A large proportion of the CD8+ T cell pool in persons chronically infected with HIV consists of cells that show features of replicative senescence, an end stage characterized by irreversible cell cycle arrest, multiple genetic and functional changes, and shortened telomeres. The objective of our research was to determine whether constitutive expression of the gene for the human telomerase (hTERT) can prevent senescence-induced impairments in human virus-specific CD8+ T cells, particularly in the context of HIV-1 disease. Our results indicate that hTERT-expressing HIV-specific CD8+ lymphocytes show both an enhanced and sustained capacity to inhibit HIV-1 replication in in vitro coculture experiments, as well as prolonged ability to produce IFN-γ and TNF-α in response to stimulation with HIV-1-derived peptides, as compared with vector-transduced controls. Loss of CD28 expression, the signature change of replicative senescence in cell culture, was retarded in those CD8+ T cell cultures that had high levels of CD28 at the time of hTERT transduction. These findings suggest that telomere shortening may be the primary driving force behind several aspects of CD8+ T cell dysfunction associated with replicative senescence. We also demonstrate reduced accumulation of the p16INK4a and p21WAF1 cell cycle inhibitors in hTERT-transduced lymphocytes, providing a possible mechanism by which stable hTERT expression is able to circumvent the senescence barrier in CD8+ T cells. Given the key role of CD8+ T cell function in controlling a variety of acute and latent viral infections, approaches to retard the functional decrements associated with replicative senescence may lead to novel types of immunotherapy.
A growing body of evidence from both humans and macaque cell depletion studies suggests that the CD8+ T lymphocyte is a crucial effector cell type responsible for controlling viremia during HIV-1 infection (1, 2, 3, 4, 5). The antiviral effects of CD8+ T cells are presumed to be mediated through multiple mechanisms, including lysis of HIV-1-infected cells via release of perforin and granzyme proteases, production of such cytokines as IFN-γ and TNF-α that exert pleiotropic antiviral and immunological effects, and secretion of soluble factors that suppress HIV-1 replication (6, 7). High frequencies of HIV-1-specific CD8+ T cells and vigorous immune responses against HIV-1 have been observed during both the acute and chronic phases of the infection, yet CD8+ T cells ultimately fail to prevent the progression to AIDS in seropositive individuals. Even the currently most successful treatment strategy, highly active antiretroviral therapy does not eradicate the virus, allowing HIV-specific CD8+ T cells to continue being stimulated for many years. Indeed, it is likely that the persistence of suboptimal (i.e., low perforin) HIV-specific CD8+ T cell responses despite prolonged therapeutic viral suppression is associated with continuous proliferation. Thus, the intrinsic cellular program of replicative senescence, which poses a strict limit on proliferation of normal T cells and is associated with shortened telomeres, altered function, and changes in gene expression, has the potential to play a role in HIV disease progression.
It is becoming increasingly clear that, in addition to viral strategies that evade immune recognition, CD8+ T cell defects that develop due to chronic cell turnover also contribute to the failure of the immune system to control HIV-1 (8, 9). Indeed, telomere length studies have demonstrated aberrantly short telomeres in the CD8+ T lymphocyte compartment of HIV+ patients, observations suggestive of a history of prolonged and continuous cellular expansion that could potentially lead to the premature exhaustion of protective antiviral responses and replicative senescence (10, 11). Consistent with this probable link between shortened telomere status and cellular dysfunction are reports documenting maintenance of T cell telomere length in HIV-infected long-term survivors and a more rapid disease progression in older HIV-infected persons (12, 13).
Telomeres are synthesized and maintained by telomerase, a ribonucleoprotein consisting of a catalytic subunit, human telomerase (hTERT),3 and an RNA component, human telomerase RNA (14). Most human somatic cells have no endogenous telomerase activity, so that the telomere shortening that accompanies each cell division is likely to be the major contributing factor to their finite replicative life span (15). By contrast, lymphocytes express telomerase during development, and transiently following activation, a unique property believed to increase their replicative potential and allow for the extensive clonal expansion necessary for an effective immune response (16). Recent studies using a dominant-negative mutant of hTERT clearly demonstrate that endogenous hTERT plays a key role in the in vitro longevity of T lymphocytes (17). Thus, the observation that telomerase activity of CD8+ T cells declines progressively with repeated antigenic stimulation in cell culture and is undetectable by the fourth encounter with Ag may explain the presence of high proportions of memory CD8+ T cells with shortened telomeres in aged persons and individuals chronically infected with HIV-1 (18).
Based on the central role of telomere shortening in the replicative senescence program, genetic manipulation of hTERT has been used to prevent senescence and increase proliferative potential in a variety of cell types (19, 20, 21, 22, 23). Additionally, constitutive expression of hTERT has also been increasingly recognized to prevent certain aspects of senescence-associated cellular dysfunction, such as response to oxidative stress (23, 24, 25, 26, 27, 28). In the present study, we apply this genetic strategy to CD8+ T cells derived from persons infected with HIV to evaluate the long-term effects of hTERT on the phenotypic, cell cycle, cytokine, and functional characteristics of virus-specific lymphocytes. Our results suggest that, in addition to increasing proliferative potential and stabilizing telomere length, gene transduction with hTERT enhances several key CD8+ T cell functions involved in the control of HIV-1.
Materials and Methods
This study included cultures derived from peripheral blood samples of eight HIV-1-seropositive HLA-A2+ participants in the University of California Multicenter AIDS Cohort Study. The study was approved by the University of California Institutional Review Board, and all subjects gave informed consent.
Synthetic peptides used in this study represented the following HIV-1-derived HLA-A*0201-restricted epitopes: Gag p17 epitope SLYNTVATL (SL9), the RT epitope ILKEPVGHV (IV9), and the Env epitope LWVTVYYGV (LV9). Peptides were obtained from the University of California Peptide Synthesis Facility.
Bulk stimulation of fresh PBMCs
To initiate the cultures, freshly isolated PBMCs (4 × 106) obtained by density gradient centrifugation over Ficoll-Hypaque (Pharmacia Biotech, Uppsala, Sweden) were stimulated with 1 × 106 autologous peptide-pulsed PBMCs. Pulsing was performed by incubating PBMCs with each peptide (10 μg/ml) for 90 min, followed by two washes. Lymphocytes were plated in a T25 flask with 15 × 106 irradiated feeder and cultured in Yssel’s medium (Gemini Bio-Products, Calabasas, CA) supplemented with l-glutamine and penicillin and streptomycin (both from Mediatech, Herndon, VA). rIL-2 (Roche, Basel, Switzerland) was added on day 4 and twice per week thereafter. After 10–14 days, the cells were tested for specificity in a standard 51Cr release CTL assay using the HLA-A2+ T2 cell line pulsed with SL9, IV9, or LV9 peptides.
Purification and establishment of CTL-enriched polyclonal CD8+ cell lines
CD8+ T cells from bulk PBMC cultures recognizing the SL9, IV9, or LV9 peptides were purified by magnetic cell sorting (Miltenyi Biotec, Auburn, CA) and plated at 4 × 106 cells/well in a 12-well plate in Yssel’s medium supplemented with 150 IU/ml IL-2. Cells were restimulated every 3–4 wk with irradiated peptide-pulsed autologous EBV-transformed B lymphoblastoid cells (B-LCLs) to enrich for CTLs. Viable cells were counted by trypan blue exclusion every 1–2 wk. Population doublings (PDs) of the cell lines were determined from the point of transduction and calculated according to the formula: PDs = log2(final cell concentration/initial cell concentration).
Cumulative PDs of a cell line is the summation of all prior doublings in response to previous restimulations.
Retrovirus and gene transduction method
The amphotrophic PA317 packaging cell lines containing the stably transfected pBABE retroviral vector with and without hTERT cDNA were provided by Geron (Menlo Park, CA). Supernatants containing retrovirus were pooled from plates at 40–60% confluence, filtered through a 0.45-μm filter, and mixed with DMEM. CD8+ T cells were plated on RetroNectin (Takara Shuzo, Shiga, Japan)-coated 24-well plates (Falcon) 2 days after stimulation and transduced with fresh retroviral supernatant containing hTERT or empty vector for 12 h. Infection was repeated a second time for optimal gene transfer. Transduced cells were then selected in pretitrated concentrations of puromycin (Sigma-Aldrich, St. Louis, MO) for 5 days. The efficiency of transduction varied from 2 to 8%.
Telomere length assay
Measurement of telomere restriction fragment length (TRF) was performed, as previously described (29). Briefly, genomic DNA was isolated from 2 × 106 CD8+ T cells using DNAzol (Molecular Research Center, Cincinnati, OH). Each DNA sample was digested with RsaI and HinfI (New England Biolabs, Beverly, MA) and electrophoresed through a 0.5% agarose gel. The gel was then dried for 1 h at 60°C, denatured, and subsequently hybridized overnight to a [γ-32P]ATP end-labeled telomeric oligoprobe (TTAGGG)4. Autoradiography was performed for at least 3 days. Mean TRF lengths were measured from phosphor imager (Packard Instrument, Downers Grove, IL) scans using ImageQuant software.
Flow cytometric analysis
mAbs used for surface staining were all purchased from BD Biosciences (San Diego, CA). To surface stain cells, Abs were added for 15 min at room temperature, after which the cells were washed in PBS, fixed in 2% paraformaldehyde, and stored at 4°C until acquisition on a FACSCalibur. For staining with HLA-A*0201 peptide-MHC tetramers, PE-labeled tetramer containing the SL9 peptide and allophycocyanin-labeled tetramer containing the IV9 peptide were obtained from the National Institute of Allergy and Infectious Diseases AIDS Reagent Program Tetramer Core Facility. LV9-bound Cy5-labeled tetramer was purchased from ProImmune (Oxford, U.K.). Pretitrated amounts of the appropriate tetramer were used to stain 5 × 105 lymphocytes for 30 min at 4°C, after which cells were washed and fixed in 2% paraformaldehyde and stored at 4°C until flow cytometry analysis.
Intracellular staining for cytokines
Analyses of intracellular IFN-γ and TNF-α productions were performed by stimulating 2 × 105 CD8+ T cells with peptide-pulsed or unpulsed autologous B-LCLs for 6–12 h in the presence of 1 μg/ml monensin (BD Biosciences). Cells were then washed, surface stained with anti-CD8 and anti-CD3 at 4°C for 20 min, fixed, and permeabilized with Cytofix/Cytoperm buffer (BD Biosciences). Intracellular staining was performed for 15 min using pretitrated concentrations of anti-IFN-γ or anti-TNF-α mAbs (BD Biosciences). Cells were then washed and stored in 2% paraformaldehyde at 4°C until flow cytometry analysis was performed.
Chromium release assay
Standard chromium release assays were performed using hTERT- and vector-transduced CD8+ T lymphocytes as effector cells. Briefly, autologous B-LCLs were labeled with 100 μCi of Na251CrO4 in the presence or absence of synthetic peptides at 10 μg/ml for 1.5 h. After three washes, cells were plated in a 96-well U-bottom plate at 10,000/well. Effector cells were then added at the indicated ratios for a 4-h incubation, after which supernatants were harvested and radioactivity measured with a gamma counter. Spontaneous release was determined in the absence of effector cells, and maximal release was determined in the presence of 2% Triton X-100 (Sigma-Aldrich). All wells were run in triplicates. The percent specific lysis was calculated according to the formula: ((experimental release − spontaneous release)/(maximum release − spontaneous release)) × 100.
Western blot analysis
Frozen pellets of hTERT- and vector-transduced cells were lysed for 30 min on ice in modified radioimmunoprecipitation buffer (50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM EDTA, and 0.5% Nonidet P-40) supplemented with proteinase inhibitors. Samples were then centrifuged for 20 min at 4°C, and aliquots of the supernatants were stored at −80°C until further use. Total protein concentration was determined using the Lowry method (Sigma-Aldrich). Proteins were separated on 7.5 or 15% SDS-PAGE (Bio-Rad, Richmond, CA), transferred to polyvinylidene difluoride membrane (Amersham, Arlington Heights, IL), and blocked overnight in PBS containing 10% dried milk and 0.1% Tween 20. Membranes were then probed with Abs against p16INK4a (BD Pharmingen, San Diego, CA), p21WAF1 (Oncogene Science, Carpenteria, CA), and actin (Calbiochem, La Jolla, CA). After washing, proteins were detected with HRP-conjugated secondary Abs (Amersham) and developed with ECL reagents (ECLplus) (Amersham), according to the manufacturer’s instructions. Densitometry on ECL films was performed using an Amersham Biosciences densitometer (Amersham Biosciences, Sunnyvale, CA).
Inhibition of HIV-1 replication by CD8+ T cells
The ability of hTERT-transduced CD8+ T cells to suppress viral replication was determined, as described previously (30). Briefly, T1 cells were acutely infected with HIV-1 IIIB strain at a multiplicity of infection of 0.01. The infected T1 cells were then mixed with telomerase- or vector-transduced CD8+ T cells in 24-well flat-bottom plates (Falcon) at a ratio of 5 × 105 T1 cells to 2.5 × 105 CD8+ lymphocytes in a final volume of 2 ml. At the indicated time points, 1 ml of medium was removed for measurement of HIV-1 replication by p24 Ag ELISA and replaced with fresh medium.
Expression of hTERT retards telomere shortening
We recently reported that CTL-enriched polyclonal CD8+ T cell lines from HLA-A2+ HIV-1-infected individuals transduced with hTERT cDNA exhibited significantly increased replicative life spans (31). We continued to expand some of these cell lines by periodic restimulations with HIV peptides, and at the time of this submission, hTERT-expressing CD8+ T cells from donor 1 have undergone 66 PDs, which is nearly three times the number at which the control cultures reached senescence (Fig. 1⇓A). Similarly, hTERT-transduced cells from donor 4 continued to show robust proliferation well beyond the PD level at which the control cells ceased dividing. Despite the dramatic increase in proliferative activity, there was no evidence of transformation; the hTERT-expressing cells maintained a normal karyotype, and required antigenic stimulation and IL-2 for continued proliferation (data not shown). The hTERT+ CD8+ T cells maintained high levels of telomerase activity throughout the culture period (data not shown), suggesting that the increased proliferative potential might be due to effects of the exogenous hTERT gene on telomere length. To test this possibility, cryopreserved lymphocytes collected at different PDs were evaluated in parallel for telomere length. Results from a representative experiment are provided in Fig. 1⇓B, which shows the pattern of TRF length in hTERT- and vector-transduced cells from donor 1. These data demonstrate that during the initial period following transduction, there was an increase in telomere length in both cultures, an observation consistent with the well-documented high levels of endogenous telomerase activity induced by antigenic restimulation (18). However, consistent with our previous demonstration that the level of endogenous telomerase activity induced in memory T cells declines with repeated antigenic stimulation, the vector-transduced culture underwent progressive telomere shortening, reaching a mean TRF of 2.4 kb at senescence. By contrast, the rate of telomere shortening was retarded in the hTERT-expressing cells, and by 45 PDs, the mean TRF had become stabilized at ∼3.0 kb.
hTERT prevents up-regulation of cell cycle inhibitors
Replicative senescence in other cell types has been correlated with elevated levels of both p16INK4A (p16) and p21WAF1 (p21) cyclin-dependent kinase inhibitors, suggesting an important role for both proteins in mediating the senescence program (32, 33, 34). We therefore compared the levels of p16 and p21 proteins in the hTERT- and vector-transduced CD8+ T cell lines. Data from Western blot analysis of the cell lines from donor 1 are shown in Fig. 2⇓, in which we compared the levels of expression of p16 and p21 in early (PD 4) and later passage (PD 21) quiescent hTERT- and vector-transduced cells. The results show that p16 and p21 accumulate in CD8+ T lymphocytes (vector transduced) with increasing numbers of cell divisions, and that expression of hTERT delayed the accumulation of both proteins. As early as 4 PDs after gene transfer, the vector-transduced cells expressed 1.5 times more p16, and 1.6 times more p21 than the hTERT-transduced cells. By 21 PDs, just a few PDs before the vector-transduced cells senesced, the levels of p16 and p21 were 4.7 and 7.9 times greater than those of the hTERT-transduced cells, respectively. hTERT- and vector-transduced cell lines analyzed from two other donors varied with respect to the degree of change in p16 and p21 expression, but in all cases, the level of one or both senescence-associated proteins was significantly reduced in the cells constitutively expressing hTERT.
hTERT retards the loss of CD28 expression
A progressive decline in the proportion of CD8+ T cells that express the CD28 costimulatory molecule has been extensively documented during aging and in disease states, and also in cell culture with increasing PDs (35). To investigate whether hTERT transduction has an effect on CD28 expression, we performed a longitudinal analysis of the CD28 expression profiles in the polyclonal CD8+ T cell lines using anti-CD28 mAb and flow cytometry (Fig. 3⇓). In the two CD8+ T cell cultures that had high proportions of CD28+ cells at the time of transduction (donors 1 and 3), constitutive expression of hTERT was associated with a retardation in the loss of cell surface CD28. Indeed, at a time point when all of the cells in their corresponding control cultures no longer expressed CD28, 35% of the hTERT-transduced cells from donor 1 and 50% of the hTERT-expressing lymphocytes from donor 3 were still CD28+. Moreover, in contrast to the control cells in which complete loss of CD28 is associated with cell cycle arrest and replicative senescence, the hTERT-transduced lines with expanded life span remained capable of further proliferation even when they became totally CD28 negative. In the case of the cell lines from donor 5 in which gene transfer was performed when the culture already contained a high proportion (60%) of cells that were CD28 negative, no effect of hTERT on the rate of loss of CD28 expression was observed. These data suggest that hTERT gene transfer is capable of retarding the loss of CD28 expression in CD8+ T cells from HIV-1-infected persons as long as there are sufficient numbers of CD28+ lymphocytes at the time of transduction.
hTERT-expressing CD8+ T cells show enhanced inhibition of HIV-1 replication in vitro
CD8+ T lymphocytes produce soluble factors that have been shown to inhibit HIV-1 replication in vitro and have been hypothesized to contribute to control of the virus in vivo (36, 37). Previously, we reported that an hTERT-transduced IV9-specific CTL clone showed markedly enhanced HIV-1-suppressive activity in vitro at a time point 8 wk after gene transduction (31). In this study, we evaluated uncloned CD8+ cell lines that have been enriched for virus-specific CTLs through repeated stimulation with peptides at multiple time points after transduction to determine the long-term kinetics of this HIV-1-suppressive activity. hTERT- and vector-transduced CD8+ lymphocytes at several stages of their in vitro proliferative life spans were cocultured with the acutely HIV-1-infected HLA-A2 cell line T1 cells. Supernatants from these cultures were collected at 2-day intervals and were then assayed by ELISA for p24 levels. Inhibition data from one of three donors analyzed (Fig. 4⇓) show that the hTERT+ cells had superior ability to inhibit viral replication at every time point tested. Indeed, the hTERT-transduced cultures maintained levels of inhibition that ranged from 68 to 82%, even as late as 50 PDs, in comparison with the more modest inhibitory levels of 35–45% of the vector-transduced cultures just before their senescence at 22 PDs. These results, together with data from other cultures tested, indicate that the ability to inhibit HIV-1 replication is a stable feature of HIV-1-specific CD8+ T cells, regardless of proliferative history. However, in all donors tested, constitutive expression of hTERT enhanced antiviral suppressive activity, consistent with our previous data on an HIV-specific CTL clone.
hTERT retards the loss of IFN-γ and TNF-α production by HIV-1-specific CD8+ T cells
We compared the capacity of hTERT- and vector-transduced HIV-1-specific CD8+ T cells to produce IFN-γ and TNF-α in response to stimulation with viral peptides. Cells at multiple stages of their in vitro proliferative life span were restimulated with peptide-pulsed APCs and evaluated 6 h later by flow cytometry for intracellular cytokines (Fig. 5⇓A). The data shown, which is representative of cultures from three donors examined, demonstrate that hTERT-transduced lymphocytes have a prolonged capacity to elaborate both IFN-γ and TNF-α in response to Ag-specific stimulation. At PD 24, which was the last time point we were able to examine the vector-transduced culture before it senesced, only 3–5% of the lymphocytes responded to stimulation by producing cytokines, whereas 15 and 19% of the cells in the hTERT+ culture still actively produced IFN-γ and TNF-α, respectively. The decline in cytokine production was not due to altered kinetics of cytokine expression in later passage cells, as no IFN-γ or TNF-α production was detected by intracellular cytokine staining performed at the 8-, 10-, and 12-h time points after stimulation (data not shown). Moreover, the decline in cytokine production by the vector-transduced cells was not due to loss of Ag-specific lymphocytes, as tetramer staining demonstrated that 19.1% were Gag (SL9)+, 16.1% were Pol (IV9)+, and 7.8% were Env (LV9)+ at the time point when the cytokines ceased being produced (Fig. 5⇓B). Indeed, our analysis indicates that despite the prolongation of Ag-induced cytokine responses in the hTERT cultures during the period between PD 6 and 24, there was nevertheless a decrease in the percentage of cytokine-producing cells (40 to 11% for IFN-γ, and 42 to 17% for TNF-α). This functional decline occurred in parallel with an increase in tetramer-binding cells from 24 to 43% during the same time frame. Similar enrichment of tetramer-binding cells accompanied the progressive decline of Ag-induced IFN-γ and TNF-α production in the hTERT cultures of two other donors examined. Taken together, these data demonstrate that hTERT transduction delays, but does not prevent the progressive loss of Ag-induced cytokine production by HIV-specific CD8+ T cells. Importantly, because the levels of exogenous telomerase activity (assayed by the telomerase repeat amplification protocol assay) (31) were maintained throughout the entire culture period, our data demonstrate a dissociation between hTERT effects on proliferation and cytokine function.
hTERT does not prevent the loss of cytolytic function of HIV-1-specific CD8+ T cells
Finally, we examined the effect of ectopic hTERT expression on the cytolytic effector function of HIV-1-specific CD8+ T cells. Before transduction, chromium release assays on in vitro expanded PBMCs from our subjects demonstrated robust SL9-, IV9-, or LV9-specific CTL responses (data not shown). The same high level of cytotoxicity was observed in the purified CD8+ T cells from the same individuals after transduction with hTERT cDNA or vector only, showing that CD8 T+ lymphocytes are cytolytic when tested in the absence of other cell types, and that hTERT does not alter the specificity of the cells as shown in the representative experiment in Fig. 6⇓A. However, with increasing PDs, cytotoxic activity of HIV-1-specific CTLs in both hTERT-expressing and control cultures from all of our donors steadily declined with parallel kinetics to the point of being undetectable. At PD 25, the noncytolytic hTERT+ culture from our representative donor still contained substantial proportions of HIV-1-specific cells, as indicated by tetramer binding (Fig. 6⇓B). However, the high proportion of perforin-expressing cells (55%) present in this culture at PD 4, a time point of high lytic activity, was almost completely absent from the same culture at PD 25, providing an explanation for the failure of CTL function (Fig. 6⇓C). Altogether, these findings suggest that the loss of HIV-1-specific CTL activity with increasing culture age is primarily due to impaired expression of perforin, and that ectopic expression of hTERT does not maintain CTL function.
The extensive analysis of long-term cultures of virus-specific CD8+ T cells established from multiple HIV-1-infected donors described in the present study has provided novel insights into the effects of ectopic telomerase expression on a variety of functional, cell cycle, and genetic characteristics of these cells. The most striking consequence of constitutively expressing telomerase was the enhanced viral suppressive activity of the CD8+ T cells. The effects of the soluble inhibitory factors released by CD8+ T cells on suppressing HIV-1 replication have been well established in vitro, and have been suggested to play an important role in HIV-1 host defense (36, 37, 38). Thus, augmentation of the suppressive activity through telomerase transduction may lead to new forms of immunotherapy, particularly during the later stages of disease when CD8+ T cell cytolytic activity has been reported to wane (reviewed in Ref. 8).
The present study underscores the notion that the proliferative block is only one facet of the replicative senescence cellular program, which is now known to also involve a variety of fundamental changes in gene expression and functional characteristics (15). Constitutive expression of telomerase, which retards/prevents replicative senescence, would therefore be predicted to influence not only the rate of telomere shortening and proliferative capacity, but also functional traits as well. Indeed, our results show that stable hTERT expression in HIV-specific CD8+ T cells derived from HIV-infected individuals, in addition to increasing longevity and stabilizing telomere length, also prevents the increase in cell cycle inhibitors, prolongs the period of secretion of cytokines associated with antiviral immune responses, and delays the loss of CD28 expression resulting from chronic Ag stimulation in vitro. These observations are consistent with reports on telomerase manipulation in other cell types, in which hTERT has been increasingly documented to affect multiple cellular functions. For example, constitutive hTERT expression enhances the ability of endothelial cells to form microvascular structures, and restores the functional capacity of senescent fibroblasts in a dermal reconstitution model (22, 25). Ectopic expression of hTERT also enhances the differentiation potential and bone-forming ability of human bone marrow stromal cells (26, 27). Similarly, exogenous telomerase prevents chromosomal instability and spontaneous immortalization of Li Fraumeni Syndrome cells (39). Finally, our own previous study showed that hTERT transduction of a selected HIV-specific CTL clone enhanced cytotoxicity and antiviral suppressive activity (31). The results of the present study showing pleiotropic and possibly beneficial effects of constitutive telomerase expression in HIV-specific CD8+ T cells derived from persons chronically infected with HIV, therefore, suggest that the functional decline associated with chronic immune stimulation may be amenable to therapeutic strategies that modulate the process of replicative senescence.
Telomere length in normal somatic cells, including CD8+ T cells, shortens with increasing PDs in cell culture (18, 40). Due to chronic activation by virus, this process of telomere shortening is actually accelerated in the CD8+ T cells from HIV-infected individuals, resulting in CD8+ lymphocytes with abnormally short telomere lengths (10). In the present study, we show that following an initial decrease in telomere length in the hTERT-expressing CD8+ T cell cultures, the rate of loss of telomeric sequences was retarded compared with the vector control, and telomere length was ultimately stabilized. This stabilization of telomeres is consistent with previous observations in hTERT-immortalized melanoma-specific CTLs (21).
A longstanding theory regarding the mechanism by which short telomeres signal senescence is that they generate a DNA damage signal that activates p53, which, in turn, is associated with induction in expression of the cell cycle inhibitory protein p21 (41, 42). Similarly, up-regulation of p16 expression can be induced by DNA damage (43). Indeed, in senescing cultures of fibroblasts, keratinocytes, and T lymphocytes, levels of p16 and p21 proteins or their corresponding mRNAs are significantly higher than in younger and more proliferation-competent cells (32, 33, 34, 44). Our observation of a diminished accumulation of p16 and/or p21 in the hTERT-transduced CD8+ T cell cultures offers further support for the notion that short telomeres are sensed as DNA damage, and also provides insight into the mechanism by which stable expression of telomerase is able to confer life span extension. The simplest scenario is that the telomerase stabilizes telomere length, and in that way, the expression of p16 and/or p21 is maintained at a low level, allowing the cells to remain in cycle. Alternatively, based on the ability of telomerase to heal damaged chromosomes in lower eukaryotes, it is possible that telomerase is acting in T cells to prevent accumulation of random oxidative DNA damage, and by this mechanism prevents the increase of p16 and p21 expression (45, 46). Current studies are directed at distinguishing between these two possibilities.
Replicative senescence in vitro is also associated with complete loss of CD28 cell surface and gene expression (47). The prolonged retention of CD28 on hTERT-transduced cell lines, therefore, suggests that loss of telomere length is somehow involved in the down-regulation of CD28 expression. The fact that stable expression of telomerase delays the loss of CD28 expression is a novel observation, and is consistent with the recently described association between loss of telomerase inducibility in CD8+ lymphocytes and the loss of cell surface CD28 expression (18). Taken together, these findings suggest that in CD8+ T cells, there is an underlying relationship between CD28 expression, telomere length, and telomerase activity. Even so, the maintenance of telomere length in the cells transduced with hTERT was ultimately unable to prevent the eventual progressive decline in CD28 expression. Based on our finding that the prolonged expression of CD28 was observed only in CD8+ cultures that were predominantly CD28+ at the time of hTERT introduction, it is possible that the efficacy of this gene manipulation technique is dependent on the age of the culture. Indeed, just as there was a lag between the introduction of hTERT and telomere length stabilization, it is possible that gene transduction at an earlier time point during the culture life span may be more efficient in retarding/preventing the CD28 expression changes. The use of lentivirus vectors, which can infect cells that are quiescent, and can therefore introduce hTERT at the earliest stage of the culture life span, may clarify this issue.
In HIV-1 disease, shortened telomeres in the CD8+ population have been suggested to play a role in the decline of virus-specific responses, leading to progression to AIDS (48, 49). Interestingly, a large-scale aging study on the relationship between telomere length and infection has recently demonstrated that in persons over 60, the mortality rate from infectious disease was 8 times higher for individuals in the bottom 25% of the lymphocyte telomere length distribution than for persons in the top 75% (50). Consistent with these findings and with the notion that HIV disease constitutes accelerated immunological aging (49), the present longitudinal analysis of the functional behavior of virus-specific CD8+ T cells derived from HIV-infected persons revealed specific impairments in antiviral activity associated with telomere shortening and senescence. With increasing number of PDs, the vector-transduced cell lines showed diminished numbers of HIV-1-specific lymphocytes that produced the effector cytokines IFN-γ and TNF-α in response to antigenic stimulation. In fact, even before the actual end stage of replicative senescence, there was almost a complete absence of cells capable of producing these cytokines, despite the presence of a substantial number of tetramer-positive cells. This finding is reminiscent of clinical studies of patients with late-stage disease, who had fewer cells that secreted IFN-γ when stimulated with HIV-1 (51, 52). Cultures that were transduced with hTERT showed a prolonged retention of significant numbers of Ag-reactive cells that produced IFN-γ and TNF-α in response to stimulation. Because of the pleiotropic antiviral effects of both IFN-γ and TNF-α, our findings suggest that ectopic expression of telomerase in virus-specific CD8+ T might be useful as a tool to increase the functional longevity of these cells.
The current study on bulk populations of CD8+ T cells yielded results on cytotoxicity that differ from our previous findings on an IV9-specific CD8+ T cell clone, although in both cases the T cells were derived from HIV-1-infected donors (31). In the present set of experiments, hTERT was unable to prevent the impaired cytotoxic capacity of the bulk-cultured senescent HIV-1-specific CTLs. By contrast, the IV9-specific CTL clone showed maintenance of cytotoxic potential even at senescence, and, further enhancement by hTERT expression. The most likely explanation for the different outcomes is that the clone was originally screened and selected based on vigorous cytotoxic activity, and may therefore not necessarily be representative of the majority of CD8+ T cells in HIV-1-infected persons. It is also possible that intrinsic differences between the particular HIV-infected donors from whom the bulk cultures and T cell clones were derived may contribute to the disparate cytotoxicity results.
In conclusion, in the present study, we have shown that transduction of the gene for the catalytic component of telomerase, hTERT, in CD8+ T cells from HIV-1-infected persons delayed or partially inhibited a number of the alterations in function and phenotype that developed as a consequence of in vitro aging and subsequent entry into replicative senescence. Our results suggest that telomere shortening is a major driving force behind the CD8+ cellular dysfunction observed in late-stage HIV-1 disease, and that the senescence-associated changes may ultimately lead to loss of control over the virus. The study thus supports the notion that replicative senescence and premature immunological aging might play a role in the functional immune deficiency induced by HIV-1 (10, 49). Our results are highly relevant to the ongoing efforts to develop therapeutic vaccines aimed at stimulating CD8+ T cell responses to alleviate symptoms after HIV-1 infection, as well as to current immunotherapy protocols based on adoptive transfer of in vitro expanded virus-specific CD8+ T cells. The ramifications and functional consequences of replicative senescence may significantly alter the effectiveness of such approaches for fighting HIV-1. For this reason, the demonstrated efficacy of ectopic hTERT expression in modulating a variety of senescence-associated changes in virus-specific CD8+ T cells from HIV-1-infected persons represents a potential new strategy for HIV-1 therapy that might result in prolonged protection by HIV-1-specific CD8+ T cells.
We thank Drs. Kenneth Dorshkind and Otto Yang for reviewing the manuscript. We are grateful to Geron for supplying the hTERT and control constructs used in this study.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
↵1 The work was supported by National Institutes of Health Grants AI47665 (to R.B.E.) and AG 05920 (to M.D.). R.B.E. holds the Elizabeth and Thomas Plott Endowed Chair in Gerontology.
↵2 Address correspondence and reprint requests to Dr. Rita B. Effros, Department ofPathology and Laboratory Medicine, David Geffen School of Medicine, 10833 Le Conte Avenue, Los Angeles, CA 90095-1732. E-mail address:
↵3 Abbreviations used in this paper: hTERT, human telomerase; B-LCL, B lymphoblastoid cell; PD, population doubling; TRF, telomere restriction fragment.
- Received March 17, 2004.
- Accepted September 8, 2004.
- Copyright © 2004 by The American Association of Immunologists