Elderly people are at higher risk for infections due to declining cellular and humoral immune responses. Central to this dysfunction is the reduced responsiveness of the naive CD4+ T cell compartment. Previous data from our laboratory suggest that although defects in the aged naive CD4+ T cell response are apparent in recent thymic emigrant populations, additional defects develop during extended post-thymic longevity in the periphery. To further investigate the factors that lead to aging defects, we took advantage of the OT-II TCR-transgenic (Tg) mouse model. We show that because of an apparent superantigen-mediated loss of naive Vβ5+ Tg CD4+ T cells from the periphery of aging OT-II mice, this compartment becomes enriched for cells of reduced post-thymic longevity, resulting in a frequency of recent thymic emigrants in aged mice that is similar to that of young mice. Purification and functional analysis of aged OT-II cells with reduced post-thymic longevity reveal that they have an age-associated decrease in expansion and IL-2 production in response to Ag in vitro. However, the in vivo expansion, IL-2 production, and cognate B cell helper ability of these cells are similar to those of cells from young mice. In contrast, T cells from aged HNT Tg mice demonstrate extended post-thymic longevity and exhibit severe defects in the same in vitro and in vivo models. These data support a correlation between the requirement for increased post-thymic longevity and the development of the most severe naive CD4+ T cell-aging defects.
The elderly are faced with a number of risk factors such as malnutrition, chronic illness, and declining immune function that increase their risk of severe morbidity and death from infectious diseases such as influenza, an otherwise mild respiratory infection. Although vaccination does provide an increased level of protection against influenza for the aged, its efficacy is limited by decreased Ab (1, 2) and cell-mediated responses (3, 4, 5, 6) to the vaccine that are indicative of the immunological defects seen in the aged. Recently, the etiology of a number of defects observed in both the humoral and cellular arms of the aged immune system has been linked to alterations in the function of the CD4+ T cell subset (reviewed in Ref. 7).
Early studies on the functions of CD4+ T cells of older animals found that they both proliferated and produced less IL-2 in response to stimulation (8, 9) and that this finding correlated with an increased proportion of cells with the CD44high memory phenotype (10, 11). Additional experiments performed with isolated aged CD4+CD44high cells demonstrated that they were refractory to ionomycin, Con A, and anti-CD3-mediated stimulation (10, 12, 13, 14). These data supported the hypothesis that the defects found in the aged CD4+ T cell compartment are due to functionally deficient memory phenotype cells that constitute an increasing proportion of the peripheral T cell population as the output of naive cells from the aging thymus decreases. The antigenic history of this developing memory population in aging animals is unclear and could in large part be a consequence of exposure to environmental Ag, as well as age-related, Ag-independent events such as continued homeostatic division.
One of the consequences of the large proportion of memory phenotype cells in the aged is the reduced “space” for naive T cells capable of responding to newly encountered Ag. In addition, studies in several model systems have revealed that the cohort of remaining naive CD4+ T cells in aged animals has decreased responsiveness to stimulation, further limiting the capacity of the elderly to respond sufficiently to new antigenic challenges. Specifically, aged CD4+CD44low cells stimulated with either ionomycin or anti-CD3 demonstrated a reduced calcium (Ca2+) response compared with young cells of the same subset (12). A defect was also seen in the capacity of anti-CD28 costimulation to enhance the proliferation of anti-CD3-stimulated aged CD4+CD44low cells (15). Comparing Ag-specific responses of naive T cells in young and aged animals has been difficult because the naive T cell repertoire changes as animals age and, thus, it is not possible in normal polyclonal situations to compare equal numbers of young and aged cells with the same antigenic specificity.
Experiments using aging colonies of TCR-transgenic (Tg)3 mice have allowed for the analysis of equivalent populations of naive monoclonal Ag-specific young and aged CD4+ T cells. Using the B10.Br AND CD4+ T cell Tg model, we (16, 18, 19) and others (17) have demonstrated a basic defect in the ability of aged naive CD4+ T cells to form immunological synapses and to proliferate and produce IL-2 in response to cognate Ag in vitro and in vivo. Aged naive CD4+ T cells are also functionally defective in their ability to provide help to a B cell response (20) and to transition into a responsive memory population (21).
Understanding the factors that lead to these defects is paramount to devising strategies to boost immunity in aging populations. Our laboratory has recently demonstrated that numerous defects observed in the aged naive CD4+ T cell response are already evident in the aged recent thymic emigrant (RTE) response to Ag (22), supporting the hypothesis that age-associated defects are at least partially due to intrinsic changes in aged T cell progenitors. Interestingly, age-associated defects are also found when young T cell progenitors develop to maturity in aged hosts (22), indicating that factors present in the aged environment can also negatively influence aged CD4+ T cell function. An additional consideration is that because the replacement of T cells in the periphery decreases as the output of new cells from the thymus declines with age, naive T cells in aged animals generally have lived longer than those in young animals (23). This increased post-thymic persistence may further jeopardize T cell function by allowing for the accumulation of oxidative damage to cellular DNA (24), mitochondria (25, 26), and lipid membranes (27), while simultaneously compromising the cellular maintenance and repair pathways that counteract such insults (27, 28). It is also possible that, with time, additional stochastic changes accumulate, such as DNA mutations and damaged proteins. Taken together, these observations have led us to hypothesize that while the genesis of aging defects in CD4+ T cells includes events occurring early during T cell development, the severity of these defects may be intensified by further differentiation and increased longevity in the aged environment (7). Our finding that strategies that favor the longevity of peripheral CD4+ T cells, such as adult thymectomy, lead to the accelerated onset of aging defects support this hypothesis. However, the interpretation of these experiments is complicated by the absence of any thymus-derived factors in thymectomized mice that could also potentially effect the function of naive CD4+ T cells (29). Therefore, we wanted to compare naive CD4+ T cells that were of shorter or longer longevity by an alternative approach.
Superantigens (SAg) expressed from mouse mammary tumor virus (Mtv) DNA intergrations can associate with MHC class II molecules (in particular I-E) to generate ligands with high affinity for certain TCR Vβ chain families (30). TCR engagement by SAg during negative selection in the thymus induces the deletion of CD4+CD8+ double-positive thymocytes expressing the appropriate Vβ chain, resulting in the absence of these T cells in the periphery. This SAg-mediated intrathymic deletion does not occur in C57BL/6 (B6) mice that lack I-E. However, a series of articles by Fink and colleagues (31, 32, 33, 34, 35) described a mechanism whereby CD4+Vβ5+ cells that encounter a weak SAg encoded by Mtv-8 or -9 in the B6 periphery are either rendered anergic and deleted or are induced to express an alternate TCR via a RAG-dependent process. CD4+ T cells from B6 OT-II TCR Tg mice express a transgene-encoded Vβ5 β-chain and could be susceptible to this process (36). In the current study, we describe the gradual decline of naive Tg CD4+ T cells from the periphery of OT-II mice, such that in the absence of new cells from the thymus nearly all of these cells are deleted or undergo TCR revision within 6 mo. We show that in 18- to 24-mo-old OT-II mice this process enriches for newer cells that have recently emigrated from the thymus, resulting in a frequency of RTE equal to that seen in young mice. These aged cells with reduced post-thymic longevity demonstrate marked functional defects in response to Ag in vitro, supporting our previous findings that aged RTE are functionally defective (22). However, upon transfer to adoptive hosts, these cells are functionally robust and they respond to Ag normally by expanding, acquiring the capacity to produce IL-2 and provide sufficient help to responding cognate B cells to generate levels of Ab equivalent to that in animals that had received young cells. We contrast these data with those from experiments using aged HNT CD4+ Tg T cells which demonstrate extended post-thymic longevity and exhibit profound functional defects in the same in vivo models. These data are consistent with a correlation between increased post-thymic longevity and the age-associated decline in critical T cell functions, and support the hypothesis that the severity of aging defects is intensified by increased persistence in the aged environment.
Materials and Methods
OT-II H-2b/b TCR Tg mice express a Vβ5/Vα2 transgene that recognizes the 323–339 fragment of OVA in the context of I-Ab (36, 37). AND TCR Tg mice bred onto C57BL/6 H-2b/b express a Vβ3/Vα11 transgene that recognizes the 88–104 fragment of pigeon cytochrome c in the context of I-Ek (16). HNT H-2d/d TCR Tg mice express a Vβ8.3 (Vα unknown) transgene that recognizes the 126–138 fragment of the PR8 influenza hemagglutinin (HA) molecule in the context of I-Ad (38). All transgenic strains, as well as C57BL/6 H-2b/b (B6) CD4+ knockout (KO), and BALB/c H-2d/d (BALB/c) CD4+KO, and B6 Thy1.1-congenic mice were bred at the Trudeau Institute Animal Facility (Saranac, NY). CbyJ.Cg-Foxn1ν (BALB/c nude) mice were purchased from The Jackson Laboratory. All mice were fed sterile standard diet ad libitum and housed in isolator cages under specific pathogen-free conditions. Mice referred to as young were 6–10 wk of age; aged mice were 17 to 24 mo old. Aged mice were inspected for gross pathology and animals exhibiting pathology were excluded from experiments. The Trudeau Institute Institutional Animal Care and Use Committee approved all experimental animal procedures.
Medium and Ags
All cells were grown in RPMI 1640 (Invitrogen Life Technologies) containing 2 mM l-glutamine, 100 IU penicillin, 100 μg/ml streptomycin (all obtained from Invitrogen Life Technologies), 50 μM 2-ME (Sigma-Aldrich), and 8% FBS (HyClone). OVA peptide 323–339 (OVA323–339; ISQAVHAAHAEINEAGR) and HA peptide 126–138 (HA126–138; HNTNGVTAACSHE) were synthesized by New England Peptide. LPS-free whole OVA was a generous gift from Dr. T. Moran (Mount Sinai School of Medicine, New York, NY). PR8 influenza virus was chemically inactivated with β-propriolactone as previously described (39).
Thymectomy and intrathymic CFSE injection
Surgeries were performed on mice while under sodium pentothal anesthesia. Wounds were closed with 9-mm wound clips. For thymectomy, thymi were removed by suction. Intrathymic CFSE injections were performed as previously described (22). Splenocytes were considered CFSE+ if their CFSE fluorescence exceeded that of CD4−CD8− splenocytes of nonthymic origin, which were not CFSE labeled during the thymic injection.
Flow cytometric analysis
For cytometric analysis, cells suspended in PBS supplemented with 2% BSA and 0.1% NaN3 were incubated with fluorochrome-conjugated Abs for 30 min on ice and in the dark. Cells were either analyzed immediately or fixed in 1% paraformaldehyde. mAbs used for these studies were specific for CD4+, CD8, Vβ2, Vβ3, Vβ4, Vβ5.1/5.2, Vβ6, Vβ7, Vβ8.3, Vβ10, Vβ11, Vα2, CD44, CD62L, CD25, mouse IL-2, and the appropriate irrelevant isotype controls. Flow cytometric data were acquired on a FACSCalibur (BD Biosciences) cytometer using CellQuest (BD Biosciences) software. Analysis of cytometric data was done using FlowJo version 6.1.1 software (Tree Star).
Lymphocytes were harvested from the spleens and peripheral lymph nodes of young and aged OT-II mice and stained to identify the CD4+Vβ5highCD44low population. Ca2+ mobilization analysis was then conducted as previously described (22).
CD4+ T cells were isolated from TCR Tg mice as previously described (18), with minor changes. Briefly, young and aged CD4+ T cells were isolated with anti-CD4+ beads according to the manufacturer’s protocol (Miltenyi Biotec) and then stained with allophycocyanin-anti-CD4+, PE-anti-Vβ5, and FITC-anti-CD44 (for OT-II), or CyChrome-anti-CD4+, FITC-anti-Vβ8.3, allophycocyanin-anti-CD44, and PE-anti-CD62L (for HNT). The CD4+Vβ5highCD44low (OT-II) or CD4+Vβ8.3+CD44lowCD62Lhigh (HNT) cells were sorted using a FACSVantage SE. Purity of all sorted populations was always ≥95%. In some experiments, sorted cells were labeled with the dye CFSE (Molecular Probes) as previously described (18). This process of positive selection for naive TCR Tg CD4+ T cells did not result in any observable Ag-independent stimulation, as measured by division and phenotypic conversion in vitro (data not shown).
B cell blasts were generated from either B6 or BALB/c mice as previously described (40) and cultured with sorted TCR Tg CD4+ T cells at a 1:1 ratio to give a final concentration of 2 × 105 cells/ml. OT-II and HNT peptides were added at a final concentration of 5 μM. IL-2 was added at a final concentration of 12 ng/ml where indicated. After 2 days, an equal volume of complete tissue culture medium was added. For [3H]thymidine incorporation assays, sorted OT-II or HNT T cells were added to 96-well plates at 10,000 cells/well with appropriate peptide and B blasts. Cells were pulsed with [3H]thymidine 18 h before the indicated day of culture and then harvested onto glass fiber filters. The cpm were determined by a scintillation counter. Culture supernatant levels of IL-2 were determined in a bioassay with NK-3 cells.
CD4+ T cell transfers
Sorted TCR Tg cells (0.2–1 × 106) from young and aged mice were adoptively transferred to the indicated hosts by i.v. injection. The following day, these hosts were immunized i.p. with 50 μg of OVA323–339 or LPS-free OVA protein in alum (Pierce) (for OT-II), or 50 μg of HA126–138, or the equivalent of 108 egg infectious units of inactivated PR8 influenza in alum (for HNT). Where indicated, spleens and peripheral lymph nodes were recovered from the individual adoptive hosts 5–10 days after immunization. Lymphocyte populations were counted, stained to identify Tg donor cells, and analyzed for phenotype or intracellular cytokine staining. Serum was collected from immunized animals as indicated for the detection by ELISA of OVA or influenza-specific Abs.
Intracellular cytokine staining (ICCS)
ICCS was performed as previously described (18, 19). Briefly, splenocytes from adoptive hosts were restimulated for 4 h with PMA (2 ng/ml) and ionomycin (200 ng/ml) (Sigma-Aldrich). Brefeldin A (10 μg/ml final concentration; Sigma-Aldrich) was added 2 h after initiation of culture. Cells were surface stained, washed, and fixed with 4% paraformaldehyde for 20 min at room temperature. Cells were washed and resuspended in saponin buffer (PBS containing 1% FBS, 0.1% NaN3, and 0.1% saponin) containing anti-mouse IL-2 or isotype control and incubated for 20–30 min at room temperature. Samples were then washed with PBS and analyzed immediately on a FACSCalibur cytometer.
OVA-specific total IgG and IgG1 and influenza- specific IgG1 and IgG2a serum Ab titers were determined by ELISA as previously described (41). End-point titers were determined by the last dilution of serum that gave an OD two times above background.
Statistical analyses were performed by Prism 4.0 software (GraphPad) using Student’s t test. Values of p < 0.05 were considered significant.
OT-II Tg CD4+ T cell numbers decline in the periphery over time
OT-II mice express a rearranged Vβ5Vα2 TCR transgene that recognizes chicken egg OVA323–339 in the context of MHC class II I-Ab (36, 37). It has previously been shown that CD4+ T cells from the B6 background mice who express a Tg Vβ5 TCRβ chain are chronically deleted from the lymphoid periphery following interaction with a weak SAg encoded by Mtv-8 and -9 (33). To determine whether OT-II CD4+ T cells in the B6 periphery were similarly deleted over time, we determined the ratio of CD4+:CD8+ T cells in the spleens of 2- to 24-mo-old OT-II mice. We found that in OT-II mice, the CD4+:CD8+ T cell ratio decreased from ∼4:1 to 1:1 by 15 mo of age, as compared with non-Tg B6 mice, whose mean CD4+:CD8+ T cell ratio remained relatively unchanged over a span of 20 mo (Fig. 1,A). These data indicate that the loss of peripheral CD4+ T cells was much greater in the cells expressing the OT-II TCR transgene. In addition to the loss of CD4+ T cells in aged OT-II mice, there was the appearance of a subset of CD4+ T cells that was intermediate/negative for Tg Vβ5 expression (CD4+Vβ5int/neg; Fig. 1,C), compared with the mostly high Vβ5 expression (CD4+Vβ5high) in young OT-II mice (Fig. 1,B). As illustrated in Fig. 1,D, by 20 mo of age the CD4+Vβ5int/neg population dominated the peripheral CD4+ T cell compartment such that only 10% of the remaining CD4+ T cells remained Vβ5high. McMahan and Fink (32, 34) described the appearance in their B6 Vβ5 Tg model of a similar Vβ5int/neg CD4+ population that expressed a diverse array of endogenous Vβs. They also found these cells to have an activated phenotype (CD44highCD62Llow) and suggested that an initial Vβ5-SAg encounter had induced a “tolerogen-driven receptor revision” mechanism that allowed escape of these cells from SAg-mediated deletion (32, 34). In accordance with a similar Vβ5-SAg interaction in OT-II mice, we observed that OT-II CD4+Vβ5int/neg T cells from both young and aged mice also express a wide range of endogenous Vβ chains (Fig. 1,E). In addition, consistent with prior SAg-TCR engagement, we found that CD4+Vβ5int/neg cells have an activated CD44highCD62Llow phenotype (Fig. 2, B and C). Therefore, similar to a previous description of SAg-induced loss of B6 Vβ5+ CD4+ T cells (31, 32, 33, 34, 35), a large proportion of the naive B6 OT-II CD4+Vβ5high T cells become activated and are lost from the periphery over time. However, this loss is not absolute in the OT-II mice and a distinct population of CD4+Vβ5high cells can still be found at 20–24 mo of age (Fig. 1,C). These cells have not been induced to express high levels of endogenous Vβs (Fig. 1,E), they are 100% positive for the OT-II Tg Vα2 TCRα chain (Fig. 2,A) and they have a mostly naive CD44lowCD62Lhigh phenotype (Fig. 2, B and C), suggesting they have not yet encountered the SAg to result in the induction of activating or tolerogenic signals (31, 32). Although these data could be interpreted as suggesting that a small population of naive CD4+Vβ5high cells persist into old age in OT-II mice, it is also possible that because of the continual SAg-induced loss of naive CD4+Vβ5high cells from the periphery, the cells found in aged OT-II mice are primarily composed of cells of reduced longevity continually emigrating into the periphery from the aged thymus.
Longevity of CD4+Vβ5high OT-II Tg T cells in the periphery
To determine whether the cohort of naive OT-II CD4+Vβ5high cells found in aged OT-II mice was likely a long-lived population or more recently contributed by the thymus, young OT-II mice were thymectomized, preventing the emigration of new cells, and then analyzed to determine the length of time that CD4+Vβ5high cells remained in the periphery. We reasoned that if the naive OT-II CD4+Vβ5high population in aged OT-II mice was dependent upon continued T cell emigration from the thymus, then this population should be lost in thymectomized mice. We found that by 26 wk after thymectomy, spleens of OT-II mice contained fewer than 2.4 × 104 (±1.5 × 104) CD4+Vβ5highCD44low Tg cells, compared with 4.3 × 106 (±.6 × 106) CD4+Vβ5highCD44low cells in nonthymectomized, aged-matched controls, a loss of >170-fold as a result of thymectomy (Fig. 3,A). This loss of CD4+Vβ5high cells from thymectomized OT-II mice is shown in histogram format in Fig. 3,B, which also reveals a remaining CD4+Vβ5int/neg population that is presumably no longer susceptible to SAg-mediated deletion. Importantly, the number of CD8 T cells in thymectomized OT-II mice was only 2-fold less than in nonthymectomized, aged-matched controls, demonstrating that there is not a general defect in the survival of peripheral T cells in OT-II mice (data not shown). For comparison, young AND CD4+ T cell Tg mice were also thymectomized and the loss of AND CD4+Vβ3highCD44low T cells was monitored. Aging defects in the AND CD4+ T cell Tg model have been well described (17, 18, 19, 20, 21, 22) and no known SAg deletion of peripheral Vβ3+ AND Tg cells has been reported. Spleens from AND mice thymectomized 26 wk earlier contained on average 2 × 105 (±3.8 × 104) CD4+Vβ3highCD44low Tg cells compared with 3.7 × 106 (±2.4 × 106) CD4+Vβ3highCD44low cells in nonthymectomized, aged-matched controls, a loss of only 18-fold (Fig. 3,A). This remaining CD4+Vβ3high population is clearly seen in Fig. 3,B and is in sharp contrast to the absence of CD4+Vβ5high cells in thymectomized OT-II mice (Fig. 3 B). These data demonstrate the nearly complete SAg-mediated loss of peripheral CD4+Vβ5high cells in OT-II mice in the absence of new cell production by the thymus and suggest that the pool of naive CD4+Vβ5high Tg cells found in aged OT-II mice is the result of the continual emigration of new cells from the thymus, and, therefore, represents cells of reduced post-thymic age. If this were the case, we would predict that the naive Tg compartment of aged OT-II mice would be enriched for RTE.
To further test this postulate, we determined the proportion of RTE in the periphery of young and aged OT-II mice using a method previously developed by this laboratory, whereby an intrathymic injection of CFSE was used to label thymocytes in situ (22). Emigration of RTE from the thymus could then be followed by the appearance of CFSE+ cells in the peripheral CD4+ T cell population. Intrathymic labeling with CFSE was specific for thymus-derived cells as demonstrated by the lack of staining in the peripheral CD4−CD8− population (see Ref. 22 and data not shown). Previously performed kinetics analysis revealed that CFSE+ RTE increased in frequency in the periphery until day 10 after CFSE injection and declined by day 14 (22). As shown in Fig. 3,C, analysis on day 10 revealed that all CD4+CFSE+ cells from the spleens of young and aged mice were restricted to the Vβ5high subset, supporting the notion that the Vβ5int/neg cells develop only after export from the thymus. The absolute number of CD4+Vβ5high RTE in the spleens of aged OT-II mice was 10-fold less than in young mice, reflecting the natural decrease in thymic output with age (data not shown). Remarkably, the frequency of RTE within the CD4+Vβ5high subsets of young and aged animals was the same, demonstrating the SAg-induced loss and subsequent replacement of equivalent proportions of CD4+Vβ5high T cells from the periphery of young and aged OT-II mice and suggesting similar longevity of all CD4+Vβ5high cells in both groups (Fig. 3, C and D). This is contrary to what is seen in aged AND Tg mice (Fig. 3 E and see Ref. 22). In aged AND mice, we find 50% fewer CD4+Vβ3+ RTE in the periphery, compared with the young, presumably due to the combination of reduced thymic output and the accumulation of naive CD4+ T cells of increased post-thymic persistence. These data provide strong evidence to suggest that the CD4+Vβ5high Tg cells of aged OT-II mice have reduced post-thymic longevity that is similar to that in young mice. We were interested to see what effect this reduced post-thymic age had on the OT-II development of age-related defects in response to Ag previously described in the aged AND CD4+ T cell Tg model (17, 18, 19, 20, 21, 22).
Functional assessment of young and aged naive OT-II CD4+Vβ5high T cell response to Ag in vitro
We first looked for aging defects that are intrinsic and already seen in RTE of aged mice (22). The rapid release of Ca2+ into the cytosol from intracellular stores following TCR cross-linking is a critical intermediate step in the cascade of signaling events that lead to full T cell activation. To examine the age-related changes in intracellular Ca2+ concentration in response to TCR cross-linking in the OT-II model, cells from young and aged mice were labeled with the Ca2+ sensor indo-1 and then stimulated with anti-CD3 and a cross-linking secondary Ab. We found that cells of the CD4+Vβ5highCD44low subset from aged animals had a reduced intracellular Ca2+ concentration flux in response to TCR cross-linking compared with the same population of cells from young mice (Fig. 4). As seen previously, this defect in the aged cells was not apparent when stimulation was provided by ionomycin (22).
To track the response of similar numbers of young and aged naive OT-II cells to cognate Ag, CD4+Vβ5+CD44low cells were sorted from the spleens and peripheral lymph nodes of young and aged OT-II mice (Fig. 5,A, young not shown). This process resulted in a very reproducible and highly enriched (>95%) population of aged naive Tg CD4+ T cells as previously described (18). Consistent with the impaired early signaling events described above, sorted aged OT-II cells mounted a significantly reduced proliferative response to Ag compared with similarly sorted young cells (Fig. 5,B). The reduced response mounted by aged cells could not be overcome by stimulation with a 10-fold higher concentration of peptide. To further study the kinetics and magnitude of this response, the number of young and aged OT-II cells on each day of a 6-day culture was determined. Although both young and aged OT-II cells responded initially, there was a 2-fold reduction in expansion by day 5 in aged cultures (Fig. 5,C). The reduced expansion by aged cells correlated with markedly decreased levels of IL-2 found in the aged cultures (Fig. 5,D). Thus, IL-2 in young cultures peaked on day 2 at 250 U/ml and gradually declined, while maximum IL-2 production in aged cultures reached only 50 U/ml on day 2 and was undetectable thereafter. Accordingly, young cells expressed higher levels of the high-affinity IL-2R α-chain, CD25 (mean fluorescence intensity (MFI) = 533), compared with aged cells (MFI = 275; Fig. 5,E). It is likely that the reduced amount of IL-2 production by the aged OT-II cells was a limiting factor in the ability of these cells to expand, since exogenous IL-2 added at the initiation of the culture rescued the aged OT-II expansion (Fig. 5 C).
The difference in the immune response of sorted young and aged OT-II cells was also evident by a lag in the Ag-stimulated division of aged cells as monitored by loss of the dye CFSE. Fig. 5 F illustrates that by day 2 of culture a greater number of young OT-II cells had undergone the maximum three rounds of division, while the majority of aged cells had only divided one to two times. This trend continued into day 3 of culture, by which time a greater number of young OT-II cells had undergone as many as five divisions, compared with the majority of aged cells that had undergone only three to four rounds of division. By day 4 of culture, most young and aged cells were CFSEneg. Therefore, despite the fact that the naive Tg compartment of aged OT-II mice was enriched for cells of reduced post-thymic age, these cells exhibited deficiencies in Ag-specific responses in vitro that are equivalent in magnitude to those previously described in the aged AND Tg model (17, 18). These data support our previous report demonstrating significant defects in the RTE population of aged CD4+ T cell Tg animals (22). However, our previous study did not determine whether the RTE population demonstrated a similar level of defectiveness when stimulated in vivo, a setting known to reveal additional profound age-related defects in the aged AND Tg model, such as the inability to provide sufficient cognate B cell help and become Ag-responsive memory cells (19, 20, 21).
Functional assessment of young and aged naive OT-II CD4+Vβ5high T cell response to Ag in vivo
To assess the capability of aged naive OT-II cells to provide B cell help, young or aged CD4+Vβ5highCD44low cells were transferred into separate syngeneic B6 CD4+ KO hosts and immunized with LPS-free OVA adsorbed to alum. CD4+KO mice exhibit no cognate B cell helper function of their own (20). As a result, B cell activation and Ab production depend entirely on help derived from the adoptively transferred OT-II cells. Mice were bled 9 days postimmunization, and their sera were assayed for OVA-specific Ab. Equivalent titers of total IgG and IgG1 isotype OVA-specific serum Ab were detected in CD4+KO mice that received either young or aged OT-II cells, illustrating the intact helper function of the aged naive OT-II Tg CD4+ T cells (Fig. 6, A and B). This intact helper function allowed equivalent titers of OVA-specific IgG1 serum Ab to be produced in recipients of young and aged OT-II cells as late as 41 days postimmunization (Fig. 6 C). Immunized animals that did not receive Tg cells failed to mount a serum Ab response above background levels of detection (data not shown).
Because the helper function of the young and aged naive OT-II Tg CD4+ T cells was equivalent, we would expect that the recovery and function of the two groups of Tg cells themselves would also be similar. To determine this, spleens were removed from CD4+KO hosts on day 9 after immunization and the numbers of young and aged OT-II cells were determined by FACS analysis for CD4+Vβ5high cells (Fig. 6,D). As expected, sorted cells that were adoptively transferred but not primed by Ag did not expand (<104 OT-II cells/spleen) and remained CD44lowCD62Lhigh (data not shown). In contrast, aged and young OT-II cells expanded to equivalent numbers in response to Ag (Fig. 6 E). In addition, both populations were CD44high and exhibited equivalent CD62L down-regulation (data not shown).
To determine whether in vivo-activated young and aged OT-II cells were poised to produce IL-2, cells were restimulated ex vivo with PMA/ionomycin and IL-2 production was then assessed by ICCS. We observed that equivalent percentages of aged and young OT-II effectors were capable of producing IL-2 (Fig. 6, F and G). Importantly, culture in PMA/ionomycin did not induce IL-2 production by the adoptively transferred OT-II cells that had not been primed by Ag in vivo (data not shown).
To determine whether the intact cellular response of aged cells was due to compensating factors derived from the empty CD4+ T cell compartment of B6 CD4+KO hosts, sorted young or aged CD4+Vβ5highCD44low (Thy1.2+) OT-II cells were adoptively transferred into intact congenic (Thy1.1+) B6 hosts and immunized i.p. with OVA323–339 adsorbed to alum. Spleens were removed on days 3, 5, and 6 after immunization and the numbers of young and aged OT-II cells were determined by FACS analysis for Thy1.2+CD4+Vβ5high cells. As seen in the CD4+KO hosts, aged OT-II cells mount an in vivo response to Ag that is equivalent in magnitude to that of the young (Table I). This is supported by the analysis of CFSE-labeled donor cells on day 3 postimmunization that revealed an equivalent early proliferative response of young and aged naive OT-II cells (data not shown). Finally, both groups of cells had the capacity to produce IL-2 as measured by ICCS following restimulation with PMA/ionomycin (Table I). Taken together, data from the aged OT-II Tg model demonstrate that an aged naive CD4+ T cell compartment that is enriched for “younger” cells with reduced post-thymic longevity only exhibits limited age-related functional defects. This is in contrast to longer-lived naive Tg CD4+ T cells from aged AND mice, which exhibit severe in vitro and in vivo defects, including a reduced capacity to provide sufficient cognate B cell help (18, 19, 20).
|Day .||Mice .||Cell No.b .||Percent IL-2c .|
|3||Young||2.94 ± 0.75 (p = 0.26)||81.9 ± 2.7 (p = 0.77)|
|Aged||1.56 ± 0.29||80.5 ± 3.3|
|5||Young||1.89 ± 0.31 (p = 0.22)||78.7 ± 3.4 (p = 0.22)|
|Aged||2.81 ± 0.56||83.8 ± 1|
|6||Young||0.77 ± 0.25 (p = 0.07)||83.1 ± 1.6 (p = 0.24)|
|Aged||1.77 ± 0.31||78.7 ± 2.6|
|Day .||Mice .||Cell No.b .||Percent IL-2c .|
|3||Young||2.94 ± 0.75 (p = 0.26)||81.9 ± 2.7 (p = 0.77)|
|Aged||1.56 ± 0.29||80.5 ± 3.3|
|5||Young||1.89 ± 0.31 (p = 0.22)||78.7 ± 3.4 (p = 0.22)|
|Aged||2.81 ± 0.56||83.8 ± 1|
|6||Young||0.77 ± 0.25 (p = 0.07)||83.1 ± 1.6 (p = 0.24)|
|Aged||1.77 ± 0.31||78.7 ± 2.6|
Sorted CD4+Vβ5highCD44low cells from young or aged OT-II Tg mice were transferred (2.25 × 105) into separate syngeneic B6 hosts that were immunized i.p. with 50 μg of OVA323–339 in alum.
Indicates the number of donor cells (×105) found in recipient spleens at the indicated time point after priming.
Indicates the percentage of donor cells producing IL-2 as observed by ICCS following ex vivo PMA/ionomycin stimulation. n = 3 mice/group per time point.
Extended longevity in the periphery correlates with the development of additional aging defects in the HNT CD4+ T cell Tg model
Since most of the previous studies used the AND TCR Tg model, we wanted to confirm that the aging defects that develop post-thymically were also seen in another TCR Tg model that involved the response to a highly effective immunogen. To this end, young and aged BALB/c HNT CD4+ T cell Tg mice were next used. HNT CD4+ T cell Tg mice recognize residues 126–138 of the HA protein from PR8 influenza (37). To compare the potential for the long-term survival of HNT CD4+ T cells to that of OT-II, young HNT and OT-II mice were thymectomized and their spleens were removed 32 wk later to determine the number of remaining naive Tg cells. As expected, the number of splenic CD4+Vβ5highCD44low cells in thymectomized OT-II mice was nearly 100-fold less than that found in the spleens of age-matched controls, reaffirming the shortened longevity of these cells in the periphery and the necessary continued contribution by the thymus to the population of CD4+Vβ5highCD44low found in aged animals (Fig. 7,A). In stark contrast, spleens from thymectomized HNT mice contained on average 4.7 × 106 (±.5 × 106) CD4+Vβ8.3+CD44low Tg T cells compared with 1.3 × 107 (±3 × 106) CD4+Vβ8.3+CD44low cells in nonthymectomized, age-matched controls, a loss of only 3-fold (Fig. 7 A). From these data, we conclude that peripheral naive Tg cells from aged HNT mice are more likely to be enriched for cells with greater post-thymic longevity than seen in the OT-II model.
To determine whether longer post-thymic longevity correlates with severe aging defects in the HNT model, the in vitro and in vivo Ag-specific responses of equivalent populations of sorted young and aged CD4+Vβ8.3+CD44lowCD62Lhigh HNT T cells were assessed. Aged naive HNT CD4+ T cells underwent significantly less proliferation in vitro 3 days after initial Ag stimulation compared with the young counterparts (Fig. 7,B). When the kinetics of expansion over the course of a 5-day culture were evaluated, young naive HNT CD4+ T cells achieved a maximum expansion of 29-fold, while aged naive HNT T cells expanded nearly 4-fold less (Fig. 7,C). In addition, at the peak of IL-2 production, naïve-aged HNT CD4+ T cells produced about one-third the level of IL-2 produced by their young counterparts (Fig. 7 D).
To evaluate the functionality of aged naive HNT CD4+ T cells in vivo, equivalent populations of young and aged cells (1 × 106) were transferred to BALB/c CD4+KO hosts that were then immunized i.p. with inactivated PR8 influenza in alum. The amount of B cell help provided by the adoptively transferred cells was then assessed by measuring influenza-specific serum Ab titers 21 days after immunization. We observed that transferred young HNT CD4+ T cells were better at providing B cell help compared with aged naive HNT CD4+ T cells as indicated by significantly higher titers of influenza-specific IgG1 (Fig. 7,E) and IgG2a (Fig. 7 F) serum Ab in hosts that received young cells.
To determine whether the impaired helper function of the aged naive HNT Tg CD4+ T cells reflected an overall weakened cellular response, spleens were harvested from CD4+KO recipients on days 5 and 9 after immunization and the numbers of young and aged HNT cells were determined by FACS analysis for CD4+Vβ8.3+ cells. Although there was an equivalent recovery of young and aged HNT cells from the spleens of immunized CD4+KO hosts at these time points (data not shown), there was a marked inability of the aged HNT cells to produce IL-2 upon ex vivo stimulation with PMA/ionomycin (Fig. 7 G). This reduced cytokine production parallels the reduced capacity of these cells to provide help to the B cell response. These data support the hypothesis that extended naive CD4+ T cell post-thymic longevity in the periphery, seen in the HNT but not in the OT-II model, correlates with the development of more severe naive CD4+ T cell-aging defects that are especially relevant in vivo.
We have previously suggested that post-thymic aging of naive CD4+ T cells is necessary for development of the most profound aging defects, including in vivo help for B cell Ab production. The support for this hypothesis came from the acceleration of aging defects following adult thymectomy, which prevents the replacement aging peripheral T cells with new cells from the thymus (29). However, thymectomized mice would also have lower levels of thymus- derived factors, such as IL-7, the absence of which could also influence naive CD4+ T cell function. In this study, we show that the SAg-induced deletion of naive CD4+Vβ5+ T cells in OT-II mice selects for naive cells with reduced post-thymic longevity even in aged mice and that these cells escape the in vivo defects seen in naive cells that have persisted longer. These observations thus provide additional support for a correlation between post-thymic longevity and the development of aging defects, while avoiding the potential pitfalls of our earlier studies that assessed the function of lymphocytes from thymectomized mice (29).
Because of the age-associated increase in memory phenotype cells as well as changes with age in the TCR repertoire of the peripheral T cell pool, TCR Tg animals have become a useful tool in the comparison of young and aged naive CD4+ T cells with the same antigenic specificity. Thus far, the bulk of the experiments studying the response of aged naive Tg CD4+ T cells have used the AND Tg model. These studies have demonstrated that aged naive CD4+ T cells demonstrate defects in the formation of immunological synapses with APCs bearing cognate Ag (42). Aging defects in this model also extend into reduced TCR signal propagation (22) and lower levels of in vitro Ag-driven expansion and IL-2 production (18). Upon transfer of naive AND CD4+ T cells to CD4+KO hosts that are then immunized with alum and cognate Ag, one finds a reduced proliferative response and a diminished capacity for the production of IL-2 (19). Most notable is the dramatic decline in the ability of naive AND CD4+ T cells to provide B cell help (20) and to generate a population of functional memory cells (21).
Recently, we found that RTE from aged AND Tg mice exhibit defects in TCR signal propagation and in vitro expansion and IL-2 production that are equivalent in magnitude to those of aged naive AND CD4+ T cells, suggesting that the age-associated defects detectable in vitro in naive CD4+ T cells are at least partially due to intrinsic changes in aged T cell progenitors (22). We also found that this set of age-associated defects is overcome by the addition of IL-2 and by the inflammation and lymphopenia produced by high-dose irradiation (22, 43). Based on these data and the previous finding that the development of aging defects is accelerated following adult thymectomy (29), we suggested that there are at least two levels of defects associated with aging: the first that develops in stem cells or early T cell precursors and is already manifest in RTE and a second level of defects that develops post-thymically as T cells in aged mice persist longer after the thymus involutes (7). Based on this hypothesis, we would expect to find a direct correlation between the severity of aging defects and the length of time that naive CD4+ T cells remain in the periphery. We demonstrate here that because of a previously characterized SAg-induced chronic loss of Vβ5 Tg CD4+ T cells from the periphery of B6 mice, CD4+Vβ5high OT-II Tg T cells have a reduced longevity in the periphery and that this results in an equivalent frequency of CD4+Vβ5high RTE in young and aged OT-II mice (Fig. 3). This similar RTE frequency suggests that although thymic output declines with age, interaction with SAg continues to remove the same proportion of CD4+Vβ5high cells from the periphery of OT-II mice. As a result, the peripheral compartment of young and aged OT-II mice contain CD4+Vβ5high cells of similar relative longevity. This provides the unique opportunity for the functional analysis of naive CD4+ T cells from aged animals with a reduced post-thymic life span. The data from the current study show that naive Tg CD4+ T cells from aged OT-II donors demonstrate a markedly depressed response to Ag in vitro (Figs. 4 and 5). These data are similar to the findings of Linton et al. (44), who found that sorted Vβ5+Vα2+ Tg CD4+ T cells from aged OT-II mice exhibit reduced proliferation and IL-2 production in response to Ag in vitro. These data further support the hypothesis that at least a subset of the aging defects are present in aged naive CD4+ T cells that have recently emigrated from the aged thymus. In contrast, however, we show here that upon transfer to CD4+KO hosts, these naive OT-II CD4+ T cells exhibit a fully functional response to Ag in vivo, including the unhindered ability to provide cognate help to responding B cells (Fig. 6). In addition, preliminary experiments demonstrate that sorted CD4+Vβ5high cells from aged OT-II Tg mice are capable of generating a memory compartment that is responsive to ex vivo Ag-specific stimulation (S. Jones and S. L. Swain, unpublished results).
A number of pieces of data suggest that intact function of aged OT-II cells in CD4+KO hosts is not due to the effects of homeostatic proliferation in this setting. First, data from our laboratory (H. Tsukamoto and S. L. Swain, unpublished data) and work by others (45) suggest that OT-II cells do not readily undergo homeostatic proliferation when transferred into a lymphopenic environment. This may be due to the low affinity of the OT-II Tg TCR for self-MHC-peptide complexes. Our finding that OT-II cells that were not primed by Ag remained CD44lowCD62Lhigh until at least day 9 after transfer also indicate that these cells are not activated by homeostatic signals in CD4+KO hosts (data not shown). Second, experiments from the Haynes laboratory (unpublished data) with the AND Tg model indicate that aged CD4+ T cells are no more likely to undergo homeostatic proliferation than young cells, suggesting that young and aged do not respond differently to this “empty” environment. Finally, the data in Table I illustrates that aged and young OT-II cells behave equivalently following transfer and immunization in intact syngeneic recipients, therefore, indicating that this result is not specific for the empty CD4+ compartment of CD4+KO hosts.
The process of aging of the CD4+ T cell compartment in AND and HNT CD4+ T cell Tg mice differs from aging in the OT-II mice in a number of ways. First, following thymectomy, a population of naive Tg CD4+ T cells persist in AND (Fig. 3,A) and HNT (Fig. 7,A) mice for at least 8 mo, suggesting that in aged mice with decreased thymic output there is an accumulation of naive CD4+ T cells with greater post-thymic longevity. In support of this, we have presented previously published data demonstrating a 50% reduction in the percentage of Tg RTE in the periphery of aged vs young AND mice (Fig. 3,D and see Ref. 22). Second, in contrast to the OT-II model, sorted naive Tg CD4+ T cells from aged HNT and/or AND donors produce less IL-2 in vivo (Fig. 7,G and see Ref. 19 , respectively), provide impaired B cell helper function (Fig. 7, E and F, and see Ref. 20 , respectively), and demonstrate weakened memory responsiveness (21), reflecting more severe defects in responsiveness that are not overcome in vivo. Taking into consideration the data from the three different aging models discussed here, we conclude that an extended longevity in the periphery correlates with the development of additional aging defects that influence in vivo responsiveness, most notably, the CD4+ T cell’s cognate helper ability.
The development of additional aging defects in the aged CD4+ T cell compartment following an extended persistence in the periphery may be caused by prolonged exposure to increased levels of oxidative stress, proposed to be one of the driving forces of the degenerative effects of aging (46). Oxidative stress is caused by an excessive production of reactive oxygen species that overwhelms the ability of antioxidant defenses to maintain equilibrium in the cell. This loss of equilibrium has been shown to damage DNA (24), interfere with mitochondrial function (25, 26), and alter the lipids of the plasma membrane (27). With regard to immune function in particular, oxidative stress has been shown to suppress human T lymphocyte IL-2 production (47) and preferentially reduce the function of naive vs memory T cells (48). In addition, antioxidant treatment of aged mice improves immune functions such as phagocytosis, NK cell activity, and the lymphocyte response to mitogenic stimulation (49, 50). Other possible explanations for the compromised function that accompanies increased cellular longevity include the accumulation of somatic DNA mutations and/or misfolded proteins. Therefore, while at least a subset of aging defects observed in vitro appears to stem from intrinsic defects in T lymphocyte progenitors, the development of additional defects that include B cell helper ability and memory seem to develop only by exposure for prolonged periods to yet to be defined environmental factors.
The loss of OT-II Tg CD4+ T cells from the periphery and the accumulation of CD4+Vβ5int/neg described here are not noticeable in 8-wk-old mice (Fig. 1). However, the deletion becomes more apparent with age, as does the accumulation of CD4+Vβ5int/neg cells. This process mirrors that described by Fink and colleagues (31, 32, 33) for the SAg-induced loss of CD4+Vβ5+ Tg cells from the B6 periphery in that in both cases there is a marked inversion of the CD4:CD8 ratio and an accumulation of CD4+Vβ5int/neg cells that bear an activated phenotype and endogenous TCR β-chain expression. Fink et al. (31) also observed that before deletion, SAg-stimulated CD4+Vβ5+ Tg cells acquired an activated CD44highCD62Llow phenotype and became anergic to plate- bound stimulation. We selected against this previously activated population by isolating CD4+Vβ5high cells with the CD44low naive phenotype, which were also uniformly CD62Lhigh (Fig. 2).
These data presented here provide further support for the hypothesis that there are at least two levels of defects that develop in aging CD4+ T cells: an intrinsic defect that is already expressed by RTE and may be present in T cell progenitors or stem cells (22) and later developing defects that arise as cells themselves age in the periphery. We suggest that these two sets of defects cooperate to produce a profound reduction in naive CD4+ T cell responses. Both proinflammatory cytokines and IL-2 added to culture can rescue in vitro responses of aged naive CD4+ T cells (18, 19) and proinflammatory cytokines also function in the same respect in vivo (19). It will be important in further studies to identify the extent to which different factors can fully reverse or compensate for both levels of aging defects.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Abbreviations used in this paper: Tg, transgenic; RTE, recent thymic emigrant; SAg, superantigen; Mtv, mouse mammary tumor virus; HA, hemagglutinin; KO, knockout; ICCS, intracellular cytokine staining; MFI, mean fluorescence intensity.