It is generally accepted that the ''clock'' that keeps track of cell divisions and signals cell cycle arrest in human cells is the telomere, a region at the end of each chromosome that consists of multiple repeats of a specific DNA sequence (Campisi et al., 2001). Due to the end-replication problem in copying the full length of the lagging DNA strand, normal somatic cells undergo progressive telomere shortening with cell division. Once the telomere reaches a certain critical length, the DNA damage signaling pathway is activated, with the concomitant up-regulation of cell cycle inhibitors. The process of replicative senescence is a stringent characteristic of human somatic cells, whereas in germ cells, in certain stem cells, and in tumor cells, telomere shortening and replicative senescence are prevented by the activity of an enzyme called telomerase, which uses its RNA template to synthesize the telomere sequence.
Human and mouse cells show fundamental differences with respect to telomere size and telomerase activity. Because of the significantly longer telomeres in laboratory strains of inbred Mus musculus, and the high levels of telomerase present in most mouse tissues, it is unlikely that mouse cells undergo telomere-based senescence (Akbar et al., 2000). Indeed, the barrier to unlimited proliferation may actually be less stringent in murine cells, since, unlike human cells, mouse cells undergo frequent spontaneous immortalization in cell culture. Thus, there is a major difference between mice and humans with respect to this important facet of cell biology. In terms of the aging immune system as well, the life-long exposure to pathogens also differentiates elderly humans from aged mice housed in barrier facilities, further underscoring the value of using a human cell culture model for analysis of the role of T cell replicative senescence in human aging.
Our cell culture analysis of human T cells has documented that telomere length undergoes progressive shortening with increasing rounds of antigen-driven proliferation, reaching 5-7 kb at senescence (Vaziri et al., 1993) Although this telomere length was similar to that of other cell types that reach senescence in cell culture, we were somewhat perplexed by the observed telomere shortening in T cells, since lymphocytes are unique among human somatic cells in that they induce high levels of telomerase activity in concert with the activation process (Weng et al., 1996). Indeed, the levels of telomerase activity in antigen or mitogen-stimulated T cells are comparable to those of tumor cells (Bodnar et al., 1996).
To address this issue, we performed a detailed kinetic analysis in cell culture of CD8+ T cell telomerase activity induced by activation. We showed that after mitogen or T cell receptor (TCR)-mediated activation, the telomerase activity peaks at 3-5 days, then undergoes a gradual decline, becoming undetectable at approximately 3 weeks. A second wave of telomerase activity can be induced by a subsequent exposure to the same antigen, and during the period of high telomerase activity, telomere length remains stable (Bodnar et al., 1996). However, in CD8+ T cells, the antigen-induced upegulation of telomerase in response to stimulation with antigen is markedly reduced by the third stimulation, and is totally absent in all subsequent encounters with antigen (Valenzuela and Effros, 2002).
One of the unexpected findings in our cell culture studies was the significant difference between helper (CD4+) and cytotoxic (CD8+) T cell subsets with respect to telomerase. This observation was made by culturing CD4+ and CD8+ T cell subsets that were isolated from the same individual, using the identical stimulatory schedule. The initial stimulation by alloantigen elicited a 45-fold and 53-fold increase in telomerase activity in the CD8+ and CD4+ cell cultures, respectively. However, by the 4th antigenic stimulation, telomerase activity was undetectable in the CD8+ T cells, whereas CD4 T cells still showed robust telomerase activity even as late as the 10th round of antigenic stimulation.
The divergent patterns of telomerase activity between CD4+ and CD8+ T cells paralleled the pattern of CD28 expression changes. By the 7th antigenic stimulation, 90% of the cells in the CD8+ culture no longer expressed CD28, whereas at that same stage, the CD4+ cultures were still 75% CD28+ (Valenzuela et al., 2002). The intimate relationship between telomerase activation and CD28 signaling was further demonstrated in experiments using antibodies to block CD28 binding to its ligand on antigen-presenting cells, which resulted in a significant reduction in telomerase activity. Finally, the distinct contribution of CD28 to telomerase activation was evident in experiments using Cyclosporin, a TCR signaling inhibitor. Cyclosporin was not able to inhibit telomerase in T cells activated by a combination of antibodies to CD3 and CD28, whereas potent inhibition was observed when only anti-CD3 was used for stimulation (Valenzuela et al., 2002).
REVERSAL/RETARDATION OF T CELL REPLICATIVE SENESCENCE IN CELL CULTURE
Given the central role of telomere shortening in the replicative senescence ''program'' in T cells, our main approach at modulating senescence has focused on strategies to enhance telomerase activity in CD8+ T cells. An excellent model system for these studies is the virus-specific CD8+ T cell response, which is known to decline with age as well as during chronic infections, such as HIV. Therefore, using the long-term culture system described above, we followed HIV-specific CD8+ T cells that had been isolated from persons infected with HIV, and tested the effect of gene transduction with the catalytic component of telomerase (hTERT). Comparisons were made between the hTERT-transduced cultures and the empty-vector-transduced cultures.
Results of these experiments showed significant effects of hTERT on proliferative and functional aspects of the T cells. Briefly, we observed that hTERT trans-duction led to telomere length stabilization and reduced expression of the p16INK4A and p21WAF1 cyclin-dependent kinase inhibitors, implicating both of these proteins in the senescence program (Dagarag et al., 2004). Indeed, the transduced cultures showed indefinite proliferation, with no signs of change in growth characteristics or of karyotypic abnormalities. In terms of protective immune function, the ''telomerized'' HIV-specific CD8+ T cells were able to maintain the production of IFNg for extended periods, and showed significantly enhanced capacity to inhibit HIV replication. The loss of CD28 expression was delayed considerably, although ultimately not prevented, suggesting that additional genetic manipulation of the CD28 gene itself may be required for full correction of this important senescence-associated alteration. Similarly, virus-specific cytolytic function was not restored by hTERT transduction (except in selected clones). Thus, hTERT corrects most, but not all, the alterations associated with replicative senescence in CD8+ T cells isolated from HIV-infected persons. Ongoing studies are addressing whether these same effects will be seen in hTERT transduced cells from healthy donors, and whether transduction at earlier timepoints along the trajectory to senescence will enhance the telomerase effects.
Telomerase enhancement may also be achieved using nongenetic strategies, which would offer more practical approaches to therapeutic interventions in the elderly.
For example, it is known that estrogen is able to enhance telomerase activity in reproductive tissues. The complex formed when estrogen binds to its receptors migrates to the nucleus and functions as a transcription factor. In normal ovarian epithelial cells, this complex actually binds to the hTERT promoter region (Misiti et al., 2000). It has been known for some time that T cells can bind to estrogen via specific estrogen receptors. Thus, we tested whether pre-incubation of T cells to 17^-estradiol prior to activation might augment telomerase activity. Our preliminary data suggest that estrogen does, in fact, enhance T cell telomerase activity. The enhancement is observed in both CD4 and CD8 subsets, and can also be seen when estrogen is conjugated to BSA, indicating that surface estrogen receptor interaction may be sufficient to mediate the telomerase effect (Effros, manuscript in preparation).
Our data on estrogen effects in T cells in vitro are reminiscent of an earlier study in which we documented the reversal of some of the age-related T cell changes in postmenopausal women treated with hormone replacement therapy (Porter et al., 2001). In another set of preliminary experiments with small molecule activators of telomerase, we have shown a significant enhancement of telomerase activity in T cells from both healthy and HIV-infected persons (Fauce et al., manuscript in preparation). Thus, therapeutic approaches that are based on telomerase modulation would seem to be promising candidates for clinical interventions in the elderly that are aimed at reversing or retarding the process of replicative senescence in T cells. The major question to be addressed is whether the process of replicative senescence, characterized so extensively in cell culture, has any relationship to events within the immune system during normal human aging. As will be described below, this certainly does seem to be the case.
Is T Cell Replicative Senescence Occurring in vivo?
SENESCENT CD8+ T CELLS ACCUMULATE WITH AGE AND WITH CHRONIC INFECTION
The results of our cell culture studies were invaluable in transitioning to in vivo analysis of the possible occurrence of replicative senescence during normal aging. Based on the absence of CD28 expression as a marker of senescence, we compared 20 adults (age 25-69) and 21 healthy centenarians for the presence of cells with a similar phenotype, using flow cytometry. We showed that there was a significant increase (p < 0.01) in the proportion of T cells lacking CD28 expression in the elderly, with some aged individuals having > 60% CD28— T cells within the CD8+ T cell subset, compared to the mean young adult value of <10%. Importantly, since the decrease in the percentage of CD28+ T cells with age was not associated with an alteration in the intensity or standard deviation of mean fluorescence, our analysis indicated that the expression of CD28 was normal on those cells that were registered as being CD28+ (Effros et al., 1994a).
There was an intriguing correlation between the presence of high proportions of CD28~ T cells and the reversal of the youthful ratio between CD4+ and CD8+ T cells. Our data showed that those elderly persons who had the highest proportion of CD28~ T cells also had CD4/8 ratios that reflected an increased proportion of CD8+ T cells (Effros et al., 1994a). Interestingly, altered CD4/8 ratios have also been reported in elderly persons who have large expansions of clonal populations of CD8+ T cells (Posnett et al., 1994). These oligoclonally expanded populations actually resemble cells in senescent cultures in terms of their increased level of activation, the absence of CD28 expressions and their inability to proliferate.
Importantly, cells lacking CD28 expression do not suddenly appear in old age; there is a progressive increase over the lifespan in the proportion of T cells that lack CD28 expression (Boucher et al., 1998). Moreover, chronological age is not unique in its association with T cell replicative senescence. Chronic infection with HIV is also associated with the progressive accumulation over time of CD8+ T cells that are CD28~ (Borthwick et al., 1994). In early reports, this unusual subset was suggested to have arisen as a distinct lineage, possibly related to some unusual aspect of HIV disease pathogenesis. However, based on our cell culture studies, we suspected that these cells were the descendants of cells that previously did express CD28, but that had undergone extensive proliferation and reached the end stage of replicative senescence. To test this idea, we performed telomere analysis on purified cell populations from persons infected with HIV. We first isolated the CD8+ T cells from peripheral blood, then sorted them further into CD28+ and CD28~ populations. Using Southern blot analysis, we showed that the telomere length of the CD28~ population was significantly shorter than that of the CD28+ population from the same donor. In fact, the mean telomere length of the CD8+CD28~ T cells from persons infected with HIV (mean age 43 years) was the same range as that of PBMC isolated from centenarians (Effros et al., 1996).
As further evidence in support of the notion that replicative senescence was actually occurring in vivo, the sorted CD8+CD28~ T cells tested immediately ex vivo showed minimal proliferative ability, even when the stimulation bypassed cell surface-mediated activation signals. Although it is impossible to formally prove that the cells with characteristics of senescence arose by the same mechanism as those in senescent cell culture, the overwhelming similarity between the two cell populations is highly suggestive that replicative senescence is not a cell culture artifact. Indeed, the cell culture model is thus far highly predictive of the functional, genetic, and phenotypic traits of certain CD8+ T cells present in vivo.
Was this article helpful?