Increases in T Cell Telomere Length in HIV Infection after Antiretroviral Combination Therapy for HIV-1 Infection Implicate Distinct Population Dynamics in CD4+ and CD8+ T Cells
Download the PDF here|
Clinical Immunology July, 1999
Sumesh Kaushal,* Alan L. Landay, Michael M. Lederman, Elizabeth Connick, John Spritzler, Daniel R. Kuritzkes, Harold Kessler, Bruce L. Levine,*,1 Daniel C. St. Louis,* and Carl H. June*,1
*Henry M. Jackson Foundation for the Advancement of Military Medicine, U.S. Military HIV Research Program, Bethesda, Maryland 20889; Departments of Immunology/Microbiology and Medicine, Rush-Presbyterian-St. Luke's Medical Center, Chicago, Illinois 60637; Division of Infectious Diseases, Case Western Reserve University School of Medicine, Cleveland, Ohio 44106; Division of Infectious Diseases, University of Colorado Health Sciences Center, Denver, Colorado 80262; and Statistical and Data Analysis Center, Harvard School of Public Health, Boston, Massachusetts 02115
"Changes in mean telomeric terminal restriction fragment (TRF) length were examined.......Increases in mean T cell TRF lengths were observed in most patients following therapy; however, the contribution of individual T cell subsets was complex.....despite potent suppression of viral replication, CD4 cell telomeres recovered in some patients, yet continued to decline in others.In summary, the present results show that potent ART therapy has marked effects on composition of the T cell compartment. However, the differential recovery rates of T cell compartments suggest that CD4 and CD8 T cell reconstitution are regulated by distinct mechanisms.......the regenerative capacity of the adult CD4 T cell repertoire is limited...... observation that CD4 cell recovery is associated with thymic enlargement while CD8 T cell recovery is not (41). McCune and co-workers found that the number of circulating naive CD4+ T cells correlates with patients who have more abundant thymic mass (42). Thus, it is possible that the large individual vari- ations in CD4 T cell telomere length changes might be relevant to individual differences in disease progression and the response to ART......Uncovering of the mechanism underlying the wide disparity in CD4 cell telomere length changes following potent ART will require further study....Reverse transcriptase inhibitors have been shown to have effects on telomeres and telomerase. For example, zidovudine has been shown to incorporate into telomeric DNA and to cause telomere shortening in HeLa cells at high concentrations (43). In human leukemic cell lines, zidovudine caused progressive telomere shortening in some cultures and inhibited telomerase activity (44). In cultures of primary human T cells, we have been unable to detect inhibition of telomerase activity or induction of telomere shortening by zidovudine (SK, unpublished). Thus, reverse transcriptase inhibitors are unlikely to have caused the increase in telomere length that we observed following the start of potent ART and it remains unclear if they might con- tribute to the telomere shortening observed in HIV- infected patients......"
Sumesh Kaushala, Alan L. Landayb, Michael M. Ledermanc, Elizabeth Connickd, John Spritzlere, Daniel R. Kuritzkesd, Harold Kesslerb, Bruce L. Levinea, 1, Daniel C. St. Louisa and Carl H. Junea, 1
a Henry M. Jackson Foundation for the Advancement of Military Medicine, U.S. Military HIV Research Program, Bethesda, Maryland, 20889
b Departments of Immunology/Microbiology and Medicine, Rush-Presbyterian-St. Luke's Medical Center, Chicago, Illinois, 60637
c Division of Infectious Diseases, Case Western Reserve University School of Medicine, Cleveland, Ohio, 44106
d Division of Infectious Diseases, University of Colorado Health Sciences Center, Denver, Colorado, 80262
e Statistical and Data Analysis Center, Harvard School of Public Health, Boston, Massachusetts, 02115
Changes in mean telomeric terminal restriction fragment (TRF) length were examined as a marker for cellular replicative history in HIV-1-infected individuals after institution of anti-retroviral therapy (ART). Increases in mean T cell TRF lengths were observed in most patients following therapy; however, the contribution of individual T cell subsets was complex. An elongation of CD8+ T cell TRF was nearly uniformly observed while changes in mean TRF length in CD4+ T cells were heterogeneous as, despite potent suppression of viral replication, CD4 cell telomeres recovered in some patients, yet continued to decline in others. Increases in CD8 cell TRF correlated with decreased memory cells, suggesting a negative selection in the periphery for CD8 cells with extensive replicative history. In contrast, increases in CD4+ T cell TRF length correlated with increases in naive cell subsets, suggesting that the CD4+ T cell TRF increase may reflect a thymic contribution in some patients. These are the first increases in somatic cell telomere length in a population of cells observed in vivo, and the findings are compatible with therapy-induced reconstitution of the lymphoid compartment with cells having a more extensive replicative potential. These findings further distinguish lymphocytes from other somatic cell populations where only decreases in TRF over time have been noted. Thus, institution of ART in persons with moderately advanced HIV-1 disease reveals distinct population dynamics of CD4 and CD8 T cell subsets and also shows that the lymphocyte replicative history is dynamic.
Telomeres are specific structures found at the ends of eukaryotic chromosomes. In somatic cells, telomere length is shortened with each cell division due to the inability of DNA polymerase to replicate the extreme 5 end of the lagging strand of DNA (1). Substantial evi- dence supports the hypothesis that telomeres function in part as a mitotic clock for the cell. In vivo, lympho- cytes have age-related reduction in telomere length (2), losing an average of 31 base pairs/year. In vitro, the rate of telomere loss from lymphocytes from normal individuals is approximately 50 to 120 bp/cell doubling, comparable to that seen in other somatic cells (2, 3). Thus, measurement of the telomeric terminal restric- tion fragment (TRF) length by Southern blot analysis is thought to serve as an index of replicative history in vivo and in vitro.
It has been suggested that replicative senescence (4) might contribute to the immunodeficiency of HIV in- fection (5). A related hypothesis is that increased clonal expansion and potential clonal exhaustion of T cells occur in response to HIV infection (6-8). Consistent with these models of immunosenescence, the mean TRF length was shown to decrease in CD8 T cells in most adult patients with HIV infection (9-11). How- ever, telomere measurements of CD4 cells from HIV- infected individuals do not shorten excessively, sug- gesting that proliferative exhaustion is not a major component of CD4 cell decline in patients with HIV infection (9, 11-13). Recent measurements of T cell turnover during SIV infection in macaques using Brdu labeling document moderately elevated turnover of both CD4 and CD8 lymphocyte subsets (14, 15).
Treatment of HIV-1-infected individuals with potent antiretroviral drug combinations results in a dramatic decline in viral load and a partial restoration of the circulating T cell compartment. Soon after initiation of combination anti-retroviral therapy an initial rapid increase in CD4 and CD8 T cell counts in the blood occurs, and this is followed by a more gradual increase in CD4 cells and a decline to below baseline in CD8 T cell counts (16-18). Further evidence for partial T cell reconstitution in patients treated with potent ART regimens includes the functional recovery of some CD4 T cell proliferative responses (17) accompanied by a late increase in proliferating CD4 T cells (19, 20). It is controversial as to whether the diversity of the T cell receptor repertoire recovers, and it is likely that this is a delayed and incomplete process (20, 21).
The mechanisms leading to incomplete T cell regen- eration are still a matter of debate (22). It is likely that impaired thymic function contributes to the failure to completely restore T cells (23). In an effort to assess possible mechanisms for T cell reconstitution after pro- found inhibition of virus replication, we examined the changes in TRF as an indicator of T cell replicative history and correlated these to longitudinal changes in lymphocyte subpopulations following institution of po- tent ART.
Increase in Mean TRF Length in Lymphocytes Following Initiation of a Potent Anti-Retroviral Therapy (ART) Regimen
A previous study has shown an accelerated rate of telomere shortening in the leukocytes of patients dur- ing the natural progression of HIV infection, with a mean loss of 114 base pairs/year in PBMC in asymptomatic patients and 175 bp/year in progressors (9). To assess the effect of potent anti-retroviral therapy on telomeres, TRF length was analyzed longitudinally in the T cells of seven men who had not had previous therapy with protease inhibitors. Four of these men had moderately advanced HIV infection (CD4 count 100 -300), and three were asymptomatic individuals with baseline CD4 counts of 350 to 500 cells per cubic millimeter of blood. Five of the seven patients had increased T cell mean TRF after 6 to 12 months of potent anti-retroviral therapy (Fig. 1). The expected wide individual variations in baseline mean telomere length were noted. Thus, while the previously docu- mented rate of telomere loss is 100 to 175 bp/year (9), we observed an average gain of 350 base pair T cells from patients treated for 6 to 12 months with protease inhibitor therapy.
Distinct Changes in CD4 and CD8 Telomeres after Institution of ART
Although the number of patients tested was small, the above results suggest that inhibition of viral repli- cation might slow the accelerated rate of loss of telomere length and may actually result in restoration of T cell telomere length in most HIV-infected individuals. To more formally test this possibility and to assess the changes in TRF length in T cell subpopulations, we performed a longitudinal study of 11 individuals en- rolled in AIDS Clinical Trials Group Protocol 315. Patient recruitment for ACTG 315 and the study design was previously described (27). Briefly, all patients had documented HIV-1 infection and CD4 T lymphocyte counts of 100-300 cells/μ and all had received pre- treatment with AZT; none had received 3TC or a pro- tease inhibitor. Eleven patients receiving triple combi- nation drug therapy with lamivudine, zidovudine, and ritonavir were selected based on sample availability for further analysis. The baseline immunologic and viro- logic indices of these patients are shown in Table 1 and are representative of the entire ACTG 315 cohort (27).
Nearly all patients had an increase in CD8+ T cell mean TRF after institution of combination ART (Fig. 2b). The median TRF length of CD8+ T cells increased 430 bp over the 48-week study period, and this increase was significant compared to baseline (P = 0.02 by Wilcoxon signed rank test; P = 0.007 by Student's t test. The univariate 95% confidence interval on the mean change in CD8 telomere length from baseline to week 48 of the study was 132 to 619 bp). The increase in CD8+ T cell TRF was evident within 12 weeks after initiation of therapy and then stabilized after 24 to 48 weeks of therapy. In contrast, the mean TRF length in CD4+ T cells did not increase (Fig. 2a). Confirming previous reports (9-11), we found that the baseline mean CD8+ T cell TRF length was 714 bp shorter than the mean CD4+ T cell TRF length in this cohort of patients (Fig. 3 and data not shown).
Significant interpatient and intrapatient heterogeneity was observed in the longitudinal analysis of the patterns of changes in telomere length in the T cell subsets from the 11 patients. Analysis of the mean TRF length changes in these treated patients revealed two primary patterns of CD4 T cell recovery. Mean TRF length in CD4+ T cells modestly increased in 5 individ- uals, but remained unchanged or were slowly lost in the 6 other patients (Fig. 2c). The pattern was complex, as 2 of the patients had transient increases in CD4+ T cell telomere length that were not sustained and were followed by subsequent decreases in length. In con- trast, the pattern was more uniform in the CD8+ T cell compartment, with 8 of 10 patients having increases in mean telomere length after 24 to 48 weeks of therapy (Fig. 2d). (One patient had no samples available after week 12). Two examples of intrapatient heterogeneity of the changes in telomere length are shown in Fig. 3. In patient No. 8, there were increases in both the CD4 and the CD8 T cell mean TRF after institution of ART, while in patient No. 9 a "split response" was noted, with a decrease in the CD4 telomere length observed while the CD8 telomere length was increasing. Of the 10 patients assayed longitudinally for 24 weeks or more of therapy, 5 patients had increases in both CD4+ and CD8+ T cell telomere length while 3 of the patients had decreases in CD4 and increases in CD8 TRF and 2 patients had TRF decreases in both T cell subsets.
The above results establish that the institution of ART is associated with changes in telomere length, and these changes differ strikingly from the expected monotonic decline observed in the PBMC from normal adults of about 30 to 50 base pairs per year (2), or from the accelerated decline of 100 to 300 base pairs per year in the CD8 subset that is reported in HIV-infected individuals (9). Perhaps most notable from these re- sults is that some patients continue to have a decline in CD4 T cell TRF length while others have increases in CD4 cell TRF length after starting ART. This does not appear to be correlated with failure of control of HIV-1 replication as only two patients had detectable plasma HIV-1 after 48 weeks of therapy. Moreover, both of these patients had increased CD8 cell TRF while one had increased CD4 cell telomere length and the other decreased CD4 cell telomere length.
We have considered a number of other potential ex- planations for the changes in telomere length observed after ART. The initial response to potent ART is an increase in CD4 and CD8 cells that may be due pri- marily to the release of cells trapped in lymph nodes.
Preferential mobilization of cells with longer telomeres such as naive T cells could contribute to the changes that we have observed. The major contributions to the increase in CD4 and CD8 cells from rebound occurs early, in the first several weeks of therapy (6, 7), and we did not observe the major increase in CD8 TRF until 24 weeks or more of therapy.
Activated T cells express telomerase (28), and hence it is possible that induction of telomerase activity could contribute to the increased telomere length during the T cell repopulation that follows institution of ART. To investigate this possibility we measured telomerase activity in the T cells from 11 patients in AIDS Clinical Trials Group protocol 315. Low levels of telomerase activity were detectable in the lymphocytes from pa- tients before therapy, confirming previous studies (9). In 10 of the patients there was no change in the telo- merase activity in relation to commencement of com- bination ART, while in 1 patient there was a temporally related increase in activity (Fig. 4). In this indi- vidual (patient 10), during the first 2 weeks after starting ART there was a substantial increase in the telomerase activity from 2% to more than 50% of the activity detected in the telomerase-positive tumor control cell line. This appears to be an uncommon occur- rence, as we have since assayed multiple other samples from a total of 22 individuals and have not observed a similar temporal induction of telomerase activity (data not shown). Together, these results suggest that it is unlikely that increased telomerase activity could ac- count for the nearly uniform increase in mean telomere TRF length in the CD8 compartment that we have noted. However, our results do not exclude the possi- bility that induction of telomerase activity limited to a compartment such as lymph nodes could contribute to the increased telomere lengths observed during cellu- lar restoration following ART.
Striking changes in the numbers and immunopheno- types of the circulating lymphocytes occurred in the 11 patients after starting ART (Table 2). There was a significant increase in CD4+ T cells at all time points. In contrast, while CD8+ T cells increased early after starting therapy confirming previous reports (16), after 36 to 48 weeks of therapy the total CD8+ T cell number was not different from the baseline number. There were, however, marked changes in the composition of the CD8+ T cell subset, with a significant decrease in activated cells bearing the CD95 antigen. In contrast, naive CD4 and CD8 cells increased with therapy. Similarly, there was an increase in T cells that express CD28, and this was significant for CD4+28+ (P = 0.002, Wilcoxon signed rank test) cells. There was, however, a significant decrease in the number of memory CD8+RO+ cells. Recent studies have documented shorter telo- meres in CD28+ T cells compared to their CD28+ counterparts (10, 29). Similarly, since memory CD45RO+ T
cells have shorter telomeres than CD45RA+ naive T cells, it is likely that the increase in TRF length in the CD8 cells of patients after starting potent ART is due to a selective persistence and/or renewal of cells with longer telomeres, such as naive and/or CD28+ T cells. Furthermore, the selective depletion of activated CD8+RO+28+ T cells with short telomeres following HAART is also likely to contribute to these findings. Thus, our results are most consistent with a mechanism for recovery of the CD8+ T cell compartment that involves the replacement of CD8+ T cell cells with high replicative history and shorter TRF length with cells having higher replicative potentials and longer TRF lengths.
In contrast to the marked decrease in CD8 cells that expressed the CD45RO or CD95 antigens, there was an increase in the absolute number of CD4 cells expressing CD45RO (P = 0.004) or CD95 (P = 0.006) (Table 2). The increase of CD4 cells with a memory phenotype likely contributes to the decline in CD4 cell telomere lengths noted in some patients.
To further clarify the distinct changes in telomeres induced by ART, the CD4 and CD8 T cell telomere lengths were correlated with the number and fre- quency of T cell subsets after 12 to 48 weeks of therapy (Table 3). It would be expected that changes in mean TRF length, an assay done on bulk populations of CD4 and CD8 cells, would correlate best with changes in the frequency of cell subsets rather than the absolute num- bers of cell subsets. Changes in the mean TRF length in the CD4+ T cell compartment correlated with changes in the frequency of CD45RA+62L= naive cells (P = 0.003) and CD28+ cells (P = 0.001), suggesting that the mean TRF length observed in the CD4 com- partment may relate to new production of CD4 T cells following combination ART (Table 3). Telomere lengths of CD4 cells also correlated with the absolute number of naive cells (P = 0.01) but not with the absolute number of CD28+ cells (P =0.14). Furthermore, elongation of CD4 T cell TRF also correlated with the frequency of CD25+CD4+ T cells (P = 0.003). It is known that naive cells have long telomeres, and it is likely that recent thymic emigrants have long telo- meres (30). In contrast, while the increase in TRF lengths of CD8+ T cells did not correlate with changes in naive cells, there was an inverse correlation of CD8 memory cells with telomere length that did not quite reach statistical significance (P = 0.06).
Previous studies have shown that HIV infection induces increased proliferation of CD4+ T cells (31), and recent studies have shown that there is a further increase in CD4+ T cell proliferation induced by potent ART (19). We found that for both CD4 and CD8 T cells, the percentage and frequency of CD25+ T cells correlated with increased telomere length. Together these results are consistent with the notion that the increase in the telomere length in the CD8+ T cell compartment was related to the selective persistence/survival of CD8+ T cells that have longer telomeres, as well as the renewal/replacement with naive cells that have longer telomeres. Finally, the a positive correlation of CD25 expression with longer telomeres in both CD4 and CD8 T cells is consistent with these cells being a marker for recently produced cells that are derived from cells with a more extensive replicative capacity.
In this report we have extended our previous studies regarding telomere lengths in HIV-1 infection. Potent ART results in T cells with longer average terminal telomeric restriction fragment lengths. Longitudinal analysis of the replicative history of T cell subsets shows that ART therapy has marked effects on the composition of the CD4 and CD8 T cell compartments which results in distinct changes in telomere length in CD4 and CD8 T cells. We found that the effect of ART in patients with moderately advanced HIV infection was to increase the median CD8+ TRF by about 430 base pairs while there was no change in the CD4+ cell TRF after 48 weeks of therapy. Thus, these are the first results to demonstrate the reconstitution of T cell com- partment with T cells having a more extensive replicative capacity, as previous studies have only demonstrated age-related declines in lymphocyte telomere lengths during steady state hematopoiesis.
There are several limitations to the use of telomere length as a marker of T cell turnover and thymic re- newal in HIV infection. First, McCune and Hellerstein have pointed out that since HIV-1 preferentially infects dividing CD4+ T cells that are about to shorten their telomeres, the T cells surviving HIV infection should have longer telomeres (22). However, this is unlikely to be a significant limitation in the present study due to the effective interruption of viral replication consequent to potent anti-retroviral therapy. Second, telomerase is a ribonucleoprotein complex that can add hexameric repeats to telomeres. Previous studies have shown that telomerase activity is strongly upregulated after T cell activation (28, 32-34). Fur- thermore, HIV-1-infected individuals have normal telo- merase activity in peripheral blood T cells (9, 11). We found that most patients do not have detectable increases in telomerase activity during the first month of HAART, and thus it is unlikely that telomerase activity accounts for the increase in telomere lengths that we have observed. However, we did find that telomere lengths were correlated with CD25 expression, and this would be consistent with telomerase-mediated lengthening of telomeres.
The central issue raised by our study concerns the nature of the differences between CD4 and CD8 T cell reconstitution. BrdU labeling studies in normal and SIV-infected macaques have shown similar CD4+ and CD8+ T cell turnover rates, and have also shown that infection with SIV induced comparable increases in the turnover rates of CD4+ and CD8+ T cells (14, 15). In studies quantifying expression of the nuclear Ki-67 antigen as a surrogate marker for cell division, CD4+ T cell proliferation was found to be increased two- to threefold in HIV-infected individuals with moderately advanced HIV infection (35, 36) but not in patients with early stage HIV infection (19). In contrast, the fraction of proliferating CD8+ T cells in HIV infection in both early and later stage HIV infection was elevated approximately sixfold compared to control do- nors (19, 35, 36). Together, these studies are compati- ble with the previous studies showing that HIV infection was associated with decreased telomere length in CD8+ T cells but not in CD4+ T cells (9-11). Thus, on the one hand, patients begin ART with more extensive depletion of telomeres in the CD8 compartment, and simply the ART-mediated relief of the ho- meostatic stress would be predicted to lead to a more rapid normalization of the CD8+ T cell telomere com- partment. In addition, the rise in mean TRF length in the CD8+ T cell compartment correlated inversely with the number of circulating CD8 memory cells, consis- tent with a negative selection in the periphery for CD8+ T cells with a less extensive replicative history.
In contrast to the relative homogeneity of the response in the CD8 compartment, the telomere response of the CD4 compartment was characterized by wide individual heterogeneity as was shown in Fig. 2. Other studies have also noted high interindividual variability in CD4 cell growth and turnover rates compared to a uniformly high CD8 turnover rate in HIV-infected in- dividuals as assessed by Ki-67 antigen expression (19, 35). There are other indications that the population dynamics of the CD4 and CD8 compartments differ.
For example, the regenerative capacity of the adult CD4+ T cell repertoire is limited (37, 38). Furthermore, following chemotherapy or marrow transplantation, CD8+ T cell regeneration occurs much earlier than CD4 recovery (39). Mackall and co-workers have sum- marized evidence that the regeneration of CD4+ and CD8+ T cell compartments is distinct (40). Most nota- ble is the observation that CD4 cell recovery is associ- ated with thymic enlargement while CD8 T cell recovery is not (41). McCune and co-workers found that the number of circulating naive CD4+ T cells correlates with patients who have more abundant thymic mass (42). Thus, it is possible that the large individual variations in CD4 T cell telomere length changes might be relevant to individual differences in disease progression and the response to ART.
Reverse transcriptase inhibitors have been shown to have effects on telomeres and telomerase. For example, zidovudine has been shown to incorporate into telomeric DNA and to cause telomere shortening in HeLa cells at high concentrations (43). In human leukemic cell lines, zidovudine caused progressive telomere shortening in some cultures and inhibited telomerase activity (44). In cultures of primary human T cells, we have been unable to detect inhibition of telomerase activity or induction of telomere shortening by zidovu- dine (SK, unpublished). Thus, reverse transcriptase inhibitors are unlikely to have caused the increase in telomere length that we observed following the start of potent ART and it remains unclear if they might con- tribute to the telomere shortening observed in HIV- infected patients.
Uncovering of the mechanism underlying the wide disparity in CD4 cell telomere length changes following potent ART will require further study. Given that we found that increases in TRF length correlated with increases in naive (CD4+45RA+62L+) cells, a marker of thymic-derived T cells in CD4 cells but not CD8 T cells, it is possible that thymic regenerative capacity will correlate with recovery of naive CD4 T cells having a more extensive replicative capacity. Recent studies of T cells containing excisional DNA products of TCR-gene rearrangement provide a measurement of thymic out- put, and these studies suggest substantial heterogene- ity of thymic function in patients with HIV infection (45). Autopsy studies also indicate substantial thymic pathology in patients with HIV-1 infection (46). In addition we found that CD4 telomere lengths corre- lated inversely with CD4 memory cells, similar to CD8 cells. Together, our data are in broad agreement with the Red Queen Model recently put forward by Haase and co-workers (36), which states that T cells, and particularly CD4 T cell replacement mechanisms, nor- mally operate in adults at close to maximum capacity just to maintain steady state. Our results add to this model by suggesting that CD8 T cells are able to more
rapidly catch up after perturbation of the steady state by HIV and that a thymic component to CD4 cell re- newal is more active in some individuals than in others. In summary, the present results show that potent ART therapy has marked effects on composition of the T cell compartment. However, the differential recovery rates of T cell compartments suggest that CD4 and CD8 T cell reconstitution are regulated by distinct mechanisms.