Some 45 years ago, the late Jesse Greenstein, author of Biochemistry of Cancer (1954) (the next major text after Warburg’s work on the biochemical characterization of neoplasia), proposed that the biochemical constitution of neoplasms tended to converge to a relatively common enzymatic pattern. Several authors, including V. R. Potter (1964) of the McArdle Laboratory, pointed out that the Greenstein hypothesis had not challenged that of Warburg but actually extended it. War- burg’s ideas of convergence were limited to the area of glycolysis and respiration, whereas those of Greenstein were extended to a number of enzymatic functions in the cell that could be ana- lyzed in the 1940s. Greenstein understood and realized the importance of using valid tissue com- parisons: he was among the first biochemists studying cancer to realize the importance of comparing liver with hepatomas, since they are relatively homogeneous cellular populations. In stating the hypothesis of convergence, Greenstein (1956) recognized the existence of exceptions to his theory, although perhaps not their importance for an understanding of the cancer problem in general. Since the enunciation of the convergence hypothesis, a number of exceptions have been reported on comparison of relatively highly differentiated neoplasms with their appropriate cells of origin (Morris, 1965; Mekhail-Ishak et al., 1989). However, like the Warburg hypothesis, the convergence hypothesis can be reconciled with our better understanding of the development of the stage of progression.
Unlike the two hypotheses mentioned above, the deletion hypothesis, first advanced by the Mill- ers (1947) more than 35 years ago, was not based on studies with many different neoplasms, but rather evolved from investigations on the production of hepatic cancer by the feeding of ami- noazo dyes to experimental animals. The basic experimental observation was that the dye be- came bound in a covalent fashion to proteins of the liver of the dye-fed animal, whereas little or no dye binding occurred in the protein of the neoplasms ultimately produced (cf. Chapter 3). The Millers thus postulated that carcinogenesis resulted from “a permanent alteration or loss of protein essential for the control of growth . . . .” The reader should note that at this time the structure of DNA was not known and the concept of the gene was rather nebulous. Later studies by Sorof et al. (1958) and others indicated that the proteins to which the dyes were bound in greatest amounts constituted an electrophoretically slow-moving class termed the h2 proteins. These proteins were found to be missing from the neoplasms that were produced by dye feeding. However, investigations with highly differentiated neoplasms have shown the presence of the h2 protein(s), although little or no dye binding occurred in the h2 fraction (Sorof et al., 1966).
It is now clear that, for the covalent linkage between amine carcinogens and liver macro- molecules to occur, aromatic amine carcinogens must be “activated” by N-hydroxylation and subsequent esterification, as described in Chapter 3. Lotlikar (1970) did demonstrate the capability for N-hydroxylation (Chapter 3) in several of the highly differentiated hepatomas. Thus, the absence of a portal blood supply and the generally lowered rate of blood perfusion through these neoplasms, as shown by Gullino and Grantham (1961), coupled with the low drug metabolism, may explain the lack of covalent interaction between the carcinogen and soluble proteins of the neoplasm. Elizabeth Miller (1951) and Heidelberger (1970) showed that a com- pletely analogous situation took place during skin carcinogenesis by hydrocarbons. A protein isolated from mouse skin binds hydrocarbons in a direct relationship to their carcinogenic activ- ity for the skin. Electrophoretically, this protein had many of the characteristics of the h2 proteins
About 10 years after the original proposal of the deletion hypothesis, Potter (1958) sug- gested that the proteins deleted during carcinogenesis may be identical to or associated with en- zymes involved in catabolic reactions, a view compatible with the Greenstein hypothesis as well as with some of the biological aspects of neoplasia; for example, rapid growth. Furthermore, several experimental hepatomas demonstrated a complete lack of many catabolic reactions char- acteristic of liver, in support of the concept of “catabolic deletion.” However, not long after the initial proposal of the catabolic deletion hypothesis, a series of hepatocellular carcinomas rang- ing from poorly differentiated to highly differentiated were produced in the laboratory of Harold Morris (Morris and Wagner, 1968), and their chemical characteristics were studied by numerous investigators. These studies demonstrated that the highly differentiated hepatocellular carcino- mas exhibited virtually all of the normal hepatic enzymatic functions investigated and, in several instances, lacked any abnormal glycolytic capacity, while the enzyme pattern of the poorly dif- ferentiated neoplasms was characteristic of the concept of “catabolic deletion.”
The discovery of the Morris hepatomas and the investigation of their biochemical charac- teristics greatly altered the biochemical concepts of neoplasia that had developed up until 1961. Studies with these neoplasms demonstrated that simple qualitative measurements of enzymes and proteins of neoplasms were not sufficient to characterize the critical difference(s) between a normal and a neoplastic cell.
Minimal Deviation Concept
As a result of investigations with the first of the transplantable Morris hepatomas (no. 5123), Potter (1961) proposed the concept of the “minimal deviation” neoplasm, at the opposite end of the spectrum from those neoplasms conforming to the original Greenstein convergence hypo- thesis. Potter’s concept was that some neoplasms were probably very closely related to or virtually identical with the initiated cell. An experimental program was proposed to examine a variety of transplantable hepatomas either to discover an essential alteration found in all examples or to disprove a generalization by finding exceptions. The phenotype of the highly differentiated hepatomas deviated only slightly from normal with respect to their growth charac- teristics and exhibited relatively few abnormalities except those necessary for the expression of neoplasia. As these cells progressed, their deviation from the cell(s) of origin increased to produce a moderately or maximally deviated neoplasm. The term minimal deviation hepatoma appeared to apply to a number of the Morris transplantable neoplasms. Although there has been some discussion of this concept in the literature, its correlation with both morphology and kary- otype appears to follow the distinction between neoplasms as to their degree of differentiation.
Molecular Correlation Concept
Shortly after the demonstration of the existence of a spectrum of transplantable hepatomas, We- ber and associates (Weber, 1983) embarked on an extensive biochemical analysis of the enzy- matic patterns of these neoplasms. These workers attempted to assemble the data into a modification of the idea of catabolic deletion, but in direct relation to cell replication and the growth rate of the neoplasm. Thus, by the molecular correlation concept, certain enzymatic ab- normalities seen in a class of neoplasms compared with their cells of origin may be closely cor- related with the growth rate of the neoplasm. Other functions, usually those more closely associated with the degree of differentiation of the organ, show little or no relation to the growth rate of the neoplasm. Weber’s investigations were extended to include several different types of experimental neoplasms, numerous enzymatic functions, and specific metabolic pathways (Weber, 1983).
Biochemical Theories of Neoplasia Revisited
Although the four theories described above are not considered in most modern-day investiga- tions of the cell and molecular biology of neoplasia, their influence on our basic knowledge of the biochemistry of neoplastic cells was very significant. Today, in retrospect, one may view the theories as related to studies of neoplastic cells in the stage of progression. Even those highly differentiated hepatocellular carcinomas forming the basis of the minimal deviation concept of Potter were almost all in the stage of progression as defined by karyotypic alterations (Nowell et al., 1967; Wolman et al., 1973). Thus, it is reasonable to suggest that the Warburg, Greenstein, and Weber concepts all involve studies of neoplasms in the stage of progression, while the dele- tion hypothesis may be more closely related to the stage of initiation, since the original theory, although ahead of its time, basically argued for the presence of a specific gene mutation as the seminal event in the conversion of a normal to a neoplastic cell. From all of these studies com- bined, however, it is reasonable to argue that there is as yet no evidence for a cellular pattern of metabolism specific to the neoplastic state, but there is evidence that in the natural history of neoplasia, there is a progression of neoplastic cells through morphological, karyotypic, and metabolic changes.
The reader should be aware that a number of other theories of carcinogenesis and of the neoplastic transformation have been proposed during the past half century. Several of these have been considered in earlier chapters, but it is not possible nor is it the function of this text to review all of the proposed theories of carcinogenesis. Rather, the rest of this chapter emphasizes several specific biochemical, cell, and molecular biological changes in preneoplastic and neo- plastic cells, making appropriate comparisons wherever possible.