Altered IL-7Rα expression with aging and the potential implications of IL-7 therapy on CD8⁺ T-cell immune responses

The proper development and maintenance of naive and memory CD8⁺ T cells are essential for host defense.^1,2  IL-7, a member of the common cytokine-receptor γ-chain (γc) family, is critically implicated in the generation and maintenance of naive and memory CD8⁺ and CD4⁺ T cells.^3-8  IL-7 is produced by multiple stromal tissues, including epithelial cells in the thymus and bone marrow.⁷  The IL-7R complex consists of 2 chains, the high-affinity IL-7Rα chain and γc.⁹  IL-7 binding to the receptor induces sequential phosphorylation of Jak1, Jak3, and STAT5, leading to the up-regulation of Bcl-2, which promotes cell survival.^10,11  IL-7 regulates peripheral homeostasis of CD4⁺ and CD8⁺ T cells by promoting cell survival. When CD8⁺ T cells from IL-7R intact and knockout (KO) mice were adoptively transferred to wild-type mice, cells from IL-7R KO mice had decreased survival compared with those from IL-7R intact mice.³  Furthermore, in mice infected with lymphocyte choriomeningitis virus, IL-7Rα^high CD8⁺ T cells had better survival and differentiation into memory cells than IL-7Rα^low CD8⁺ T cells.⁵ 

Alterations in T-cell immunity occur with aging. These alterations include thymic atrophy and changes in T-cell subsets and function, which likely contribute to increased risk for infection in the elderly.^12,13  For example, in 1998, 82 989 deaths in persons 65 and older in the United States were attributed to influenza and pneumonia, compared with 5969 deaths in persons aged 15 to 64.¹⁴  In addition, CD4⁺ and CD8⁺ T-cell responses to the influenza vaccine were decreased in elderly subjects compared with young subjects.^15,16  Although the underlying mechanism for altered T-cell immune responses with aging is largely unknown, IL-7 may have a role in developing such immune changes given that IL-7 is critically involved in the development and maintenance of T cells^7,8  and that IL-7 levels in the thymus and serum are reduced with aging.^15,17 

A potential role for IL-7 therapy in rejuvenating T-cell immunity has been suggested.^7,18,19  In fact, a recent mouse study demonstrates that IL-7 could serve as an effective vaccine adjuvant by augmenting responses to subdominant antigens and improving the survival of memory CD8⁺ T cells.¹⁹  However, before such therapy is initiated in an elderly patient, it is critical to investigate whether the patient has a defect in other steps involved in the IL-7-mediated CD8⁺ T-cell survival pathway because this defect may not be corrected by providing exogenous IL-7. We addressed this issue by measuring IL-7R expression, signaling, and cell survival in different subsets of CD8⁺ T cells in response to IL-7 in young and elderly subjects. In addition, the characteristics of CD8⁺ T-cell subsets, as defined by markers including IL-7Rα, was investigated by measuring cell-surface molecules, function, and TCR repertoire. The results of our study show that aging affects IL-7Rα expression by EM_CD45RA+ CD8⁺ T cells, leading to impaired signaling and survival responses to IL-7, and that IL-7 therapy may selectively improve the survival of EM_CD45RA+ CD8⁺ T cells with a diverse TCR repertoire in the young but not in the elderly.

Healthy elderly (65 years and older; n = 29) and young (40 years and younger; n = 27) subjects were recruited for this study (mean ± SD, 74.6 ± 4.95 and 31.7 ± 5.87 years). There was no sex difference between the 2 groups (P = .343, χ² test). Subjects who were taking immunosuppressive drugs or who had a disease potentially affecting the immune system were excluded.^15,20  Informed consent was obtained from all subjects. This work was approved by the institutional review committees of Yale University and the Veterans Administration New England Health Care System, West Haven Campus.

Peripheral-blood mononuclear cells (PBMCs) were stained with goat anti-human IL-7Rα and with mouse anti-human IL-15Rα, -IL-2/15Rβ, -γc, -IL-21R, or isotype antibodies (R&D Systems, Minneapolis, MN). Cells were washed and stained with donkey anti-goat IgG and donkey anti-mouse IgG antibodies (Molecular Probes, Carlsbad, CA) and with antibodies to CD8, CD45RA, and CCR7 (BD PharMingen, San Jose, CA). Some cells were additionally stained with antibodies to CD27 (e-bioscience, San Diego, CA), CD28, CD57, perforin, or isotype controls (all from BD PharMingen). In measuring intracellular STAT5 and phospho-STAT5 (P-STAT5), PBMCs were first stained with antibodies to CD8, CD45RA, and CCR7. Cells were then stimulated for 10 minutes with recombinant human IL-7 (10 pg/mL-10 ng/mL; BD PharMingen) or PBS and were stained with antibodies to STAT5 (Santa Cruz Biotechnology, Santa Cruz, CA), P-STAT5 (BD PharMingen), or isotype antibodies.

In determining Bcl-2 and cell survival, PBMCs were sorted into naive (CD45RA⁺CCR7⁺), central memory (CM, CD45RA^-CCR7⁺), effector memory (EM, CD45RA^-CCR7^-), and EM_CD45RA+ (CD45RA⁺CCR7^-) CD8⁺ T cells using a FACSAria (BD Immunocytometry, San Jose, CA). Cells were incubated for 6 days in the presence of recombinant human IL-7 (100 ng/mL) or PBS. Cells were stained with antibodies to Bcl-2 (BD PharMingen) and with annexin V and 7-AAD using a commercially available kit (BD PharMingen). In measuring cell proliferation, sorted cells were labeled with carboxyfluorescein diacetate (CFSE; Molecular Probes, Eugene, OR), as described previously.²¹  Cells were incubated for 6 days in a 96-well tissue culture plate coated with anti-CD3 antibodies (BD PharMingen) at 10 μg/mL or with PBS in the presence of anti-CD28 antibodies (10 μg/mL; BD PharMingen). Stained cells were analyzed on a FACSCalibur or an LSRII (BD Biosciences Immunocytometry Systems). Collected flow cytometry data were analyzed using FlowJo software (Tree Star, Ashland, OR).

Plasma levels of IL-15 and IL-21 were determined using commercially available enzyme-linked immunosorbent assay (ELISA) kits (R&D Systems and Caprologics, Hardwick, MA, respectively) according to each manufacturer's instructions.

Total RNA was isolated from sorted cells and used for cDNA synthesis. Polymerase chain reaction (PCR) was performed by 1 × SYBR Green mix (Qiagen, Valencia, CA). All results were normalized to β-actin mRNA. Primers used in real-time PCR were as follows: IL-7Rα forward, 5′-TGGACGCATGTGAATTTATC-3′, and reverse, 5′-CATTCACTCCAGAAGCCTTT-3′; GABPα forward, 5′-AGCATCAGTGCAATCTGCTA-3′, and reverse, 5′-TTCCCAGGTGAGCTTCTATC-3′; GFI1 forward, 5′-TGACTTGGGGAAGGAATTTA-3′, and reverse, 5′-CCAGTGATGAGGTTTTCACA-3′.

CDR3 length distribution analysis was performed using approximately 1.5 × 10⁵ cells from each sorted CD8⁺ T-cell subset, as previously described.²²  Briefly, cDNA was amplified by PCR through 35 cycles (94°C for 1 minute, 55°C for 1 minute, and 72°C for 1 minute) with primers specific to 24 different Vβ families and a 6-FAM-labeled BC primer.^23,24  Fluorescent PCR products were analyzed using an automatic sequencer with GeneMapper software (Applied Biosystems, Foster City, CA).

Mann-Whitney U and Student 2-tailed t tests were used for analyses of differences between the young and the elderly. A correlation between the frequencies of IL-7Rα^high cells and P-STAT5^high cells was determined using Pearson correlation. All statistical analyses were performed using SPSS 12.0 (SPSS, Chicago, IL).

We determined the expression of IL-7Rα and γc by CD8⁺ T-cell subsets using flow cytometry. In accordance with our published data,²⁰  the frequency (mean ± SEM) of naive CD8⁺ T cells was decreased and the frequency of CM CD8⁺ T cells was similar between elderly and young subjects (naive, 5.32% ± 1.0% vs 35.4% ± 2.6%; CM, 5.31% ± 0.7% vs 5.4% ± 0.6%; P < .001 and P = .879, respectively, by Student t test). The frequencies of EM and EM_CD45RA+ CD8⁺ T cells were higher in the elderly than in the young (EM, 52.7% ± 3.6% vs 36.1% ± 2.6%; EM_CD45RA+, 36.7% ± 3.6% vs 23.1% ± 1.9%; P = .003 and P = .001, respectively, by Student t test). Young and elderly subjects had homogeneous expression of IL-7Rα by naive and CM CD8⁺ T-cell subsets (Figure 1A). In contrast, such receptor expression by EM and EM_CD45RA+ CD8⁺ T-cell subsets was heterogeneous (Figure 1A). Thus, we compared the median fluorescence intensity (MDFI) of IL-7Rα expression by the 4 cell subsets between the young and the elderly. Elderly subjects had decreased IL-7Rα expression by naive and EM_CD45RA+ CD8⁺ T-cell subsets compared with young subjects (Figure 1B; MDFI ± SEM, 215.8 ± 8.9 vs 252.6 ± 8.6 and 77.6 ± 18.6 vs 176.5 ± 28.6, respectively). However, the 2 groups had similar IL-7Rα expression by CM- and EM-cell subsets (Figure 1B). The expression of γc by the 4 CD8⁺ T-cell subsets was also similar between the young and the elderly (Figure 1C). We analyzed the frequency of IL-7Rα^high and IL-7Rα^low cells in EM and EM_CD45RA+ CD8⁺ T-cell subsets in both groups because these 2 subsets had heterogeneous expression of IL-7Rα (Figure 1A). Of interest, the frequency of IL-7Rα^high cells was decreased and the frequency of IL-7Rα^low cells was increased in elderly subjects compared with young subjects (Figure 1D; IL-7Rα^high cells, 37.5% ± 7.4% vs 59.1% ± 5.4%). However, in the EM CD8⁺ T-cell subset, the frequency of IL-7Rα^high and IL-7Rα^low cells was not different between the 2 groups (data not shown). These findings indicate that an age-associated alteration in IL-7Rα expression occurs in naive and EM_CD45RA+ CD8⁺ T-cell subsets and that such a change in the latter subset stems from an altered frequency of IL-7Rα^high and IL-7Rα^low cells with aging.

We next addressed whether altered IL-7Rα expression by naive and EM_CD45RA+ CD8⁺ T-cell subsets in the elderly had a physiologic consequence in cell signaling and survival. We stimulated PBMCs with IL-7 and measured P-STAT5 in the CD8⁺ T-cell subsets using flow cytometry. In naive and CM subsets, the phosphorylation of STAT5 was homogeneous (Figure 2A). In contrast, EM and EM_CD45RA+ subsets with IL-7Rα^high and IL-7Rα^low cells had 2 different T-cell populations with P-STAT5^high and P-STAT5^low, reflecting their IL-7Rα expression (Figure 2A). In the latter 2 subsets, a strong correlation between the frequencies of IL-7Rα^high cells and P-STAT5^high cells was observed in young and elderly subjects (Figure 2B). A similar correlation was observed between the frequencies of IL-7Rα^low cells and P-STAT5^low cells in the same subsets (data not shown).

Elderly subjects had significantly fewer P-STAT5 in the EM_CD45RA+ subsets than did young subjects (Figure 2C). This difference was greater when higher doses of IL-7 were used for cell stimulation. A similar trend was observed in the naive subset but did not reach statistical significance (Figure 2C). In contrast, for CM and EM subsets, both groups had comparable levels of P-STAT5. There was no difference in total STAT5 among the 4 subsets in young and elderly subjects as measured by flow cytometry (data not shown). These findings demonstrate that the magnitude of IL-7Rα expression by CD8⁺ T cells dictates the extent of subsequent cell signaling and that an age-associated alteration in IL-7Rα expression by the EM_CD45RA+ CD8⁺ T-cell subset impairs IL-7 signaling.

To investigate whether an alteration in IL-7Rα expression and STAT5 phosphorylation in CD8⁺ T-cell subsets could affect Bcl-2 regulation and cell survival in elderly subjects, Bcl-2 up-regulation and cell survival were also determined in sorted CD8⁺ T-cell subsets in response to IL-7. As with IL-7Rα expression and STAT5 phosphorylation, Bcl-2 was up-regulated homogeneously in naive and CM subsets in response to IL-7 (Figure 2D). However, in EM and EM_CD45RA+ CD8⁺ T-cell subsets, Bcl-2 up-regulation was heterogeneous, reflecting IL-7Rα expression by these cell subsets. The up-regulation of Bcl-2 in EM_CD45RA+ CD8⁺ T-cell subset was lower in elderly subjects (Figure 2D). We next measured cell survival in young and elderly subjects from sorted CD8⁺ T-cell subsets incubated for 6 days with IL-7 (100 ng/mL) or PBS. Cells in the 4 subsets incubated with IL-7 had higher numbers of live cells than cells incubated with PBS. In the EM_CD45RA+ CD8⁺ T-cell subset, the increase in the frequency of live cells was lower in elderly than in young subjects (Figure 2E; 19.9% ± 2.4% vs 26.8% ± 1.01%), whereas the frequency of such cells in the other 3 subsets increased similarly in both groups. These findings indicated that elderly subjects had decreased cell survival responses to IL-7 in the EM_CD45RA+ CD8⁺ T-cell subset that stemmed from decreased IL-7Rα expression, STAT5 phosphorylation, and Bcl-2 up-regulation.

To investigate the mechanisms for increased frequency of EM_CD45RA+ IL-7Rα^low CD8⁺ T cells with aging, we determined the cellular characteristics of the CD8⁺ T-cell subsets. In young and elderly subjects, EM and EM_CD45RA+ IL-7Rα^low CD8⁺ T cells were largely antigen experienced and perforin^high (CD27^-CD28^- perforin^high) with replicative senescence (CD57⁺)^25,26  (Figure 3A-C). Such molecular changes were more notable in elderly subjects (Figure 3A-C). In measuring the proliferation of EM_CD45RA+ IL-7Rα^high and IL-7Rα^low CD8⁺ T cells in response to anti-CD3 and anti-CD28 antibodies, IL-7Rα^low cells proliferated minimally whereas IL-7Rα^high cells proliferated to the same extent as naive CD8⁺ T cells in elderly (Figure 3D) and young (data not shown) subjects. These findings suggest that the increased frequency of EM_CD45RA+ IL-7Rα^low CD8⁺ T cells likely stems from the age-related accumulation of antigen-experienced and replicatively senescent CD8⁺ T cells. This notion is further supported by studies showing the down-regulation of IL-7Rα by TCR triggering in T cells,²⁷  which we also observed (Figure 4A-B). Of interest, CD57⁺ cells in the EM_CD45RA+ IL-7Rα^high CD8⁺ T-cell subset proliferated significantly in response to anti-CD3 and anti-CD28 antibodies, whereas cells expressing the same molecule in the EM_CD45RA+ IL-7Rα^low CD8⁺ T-cell subset proliferated poorly (Figure 3D). This finding suggests that some CD57⁺ CD8⁺ T cells are functional in terms of cell proliferation and that IL-7Rα can be a marker for identifying such cells. We also measured phosphorylated STAT5 signaling and survival rates of IL-7Rα^high and IL-7Rα^low cells in the EM_CD45RA+ CD8⁺ T-cell subset in young and elderly subjects after stimulating cells with IL-7. As expected, IL-7Rα^high cells had better survival rates (Figure 3E) and phosphorylation signaling (data not shown) in response to IL-7 than did IL-7Rα^low cells, which further supports the role of IL-7Rα in promoting CD8⁺ T-cell survival in response to IL-7.

IL-7Rα^low cells in the EM and EM_CD45RA+ CD8⁺ T-cell subsets were maintained despite these cells having impaired proliferation to TCR stimulation (Figure 3D) and decreased IL-7Rα expression, critical for cell survival. IL-15 and IL-21 are also members of the γc cytokine family and potently promote the proliferation of CD8⁺ T cells.^28-30  Thus, we investigated whether IL-7Rα^low cells in the EM and EM_CD45RA+ CD8⁺ T-cell subsets had altered expression of IL-15Rα, IL-2/15Rβ, and IL-21R compared with IL-7Rα^high cells. In young subjects, the expression of IL-15Rα by IL-7Rα^high cells in the EM and EM_CD45RA+ CD8⁺ T-cell subsets was higher than by IL-7Rα^low cells, but the difference was smaller in elderly subjects (Figure 5A). This suggests that the use of IL-15 by IL-7α^high and IL-7α^low cells through IL-15Rα is similar in the elderly, though such cytokine use by IL-7Rα^high cells can be higher than by IL-7Rα^low cells in the young. The expression of IL-2/15Rβ by EM_CD45RA+ IL-7Rα^low CD8⁺ T cells was higher (young subjects) and similar (elderly subjects) compared with IL-7Rα^high cells (Figure 5B), suggesting that the use of IL-15 by EM_CD45RA+ IL-7Rα^low cells through IL-2/IL-15Rβ is at least comparable to that of IL-7Rα^high cells. The expression of IL-21R by IL-7Rα^high and IL-7α^low cells in EM and EM_CD45RA+ CD8⁺ T-cell subsets was similar in the 2 groups (Figure 5C). Because any alteration in IL-15 and IL-21 levels can affect CD8⁺ T-cell homeostasis, we also measured plasma levels of these cytokines in both groups of subjects. They had comparable plasma levels (IL-15, 1.75 pg/mL ± 0.15 vs 1.49 pg/mL ± 0.20; IL-21, 225 ng/mL ± 77 vs 272 ng/mL ± 73; P = .322 and P = .676, respectively), indicating the unaltered age-related availability of such cytokines in blood. These findings suggest that IL-7Rα^low CD8⁺ T cells in the elderly have unaltered IL-15- and IL-21-mediated CD8⁺ T-cell homeostasis than do IL-7Rα^high cells, which may account for the maintenance of IL-7Rα^low CD8⁺ T cells in vivo.

We investigated how EM_CD45RA+ IL-7Rα^low CD8⁺ T cells maintain low levels of IL-7Rα expression in homeostasis by measuring the gene expression of IL-7Rα and 2 transcriptional factors, guanine and adenine binding protein α (GABPα) and growth factor independence-1 (GFI1), in sorted CD8⁺ T-cell subsets. GABPα is required for the expression of IL-7Rα in T cells,³¹  whereas GFI1 is involved in suppressing IL-7Rα expression in mouse lymph node CD8⁺ T cells in response to IL-7.³²  In our study, IL-7 withdrawal by the maintenance of cells for 24 hours in tissue culture medium—a known condition up-regulating the expression of IL-7Rα—increased IL-7Rα and GABPα mRNA expression by naive and EM_CD45RA+ IL-7Rα^high CD8⁺ T cells (Figure 4A-B). In contrast, TCR stimulation for 24 hours with PHA suppressed IL-7Rα and GABPα mRNA expression by the same cell subsets compared with samples stimulated for 24 hours with PBS (Figure 4A-B). These findings indicate that GABPα is likely involved in the regulation of IL-7Rα expression in human CD8⁺ T cells in response to stimulations altering IL-7Rα expression. Surprisingly, IL-7 withdrawal increased the expression of GFI1, a suppressor of IL-7Rα expression in mice, by naive and EM_CD45RA+ IL-7Rα^high CD8⁺ T cells. However, TCR stimulation did not change or down-regulated the expression of GFI1 by the same cell subsets compared with samples treated for 24 hours with PBS in tissue culture media (Figure 4A-B). These findings suggest that the role of GFI1 in regulating IL-7Rα expression in human CD8⁺ T cells may not be the same as in mouse lymph node T cells.³² 

We next determined mRNA expression of IL-7Rα, GABPα, and GFI1 in freshly sorted naive EM_CD45RA+ IL-7Rα^high and IL-7Rα^low cells in elderly subjects. The expression of IL-7Rα mRNA was lower in EM_CD45RA+ IL-7Rα^low CD8⁺ T cells than in other subsets (Figure 6). However, GABPα mRNA expression was not statistically different among naive, EM_CD45RA+ IL-7Rα^high, and IL-7Rα^low cells, suggesting that this transcriptional factor was not accountable for the suppressed IL-7Rα mRNA expression in EM_CD45RA+ IL-7Rα^low cells in homeostasis. Of interest, GFI1 mRNA expression was higher in EM_CD45RA+ IL-7Rα^high cells than in naive and EM_CD45RA+ IL-7Rα^low cells, which contradicted the previous study reporting the suppressor role of GFI1 in regulating IL-7Rα expression in mouse lymph node T cells.³²  These observations suggest that the mechanism involved in maintaining the low levels of IL-7Rα mRNA expression in EM_CD45RA+ IL-7Rα^low CD8⁺ T cells in homeostasis are different from the known IL-7Rα regulatory mechanisms.

Data, including ours, indicate that survival of the CD8⁺ T-cell subsets can be promoted by IL-7.⁷  Therefore, IL-7 therapy can be considered to improve the development and survival of naive and memory CD8⁺ T cells. However, before attempting such therapy, it is important to determine the diversity of the TCR repertoire in IL-7Rα^high and IL-7Rα^low CD8⁺ T cells because a restriction in the TCR repertoire with oligoclonal CD8⁺ T-cell expansion occurs with aging¹³  and such changes can impair the development of CD8⁺ T-cell immune responses to new antigens.³³  In our study, young and elderly subjects had diverse TCR repertoires in naive CD8⁺ T cells but limited TCR repertoires in EM_CD45RA+ IL-7Rα^low CD8⁺ T cells (Figure 6A-B). However, elderly subjects had a more limited TCR repertoire in EM_CD45RA+ IL-7Rα^high CD8⁺ T cells than did young subjects (Figure 7A-B). These findings suggest that IL-7 therapy may enhance the survival of EM_CD45RA+ IL-7Rα^high CD8⁺ T cells with a diverse TCR repertoire in the young but not in the elderly, though such therapy may increase the survival of naive CD8⁺ T cells with a diverse TCR repertoire in both groups.

This is the first study demonstrating an age-associated alteration in the frequency of EM_CD45RA+ IL-7Rα^high and IL-7Rα^low CD8⁺ T cells, leading to decreased cell signaling and survival responses to IL-7. In measuring IL-7Rα expression, naive and CM CD8⁺ T-cell subsets had homogeneous expression of IL-7Rα (Figure 1A), whereas 2 different cell populations, IL-7Rα^high and IL-7Rα^low, were identified in EM and EM_CD45RA+ CD8⁺ T-cell subsets. Of interest, elderly subjects had decreased expression of IL-7Rα by naive and EM_CD45RA+ CD8⁺ T cells. Such a change in the latter subset was secondary to an increased frequency of IL-7Rα^low cells (Figure 1B, D), which were largely late differentiated (CD27^-CD28^-) and replicatively senescent cells (CD57⁺) (Figure 3A-C).^25,26  Consistent with the expression of CD27 and CD28, costimulatory molecules for T-cell proliferation,²⁵  the proliferative response was markedly reduced in EM_CD45RA+ IL-7Rα^low cells compared with EM_CD45RA+ IL-7Rα^high cells (Figure 3D). These findings indicate that IL-7Rα^low cells are generated as a result of antigenic stimulation because TCR stimulation suppresses CD27 and CD28 expression.²⁵  This notion is further supported by the results of studies, including ours, showing the down-regulation of IL-7Rα by TCR stimulation (Figure 4A-B)^27,34  and the expansion of IL-7Rα^low CD8⁺ T cells in persons infected with HIV who have chronic immune activation.³⁵  Overall, the results of our study demonstrate that the increased frequency of EM_CD45RA+ IL-7Rα^low CD8⁺ T cells in the elderly is likely related to an accumulation of antigen-experienced and replicatively senescent CD8⁺ T cells with aging.

An intriguing question is how IL-7Rα^low cells in EM and EM_CD45RA+ CD8⁺ T-cell subsets can be maintained in vivo despite low levels of IL-7Rα expression that is critical for cell survival. We addressed this question by measuring the expression of IL-15Rα, IL-2/15Rβ, and IL-21R by IL-7Rα^high and IL-7Rα^low cells. IL-15 and IL-21 potently induce the proliferation of memory CD8⁺ T cells,^30,36-39  whereas IL-7 predominantly promotes the survival of these cells.^3-5,7,8  In our study, IL-7Rα^low cells in the EM_CD45RA+ CD8⁺ T-cell subsets had increased (young) or similar (elderly) expression of IL-2/15Rβ compared with IL-7Rα^high cells (Figure 5B). This finding suggests that IL-7Rα^low cells can be maintained in vivo through the use of IL-15 despite the decreased IL-7Rα expression on these cells. In fact, the crucial role of IL-2/15Rβ in IL-15-mediated cell signaling and in maintaining memory CD8⁺ T cells has been reported.^40,41  This notion is further supported by the finding that IL-7Rα^low CD8⁺ T cells survive better in IL-15 intact mice than in IL-15 KO mice (S. Kaech and Nikhil Joshi, Yale University, New Haven, CT; unpublished observation, June 2005).

To measure the physiologic significance of IL-7Rα expression changes with aging, we investigated CD8⁺ T-cell signaling and survival in response to IL-7. First, we measured the phosphorylation of STAT5 in CD8⁺ T-cell subsets in response to IL-7 (Figure 2A). The effect of IL-7 on STAT5 phosphorylation in CD8⁺ T cells appears to be dependent on 2 factors, the quantity of IL-7Rα expressed on cells and the strength of IL-7 stimulation. The importance of the former factor is demonstrated by the correlation between the frequencies of IL-7Rα^high cells and P-STAT5^high cells in EM and EM_CD45RA+ CD8⁺ T-cell subsets (Figure 2A-B). The quantitative role of IL-7Rα expression in determining the extent of STAT5 phosphorylation is further supported by the results of our experiment showing reduced P-STAT5 in IL-7-stimulated EM_CD45RA+ CD8⁺ T cells in elderly subjects that contained a higher number of IL-7Rα^low cells than in young subjects. The phosphorylation of STAT5 in the CD8⁺ T-cell subsets also correlated with the strength of IL-7 stimulation in the young. Although a similar correlation between the 2 parameters in naive, CM, and EM CD8⁺ T-cell subsets was observed in the elderly, the phosphorylation of STAT5 in the EM_CD45RA+ CD8⁺ T-cell subset reached a plateau at 1 ng/mL IL-7 (Figure 2C). This suggests that even high-dose IL-7 cannot overcome the physiologic effect of the decreased IL-7Rα expression by EM_CD45RA+ IL-7Rα^low CD8⁺ T cells in the elderly.

Bcl-2 is an antiapoptotic molecule that is critically involved in the IL-7-mediated cell-survival pathway.⁷  The results of our study demonstrate that this correlation of IL-7Rα and bcl-2 up-regulation also occurs in human peripheral CD8⁺ T cells. Similar to STAT5 phosphorylation, Bcl-2 up-regulation in EM_CD45RA+ CD8⁺ T cells was lower in elderly subjects than in young subjects (Figure 2D). Furthermore, the survival response of EM_CD45RA+ CD8⁺ T cells on IL-7 stimulation was lower in elderly than in young subjects (Figure 2E). These findings indicate the physiologic impact of the decreased IL-7Rα expression and STAT5 phosphorylation by this cell subset with aging on subsequent Bcl-2 up-regulation and cell survival.

The regulatory mechanisms for IL-7Rα expression in T cells are largely unknown. Recent studies reported a role of 2 transcriptional factors, GABPα and GFI1, in regulating IL-7Rα expression in human and mouse T cells.^31,32  GABPα regulated IL-7Rα expression in T cells by binding to the GGAA motif in the IL-7R promoter in mouse splenic T cells.³¹  In addition, IL-7Rα expression in human and mouse T cells was suppressed by siRNA targeting of GABPα. In contrast to GABPα, GFI1 suppressed the expression of IL-7Rα in mouse lymph node T cells stimulated with prosurvival cytokines including IL-7.³²  Based on these observations, we asked how IL-7Rα expression was kept low in EM_CD45RA+ IL-7Rα^low cells in vivo. First, we confirmed the correlation between IL-7Rα mRNA and protein expressions in naive and EM_CD45RA+ IL-7Rα^high and IL-7Rα^low CD8⁺ T-cell subsets, indicating that decreased IL-7Rα protein expression by IL-7Rα^low cells was directly related to the suppression of gene expression (Figure 6). Of interest, GABPα expression was not different among the 3 CD8⁺ T-cell subsets despite the difference in IL-7Rα mRNA expression (Figure 6). This suggests the limited role of GABPα in regulating IL-7Rα mRNA expression in primary human CD8⁺ T cells in homeostasis. The result of measuring GFI1 in sorted naive, EM_CD45RA+ IL-7Rα^high, and IL-7Rα^low cells was surprising. GFI mRNA expression was higher in EM_CD45RA+ IL-7Rα^high cells than in IL-7Rα^low cells (Figure 6), which contradicted the previously reported suppressor role of this transcriptional factor in regulating IL-7Rα expression in mouse lymph node T cells.³²  In fact, in our study, the expression of GFI1 was also unaltered or suppressed in naive and EM_CD45RA+ IL-7Rα^high CD8⁺ T cells after 24-hour TCR stimulation, a known condition that down-regulates IL-7Rα expression,²⁷  compared with cells stimulated for 24 hours with PBS (Figure 4A-B). This finding further questions the suppressor role of GFI1 in human peripheral CD8⁺ T cells. There are several possible explanations for such a difference between our findings and the results of the previous studies on GABPα and GFI1. This difference could be secondary to using T cells from different species or activation status. More likely, in homeostasis, the low levels of IL-7Rα mRNA expression in EM_CD45RA+ IL-7Rα^low CD8⁺ T cells are maintained by an unknown factor(s) involved in the IL-7Rα regulatory mechanism.

The reduced TCR repertoire of CD8⁺ T cells with oligoclonal expansions has been observed with aging.¹³  In our study, EM_CD45RA+ IL-7Rα^low CD8⁺ T cells had reduced TCR diversity compared with naive cells in young and elderly subjects. However, the TCR diversity of EM_CD45RA+ IL-7Rα^high CD8⁺ T cells was different between the young and the elderly groups. The diversity of such cells in elderly subjects was limited, suggesting oligoclonal expansions. This finding prompts caution in attempting IL-7 therapy in the elderly because such therapy may increase the survival of EM_CD45RA+ IL-7Rα^high CD8⁺ T cells with a limited TCR repertoire and possible oligoclonal expansions that may not be beneficial to the immune system,³³  although it can still increase the survival of naive CD8⁺ T cells with a diverse TCR repertoire. Indeed, a recent study³³  demonstrated the impaired development of CD8⁺ T-cell immune responses to herpes simplex virus-1 (HSV-1) infection in aged mice with oligoclonal CD8⁺ T-cell expansions. However, in the young, IL-7 therapy can be useful in promoting the survival of naive and EM_CD45RA+ IL-7Rα^high CD8⁺ T cells with a diverse TCR repertoire, leading to improved immune responses to new antigens. In fact, this notion is supported by a recent mouse study¹⁹  reporting an adjuvant role of IL-7 in augmenting CD8⁺ T-cell immune responses to subdominant antigens and improving the survival of memory CD8⁺ T cells.

In summary, our study demonstrates an age-associated alteration in the frequency of EM_CD45RA+ IL-7Rα^high and IL-7Rα^low CD8⁺ T cells that impairs cell signaling and survival responses to IL-7. Such an alteration stems likely from age-long antigenic CD8⁺ T-cell stimulation, leading to the down-regulation of IL-7Rα mRNA synthesis and maintained by an unaltered use of IL-15 through IL-15Rα and IL-2/15Rβ expression. Our findings also suggest that IL-7 therapy in the elderly may not be as effective and beneficial as in the young because a large proportion of CD8⁺ T cells in the former group has decreased IL-7Rα expression and limited TCR repertoire compared with the latter group. Overall, the results of our study are unique in that they provide new insight into aging and IL-7-mediated CD8⁺ T-cell homeostasis and into the potential implications of IL-7 therapy in the elderly.

We thank Drs Susan Kaech and Alexia Belperron for critical reviews of this manuscript and Lynne Iannone, Christine Bailey, and Barbara Foster for assistance in the recruitment of human subjects.