866 THE CARCINOGENIC STIMULUS II 



Methionine deficiency per se induced by feeding ethionine (Popper et al., 1953) 

 and choline deficiency (Copeland and Salmon, 1946) may be responsible for the 

 induction of hepatic neoplasms. These are also cirrhotogenic agents; the correla- 

 tion of cirrhosis with hepatic carcinogenesis in man is well known. Low riboflavine 

 content of the liver can be associated with the induction of liver tumors by azo 

 dyes; maintenance of a high riboflavine content is tumor-inhibitory (Griffin and 

 Baumann, 1948). Dye-binding is facilitated by riboflavine deficiency. 



Mice on a cystine-deficient diet resist the leukemogenic action of methyl- 

 cholanthrene (White et al., 1947); the body growth of mice on a lysine-deficient 

 diet is similarly inhibited, but susceptibility to the induction of leukemia is not 

 altered, suggesting that cystine-deficiency per se is a potent inhibiting influence. 



Caloric restriction may inhibit and high fat diet favor the development of 

 certain types of neoplasms. Tannenbaum (1953) has done crucial experiments to 

 demonstrate the stage of carcinogenesis aflfected by diet. Apparently initiation 

 proceeds, but the "developmental" or "promotional" stage is inhibited by caloric 

 restriction. 



VI. GENETIC STUDIES 



The possibility that genie alteration is basic to carcinogenesis has been dis- 

 cussed above. Since cancer cells pass on the abnormality responsible for uncon- 

 trolled proliferation to their progeny, it seems reasonable to regard cancer as a 

 heritable cellular anomaly, heritable as used here in the "somatic" sense, since 

 the neoplasm might arise in an individual whose germ cells are not similarly 

 deviated. Upon transplantation, cancer cells inay proliferate in a histocompatible 

 normal (non-cancerous) host (Figs. 60, 62-65). Carcinogenic agents might act 

 either by entering or entrance into cells altering cellular constituents, or by altering 

 the cellular environment. Cellular mutation might result from either change. 



Certain inbred strains of mice are susceptible and others resistant to the induction 

 of tumors by specific agents. Susceptibility is determined by dominant genes; Fi 

 hybrids between susceptible and resistant strains exhibit susceptibility to car- 

 cinogenic induction of neoplasms. If the target tissue for the carcinogenic agent is 

 grafted into Fi hybrids, parent stock tissue serving as donors, and if the Fi hybrid 

 is treated with the carcinogen, tumors appear in the Fi hybrid and grafted sus- 

 ceptible parent tissue, but not in the grafted tissue derived from the resistant 

 parent. This phenomenon of control of tumorigenesis by target tissue applies to 

 mammary (Prehn, 1953), pulmonary (Heston and Dunn, 1951; Shapiro and 

 Kirschbaum, 1 951), thymic (Kaplan ^i a/., 1956a; Law andMiller, 1950b) testicular 

 (Trentin and Gardner, 1957) and adrenal cortical (Huseby and Bittner, 1951) 

 tissue of mice (Fig. 66) . 



When gonadectomized-adrenalectomized Fi hybrids between stocks sus- 

 ceptible and resistant to the induction of post-castrational cortical carcinomas 

 received adrenal grafts from both, castration-induced carcinomas developed only 

 in the adrenal cortices supplied by the susceptible donors (Huseby and Bittner, 

 1 95 1 ) . Since the same trophic hormonal stimulus acted upon both adrenals in the 

 common host, the adrenal cority. per se determined tumorigenesis. Similarly, when 



