Aflatoxin BInduced Liver Cancer in Experimental Animals and Humans

In experimental hepatocarcinogenesis, aflatoxin I*! (AFB,) is a poor carcinogen in the mouse, but a powerful carcinogen in the rat, provided the rat has not been treated with inducers. The carcinogenic electrophile produced by AFB! metabolism is now known to be AFB, e;to-8,9-oxide. In humans, and to a lesser extent the rat, the isomeric endo-oxide is also formed but this does not appear to react significantly with DNA and may have little importance in carcinogenesis (Raney et al., 1992; Iyer et al., 1994).

The AFB,-oxides are potential substrates for GSTs. The low carcinogenicity of AFB, in the mouse, relative to the rat, is probably due to the 100-fold greater constitutive activity of its hepatic GSTs, which are two orders of magnitude greater than those in rat liver. The principal activity in mice is the alpha enzyme (mouse GST4-4), which has an activity of 300 nmolmin'mg 1 (Ramsdell and Eaton, 1990). Male rats exposed to appropriate inducers become much less susceptible to AFB,-induced carcinogenesis. Kensler and colleagues (1991) have shown that when olti-praz [5-(2-(pyrazinyl)-4-methyl-l ,2-dithiol-3-thione] is added to a diet containing levels of AFB, sufficient to produce 20% tumorigenesis DNA-bound AFB, and the development of preneoplastic nodules were greatly reduced and no tumors developed. Analysis of hepatic GSTs shows that administration of oltipraz and the parent compound 1,2-dithiole-3-thione induces all GSTs present in the normal male adult rat (particularly GST lb) and, in addition, brings about large inductions of GST7-7 and GST10-10, which are associated more with the fetal than the adult state (Fig. 6, Meyer et al., 1993b). However, the induction that is most important in AFB, anticarcinogenesis is that of GSTI0-10, which has activity toward AFB,-e;t0-oxide similar to that of the mouse GST4-4. GST10-10 has yet to be cloned but the sequence available shows strong structural homology to mouse GST4-4 (Hayes et al., 1991; Meyer et al., 1991b; Buetler and Eaton, 1992).

Hepatocarcinogenesis in man has been associated with the contamination of foodstuffs with AFB, and hepatitis B (HBV) infection (Palmer

Beasley et al., 1981; Groopman et al., 1988; Kensler et al., 1991). In a prospective epidemiological study it has been shown that although AFB, and HBV might be independent hepatacarcinogens, they are much more powerful in combination (Ross et al., 1992). In this respect it is interesting to note that inhabitants of Chongming, an island off the coast of Shanghai, who originally consumed a maize diet susceptible to Aspergillus flavus infestation, had the expected high incidence of hepatocellular carcinoma. When, however, wheat, which is much less readily infected, became the principle grain in the diet, the incidence of liver cancer fell rapidly, a phenomenon associated with the removal of promoting activity (Tu et al., 1985). It is therefore important to know the activity of human hepatic GSTs toward AFB,-oxides and their inducibility. Human liver contains predominantly the alpha family enzymes GSTA1 and GSTA2 and also, in about 50% of individuals the mu family enzyme, GSTM1. The activity of human liver supernatant fraction is an order of magnitude lower than that of the normal rat, and none of the enzymes present has an activity of the level associated with GST 10-10 or the mouse GST4-4.

It is possible to gain some idea of the value of inducers for human anticarcinogenesis with isolated human hepatocytes obtained from biopsy material in culture. Dithiole thiones are found to increase levels of the alpha enzymes, particularly GSTA1, over those in control cells in culture and also to have some effect on GSTM1 (Morel et al., 1993).

The protective effect of inducers is worth investigation in populations at risk. An epidemiological survey has been proposed to study their effect on urinary markers of AFB, toxication and detoxication, such as AFB,-guanine and AFBrmercapturates respectively. Induction may be particularly valuable in the case of oltipraz. This compound not only induces GSTs but also cytochromes P450 catalyzing pathways of AFB, metabolism, alternative to epoxidation and therefore detoxifying.

B. GSTs and Human Lung Cancer—Effect of GSTM1 Polymorphism and Its Possible Mechanism

Tobacco smoke, the major cause of lung cancer, contains numerous carcinogens (the number 43 has been quoted, Hoffman and Hecht, 1990), but so far only one group has been shown to give rise to substrates for GSTs, namely the polycyclic aromatic hydrocarbons (Ketterer et al., 1992). They are a diverse group, but in general have similar metabolic pathways.

Benzo[a]pyrene is one of the most commonly occurring and is the most studied tobacco-smoke carcinogen. The genotoxin, associated with its carcinogenesis ( + )-i/nii'-benzo[rt]pyrene-7,8-diol-9,10-oxide (Harvey, 1991), which is a good substrate for GSTs Ml-1 (Robertson et al., 1986), M2-2

(B. Jernstrom and B. Mannervik, unpublished data), and M3-3 (B. Jern-strom, B. Ketterer, and D. J. Meyer, unpublished data) members of the human mu family and better still for GSTP1-1 of the pi family, but very poor for the alpha family (Robertson et al., 1986). Unlike AFB^ejco-oxide, (+ )-tfrt//-benzo[tf]pyrene-7,8-diol-9,10-oxide is utilized equally well by rat and human enzymes, possibly because homology both within and across species is greater in the mu and pi families than in the alpha family.

When an epidemiological study was made comparing 125 smokers with lung cancer with 114 controls, there appeared to be a significantly greater susceptibility to lung cancer in individuals with the GSTM1 null phenotype (Seidegard et al., 1986, 1990). This association supports the general belief that polycyclic aromatic hydrocarbons are important causative factors in smoking-induced lung cancer.

The GSTM1 polymorphism has been known for some years to be genetic in origin (Board, 1981), and recently Seidegard et al. (1986) showed that it was due to a gene deletion. At least five additional epidemiological studies have been made, involving either genotyping or phenotyping (Zhong et al., 1991; Heckbert et al., 1992; Brockmoller etal., 1993;Nazar-Stewart et al., 1993; Hirvonen et al., 1993). Despite differences in experimental design the overall conclusion is (a) that the GSTM1 null phenotype increases susceptibility to lung cancer, (b) that this is more evident when smokers with lung cancer are compared with smoking controls, and (c) that it is even more evident among heavy smokers with lung cancer.

When, in investigations of the origins of the effect of the GSTM1 phenotype, pulmonary GSTs were analyzed, it was shown that GST subunits Al, A2, Ml, M3, and PI were present, GSTP1 being by far the most abundant. Thus, if the detoxication of polycyclic aromatic hydrocarbon diol-epoxides derived from tobacco smoke occurred solely in the lung, little effect of the GSTM1 null phenotype would be expected. The liver, in terms of xenobiotic metabolism, is much more powerful than the lung (the liver is about four times larger than the lung and is much richer in drug metabolizing enzymes). Since the liver contains no GSTP1 but only GSTs Al andA2, and Ml, in GSTM 1-positive individuals, it would appear to be the most likely origin of an effect of GSTM1 polymorphism on smoking-induced cancer (Ketterer et al., 1992).

If the liver is to be important, a large proportion of the polycyclic aromatic hydrocarbon burden in tobacco smoke should pass unchanged through the peripheral lung to reach the liver through the systemic circulation. The transmission of any hepatic effect to the lung assumes that activated carcinogens are released into the hepatic bloodstream and transported to the lung without inactivation. There is good evidence that this occurs in the mouse, where benzo[«]pyrene metabolism by the liver results in the release of ( + )-a»r/-benzo[tf]pyrene-7,8-diol-9,10-oxide into the blood stream where it is transported, apparently protected, to affect other parts of the body (Ginsberg and Atherholt, 1990). That (+ )-anti-bznzo[a]-pyrene-7,8-diol-9,10-oxide also enters the circulation in humans is shown by its presence as a hemoglobin adduct in samples of human blood (Jankowiak et al., 1990). Since GSTM1 null individuals are presumably much less able to detoxify polycyclic aromatic hydrocarbon diol epoxides by glutathione conjugation, much greater amounts of these toxins should pass into the blood and be dispatched to the lung.

Thus polycyclic aromatic hydrocarbons from tobacco smoke are dealt with by the lung in several passes. In the first pass, presumably a small amount is retained and metabolized there. Present knowledge indicates that the distribution of CYP1A1, important for their epoxidation, tends to be peripheral (Antilla et al., 1992). In contrast, the GSTs tend to be bronchial in disposition and therefore encountered first (Antilla et al., 1993). In the second pass, diol-epoxides originating from the liver reach the lung via the pulmonary artery and therefore enter at the proximal end of the respiratory tree and encounter the bronchial region first where GSTs PI, Ml, M2, and M3 are present, but are not sufficient to eliminate the diol-epoxides completely. The load of diol-epoxide entering a lung on the second pass and therefore the potential for damage presumably depends on whether the individual has a positive or negative GSTM1 phe-notype.

It is possible that the stable diol precursor of the diol-epoxide also reaches the lung from the liver. In this case its activation may be in the periphery, where CYP1A1 but not GST activity is most abundant.

The GSTM1 polymorphism is not the only one to affect smoking induced lung cancer and the coincidence of two or more should have a cumulative effect. This has already been shown in a CYP1A1 polymorphism, which, combined with the GSTM1 null phenotype, gives rise to a susceptibility 10 times that of controls (Hayashi et al., 1992). A cumulative effect might also be seen with the concurrence of a GSTM1 phenotype with the debriso-quine hydroxylase "extensive metabolizer" (Caporaso et al., 1990).

Kicking the Habit

Kicking the Habit

Stop Thinking About How You're Going To Quit Smoking When You Can Instantly Stomp It In Less Than 30 Days With A Proven Set of Techniques. When your

Get My Free Ebook

Post a comment