Over the past several years, the scientific and popular available for studying tumor development in the mouse), press has been bubbling over with experimental findings there are two important differences between mice and that would seem to be setting the stage for a magic humansregardingtelomeresandtelomeraseregulation. bullet against cancer. The focus of this fervor is the First, telomerase activity is less stringently repressed in enzyme telomerase, the reverse transcriptase responsithe somatic tissues of mice compared to humans; sevble for maintaining the telomeric DNA at the ends of our eral tissues in the adult mouse show measurable telochromosomes. Telomeres are essential elements that merase activity (Prowse and Greider, 1995). Second, the protect chromosome ends from degradation and ligatelomeres of mice are about 20 kb longer than human tion. Without their telomeric caps human chromosomes telomeres, whichareonly5–15kb. Sincetelomeresonly undergo end-to-end fusions, forming dicentric and shorten by about 100 bp per cell division, mouse cells multicentric chromosomes that break in mitosis, activatlacking telomerase would require many more doublings ing DNA damage checkpoints and otherwise wreaking to exhaust telomeric sequences compared to human havoc on the genome (Harley et al., 1990; Counter et cells. Given these crucial differences, it is unlikely that al., 1992; Harley, 1995; van Steensel et al., 1998). Telomice normally use telomere erosion as a tumor suppresmerase is not active in most somatic tissues but is widely sor mechanism. To circumvent this problem, the teloactivated in cancer cells (Kim et al., 1994; Shay and merase-negative mice had to be bred for six generations Bacchetti, 1997). Its activity in tumors has been exuntil their telomeres had been brought down to human plained by the need of continuously dividing cells to size. overcome the progressive loss of telomeric sequences In the first report on the mTR mutant mouse in 1997, that results from the fact that DNA synthesis leaves Blasco et al. reported that mice could tolerate the aba terminal stretch of unreplicated DNA (Watson, 1972; sence of telomerase activity, not just in the first genera-Olovnikov, 1973). tion but through at least five successive generations of The prevailing model holds that if not for activation of interbreeding telomerase-deficient animals. Using emtelomerase, incipient tumor cells would either undergo bryo-derived fibroblasts (MEFs) from mice at the differterminal growth arrest (senescence), or die, when their ent generations (G2, G3, G4, etc.), they could show that telomeres were whittled away to an unacceptably short telomere length declined over successive generations, length (Counter et al., 1992). Thus, telomere shortening ultimately giving rise to cells with reduced telomeric has been viewed as a tumor suppressor mechanism, signal by in situ hybridization and some of the anticiand the near-universal phenomenon of telomerase acti- pated chromosomal effects such as end-to-end fusions. vation in cancer would seem to define an Achilles’ heel Therefore, these experiments confirmed that telomerase of the tumor cell. According to the many proponents of was necessary for maintenance of telomeric sequences this scenario, inhibition of this inappropriately expressed in mice, but it was not required for the initial developenzyme would render the cells unable to maintain their ment and health of the animals, at least out to generation telomeric DNA, resulting in tumor stasis or regression six (G6). due to the growth inhibitory effects of excessive telo- Furthermore, these investigators found that the mTRJ/J mere erosion …