Although the mechanism of action of hormones involves few if any structural alterations in these chemical agents, in analogy to chemical carcinogenesis one might expect that some hormones, or more likely their metabolites, can interact directly with macromolecules, especially DNA, as has been described for nonhormonal chemical carcinogens (see above). One of the earliest stud- ies demonstrating a low level of covalent interaction of hormones with proteins was that reported by Riegel and Mueller (1954), who demonstrated the formation of a protein-bound metabolite of estradiol in liver homogenates of female rats. Kappus and Remmer (1975) also described the irreversible binding of a synthetic estrogen to proteins of liver microsomes, possibly through an intermediate epoxide form. More recently Epe et al. (1990) have described the covalent binding of quinoid metabolites of diethylstilbestrol to microtubular protein in vitro and have proposed a role for such binding in the development of aneuploidy and the development of neoplasia in- duced by this synthetic estrogen.
The binding of both synthetic estrogens and their natural forms to DNA or the induction of DNA adducts revealed by 32P-postlabeling (see above) has also been reported. Lutz and co- workers (1982) reported the covalent binding of diethylstilbestrol to DNA of rat and hamster liver and kidney, albeit at extremely low doses compared with the binding of nonhormonal chemical carcinogens. These workers (Jaggi et al., 1978) had earlier described the covalent bind- ing of ethinylestradiol and estrone to liver DNA at levels four orders of magnitude lower than that of dimethylnitrosamine. Since these reports, Liehr and his associates (1986, 1987) have re-ported the appearance of new 32P-postlabeling spots (DNA adducts) in hamster kidney, in which such hormones induce malignant neoplasms. Although it is not clear that the new adducts are formed from the hormones in most instances, Gladek and Liehr (1989) presented evidence that 32P-postlabeled spots (adducts) appearing after diethylstilbesterol treatment of animals were likely to be generated from the reaction of the 4′,4″-quinone metabolite of diethylstilbesterol with DNA. Since steroid estrogens are metabolized to catechol estrogens, which are capable of forming quinones, Liehr (1990) has postulated a major role for such metabolites in estrogen- induced carcinogenesis. Interestingly, administration of the antiestrogen tamoxifen inhibited the development of renal neoplasms after 17β-estradiol administration but did not alter the DNA adduct levels (Liehr et al., 1988). Further support of a role for metabolic activation of synthetic and natural estrogens in their carcinogenic action is seen from the generation of unscheduled DNA synthesis (DNA damage) in cultured hamster cells by diethylstilbestrol and related com- pounds (Tsutsui et al., 1984) and the induction of sister chromatid exchanges both in vivo and in vitro by diethylstilbestrol (Mehnert et al., 1985). A similar increase in sister chromatid ex- changes in human lymphocytes induced in vitro by diethylstilbestrol was not accompanied by a detectable change in 32P-postlabeling patterns (Lundgren et al., 1988). That estrogens may in-duce neoplasia by other pathways that may be important for chemical carcinogenesis has been shown by Roy et al. (1991), who presented evidence for a role of active oxygen in DNA damage resulting from diethylstilbesterol treatment of hamsters in vivo. In addition, Lu et al. (1988) re- ported the hypomethylation of DNA in estrogen-induced hamster renal neoplasms. Thus, at least for synthetic and in one system natural estrogens, there is some evidence that metabolic activa- tion may play a role in their carcinogenic activity, exemplified predominantly in the hamster. There is no evidence, however, that polypeptide hormones such as thyrotropin or gonadotropins, which exert their action at the cell surface and in all likelihood never reach the cell nucleus in an undegraded form, would exert their carcinogenic action by covalent adduction to DNA.
Furth suggested that the effect of hormones is to increase the rate of cell replication and that the rapidly dividing cell thus becomes more susceptible to “endogenous” carcinogenesis from genetic “mistakes” or mutations; the chance of such formations is increased by the more rapid rate of DNA synthesis. On this basis one should expect that carcinoma of the small intestine would be a very common neoplasm because of the extremely rapid rate of replication of intestinal epithelial cells. This is not the case either in humans or in animals. An earlier report by Málková and associates (1977) reported that the chronic administration of estrogen, as estradiol benzoate, induced hyperplasia in the cells of the pituitary of rats. For the first 2 months of the administra- tion, hyperplastic cells exhibited normal karyotypes, but after that time the number of aneuploid cells increased, followed by neoplasms. In contrast, in the bone marrow, a rapidly replicating tis- sue, no detectable changes were seen in chromosomal morphology, although the mitotic index was lowered, as also occurred with the pituitary after four weeks of estrogen administration. Thus, rapid cell replication by itself does not appear to directly induce the neoplastic change, since rapidly replicating normal cells such as intestinal epithelial cells do not show an inordinate tendency to develop neoplasia. As discussed further on (Chapter 7), it seems likely that hormones may not act directly on cells to induce the neoplastic transformation; rather, they may act to en- hance the replication and progression of a few cells already potentially neoplastic as a result of ambient environmental factors such as dietary contaminants, background radiation, and so on.