderma pigmentosum, ataxia telangiectasia, and Fanconi's anemia are about 10-5, and for those who are heterozygous, about 10-2. Those who are heterozygous for ataxia telangiectasia and are less than 45 yr old have a fivefold increase in the risk of cancer,180 and those heterozygous for Fanconi's anemia may account for 5% of all leukemia deaths (approximately a fivefold increase in susceptibility).179 Because these people are deficient in the ability to repair radiation- induced DNA damage and chemical-induced DNA damage, 169 it has been suggested that alteration in DNA-re~air capacity may put them at greater risk of chemically induced cancers. 68 It must be pointed out that many of these diseases, especially ataxia telangiectasia, are also associated with abnormalities of the immune system. Thus, genetic disease may result in higher risks of cancer via deficiencies in DNA-repair capacity or immunocompetence. Among the normal population of humans, there are probably subtle variations in DNA-repair capacity, but whether these variations are genetically controlled or are related to cancer risk remains to be determined. PROMOTION AND COCARCINOGENESIS Many s tudies have shown that a number of modifying fac tars can increase the effect of low-dose or low-potency carcinogens that by themselves would be insufficient to induce malignancies.l09~200 Many PAHs are complete carcinogens; that is, they have both initiating and promoting activities. Others--such as pyrene, benzote~pyrene, fluoranthene, and benzotghi~perylene--are weak complete carcinogens and weak cocarcinogens.l9l'192 It is difficult to determine what role PAHs might have in tumor promotion in humans, because there are no good methods for measuring this activity in the human population. Such end points as induction of ornithine decarboxylase activity,13 phosDholipid synthesis,l59~178 inflammation,~73 protease activit ,~87 cellular proliferation,57 decrease in differentiated states,;6~201 and formation of "dark cells''l48~172 are manifestations of man romoters, and many PAHs can induce at least some of these changes.l95~500 But no single end point correlates with the promoting activity of all the different chemicals that have promoting activity. In animal systems, there seems to be a genetic basis for promota- bility, in that different strains of mice express different suscepti- bility to promotion during the standard two-stage carcinogenesis assay. Such strains as CD-1 and BALB/c are relatively resistant, whereas the specifically derived SENCAR strain (i.e., sensitive to carcinogenesis) is very sensitive to promotion of skin cancer~78~58 The molecular basis of this difference has not been defined, but recent informal ion suggests that the skin itself has the sensitivity, inasmuch as skin from SENCAR mice remains sensitive to promotion even after grafting to BALB/c mice.203 7-7

OCR for page 273
No genetic variation in promotability in humans has been described. However, the fact that pyrenes may have promoting and cocarcinogenic activity, the possibility that such activity plays a major role in cancer formation in humans, and the absence of effective end points in the human population all suggest that much more work is necessary before the role of PAHs in promotion can be understood. I~4UNOCOMPETENCE _ Substantial interest has centered on the role of the immmune system in preventing the expression of malignancy by recognition and destruc- tion of newly formed malignant cells. The concept of "immunosurveil- lance," however, has not been well supported, and, in fact, "stimula- tion" of malignant cells may even occur.75 Immunodeficient persons do have a greatly increased risk of develop- ing a malignancy of the lymphoreticular system.17~60~93~126 The exact mechanism responsible for the increase, however, is not clear. A number of genetic disorders in humans are associated with immuno- deficiencies. These disorders include ataxia telangiectasia, Wiskott- Aldrich syndrome, Bloom's syndrome, common variable immunodeficiency, selective IgA deficiency, Bruton's agammaglobulinemia, severe combined immunodeficiency, selective IgM deficiency, and immunodeficiency with normal or increased immunoglobulins.73 These immunodeficient genetic disorders are usually heterogeneously linked with a variety of other distinct underlying defects. For example, persons with ataxia telangiectasia and Bloom's syndrome have severely impaired DNA-repair capacities,169~195 and those with severe combined immunodeficiency also have adenosine deaminase deficiency.73 Therefore, it is difficult to determine the reasons for the increased cancer susceptibility of these persons. Epidemiologic evidence fails to support the idea that immunosurveillance mechanisms are generally involved in carcinogenesis 5 but does provide clues to immunologic processes that may predispose to particular neoplasms.38 In animal-model systems, PAHs can cause tumors of the lympho- reticular system, and association with the Ah locus has been suggested.29~113 In humans, exposure to some hydrocarbons, such as benzene, has been repeatedly associated with leukemia. Whether variations @ ~ Ln 1mmunocompetence occur naturally in the normal population and whether PAHs, as a group of environmental contaminants, pose a special risk to persons with such variations are not known. STAGE OF DEVELOPMENT . Some cell types undergo periods of heightened sensitivity to chemicals during their normal growth cycles. For example, in animal- model systems there are striking differences between geru-cell stages in

OCR for page 273
the chemical induction of dominant lethals, translocations, and specific- locus mutations. l4, 160 Moreover, the fetus is at greater risk than the mother, owing to high doses of environmental chemicals; the permeability of the blood-brain barrier is greater, and liver-enzyme conjugating function is poorer.61 The greater the lipid solubility of a chemical, the greater its placental transfer; and the placenta is readily permeable to chemicals with molecular weights less than 600. Most PAHs fit into these categories, and in animal-model systems such PAHs as BaP, 3-methylcholanthrene, and 7,12-dimethylbenz~aJanthracene cause oocyte and follicle destruction and embryo lethality and resorption and have a greater incidence of malformation and even cancer in surviving embryos.88~103~171~190 In humans, gross congenital abnormalities occur in some 2t of all infants and are the cause of about 15t of the deaths of infants less than a year old. Exposure to such agents as viruses, mercury, DUT, CO, and polybrominated biphenyls probably accounts for 5-10t of the birth defects; genetic abnormalities cause 25%; and the causes of the remainder are largely unknown.61 Interactions in the intrauterine environment between genetic predisposition and chemical and biologic factors are probably responsible for these birth defects. Although occupational exposure of human males52 and both paren-ts204 to PAHs was not associated with increased cancer incidences in the offspring, recent work has suggested that a combination of chemical exposures of both parents (especially the mother) resulted in higher incidences of brain tumors in the offspring. 137 Maternal cigarette-smoking is associated with decreased birthweight, increased perinatal morbidity and mortality, and other harmful effects on the newborn.l41 The PAHs in cigarette smoke may account for some of its biologic activity, inasmuch as a relationship has been shown between cigarette-smoking, induction of AHH activity in human placental tissue,ll4'l98 and a decrease in placental size;135 PAHs are the major class of AHH inducers found in cigarette smoke, 83 and thus it is important to note that BaP, which is in cigarette smoke, can cross the placental barrier.92 Because PAHs must be metabolized before they produce a biologic effect, the impact of PAHs on maternal and fetal tissues can be quite complex. Some examples of these complexities are differences in developmental patterns of specific enzymes, the relative importance of maternal and fetal metabolism, the role of metabolism in placental tissue, the relative importance of hepatic and extrahepatic metabolism, and sex differences in developmental patterns. The induced and control forms of AHH and acetanilide 4-hydroxylase are temporally regulated both before and after the birth of animals.68 The deactivation of conjugating enzymes (e.g., UBP-glucuronyltransferase, sulfotrans- ferase, and N-acetyltransferase) is also temporally regulated both before and after birth, but this regulation can be quite different from that of AHH.94 The relation between activation and inactivation can be influenced by the sex of animals.70 Shum et al.171 showed that both the fetal and maternal enzymes play an active role in determining the ultimate fetal toxicity of BaP. Using specific crosses between AHH-responsive and AHH-nonresponsive strains, these authors could show 7-9

OCR for page 273
that when the mother was nonresponsive the enzyme capacity of the fetal tissue determined the toxicity of BaP, but that when the mother was AHH-responsive there was no difference in fetal toxicity between nonresponsive and responsive fetuses. Mice seem to have AHH activity as early as about 7.5-8.5 d of gestation.40 This activity slightly increases before birth, but increases greatly in the first few days after birth94 and then slowly decreases as the mouse ages.68 It should be pointed out that in vivo exposure to BaP, in addition to inducing higher AHH activity in mouse fetal tissue, can suppress bumoral immunity in animals that survive and can cause about a 10-fold increase in the incidence of various tumors in surviving animals.l90 It seems likely that, in rodents (and perhaps in bumans), PAHs can be taken up and distributed through the placenta intact or in the form of metabolites, that the metabolites themselves can cause fetal toxicity or the delayed effects of immune suppression or cancer, and that intact PAHs can cause fetal enzyme induction, metabolism, and the sequelae mentioned earlier. MODIFYING FACTORS - A variety of environmental factors can mitigate or exacerbate the inherent sensitivity of mammalian tissues to PAHs. These factors are probably at least as important as some of the genetically controlled differences discussed earlier and tend to make genetic differences less distinct. Two factors known to modify PAH carcinogenesis, at least in animal-model systems, are the physical state of the PAH and the nutritional state of the exposed organism. PHYSICAL STATE OF PAH . The sources and the formation of PAHs in the environment are dis- cussed in Chapters 1-3. Most of them are found as mixtures and many are found in association with particles, such as cigarette-smoke particles, i74 fossil-fuel combustion products, 9 coal flyash, 77 and asbestos fibers.86~105 This association can be important, because PAHs in the presence of or adsorbed on particles are transported through membranes more e f f ic iently,75 are cleared from tissue more slowly, 25 and have a different tissue distribution--that determined by the particle size, rather than by the size of the free PAH.147 The increased uptake results in more efficient induction of AHH activity at low PAH concentrations.l5 Those exposed to particles containing PAHs are probably at greater risk of various cancers.166 Uptake, distribution, and metabolism of PAHs can be so altered by particles that those who normally would be unaffected by the PAHs may be adversely affected. NUTRITIONAL STATE OF HOST Nutritional status can substantially modify the toxicity of some environmental pollutants. For example, specific dietary 7-10 ~... . .