IL-7 and Not Stem Cell Factor Reverses Both the Increase in Apoptosis and the Decline in Thymopoiesis Seen in Aged Mice¹

Atrophy of the thymus has been identified as one of the key events that precedes inefficient functioning of the immune system in later life. As thymic T cell production diminishes with age, a decline in contribution made by thymic emigrants to the naive T cell pool occurs (1, 2). Residence within the naive T cell pool is finite (3), and naive T cells depend on TCR ligation with MHC for their survival, whereas their expansion requires Ag (4). Diminution in the size of the naive T cell pool is a common finding with aging (5, 6) and is a consequence of reduced thymic output. Compensatory maintenance of T cell numbers is thought to come from expansion of the renewable memory T cell pool.

Stem cells from the bone marrow continually supply the thymus well into old age (7, 8). The αβ T cells are produced through a series of developmental steps involving survival, proliferation, and differentiation of these stem cell precursors driven by cell-cell interactions, growth factors, and cytokines provided by the thymic microenvironment (9, 10). In the adult mouse thymus, the earliest precursors of the T cell pathway are within the population defined phenotypically as CD3⁻CD4^lowCD8⁻ (11). Because CD4 knockout mice show normal progression through this early phase of T cell development, the expression of CD4 does not appear to have a fundamental role; therefore, these early precursors can be considered functionally triple negative (TN)³ (12). This TN population has been subdivided on the basis of expression of CD44 and CD25, with the most immature stage of development identified with the phenotype CD44⁺CD25⁻ TN (13). The following stage is defined by acquisition of CD25 and represents a highly proliferating population (14). The subsequent loss of CD44 expression is accompanied by extensive TCR β-chain rearrangement (15). The TCR β-chain forms and associates with an invariant pre-Tα molecule forming the pre-TCR complex (16). This pre-TCR complex drives expansion via the CD44⁻CD25⁺ TN through the CD44⁻CD25⁻ TN stage to the CD4⁺CD8⁺ double-positive stage (17).

Analysis of aged mice reveals that there is a bottleneck in thymocyte production between the multipotent stem cell progenitor stage (CD44⁺CD25⁻ TN) and their progeny (CD44⁺CD25⁺ TN) that have become committed to the T cell lineage. The number of CD44⁺CD25⁻ TN thymocytes does not alter with age; however, all subsequent subsets show markedly reduced numbers. This, along with evidence from F₅-transgenic mice and recombination-activating gene (RAG)^−/− F₅-transgenic mice, suggests that this age-associated decline in commitment to the T cell lineage is a result of problems with rearrangement of the TCR β-chain (18).

IL-7 is a cytokine provided by the thymic cortical epithelial cells, and interaction with the IL-7R is important for TN cell development (19, 20). Evidence suggests that the decline in IL-7 expression may limit thymocyte development by restricting combinations of survival, proliferation, and rearrangement of the TCR β-chain (reviewed in Ref. 21). Stem cell factor (SCF) is also present in the thymus, produced by stromal cells (20) as a transmembrane protein on the stromal cell surface and as a secreted soluble molecule generated by differential splicing (22). Interactions between SCF and c-kit also play a major role in TN thymocyte development and have been shown to promote the proliferation of immature thymocytes in vivo (23). Thymopoiesis is limited in IL-7^−/−, IL-7Rα^−/−, IL-7Rγ^−/−, and c-kit^−/− mice (24, 25, 26, 27), and c-kit^−/−common γ-chain (γ_c)^−/− mice demonstrate complete abrogation of T cell development (27). The reduced development of the early stages of T cell development observed in IL-7^−/−, IL-7Rα^−/−, IL-7Rγ^−/−, and c-kit^−/− mice is similar to that seen with aging. Clearly, both IL-7 and SCF have an important role in the early stages of the developmental pathway, and both may be required to renew thymopoiesis in old animals.

Therefore, the aim of this study was to investigate whether age-associated alterations in TN survival could account for the decline in thymocyte numbers with age and identify the effect of IL-7 and SCF on the survival of TN cells from aged mice. Finally, this paper investigated the effect of IL-7 and SCF therapy on TN survival in vivo and their impact on thymopoiesis in aged mice.

Normal C57BL/10 mice were obtained from Harlan Olac (Oxfordshire, U.K.) and were maintained in the animal house at the Imperial College School of Medicine in accordance with local rules and regulations.

Young (2–3 mo) or old (22–26 mo) mice were sacrificed by CO₂ asphyxiation, their thymi were removed, and thymocyte cell suspensions were prepared by pressing the tissue through a 100-μm cell strainer (Becton Dickinson, Oxford, U.K.) into RPMI 1640 medium (Life Technologies, Paisley, U.K.) supplemented with 10% FCS (Sigma, Dorset, U.K.). Erythrocytes were lysed using Ortholyse (Ortho, Amersham, U.K.), and the total thymocyte number was counted using a hemocytometer.

Thymocytes were incubated for 15 min at 6–12°C with primary Abs; anti-CD3-biotin (clone KT3; Serotec, Oxford, U.K.), anti-CD4 (clone YTS 191.1; Serotec), anti-CD8 (clone YTS 169.4; a kind gift of Dr. B. Roser, Anglia Polytecnic University, Cambridge, U.K.), anti-CD19 (clone 6D5; Serotec), and F4/80 (clone C1:A3-1; Serotec) at a concentration of 1 μg/10⁶ cells in ice-cold MACS buffer (5 mM EDTA, 1% BSA in PBS). Cells were washed in MACS buffer and then indirectly labeled with goat anti-rat IgG MACS microbeads and streptavidin MACS microbeads (Miltenyi Biotec, Bisley, U.K.) for 15 min at 4°C and then negatively selected on a MACS system. The CD3⁻CD4⁻CD8⁻ (TN) thymocytes were counted and resuspended in RPMI 1640 medium (Life Technologies) supplemented with 10% FCS (Sigma), l-glutamine (200 mM), penicillin (100 IU/ml), streptomycin (100 μg/ml) (Sigma), and 5 × 10⁻⁵ M 2-ME (Life Technologies). TN thymocytes (1–2 × 10⁵ cells/200 μl) were cultured at 37°C in a 96-well U-bottom plate (Greiner Labortechnik, Glos, U.K.) in the presence of either PBS, IL-7, SCF, or both IL-7 and SCF (PeproTech EC, London, U.K.). Titration experiments were performed showing the percentage of live cells with IL-7 or SCF within a 0–500 ng/ml concentration range, and a concentration of 50 ng/ml, a concentration previously used by Kim (28), was shown to be mid-plateau for IL-7. SCF did not improve the percentage of live cells from control experiments at any concentration within this range; therefore, 50 ng/ml was chosen for convenience.

For the analysis of apoptosis within the TN thymocyte subpopulations, a four-color FACS analysis was performed at 0 h, 24 h, and 5 days of culture. Cells were harvested, washed, resuspended in PBS, and stained with anti-CD44-APC (clone IM-7; PharMingen, Oxford, U.K.), anti-CD25-R-PE (clone AMT-13; Sigma), and control Abs conjugated to APC and R-PE (PharMingen) for 20 min on ice. Cells were washed and resuspended in annexin V binding buffer (PharMingen) and stained with annexin V-FITC and 7-amino actinomycin D (7-AAD) viability probe (PharMingen) for 15 min at room temperature in the dark. Cells were analyzed on a Becton Dickinson FACSCalibur within 1 h on a program acquiring 10,000 cells.

For analysis of IL-7R expression, four-color FACS analysis was performed immediately post-TN purification. Cells were washed, resuspended in PBS, and stained with anti-CD44-CyChrome (clone IM-7; PharMingen), anti-CD25-FITC (clone AMT-13; Sigma), anti-IL-7Rα-biotin (Research Diagnostics, Flanders, NJ), streptavidin-R-PE (Serotec), and control Abs conjugated to CyChrome, FITC, and biotin (PharMingen) for 20 min on ice. Cells were washed in PBS, resuspended in 1% paraformaldehyde, and analyzed on a Becton Dickinson FACSCalibur within 5 days of fixation on a program acquiring 10,000 cells. All results were analyzed with WinMDI v.2.7 (Scripps Research Institute, La Jolla, CA).

Groups of male mice aged 22–23 mo were injected s.c. once a day with one of the following treatment regimens: 1) 1 μg of carrier-free recombinant murine IL-7 (R&D Systems Europe, Oxford, U.K.) per day (an approximate concentration of 25 μg/kg/day) in PBS, representing a per diem dose similar to that used previously (29); 2) 4 μg of carrier-free recombinant murine SCF (PeproTech EC) per day (an approximate concentration of 100 μg/kg/day) in PBS, a dose used previously (30); 3) 1 μg IL-7 and 4 μg SCF per day in PBS; and 4) PBS alone. This regimen was followed for 4 days, and on day 5 the animals were sacrificed and the thymi were removed and analyzed.

Mice were sacrificed, thymi were removed, and thymocytes were counted as described above. Apoptotic cells were selected using the annexin V microbead apoptotic cell isolation kit (Miltenyi Biotec, Bisley, U.K.) according to the manufacturer’s instructions. For the analysis of the TN thymocyte subpopulation, cells were stained with biotin-conjugated anti-CD4 (clone KT6), biotin-conjugated anti-CD3 (clone KT3), streptavidin conjugated to R-PE, anti-CD8-PE (clone 53-6.7), anti-CD44-CyChrome (clone IM7), and anti-CD25-FITC (clone7D4). Control Abs were conjugated to PE, FITC, biotin, or CyChrome. Cells were fixed poststaining with 1% paraformaldehyde in PBS and analyzed on a Becton Dickinson FACSCalibur within 5 days of fixation on a program acquiring 50,000 cells. The results were analyzed using WinMDI v.2.7 (Scripps Research Institute). The number of live cells in each TN subpopulation was calculated by subtracting the number of apoptotic cells from the total number of each subpopulation.

Comparison of samples was conducted using a two-tailed t test for samples with unequal variance using Microsoft Excel software (Redmond, WA). Differences were considered significant for p < 0.05.

To establish whether an increase in apoptosis within the TN pathway occurs with age, TN thymocytes were purified from young and old thymi and identified as follows. Apoptotic cells were detected by binding annexin V to their plasma membrane, live cells by exclusion of annexin V and 7-AAD and dead cells by binding annexin V and 7-AAD. As shown in Fig. 1, there was a significant increase in the percentage of apoptotic TN thymocytes with age. To identify whether this increase could be attributed to a particular stage in TN development, the apoptotic profile within each of the four TN subpopulations was analyzed. Fig. 2 reveals that only the apoptotic profile of the CD44⁺CD25⁺ and CD44⁻CD25⁺ subpopulations significantly differed between young and old animals. The CD44⁺CD25⁺ and CD44⁻CD25⁺ populations demonstrated a significant age-associated decrease in live cells and an increase in apoptotic cells, thus locating the age-associated increase in apoptosis to the stages of the TN pathway associated with initiation of TCR β-chain rearrangement.

To examine whether this age-associated increase in apoptosis could be reversed with cytokines known to have a central role at this stage of thymocyte development, TN thymocytes from young and old donors were cultured with IL-7, SCF, or a combination of IL-7 and SCF for 1 and 5 days. Fig. 3 reveals that TN cells from both young and old donors were protected from apoptosis in the presence of IL-7 after 24 h, with protection clearly still evident after 5 days in culture. Cultures with IL-7 established from young donors showed significant increases in live TN cells at 1 (p < 0.05) and 5 days (p < 0.01). This was paralleled by significant decreases in total annexin V⁺ (cells that have entered apoptosis or died via apoptosis) TN cells at both time points. Cultures established from young donors with IL-7 in combination with SCF also showed significant increases in live TN cells at both time points. This again was paralleled by significant decreases in total annexin V⁺ TN cells at both time points. Cultures established from old donors followed this same pattern at both time points. However, analysis of both young and old reveals that IL-7 and SCF together did not significantly improve upon IL-7 alone, and culture with SCF alone had no detectable antiapoptotic effect. Tables I and II show the percentage of live TN and annexin V⁺ TN cells for all individual experiments and clearly demonstrate that all IL-7 cultures show this increase in live TN and decrease in annexin V⁺ cells compared with control cultures.

Because culturing with SCF did not differ from culturing with PBS, and a combination of IL-7 and SCF did not differ from IL-7 alone, we have displayed the pattern of apoptosis within the four TN subpopulations in IL-7 and control cultures after 24 h as examples of each to establish whether the prosurvival effect of IL-7 could be located to a certain stage of TN thymocyte development. Fig. 4 reveals that IL-7 added to cultures of TN cells from both young and old animals reduced the percentage of apoptotic cells and increased the number of live cells in all four subpopulations. Most notably, a significant decrease in apoptosis and a significant increase in live cells was observed after 24 h in culture in the CD44⁺CD25⁺ and CD44⁻CD25⁺ TN populations in old animals, earlier revealed as the location of the age-associated increase in apoptosis. Although young donors revealed an increase in live cells at the CD44⁺CD25⁺ and CD44⁻CD25⁺ TN stages, unlike the old, these values were not significant. Young donors showed significant increases in live cells at the CD44⁺CD25⁻ and CD44⁻CD25⁻ stages only. SCF did not demonstrate detectable antiapoptotic effects on any of the four subpopulations, and IL-7 and SCF together did not differ from culture with IL-7 alone (data not shown).

After 5 days of TN culture, IL-7-mediated protection from apoptosis was still clearly evident (Fig. 5). All four TN subpopulations from young and old showed increases in live thymocytes. In the absence of IL-7, live cells are virtually undetectable. Again, SCF failed to demonstrate an antiapoptotic effect on any of the four subpopulations (no live cells were detectable), and all four subpopulations cultured with IL-7 and SCF together did not differ from those cultured with IL-7 alone (data not shown).

To establish whether the age-associated increase in apoptosis within the TN compartment was the result of reduced IL-7R expression, TN from young (n = 6) and old mice (n = 6) were examined for IL-7R α-chain expression. The percentage of CD44⁺CD25⁻ TN thymocytes expressing the IL-7R was 58 ± 7% in young and 74 ± 6% in old mice, revealing a significant increase in expression with age (p < 0.01). Expression in CD44⁺CD25⁺ TN thymocytes was 39 ± 20% in young and 61 ± 22% in old mice. Although expression in this population is 1.5 times higher in old animals, this was not significant (p = 0.09). Expression within the CD44⁻CD25⁺ and CD44⁻CD25⁻ populations did not significantly change with age (22 ± 19% and 15 ± 10% in young, 20 ± 15% and 10 ± 4% in old mice). This result reveals that a decline in IL-7R expression is not responsible for the age-associated increase in apoptosis within the TN compartment. Fig. 6 shows a representative experiment in this series.

Groups of old mice were treated with IL-7, SCF, IL-7 plus SCF, or PBS to determine whether the results in vitro were echoed in vivo. Treatment of old mice with IL-7 alone revealed a significant increase in the total TN number (Table III). In comparison, treatment with SCF resulted in no change in total TN number, and treatment with IL-7 and SCF together did increase total TN thymocyte number, although not significantly. There was no significant difference between IL-7 and combination treatment in total TN number.

The total number of live cells was calculated by subtracting the numbers of annexin V⁺ cells from the total numbers of TN cells and from the numbers of each of the four subsets. Compared with PBS-treated controls, IL-7 therapy significantly increased the number of live cells in the TN population (Table III). Because the total TN number and total live TN number after treatment with IL-7 plus SCF did not significantly differ from IL-7 alone, we have displayed the numbers of live cells within the four TN subpopulations in IL-7-treated mice to establish the prosurvival effect of IL-7. IL-7 increased the number of live cells in the CD44⁺CD25⁻, CD44⁺CD25⁺, and CD44⁻CD25⁺ subsets (Fig. 7), although only the increases in CD44⁺CD25⁻ and CD44⁻CD25⁺ were significant. Treatment with SCF did not significantly increase the total number of live cells in the TN population or in any of the four subsets. However, the percentage of live cells within each subpopulation, did not significantly differ between IL-7- and PBS-treated mice. The percentage of live cells within the CD44⁺CD25⁻, CD44⁺CD25⁺, CD44⁻CD25⁺, and CD44⁻CD25⁻ TN thymocytes of IL-7-treated mice were 66 ± 4%, 73 ± 14%, 73 ± 14%, and 54 ± 37%, and for the PBS-treated mice 69 ± 11%, 70 ± 22%, 77 ± 11%, and 74 ± 21%, respectively.

Control over the production of thymocytes is exerted by phases of expansion, selection, and apoptosis during development in the thymus, and any changes in these controlling processes will consequently lead to an alteration in the numbers produced. The reduction in thymocyte number and subsequent decline in thymic output seen with age suggested that a change in these controlling processes had occurred.

Here we show for the first time that there is a significant age-associated increase in apoptosis within the TN population that can be pinpointed to the CD44⁺CD25⁺ and CD44⁻CD25⁺ stages of thymocyte development. These stages express growth factor receptors with an essential role in thymocyte development, the most important of which are c-kit and the IL-7R complex (IL-7Rαγ_c). c-kit is expressed on CD44⁺CD25⁻ and CD44⁺CD25⁺ TN populations, and expression is lost by the CD44⁻CD25⁺ stage (31), whereas the IL-7R is expressed on all four TN populations (32). Evidence for the obligatory requirement of both SCF and IL-7 comes from mice doubly deficient in both receptors. These c-kit^−/−γ_c^−/− mice show complete abrogation of T cell development, which is not apparent in the single-deficient mutants (27). The limited thymopoiesis present in IL-7^−/−, IL-7Rα^−/−, IL-7Rγ_c^−/− (24, 25, 26), and c-kit^−/− mice (27) has led to the conclusion that SCF and IL-7 act synergistically at the early stages of T cell development (26, 27). This conclusion provided our approach in analyzing the effect of both IL-7 and SCF on the in vitro survival of TN cells from old mice and identifying the effect of IL-7 and SCF treatment on TN survival in vivo as a potential means of reversing age-associated thymic atrophy. Analysis of the thymi in aged mice reveals a bottleneck in thymocyte production between the multipotent stem cell progenitor stage (CD44⁺CD25⁻ TN) and their progeny (CD44⁺CD25⁺ TN) that have become committed to the T cell lineage. The number of CD44⁺CD25⁻ TN thymocytes does not alter with age; however, all subsequent subsets show markedly reduced numbers (18). This age-associated increase in apoptosis at the CD44⁺CD25⁺ and CD44⁻CD25⁺ TN stages provides some explanation for the decrease in population numbers after the CD44⁺CD25⁻ stage of development.

Our results reveal differences in the comparative ability of IL-7 and SCF to maintain TN thymocytes in vitro. IL-7 alone or in combination with SCF promotes TN survival, reducing apoptosis in vitro. Furthermore, our results showed that IL-7 significantly increased live thymocytes and decreased apoptotic thymocytes at the CD44⁺CD25⁺ and CD44⁻CD25⁺ TN stages from old donors after 24 h in vitro, thus reversing the age-associated increase in apoptosis. SCF failed to demonstrate any detectable antiapoptotic properties in cultures from young or old donors, and a combination of SCF and IL-7 did not improve upon the effect of IL-7 alone. Finally, the ability of IL-7 and IL-7 plus SCF to maintain live cells in culture was more clearly apparent after 5 days in vitro, revealing the essential nature of IL-7 and not SCF as a survival factor. Although SCF/c-kit and IL-7/IL-7Rαγ interactions have been reported to compensate for each other functionally (26, 27), we did not observe synergy with respect to an antiapoptotic effect on TN cells. In the absence of IL-7, SCF does not replace IL-7 as a prosurvival factor in culture. However, we did not investigate other aspects of synergy, such as thymocyte proliferation.

To reinforce the in vitro observation that IL-7 reduces the age-associated increase in TN apoptosis and to identify the impact on survival in vivo, we treated aged mice with IL-7, SCF, and a combination of IL-7 and SCF. IL-7 treatment proved an effective therapy, clearly shown by the significant increase in live TN cell numbers following treatment. The increase in live CD44⁺CD25⁻ TN numbers may be due to enhanced survival, enhanced intrathymic proliferation, or increased production of progenitors by the bone marrow and hence increased thymic entry. The increased number of live CD44⁺CD25⁺ and CD44⁻CD25⁺ TN cells mirrored the changes observed in vitro. Recent work by Kim et al. revealed that the trophic action of IL-7 stopped at the CD44⁻CD25⁺ TN stage of fetal thymocyte development and that cell death at the CD44⁻CD25⁻ TN stage was independent of IL-7 (28). Our results support this, revealing that IL-7 did not increase live CD44⁻CD25⁻ TN cells in aged mice, suggesting that the trophic action of IL-7 at this stage is minimal. The observation that the percentage contribution of live cells to each of the TN subpopulations did not increase with IL-7-treated mice implies that the significant increase in live numbers in vivo is the result of a complex interaction between increased thymopoiesis and the rate of apoptosis.

Having shown that IL-7 induced thymopoiesis in aging animals, we asked whether SCF treatment could induce the same effect. SCF has been shown to prevent apoptosis in various cell types (33, 34), but not in thymocytes. Here we clearly demonstrate that SCF treatment does not increase numbers of live cells in vivo. This result mimics our observations in vitro. The only population to show an increase after SCF treatment in vivo was the CD44⁻CD25⁻ TN population, probably reflecting a change in the kinetics of TN development. These results suggest that IL-7 and not SCF is the important contributory factor to the changes seen in the aging thymus, whose deficit may be central to the decline in thymocyte production. TN thymocytes showed no decline with age, either in IL-7Rα expression or in ability to respond to IL-7 in vitro, strongly suggesting that survival of TN cells in the aging thymus is more likely linked to a reduced availability of IL-7 rather than an inability of the cells to respond. In support of this, recent work in our laboratory has demonstrated that expression of intrathymic IL-7 declines in old mice.⁴

IL-7 plus SCF treatment increases the total number of live TN thymocytes, although not significantly, but failed to improve upon treatment with IL-7 alone. Despite the large increases in total and live TN numbers, the lack of significance in the combination-treated group can be explained by the large SE in this group. There was no significant difference between IL-7 and combination treatment, with the exception of the CD44⁻CD25⁺ subpopulation. Here, live cells were significantly lower in the combination-treated mice and may reflect either a change in the kinetics, transit through the group, or the induction of an additional effector molecule affecting this stage of the pathway alone.

The hypothesis explaining that the bottleneck in TN development is due to problems with TCR β rearrangement as a result of reduced intrathymic levels of available IL-7 (18) was questioned recently by a paper showing age-associated thymic atrophy in three strains of mice carrying differing TCR αβ transgenes (35). Unlike the original observations (18), this later study failed to control for the effect of the endogenous transgene by not observing TCR αβ transgenes on a RAG knockout background. The H-Y-transgenic mouse strain mentioned by Lacorazza et al. (35) showing age-associated thymic atrophy does not show atrophy when present on a RAG-2 knockout background (B. Rocha, unpublished observation). Similarly, when mice express the AND transgene (anti-pigeon cytochrome c, class II restricted) on a RAG-2^−/− background, age-associated thymic atrophy does not occur (B. Lucas, unpublished observation).

In conclusion, the increase in TN apoptosis, the ability of IL-7 to reverse this apoptosis, no decline in IL-7Rα-chain expression, and finally, the fact that intrathymic IL-7 expression declines with age⁴ all lend support to the concept that thymic atrophy is due to a deficiency of available intrathymic IL-7 (18). The implication of this work is that therapy with IL-7 is capable of renewing thymopoiesis in old animals, suggesting the possibility that manipulations undertaken to increase the level of intrathymic IL-7, thereby enhancing thymopoiesis, may hold therapeutic potential even in very aged hosts.

We thank Prof. Frances Gotch and Dr. Nesrina Imami for reviewing the manuscript and for fruitful discussion; and Dr. N. Wells at the Imperial College of Science and Medicine for kindly providing young C57BL/10 mice.