Telomerase insufficiency in rheumatoid arthritis

Abstract

In rheumatoid arthritis (RA), chronically stimulated T lymphocytes sustain tissue-destructive joint inflammation. Both naïve and memory T cells in RA are prematurely aged with accelerated loss of telomeres suggesting excessive proliferative pressure or inadequate telomeric maintenance. Upon stimulation, RA naïve CD4 T cells are defective in up-regulating telomerase activity (P < 0.0001) due to insufficient induction of the telomerase component human telomerase reverse transcriptase (hTERT); T cell activation and cell cycle progression are intact. Telomerase insufficiency does not affect memory T cells or CD34 hematopoietic stem cells and is present in untreated patients and independent from disease activity. Knockdown of hTERT in primary human T cells increases apoptotic propensity (P = 0.00005) and limits clonal burst (P = 0.0001) revealing a direct involvement of telomerase in T cell fate decisions. Naïve RA CD4 T cells stimulated through the T cell receptor are highly susceptible to apoptosis, expanding to smaller clonal size. Overexpression of ectopic hTERT in naïve RA T cells conveys apoptotic resistance (P = 0.008) and restores proliferative expansion (P < 0.0001). Telomerase insufficiency in RA results in excessive T cell loss, undermining homeostatic control of the naive T cell compartment and setting the stage for lymphopenia-induced T cell repertoire remodeling. Restoring defective telomerase activity emerges as a therapeutic target in resetting immune abnormalities in RA.

In concert with other cell types, T lymphocytes hold a critical position in rheumatoid arthritis (RA), mediating multiple pathogenic pathways (1, 2). In the inflamed synovium memory/effector CD4 T cells form organized ectopic lymphoid structures (3). RA memory CD4 T cells are characterized by increased CD28⁻ loss, oligoclonal populations, and contraction in T cell receptor (TCR) diversity (4), possibly reflecting chronic antigenic stimulation. However, T cell dysfunction in RA extends to the naïve CD4 T cell pool, implicating antigen-independent mechanisms in deviating T cell function. Unprimed CD4 CD45RA⁺ T cells from RA patients have shortened telomeres and, when driven into clonal expansion, fail to reach similar clonal sizes as controls (5, 6). Telomeric shortening in naïve T cells has been associated with the HLA-DR4 haplotype, the major genetic risk factor predisposing to RA (6, 7).

Mechanisms underlying the premature aging of the RA T cell pool are not understood. Prime candidates include chronic stimulation by arthritogenic antigens or continuous activation in a cytokine-rich microenvironment. Even in healthy individuals, progressive age is associated with the loss of immunocompetence. Akbar and colleagues have emphasized the role of T cell proliferative capacity to sustain memory function and have implicated telomere loss in losing immune memory (8). Memory CD4 T cells encountering antigen in the skin incur telomere erosion, partially due to type I IFN-mediated suppression of telomerase induction (9). Persistence of virus-specific CD8 T cells requires rescue from apoptosis and up-regulation of telomerase activity (10). As memory T cells differentiate, telomerase activity progressively declines (11). Decreased AKT signaling has been implicated in diminishing telomerase activity in end-differentiated CD8 CD28⁻CD27⁻ T cells (12). Thus, T cell memory is closely linked to preserving telomeres.

In RA patients, telomere attrition in CD4 T cells is accelerated (5, 6) by either proliferative stress or insufficient telomeric repair. Telomeres are not critically short and should not force T cells into cell cycle arrest or cell death. Nevertheless, defects in telomeric maintenance could affect broader cellular functions. An important consequence of telomere shortening is the induction of replicative senescence, a state in which the cell is viable but prohibited from further cell divisions (13).

Telomeric lengthening and maintenance is facilitated by telomerase, an enzyme composed of a catalytic protein unit known as human telomerase reverse transcriptase (hTERT) and an RNA template complementary to the telomeric DNA (hTR) (14). In most tissues, telomerase is strongly suppressed, but in T cells telomerase activity is dynamically regulated and coincides with periods of cellular expansion (15, 16). Telomerase induction allows for telomere elongation, translating into lengthening of life span (17–20). How ongoing telomeric maintenance affects T cell proliferation and function before the state of a short and dysfunctional telomere is reached is currently not understood.

Here we report that in RA, naïve CD4 T cells fail to up-regulate telomerase when primed through the TCR. Knockdown of hTERT in primary human T cells revealed a direct effect on cell survival, with telomerase insufficiency rendering T cells apoptosis susceptible. Naïve RA CD4 T cells were prone to die when driven into clonal expansion, impairing their clonal size. Apoptosis during this early phase in the T cell life cycle was Fas independent and mediated through the mitochondrial pathway. Ectopically expressed hTERT repaired apoptotic propensity of RA T cells. In essence, the enzyme telomerase is critically involved in determining life/death decisions in proliferating CD4 T cells. Telomerase insufficiency in RA T cells may lead to a defect in the homeostatic regulation of the T cell pool.

Results

RA Naïve CD4 T Cells Fail to Up-Regulate Telomerase Activity.

To assess telomeric maintenance and repair in naïve CD4 T cells, we examined the kinetics of telomerase induction during priming. CD4⁺CD45RO⁻ and CD4⁺CD45RA⁻ populations were stimulated with bead-immobilized anti-CD3/CD28. In both T cell populations, TCR triggering up-regulated enzyme activity, but did so distinctly more in the naïve than in the memory population. Induction of telomerase activity was similar in memory T cells from patients and controls; the naïve T cell populations differed significantly (Fig. 1A). Maximal enzyme activities in RA T cells on day 3 reached only 60% of the levels in controls (P = 0.00000005) and continued to be significantly lower on day 6 (P = 0.01). Telomerase activity on day 3 after stimulation correlated with telomeric lengths in fresh CD4 T cells (supporting information (SI) Fig. S1). In contrast, circulating CD34 hematopoietic stem cells (HSC) telomerase activities were significantly higher from RA patients than controls, indicating cell specificity in telomerase deficiency (Fig. 1B). Stimulation-induced telomerase activities correlated closely with hTERT transcript concentrations (Fig. 1C), suggesting that the activity of this enzymatic system in T cells is primarily regulated through hTERT transcription.

Fig. 1.

Impaired induction of telomerase activity in naïve CD4 T cells in RA. (A) Naïve (CD45RO⁻) and memory (CD45RA⁻) CD4 T cells were isolated from RA patients (n = 38, black bars) and age-matched healthy controls (n = 26, gray bars) and stimulated with anti-CD3/CD28-coated beads. Telomerase activity was measured by TRAP assay on days 0, 3, and 6 following stimulation. Data are presented as means ± SEM. (B) CD34⁺ cells were purified from the blood of RA patients (n = 10) and age-matched healthy controls (n = 10), and telomerase activity was determined by TRAP assay. (C) hTERT transcripts were quantified by qPCR on day 3 after T cell stimulation, normalized to 18S ribosomal RNA, and correlated with enzyme activity (r = 0.87, P < 0.001). (D) Naïve (CD45RO⁻) CD4 T cells were isolated from 6 RA patients (dashed lines) and 6 controls (solid lines) and stimulated with anti-CD3/CD28-coated beads. Cell surface expression of CD69, CD71, and CD25 was analyzed by flow cytometry 0, 4, 12, 24, 48, and 72 h after stimulation. Frequencies of positive cells are given as means ± SEM.

To exclude that the failure to up-regulate hTERT represented a global activation defect in RA T cells, induction of activation markers was quantified (Fig. 1D). The kinetics of CD69, CD71, and CD25 induction after anti-CD3/CD28 stimulation were identical in RA and control naïve CD4 T cells. To rule out contamination of CD45RA⁺CD4⁺ T cells with memory cells that had converted back to the CD45RA⁺ phenotype, we analyzed CD4CD45RA⁺ and CD4CD45RA⁻ T cells for the expression of CCR7 and CD28 (Fig. S2). Almost all CD4CD45RA⁺ cells in patients and controls were positive for CCR7, identifying them as naïve T cells. CD4CD45RA⁺CD28⁻ T cells were explicitly infrequent.

Telomerase Deficiency Is Independent from the Inflammatory Milieu and Present in Vivo.

To examine whether the RA inflammatory milieu affects hTERT transcription, telomerase activities in naïve CD4 T cells were compared in patient subsets. Neither disease activity nor disease duration influenced telomerase induction (Fig. 2A). Most remarkably, patients with newly diagnosed and untreated RA had suppressed telomerase levels, indistinguishable from those in treated patients (Fig. 2A). Thus, the defect in up-regulating telomerase is unlikely an immediate consequence of inflammation or treatment. To understand whether lacking telomerase induction was relevant in vivo, we measured telomerase activity in freshly harvested samples. Telomerase activity was mostly restricted to CD3⁺CD4⁺CD71⁺ T cells (Fig. 2B). When adjusted for the frequencies of CD71⁺ T cells, sorted naïve CD4 T cells from RA patients had significantly lower telomerase activity compared to age-matched controls (P = 0.01) (Fig. 2C). Enzyme activities in freshly isolated memory CD4 T cells were generally low.

Fig. 2.

Telomerase insufficiency is independent from disease activity and treatment and present in vivo. (A) Patients described in Fig. 1 were assigned to clinical categories [active vs. inactive disease, treated vs. untreated disease, early (≤ 2 years duration) vs. established disease (>2 years)], and telomerase activity was measured in naïve CD4 T cells 3 days after TCR-mediated stimulation by TRAP assay. Activities are expressed as arbitrary units (A.U.) per 1,000 cells after normalization to an internal standard. Results are shown as means ± SEM. (B) Telomerase-expressing T cell populations in freshly harvested blood samples were identified by isolating CD3⁺CD4⁺CD71⁻ and CD3⁺CD4⁺CD71⁺ subsets. Results are shown as means ± SEM per 1000 cells. (C) Blood samples were collected from RA patients (n = 11) and controls (n = 9), and naïve and memory CD4 cells were purified. Telomerase activity measurements in each T cell subset were normalized to the frequency of CD71⁺ cells and are given as means ± SEM per 20,000 cells.

Knockdown of Telomerase Activity Renders Naïve CD4 T Cells Apoptosis Susceptible.

To assess the functional impact of diminished telomerase activity, hTERT transcription in naïve CD4 T cells was suppressed through siRNA (Fig. 3). Twenty-four hours later, T cells were primed with anti-CD3/CD28-coated beads. RNA interference reduced hTERT transcripts to 50–60% (Fig. 3A) and enzyme activity to about half of the levels in control T cells (Fig. 3B), mimicking conditions in RA T cells. T cells with impaired telomerase induction failed to expand to similar clonal size (Fig. 3C), and survival rates were significantly lower than in control transfected cells (P = 0.04) (Fig. 3D). The shortfall of clonal burst could be attributed to increased apoptosis. Frequencies of propidium iodide (PI)⁺ (Fig. 3E, P = 0.00005) and TUNEL⁺ (Fig. 3F, P = 0.00005) cells were significantly higher in hTERT-deficient T cells. Cell division over the short period of 4 days was not sufficient to cause measurable telomeric loss (Fig. 3G). Thus, the role of telomerase in influencing apoptotic susceptibility during clonal expansion of naïve T cells was independent from telomeric length, suggesting additional functions of this enzyme.

Fig. 3.

Knockdown of hTERT in naïve CD4 T cells increases apoptotic susceptibility and restricts clonal expansion during the priming response. CD4 T cells were purified from 5 normal donors by positive selection and transfected by nucleofection (AMAXA) with negative control or validated siRNA oligonucleotides specific for hTERT. Twenty-four hours later the transfected cells were stimulated with anti-CD3/CD28-coated beads. hTERT transcripts (A) and telomerase activity (B) were measured on day 3. The following parameters were determined on day 4: cell numbers (C), survival rates (D), apoptosis rates measured as PI-positive cells (E), TUNEL-positive cells (F), and lengths of telomeric sequences (G). Results are presented as means ± SEM.

RA Naïve CD4 T Cells Are Defective in Clonal Expansion and Are Prone to Undergo Apoptosis.

To determine whether telomerase insufficiency in RA T cells affects life-death decisions, CD4 CD45RO⁻ naïve and CD45RA⁻ memory cells from patients and controls were stimulated with anti-CD3/CD28-coated beads (Fig. 4). Over 6 days, control naïve CD4 T cells multiplied 30- to 40-fold, expanding mostly between days 3 and 6; RA T cell expansion was curtailed, reaching only 50% of the population size on day 6 (Fig. 4A, P = 0.000008). Growth kinetics of memory T cells were indistinguishable between patients and controls. (P = 0.95, Fig. 4B).

Fig. 4.

Defective clonal proliferation and increased apoptotic susceptibility in naïve CD4 T cells of RA patients. (A) Naïve and (B) memory CD4 T cells were purified from RA patients (n = 33, dashed line) and controls (n = 23, solid line) and stimulated with anti-CD3/CD28-coated beads. Cell numbers were assessed on days 3 and 6 following stimulation; results are given as means ± SEM. (C) Cell cycle progression was monitored by CFSE dilution on days 3 and 6 after stimulation. Representative FACS profiles of unstained T cells (dashed line), activated and CFSE stained T cells (gray filling), and non-stimulated and CFSE-stained T cells (solid line) are shown. (D) Mean numbers of cell cycles completed by days 3 and 6 after stimulation. Results from 10 RA patients (gray bars) and 10 controls (black bars) are presented as means ± SEM. (E) Based on number of cell cycles and actual recovery rates, survival rates by day 6 were calculated and compared in RA patients and controls (P < 0.001). (F) Telomerase activities in bead-stimulated CD4⁺CD45RO⁻ T cells were measured by TRAP assay and correlated with survival rates reached by day 6 (r = 0.58; P < 0.001). (G and H) CD4⁺CD45RO⁻ T cells from patients and controls were stimulated with anti-CD3/CD28-coated beads and frequencies of PI-positive cells (G, P < 0.001) and TUNEL-positive cells (H, P < 0.001) were quantified by FACS analysis on day 6 following stimulation.

When assessed by carboxyfluorescein diacetate succinimidyl ester (CFSE) dilution analysis, responsiveness and cell cycle progression was intact in RA T cells. Upon TCR ligation, essentially all CD4 CD45RO⁻ naïve T cells entered the cell cycle (Fig. 4C). By day 3, control T cells had passed through 2.3 ± 0.03 cell cycles as compared to 2.6 ± 0.12 in patients (Fig. 4D) (P = 0.06). By day 6, CD4 T cells from patients and controls had completed 6.37 ± 0.25 and 5.99 ± 0.14 cell cycles, respectively (P = 0.2).

By comparing actually recovered cell numbers to those predicted by the number of cell divisions, survival rates were calculated (Fig. 4E). After 6 days of TCR-induced proliferation, control T cells had generated 25% of the expected cell numbers. Thus, a large proportion of dividing T cells did not survive. Survival rates were significantly lower in the RA naïve CD4 T cells (P = 0.0001) reaching only 13% of the predicted cell recovery (Fig. 4E). In essence, the impaired expansion of RA T cells resulted from disproportionate cell loss. A mechanistic link between defective T cell survival and failure of up-regulating hTERT was suggested by a close correlation between telomerase activities on day 3 and survival rates on day 6 (r = 0.58, P < 0.001, Fig. 4F).

Naïve CD4 T Cells in RA Are Apoptosis Susceptible and Die Through the Intrinsic Death Pathway.

Impaired clonal expansion in RA could result from increased susceptibility to apoptosis. Apoptotic T cells, identified by PI and TUNEL positivity were distinctly rare for the first 3 days. On day 6, 2–3% of control T cells were apoptotic (Fig. 4 G and H). Rates were doubled in the patients' T cells (P < 0.001).

Pathways facilitating clonal contraction during the priming responses of naive CD4 T cells are incompletely understood. Blocking the extrinsic death pathways by disrupting Fas-Fas ligand (FasL), TNFα-TNF receptor, and TRAIL-TRAIL receptor interactions did not affect cell recovery (Fig. 5A) or apoptosis rates (Fig. 5B). Resistance to exogenous death ligand-receptor interactions was a unique feature of naïve T cells undergoing stimulation. Following repetitive restimulation, T cells became susceptible to FasL (Fig. S3). To explore the contribution of Bcl-2 in regulating apoptotic sensitivity in naïve CD4 T cells, CD4⁺CD45RO⁻ T cells were transfected with empty-IRES-GFP or Bcl-2-IRES-GFP 2 days after stimulation (Fig. S4) Bcl-2 overexpression rescued proliferating CD4 T cells from apoptotic death and improved cell recovery markedly (P < 0.0001) (Fig. S4).

Fig. 5.

Naïve RA CD4 T cells undergoing priming overexpress the apoptosis initiator Bim and lack expression of the pro-survival protein Bcl-2. (A) CD4⁺CD45RO⁻ T cells from 5 RA patients were stimulated with anti-CD3/CD28-coated beads. Isotype control antibodies or blocking antibodies (10 μg/ml) to FasL (NOK-1), TNF-α (infliximab), and TRAIL (RIK-1) were added on days 0 and 3. Cell numbers were measured on day 6. (B) PI⁺ cells measured by FACS analysis on day 6. (C-F) Naïve CD4 T cells from RA patients (n = 7, black bars) and age-matched healthy controls (n = 7, gray bars) were stimulated with anti-CD3/CD28-coated beads. On day 3 after stimulation, Bcl-2 and Bim transcripts were quantified by qPCR (C and D), Alternatively, cells were fixed, permeabilized, and stained with FITC-conjugated anti-Bcl-2 or Alexa Fluor 488-labeled anti-Bim antibody and analyzed by flow cytometry (E and F). Results are presented as mean fluorescence intensity (MFI) ± SEM of 7 independent experiments.

To test whether proteins of the intrinsic death pathway, specifically Bcl-2 and Bim, were involved in rendering RA CD4 T cells apoptosis susceptible, expression of Bcl-2 and Bim were quantified (Fig. 5). On day 3 after TCR triggering, Bcl-2 transcript concentrations were similar in patients and controls but RA T cells produced significantly more Bim sequences (P = 0.02). Flow cytometric analysis showed significantly less Bcl-2 (P = 0.02) and significantly more Bim (P = 0.01) in naïve RA CD4 T cells undergoing priming. FACS analysis for mitochondrial membrane potential and active caspase was compatible with T cell apoptosis mediated by the intrinsic pathway (Fig. S5).

Ectopically Expressed hTERT Restores Apoptosis Resistance in RA Naïve CD4 T Cells.

To support the hypothesis that apoptotic propensity during the priming response was influenced by telomerase, we overexpressed hTERT through lentiviral transfer to repair the induction defect in RA T cells (Fig. 6). Aberrant expression of hTERT markedly enhanced enzymatic activity of telomerase (Fig. 6B). Overexpression of hTERT had a direct effect on cell survival (Fig. 6C). On day 6 (P = 0.008) and day 9 (P = 0.003) following CD3/CD28 cross-linking, cell recovery was significantly improved in the T cells with ectopic hTERT. Thus, telomerase directly or indirectly affects life-death decisions of CD4 T cells undergoing clonal expansion.

Fig. 6.

Overexpression of hTERT repairs the survival defect of RA naïve CD4 T cells. Naïve CD4 T cells from RA patients were transduced with viral supernatants containing hTERT-IRES-puro, GFP-IRES-puro, or Empty-IRES-puro constructs. Cells were infected over 4 cycles, selected with 2 μg/ml of puromycin, depleted of dead cells, and stimulated with anti-CD3/CD28 beads. (A) GFP expression of transduced cells 2 days after puromycin selection. The solid line shows GFP-IRES-puro transduced cells; the dotted line shows Empty-IRES-puro transduced cells. (B) Telomerase activity of transduced T cells expressed as A.U./1,000 cells is shown as means ± SEM. (C) Cell recovery determined 6 and 9 days after stimulation. Black bars: hTERT-IRES-puro; gray bars: Empty-IRES-puro. Results are given as means ± SEM.

Discussion

The current study has examined mechanisms and functional consequences of accelerated telomeric loss in T cells in RA. Upon antigenic priming healthy T cells robustly induce telomerase, a process that is defective in RA T cells. While telomeric length is not compromised during this early phase of T cell expansion, telomerase insufficiency renders RA T cells apoptosis susceptible. Naïve RA CD4 T cells express higher levels of the apoptosis initiator Bim and reduced levels of the pro-survival protein Bcl-2 and fail to reach a similar clonal size as control T cells. Ectopically expressed hTERT restores apoptotic resistance and improves survival rates, establishing a direct mechanistic link between telomerase and apoptosis threshold of T cells. Conversely, knockdown of hTERT enhances apoptosis sensitivity assigning gate keeper function to the telomeric repair machinery in regulating T cell fate decisions.

Mutations in proteins assembling/stabilizing the telomerase complex have been highly instructive in understanding telomere biology and its role in restricting proliferation potential. The premature aging syndrome dyskeratosis congenita (DC) exemplifies this; it shows aberrant telomeric shortening in somatic cells lineages, including lymphocytes (21). TERC and TERT gene mutations occur in DC autosomal forms (22, 23). DC clinical manifestations include bone marrow failure, mucocutaneous abnormalities and predisposition to cancer; immunodeficiency becomes obvious in the more severe variant Hoyeraal-Hreidarsson syndrome (24). Like RA T cells, DC patient T cells are competent to undergo activation, but proliferation is diminished (25). DC is rare; the risk of patients to develop RA is unknown, but Fudenberg (26) described a DC patient whose grandmother and sister suffered from RA. In RA, telomeric shortening predominantly affects naïve CD4 T cells and telomerase insufficiency is selective for naïve T cells; telomerase induction is well maintained in memory T cells and even enhanced in CD34 HSC (Fig. 1). This constellation makes a genetic defect of telomerase components unlikely.

CD34⁺ HSC from RA patients compared to those of age-matched controls display telomeric shortening by 1,600 bp (27), yet telomerase activity is enhanced in this cell lineage. These data suggest cell-type specific regulation of telomerase in the hematopoietic system. In RA naïve CD4 T cells, telomerase activity correlated strongly with hTERT transcripts (Fig. 1) suggesting a defect in transcriptional control of the hTERT gene. We could not associate hTERT repression with excessive production of type 1 IFN (9), a cytokine typically overproduced in systemic lupus erythematosus (SLE), but not RA, nor was hTERT repression associated with defective AKT phosphorylation (12).

Data presented here suggest that RA patients have difficulties generating sufficient T cells when their CD4⁺CD45RA⁺ cells respond to antigen. Priming of naïve CD4 T cells causes 35-fold expansion with 25% of that population surviving (Fig. 4). It thus represents a powerful mechanism of repopulating the T cell pool. Increased apoptosis during T cell priming may restrict peripheral T cell generation. Homeostatic proliferation of naïve T cells follows similar rules as antigen-induced T cell expansion. A change in the threshold setting of apoptotic susceptibility amongst naïve T cells will thus fundamentally affect repopulation of the T cell pool. As a consequence, T cells are forced into accelerated autoproliferation. In line with this model, TREC⁺ recent thymic emigrants are decreased in RA patients, indicating more pronounced dilution of that unique T cell subset (5, 6).

In contrast to most somatic cells, T cells possess the ability to up-regulate telomerase robustly (15, 16, 28). As a highly proliferative tissue, T cells are prone to undergo replicative senescence. If arrested in cell cycle, they are of limited value for the immune system, and elongating telomeres clearly gives them a competitive advantage. CD4⁺CD45RA⁺ T cells from RA patients do not display classical characteristics of senescence (Fig. S2). When signaled through the TCR, they progress through the cell cycle (Fig. 4). Also, proximal and distal events in the T cell activation cascade are intact (Fig. 1), identifying the restricted induction of hTERT as a selective defect. hTERT levels correlated closely with enzyme activity (Fig. 1), confirming that transcription of the catalytic subunit is the limiting factor for telomerase activity. Diminished induction of telomerase remains relevant in vivo (Fig. 2), suggesting that RA patients have fundamental abnormalities in priming T cell responses and maintaining T cell homeostasis. Patients with active and inactive disease had similar telomerase activities excluding the inflammatory milieu as the primary regulator of T cell telomerase. Even untreated patients lacked the ability to induce sufficient hTERT levels, raising the possibility that this is an inherent deficiency in this autoimmune syndrome.

Telomerase insufficient T cells died from excessive apoptosis despite stability in telomeric length. By reinstating lost DNA telomeric sequences, telomerase protects cells from replicative senescence (29) and may confer cellular immortality (30). Transfer of hTERT into human mammary epithelial cells promotes spontaneous growth (31). In a recent report (31), the growth-promoting ability of telomerase was p53 dependent, suggesting that the cell recognizes a shortened telomere as damaged DNA. 53BP1/phosphorylated histone H2AX foci appeared at chromosome ends long before telomeres were critically shortened indicating that telomeric dysfunction precedes erosion of telomeric sequences and functionally affects the cell long before senescence. DNA double-stranded breaks induce recruitment of Ku, DNA-PKcs, Nbs1, and Mre11; all of these molecular components have also been associated with telomeres (32). Gamma-H2AX has a role in early signaling of DNA damage (33) and, again, has been observed in foci at the telomeric ends (34). The detailed structure of telomeres in T cells as they pass through different life-cycles is unknown. Besides the premature loss of telomeric ends, telomere structure in RA T cells may be sufficiently disturbed to physically disrupt binding sites for essential proteins of the shelterin complex (35) or may interrupt T-loop formation (36).

The question remains of how impaired telomerase induction and apoptotic hypersensitivity in naïve CD4 T cells is related to RA. Most important to consider is the impact of chronic T cell attrition on homeostatic control of the T cell compartment. Notably, the major risk factor for RA is advanced age; most patients are diagnosed during the second half of life. During this life period, naïve T cells are not replenished through thymic production but rather through autoproliferation. Failure to reach appropriate clonal size during priming must inevitably lead to smaller clonal sizes of memory cells. This scenario predicts that RA patients have difficulties maintaining a filled T cell pool and expose memory T cells to more and more replicative turnover. In essence, the entire T cell pool in RA is overaged, forcing the patient to generate immune responses with T cells that have essentially reached the end of their life span. Interestingly, senescent CD4 T cells acquire apoptotic resistance (37), endowing them with a survival advantage. If such end-differentiated memory T cells stay alive, they will compete for space and further disadvantage incoming new cells. T cell senescence has also been identified as a pathway in atherosclerosis, particularly the unique inflammatory response precipitating plaque instability (38, 39). Cardiovascular complications are now recognized as extra-articular manifestations of RA (40). Finally, dysfunction of naïve CD4 T cells puts the patient at risk for inadequate anti-pathogen responses, a complication well recognized within the spectrum of the rheumatoid syndrome (41). Implicating telomerase in T cells as an element of the pathogenic network in RA provides novel and exciting opportunities for therapeutic approaches in this chronic and as yet incurable disease.

Materials and Methods

Patients and Controls.

Demographic characteristics of 69 patients and 60 healthy control subjects are given in Table S1. All patients fulfilled the diagnostic criteria for RA and were positive for rheumatoid factor. Control individuals had no personal or family history of autoimmune disease, cancer, or any other inflammatory syndrome. Selection of patients and controls and clinical data analysis and assessment of disease activity has previously been described (27). The study was approved by the Emory University Institutional Review Board and written consent was obtained from all participants.

Cells and Culture.

T cell subsets were purified from PBMC using CD45RO, CD45RA, CD4, and CD71 (Miltenyi) microbeads. Subset purity was >95% by FACS. Circulating CD34⁺ hematopoietic stem cells were purified as described (27). CD4⁺CD45RO⁻ or CD4⁺CD45RA⁻ cells (1.0 × 10⁵/well) were stimulated with CD3/CD28 Dynabeads (Dynal) at a 2.5:1 ratio. IL-2 (20 IU/ml) was added on days 0, 3, and 6. Cells were counted by flow cytometry or trypan blue exclusion. For blocking of death receptors, cells were cultured in the presence of 10 μg/ml of anti-Fas ligand mAb (NOK-1; BD), anti-TRAIL mAb (RIK-2; BD), anti-TNFα mAb (Centocor) or isotype control antibody. To assess T cell activation, CD4 T cells were stained with FITC-anti-CD69, PE-anti-CD71, or PE-anti-CD25 monoclonal antibodies (BD). CFSE staining and analysis was completed as described by Quah (42). Frequencies of apoptotic T cells were measured by cytometric analysis of PI or TUNEL expression. Apoptotic cells with high caspase activity were detected with MitoCasp kits (Cell Technology).

RT-PCR.

Total RNA was extracted with TRIzol, and cDNA was synthesized with AMV-reverse transcriptase (Roche) and random hexamers (Roche). PCR was performed with 200 nM of sense and anti-sense primer using a MX3000P (Stratagene); 3 mM MgCl₂, 0.2X SYBR Green; and 0.75U Platinum Taq (Invitrogen). See Table S2 for primer sequences. Copy numbers were calculated by comparing each sample with a standard curve generated by amplifying serially diluted plasmids containing relevant sequences.

Telomeric Repeat Amplification Protocol (TRAP) assay.

Telomerase activity was quantified as described (43). Enzyme activities were calculated by comparing Ct values of each sample with a standard curve of serially diluted Jurkat cell extracts and expressed as arbitrary units (AU).

Measurement of Telomeric Sequences.

T cell telomere lengths were measured by real-time PCR modified from Cawthon (44).

siRNA-Mediated Knockdown of hTERT Sequences.

To knock down hTERT expression, 6 μg of small interference RNA oligonucleotides were transfected into resting purified CD4 T cells using the Amaxa Nucleofector system and the Human T cell Nucleofector kit (Amaxa). Oligo duplex RNA specific for hTERT was purchased from Qiagen (Hs-TERT 3 HP siRNA, target sequence; 5‘-CTGGAGCAAGTTGCAAAGCAT-3‘). As a negative control, AllStars Negative Control siRNA (Qiagen) was used. Twenty-four hours after transfection, cell numbers were adjusted, and CD4 T cells were stimulated as described.

Stable Transfection of hTERT in Naïve CD4 T Cells.

Replication-defective lentiviral particles were produced by transfection of 3-plasmids (pHR'CMV-GFP-IRES puro/pHR'CMV-hTERT-IRES puro/pHR'CMV-empty-IRES puro, pCMV delta 8.91 and pMD.G) to 293T cells. Activated naïve CD4 T cells were cultured with virus containing supernatants for several cycles and selected with 2 μg/ml puromycin.

Flow Cytometric Analysis of Bcl-2 and Bim Protein Expression.

Seventy-two hours after stimulation, T cells were harvested, fixed with 2% formaldehyde for 10 min at RT, incubated in 100% methanol, and stained with FITC-conjugated anti-Bcl-2 (BD) or Alexa Fluor 488-labled anti-Bim antibody (Cell Signal Technology). Protein expression was detected by a LSR II flow cytometer.