Telomeres are repeat sequences found at the distal ends of chromosomes, forming a cap or end to the chromosome and preventing interactions between chromosomes that may threaten their structure and/or stability (Greider and Blackburn, 1996). While the repeated sequence occurring in telomeres may vary in length from 5 to 15 kb, the maintenance of telomere length presents a problem in cell replication because lagging-strand synthesis is not able to fully replicate the te- lomere end (Figure 15.8). From the figure, it becomes obvious that cells must possess one or more mechanisms to solve the problem noted in the figure—i.e., to prevent telomeres from be-
Figure 15.8 The telomere end-replication problem seen diagrammatically. During DNA replication, synthesis of the lagging strand requires an RNA primer (lined box). The primers that are extended by DNA polymerase are subsequently removed, leaving gaps which the polymerase can repair as long as they do not occur at the end of the telomere because there is the absence of a primer 5′. As a result, one daughter strand will lack the DNA previously encoded by the terminal sequence. If no correction is made, then the DNA sequence will decrease in length each time cell replication occurs. (Adapted from Hamilton and Corey,
1996, with permission of the authors and publisher.)
coming progressively shorter during successive cell divisions, as such a process ultimately leads to a loss of viability. The solution of this problem by the cell was reported by Blackburn and associates (Greider and Blackburn, 1985) in the discovery of telomerase, an enzyme that has the function of extending telomeres.
Telomerase is actually an RNA-dependent DNA polymerase of specialized function. Figure 15.9 presents a model for telomerase action in which the enzyme possesses an RNA hav- ing sequences complementary to the telomeric DNA of the region. As shown in the figure, te- lomerase may extend the DNA sequences by the telomeric repeats for the appropriate number of species and tissue. In this way, telomeres would not decrease in size, as shown in the problem diagrammed in Figure 15.8. Recent studies by Griffith and associates (1999) indicate that pro- teins binding to telomeres may also be involved in the formation of a large duplex loop of the
telomere, which may be concerned with the protection and replication of telomeres. There is also evidence that cells may use other mechanisms besides polymerase to solve the end- replication problem, specifically mechanisms involved in the repair of double-strand DNA breaks (cf. Colgin and Reddel, 1999). Mice whose telomerase gene has been eliminated by gene targeting (Chapter 5) did not appear to show any significant phenotypic differences from normal animals when first reported (Blasco et al., 1997), but later investigations of a similar strain of telomerase-negative mice indicated that these animals exhibited a somewhat shortened life span and increased incidence of spontaneous neoplasia (Rudolph et al., 1999).
Following Blackburn’s discovery, a number of observations noted that chromosome ends became shorter in cells during aging, both in vivo and in vitro (cf. Shay, 1995). In addition, te- lomerase activity in the majority of somatic tissues in adult animals was found to be essentially nonexistent. Exceptions were in the germ cells and also in some stem cell populations (cf. Bac- chetti, 1996). Investigation of neoplastic cells demonstrated that, in contrast to normal cells, most neoplastic cells do possess telomerase activity, although they usually have relatively short telomeres (cf. Ishikawa, 1997). More detailed studies, however, have not shown a consistent ab- sence of telomerase in normal tissues and an increased activity in the neoplastic tissues (e.g., King et al., 1999; Kojima et al., 1997). An example of one study of a variety of human neo- plasms and the activity of telomerase in these tissues is seen in Table 15.11. Note that many normal tissues do not exhibit activity of the enzyme, while others do. In some instances, such as large cell carcinoma of the lung, the telomerase activity is essentially equal to that of normal tissue. Preneoplastic lesions of liver (Kitamoto and Ide, 1999; Tsujiuchi et al., 1996), prostate (Zhang et al., 1998), lung (Yashima et al., 1997), and ovary (Wan et al., 1997) did exhibit telo- merase activity, but usually at a lesser level than that seen in neoplastic tissues. Other studies, however, have indicated that there may be no general tendency toward telomere reduction in malignant tissues (Schmitt et al., 1994) and that telomerase activity may well be a biomarker of cell proliferation rather than malignant transformation (Belair et al., 1997). In some studies, pro- gressive telomere shortening and telomerase reactivation are associated with tumor progression (Miura et al., 1997; Sawyer et al., 1996) and the cell cycle (Harley and Sherwood, 1997). Thus, while many neoplasms exhibit increased telomerase activity and some degree of telomeric pres- ence in their cells, it does not appear that all neoplasms exhibit this interesting change in vivo. The next chapter takes up telomerase loss and reactivation during cell culture as a characteristic of the neoplastic transformation in vitro.
Isozymes, Mutant or “Unique” Proteins, and Fetal Gene Expression in Neoplasia
One of the anticipated results of biochemical studies of neoplasia was the discovery of an en- zyme or protein unique to the neoplastic cell. However, an apparently unique protein species could be the result of the derepression of a normally quiescent segment of the genome expressed in fetal but not in adult tissue of the same lineage, or it could be genuinely unique, as the result of one or more mutations in a gene for an enzyme or protein of the normal cell of origin. The presence in tumor tissue of “new” proteins not found in their cell of origin but resulting from expression of genetic information normally repressed in the cell of origin has been described repeatedly. This latter mechanism is directly related to the “relative autonomy” of the neoplastic cell in the sense of its inability to regulate the expression of its own genetic material, as is done by its cell of origin.
In 1959, Markert and Moller coined the term isozyme, which was generally defined as a member of a family of multiple, separable forms of enzymes occurring within the same organ- ism and having the same or very similar catalytic activity. Examination of such isozyme families in neoplasms, especially those of rodent liver, demonstrated that several isozymic forms of en- zymes occurred in experimental hepatomas but were either absent or at very low levels in normal liver. In 1982, Weinhouse reviewed much of the literature to date on this subject in relation to murine hepatic lesions, demonstrating that the new and unique isozymic forms seen in neo- plasms, as compared with their cell of origin, could almost always be found in other normal tissues in the adult or in the cell of origin of the neoplasm during fetal life. A “unique” aldehyde dehydrogenase was found in rat hepatomas by Lindahl (1979) but was later found to be induced in normal adult tissues by some carcinogenic agents, especially 2,3,7,8-tetrachlorodibenzo-p- dioxin (Hempel et al., 1989). Some enzymes and proteins in tumors were found to differ in the nonprotein portion of the molecule, thus leading to an apparent new “isozyme,” as reported with γ-glutamyltranspeptidase (Tsuchida et al., 1979), and also with the iron storage protein ferritin (Linder et al., 1975).
The altered isozyme composition of neoplasms is not ubiquitous but appears to be more characteristic of the poorly differentiated, rapidly growing tumors than of the highly differenti- ated, slowly growing neoplasms. Thus, the appearance of fetal or other isozymes in neoplasms may be related to tumor progression, as suggested by Weinhouse (1973). On the other hand, the appearance of fetal forms of proteins in neoplastic cells is not unique to isozymes. In fact, the expression of fetal genes in neoplasms in the adult is the rule rather than the exception. As with other biochemical characteristics of neoplasms, however, this expression also appears to be quite
Figure 15.9 Model for telomerase action. a. Binding of telomerase to the end of telomeric DNA via the RNA component of the enzyme. b. The RNA component of telomerase serves as a template for the exten- sion of the telomere 3′ end. c. Translocation of telomerase along the newly synthesized strand. d. Further extension of the telomere 3′ end occurs, and the process continues with subsequent translocation-extension variable and heterogeneous. Furthermore, expression of a fetal gene is not unique to the neoplas- tic state but may occur in vivo in certain pathological conditions such as inflammation and re- generation. Cultured cells may also express fetal genes in vitro, as exemplified by the investigations of Sirica and associates (1979).
While these earlier studies laid the foundation for identifying and understanding differ- ences in phenotypes of various neoplasms, it was not until the advent of the “explosion” in mo- lecular biological techniques and knowledge in the last two decades that structural differences in a variety of proteins in neoplasms resulting from mutational alterations in the genes for such proteins became well documented. Unlike the situation 25 years ago, when few if any docu- mented alterations in protein structure could not be accounted for by altered gene expression in neoplasms, today there is considerable evidence for the presence of mutant proteins and their genes within neoplastic cells. The majority of such examples occur within cellular oncogenes,
steps. e. Synthesis of the complementary strand (C-rich) then occurs by extension of an RNA primer by a conventional DNA polymerase. Removal of the RNA primer leaves a 3′ overhang at the telomeric end. hTR, human telomerase RNA component; hTERT, human telomerase reverse transcriptase. (Adapted from Urquidi et al., 1998, with permission of the authors and publisher.)
alterations have been discussed in this chapter and earlier chapters, but by way of review some of the better-documented structural alterations in such genes are listed in Table 15.12. Although the p53 tumor suppressor gene has been discussed earlier, it is important to emphasize the ex- tremely high rate of mutations seen in this gene in human neoplasia (cf. Hainaut and Hollstein,
2000). In Figure 15.10 may be noted the percentage and relationship of transitional and transver- sional mutations in this gene in a variety of different human neoplasms.
The importance of p53 function has been previously noted in Figure 15.9 in a highly sim- plified diagram. The role of p53 as “guardian of the genome” is implied from that figure in that, in the presence of DNA, damage with subsequent activation of p53 by the ATM kinase mediated by a checkpoint kinase (Hirao et al., 2000; Morgan and Kastan, 1997) can lead to arrest of the cell cycle by transactivation of the cyclin inhibitor p21, allowing repair to occur. This is paral- leled by activation of an apoptotic pathway which, if the cell damage is not repaired in a timely fashion or cannot be repaired, leads to death of the cell. Mutations in p53 to a great extent alter its transactivation capabilities, thus interrupting its guardian function and allowing the cell to survive and replicate with major DNA damage. Although mutation of the p53 gene may be con- sidered an initiation event, most studies both in humans (e.g., Navone et al., 1993; Tanaka et al.,
1993; Kuwabara et al., 1998) and animals (Miller, 1999) indicate that such mutations are identi- fied predominantly during the stage of progression. Furthermore, during the stage of progression there is ample evidence that elimination of wild-type p53 activity leads to gene amplification (Livingstone et al., 1992), chromosome instability (Shao et al., 2000), and genomic instability due to alteration in checkpoint function (cf. Smith and Fornace, 1995). Thus, loss of p53 func- tion may contribute in a major way to the evolving karyotypic instability characteristic of the stage of progression.
The PTEN tumor suppressor gene appears to be a regulator of specific kinases involved in signal transduction (Sun et al., 1999). Somatic mutations of the PTEN gene have been seen in acute myeloid leukemia (Liu et al., 2000) and hepatocellular carcinomas (Kawamura et al.,
1999). The PTEN gene appears to function in neoplasms of the endometrium as a “gatekeeper,” as does the p53 gene for a number of other tissues as well (Ali, 2000). Genes having “caretaker” functions may be those such as the mismatch repair genes found defective in hereditary non- polyposis colorectal carcinoma (Chapter 5). The concept of gatekeeper and caretaker has been proposed and discussed by Kinzler and Vogelstein (1997). Allelic imbalances and somatic muta- tions of the p16 tumor suppressor gene, an inhibitor of CDK4 (Table 15.7), were seen in soft tissue sarcomas (Schneider-Stock et al., 1998).
Mutations in cellular adhesion molecules involved in invasion and metastasis have also been described, as indicated previously in Chapter 10. In particular, genes of the cadherin fam- ily, which are considered tumor suppressor genes, are mutated in a fairly high percentage of breast cancers (Berx et al., 1996) and very likely in a variety of other human neoplasms (cf. Semb and Christofori, 1998). Mutations in the catenin gene, which is involved in the function of
Table 15.12 Genes and Gene Families Commonly Exhibiting
Mutations in Neoplasia
Cell cycle inhibitors, e.g., p16
Cadherins, catenins, and other genes involved in cell-cell interactions
Receptors G proteins Transcription factors
Figure 15.10 Relationship between CpG transitions and GC to TA transversions in the p53 cDNA. Neoplasms have been listed according to the frequency of GC to TA transversions (increasing from top to bottom). Note that those neoplasms exhibiting a low frequency of GC to TA transversions show a high frequency of CpG transitions and vice versa. Generally it is considered that neoplasms exhibiting a low level of p53 CpG transitional mutations and high GC to TA transversion mutations are associated with or caused by exogenous risk factors. (Adapted from Hainut and Hollstein, 2000, with permission of the au- thors and publisher.)
the APC tumor suppressor gene (Chapter 5), have been noted in both human neoplasms (Koch et al., 1999) and those induced in animals (Ogawa et al., 1999). Other proteins involved in such cell adhesion function include integrins and the CD44 glycoprotein family. In both of these in- stances, neoplasms express isoforms that their cells of origin frequently do not (cf. Rudzki and Jothy, 1997; Schapira, 1981). The other three types of protein listed in the table have been dis- cussed rather extensively in this chapter and will not be discussed again here. It is sufficient to say that mutations in proto-oncogenes, resulting in their conversion to cellular oncogenes, and in tumor suppressor genes are undoubtedly ubiquitous in neoplasia although it is not yet clear in exactly which stage of neoplastic development the predominant mutations occur, nor, with the exception of specific hereditary conditions involving single tumor suppressor genes (Chapter 5), what the actual initiation mutation is.
Secondary Biochemical Changes Resulting from Alteration of
It appears that the principal mechanisms resulting in the alteration of genetic expression in neo- plastic cells concern themselves with the processes of transcription or translation or both, but changes in the concentration and/or presence of specific enzymes have distinct ramifications in
the ultimate phenotype of the neoplastic cell. Such changes are seen in structural lipids, espe- cially the glycolipids, and in their presence or absence on the surface membrane of the cell. Ha- komori and Kannagi (1983) have summarized many of the data concerning changes in the expression of these molecules that result from alterations in levels of enzymes responsible for their synthesis in a variety of different neoplasms. Such surface changes may affect the anti- genicity of the neoplastic cell (Chapter 15) or its metastatic potential. The appearance or disap- pearance of specific glycolipid structures on the surface of the cell may also express a fetal phenotype.
In addition to glycolipids, the lipid composition of neoplasms, especially hepatomas, dif- fers from that of normal liver (cf. Ruggieri and Fallani, 1979), as does the fatty acid composition of phosphatides of normal and malignant epidermis (Carruthers, 1967). Such changes would be expected because of the altered regulation of enzymes synthesizing sterols as well as fatty acids (Sabine, 1975). In particular, Van Hoeven and his associates (1975) demonstrated in a study of several different mouse and rat hepatomas that the plasma membranes of such cells exhibited increased cholesterol content and a decrease in the degree of unsaturation in the fatty acids of most lipid classes in the neoplasms.
Illustrating further ramifications of such structural alterations resulting from changes in the regulation of genetic expression, a number of studies have indicated both altered and de- creased interaction of hormones with their receptors and the activation of adenylate cyclase in neoplasms of both the human and lower animals (cf. Hunt and Martin, 1979). Pezzino et al. (1979) have shown that five malignant hepatomas of varying growth rates have lowered capaci- ties for binding insulin and glucagon to membrane receptors; this may be due to a decrease in receptor number, lowered affinity, or site-site interactions of the hormones, receptors, and plasma membrane structures. These findings further demonstrate the ramifications of altered ge- netic expression, which can, in turn, lead to alterations in the ability of environmental factors to initiate intracellular mechanisms leading to the control of gene expression.
The importance of structural alterations in proteins resulting from changes in the regula- tion of genetic expression is probably not completely appreciated at the present time. The pio- neering studies of Hakomori and associates pointed out the importance of glycolipid and other structural component changes which in turn alter the function of the cell. Olden (1993) has em- phasized that the expression of aberrant oligosaccharide moieties of glycoproteins and glycolip- ids is a typical characteristic of essentially all animal and human neoplasms regardless of their causative mechanisms. Thus, it is obviously not necessary to have mutations in specific proteins in order to alter functional characteristics of cells; rather, alterations and the regulation of ge- netic expression may be of equal or greater importance than mutational alterations in specific proteins and groups of proteins.