Validation of the Hypothesis

Since the initial proposal of the telomere hypothesis for cell aging and immortalization, all of the predictions of this hypothesis have been shown, by many independent studies, to be correct.

The first prediction, that telomeres will shorten in human somatic cells during replicative aging, has not only been demonstrated in HDFs (Harley et al. 1990, Allsopp et al. 1992), but also in other types of primary human cell cultures, including, but not limited to, keratinocytes (Ramirez et al. 2001), melanocytes

(Bandyopadhyay et al. 2001), endothelial cells (Chang and Harley 1995), vascular smooth muscle cells (Minamino et al. 2001), muscle satellite cells (Decary et al. 1997), mammary epithelial cells (Stampfer et al. 1997), retinal pigment epithelial cells (RPE) (Bodnar et al. 1998), lymphocytes (Hastie et al. 1990; Weng et al. 1995), and hematopoietic stem cells (HSC) (Vaziri et al. 1994). The rate of telomere shortening is variable (-30-250bp/PD), not only for different cell types, but among strains of the same cell type established from different donors. The only types of cultured human cells in which telomeres do not shorten are, as mentioned above, human embryonic stem cells (Thomson et al. 1998), which express telomerase and are immortal, and cells which are maintained in a quiescent (nonproliferating) state for extended periods of time, such as neuronal cells, or quiescent HDF (Allsopp, et al. 1995). The lack of telomere shortening over time in quiescent cells is exactly what is predicted by the end replication problem—if DNA does not replicate, then the amount of telomeric DNA should remain constant.

The second prediction, that telomere length is maintained in the germ line by telomerase, has also been confirmed. Studies have shown that telomerase is readily detectable in human testes and ovaries (Kim et al. 1994, Wright et al. 1995). Furthermore, telomeres in mature spermatozoa are relatively long, and do not shorten during aging in vivo (Allsopp et al. 1992). However, whether telomere length shortens during aging in mature oocytes has yet to be confirmed.

The third prediction, that shortening of telomeres below a critical size signals the senescence checkpoint, has also been shown to be correct in a number of elegant studies. The principal demonstration that telomere shortening causes replicative senescence and limits the proliferative life span of human somatic cells was demonstrated once the catalytic component of telomerase, telomerase reverse tran-scriptase (Tert), was cloned in the mid-1990s (Nakamura et al. 1997; Meyerson et al. 1998). The human Tert gene was first cloned independently by two groups, who searched a then newly established human EST data base for sequences homologous to the previously cloned Tert gene in Saccharomyces cerevisiae (Nakamura et al. 1997, Meyerson et al. 1998). Shortly after the human Tert gene was identified and cloned, experiments were performed to assess the affects of ectopic expression of Tert in a number of different primary human cell cultures, including HDF, endothelial cells, and RPE (Bodnar et al. 1998). In all of these different types of primary human cells, which were also shown to normally lack tel-omerase activity and detectable expression of endogenous Tert, ectopic expression of Tert was sufficient not only to restore active telomerase and prevent telomere shortening during replicative aging, but also to endow these cells with replicative immortality. Since these initial studies, many other types and strains of primary human cells have been immortalized by ectopic expression of Tert (Harley 2001).

Studies have also been performed to assess whether the shortening of just one or a few telomeres below length that is critical for telomere function, or the reduction in the average size of all telomeres in a cell below a certain length, is required to signal replicative senescence. Together, the data suggest that the former event is correct. First, analysis of karyotype stability during replicative aging of HDF has shown that the frequency of dicentric chromosomes involving end-to-end fusion events, which are presumed to arise due to the critical shortening and loss of the protective cap structure of a telomere, are elevated in HDF near senescence (Benn 1976). Second, a number of studies have observed, based on the analysis of signal intensity at the ends of metaphase chromosomes stained with a fluorescent telomeric probe using fluorescent in situ hybridization (FISH), interchromosomal variability in telomere length within primary human cells (Henderson et al. 1996. Lansdorp et al. 1996). This finding demonstrates that some telomeres are shorter than others within a cell, and therefore, if all telomeres shorten at the same rate, the shortest telomere in a cell will become critically short before the rest. Third, experiments examining karyotype stability in human transformed cell lines have demonstrated the presence of dicentric chromosomes at a low intracellular frequency (~0-2 per cell) in transformed human cell lines that are approaching crisis (Counter et al. 1992). Fourth, and perhaps most convincingly, a detailed analysis of individual telomeres in primary mouse cells using a combination of spectral karyotyping (SKY) and quantitative telo-FISH (q-FISH) not only confirmed intracellular variability in the size of telomeres, but also showed that cells from two different groups of mice, both of which had critically short telomeres but differed considerably in average telomere length, had attenuated replicative life spans (Hemann et al. 2001). This experiment was done using a mouse strain deficient in the telomerase RNA component (mTR), which is essential for activity (see Chang, this volume). Specifically, when late generation (G6) mTR knockout (-/-) mice, which have an accelerated aging phenotype and in which the average telomere length is short and where most cells contain one or more chromosomal ends that lack a detectable telomere, were bred with mTR heterozygous (+/-) mice, in which the average telomere length is considerably longer, the F1 mTR-/- progeny had a similar accelerated aging phenotype as the G6 mTR-/- mice despite a marked difference in average telomere length between these two groups of mice. Furthermore, a normal healthy phenotype was restored to the F1 mTR+/- progeny from this cross, even though the average telomere length for these mice was essentially identical to that of the F1 mTR-/- mice. This latter observation is very important since it suggests that telomerase functions in a cell in vivo to selectively maintain or heal telomeres that are, or are becoming, critically short.

The fourth prediction, that telomeres will continue to shorten as transformed primary human cells continue to proliferate beyond the Hay flick limit, or M1 checkpoint, until most telomeres become critically short, has also been demonstrated. In the initial study by Counter et al., which confirmed this prediction, primary cultures of human embryonic kidney cells were transformed with SV40 large T antigen, and telomere length and chromosome stability were assessed by Southern analysis of terminal restriction length (TRF) and karyotype analysis respectively (Counter et al. 1992). It was found that these transformed cells were able to proliferate an additional finite number of population doublings beyond the M1 checkpoint before hitting crisis (the M2 checkpoint), during which time telo-meres continued to shorten. More importantly, end-to-end chromosomal fusions, as well as signal free ends, both indicators of the loss of telomeric DNA from the ends of chromosomes, could be detected throughout the extended life span phase of these cells, and increased in frequency as these cells approached crisis.

Finally, the fifth prediction, that telomeres will cease shortening and telomerase will be activated in the rare immortalized clonal lines that emerge from crisis has also been validated. This prediction was also initially demonstrated by Counter et al., who found that 293 cells (a human embryonic kidney cell line immortalized with SV40 large T antigen) maintained telomere length over a large number of population doublings (> 200), and also contained high levels of telomerase activity (Counter et al. 1992). A more thorough analysis of telomerase activity in many different types of human tumor cell lines found that >90% of human tumors expressed readily detectable levels of telomerase (Kim et al. 1994). Interestingly, while most human tumor cell lines maintain telomere length at a relatively short length (close to the average telomere size for cells when they hit crisis, ~1-2 Kb), the average telomere size in some cell lines is quite long (Takada et al. 1992, Klingelhutz, et al. 1994). The reason why only some tumor cell lines extend and maintain telomeres at a relatively large size is still poorly understood.

0 0

Post a comment