|Home | About | Journals | Submit | Contact Us | Français|
Persistent CMV infection has been associated with immune senescence. To address the causal impact of life-long persistent viral infection on immune homeostasis and defense, we infected young mice systemically with HSV-1, MCMV or both viruses, and studied their T-cell homeostasis and function. Herpesvirus+ mice exhibited increased all-cause mortality compared to controls. Upon Listeria-OVA infection, 23 month-old animals that had experienced life-long herpesvirus infections showed impaired bacterial control and CD8 T-cell function, along with distinct alterations in the T-cell repertoire both before and after Listeria challenge, compared to age-matched, herpesvirus-free controls. Herpesvirus infection was associated with reduced naïve CD8 T-cell precursors, above the loss attributable to aging. Moreover, the OVA-specific CD8 T-cell repertoire recruited after Listeria challenge was entirely non-overlapping between control and herpesvirus+ mice. Our study for the first time causally links life-long herpesvirus infection to all-cause mortality in mice and to disturbances in the T-cell repertoire, which themselves correspond to impaired immunity to a new infection in aging.
Aging is associated with a pronounced impairment in immune defense to new pathogens. Factors believed to contribute to this progressive weakening of the immune system include cell-intrinsic defects (1), but also may include life-long dietary, metabolic and microbial influences and other environmental stressors (2-5). Understanding the relative weight of each of these potential contributors is a complex task, requiring careful examination of each individual factor in longitudinal studies in both humans and experimental model systems.
Nearly every human carries multiple latent persistent viral infections (6), including HSV, VZV, EBV and, above all, CMV. Over a lifetime, repeated interactions between CMV and antigen-specific T-cells leads to “memory inflation” (7, 8) of the antiviral T-cell populations in both mice and humans, which can occupy up to 50% of the human T-cell pool in late life (9-11). While other herpesviruses can sometimes produce similar effects, they are much less pronounced than those of CMV, likely due to a combination of CMV's exquisite immune evasion and reactivation properties.
At the time memory inflation was discovered, several studies have clinically associated CMV positivity with the manifestations of immune aging (rev in.12). It has therefore been proposed that antiviral memory T-cell inflation comes at a cost to the immune system as a whole and that CMV may produce many of the signs and symptoms of immune aging (12). The competing hypothesis, that CMV-positive individuals may contain a deficiency that simultaneously predisposes them to CMV infection and to pronounced immune aging, could not be tested in humans and was not addressed so far in experimental animals.
Moreover, if the relationship between CMV and any of the components of the immune aging were to be causal, one would need to posit and prove the hypothesis about the precise mechanisms how that would occur. A homeostatic hypothesis could posit that there is competition between memory and naïve T-cells for homeostatic survival signals with aging, which could impair the maintenance of a diverse naïve T-cell pool. There is evidence that age-related T-cell clonal expansions (TCE) in unimmunized mice result in holes in the naïve T-cell repertoire, particularly for new pathogens whose response would be dominated by T-cells of the same TCR Vβ family to which the TCE belong (13). The implication was that the presence of large populations of (memory) T-cells within a particular TCR Vβ family somehow impairs new immune responses to pathogens also dominated by that TCR Vβ family. This may be due to 1) loss of naïve clonotype diversity within that TCR Vβ family, or 2) impaired recruitment of naïve clonotypes into the new response, which could occur through several mechanisms (14), and where the immune response could be affected by the general or specific accumulation of TCE. All of this would occur alongside the general loss of naïve T-cell precursors that occurs as a consequence of aging, resulting in an even less complete T-cell repertoire and even more impaired responsiveness to infectious challenge (15, 16).
We report here that life-long persistent herpesvirus infection in mice erodes CD8 T-cell responses to a new pathogen encountered in late life, over and above the effects of aging itself. Beyond the previously reported age-related impact on the CD8 T-cell compartment, infection with both HSV (systemic, MCMV-like infection) and/or MCMV in early life exacerbated functional defects in the CD8 response to challenge with recombinant Listeria monocytogenes expressing the OVA surrogate antigen (Lm-OVA) in late life. Aged animals demonstrated a reduced number of naïve CD8 T-cell precursors specific for either the H2Kb-presented B8R20-27 or OVA257-264 epitopes relative to young mice, and life-long MCMV infection further eroded the number of naïve B8R20-27-specific precursors. Further, the TCR Vβ repertoire of CD8 T-cells recruited into the Lm-OVA response was completely different in life-long MCMV-infected mice relative to aged controls. This is the first evidence that life-long, persistent infection with herpesviruses results in changes to the naïve CD8 T-cell repertoire, with functional consequences to the immune response to new pathogens in late life.
8 week-old C57BL/6 (H-2b) mice were purchased from Jackson Laboratory (Bar Harbor, ME). Mice were rested for 1 week prior to herpesvirus infection. At 9 weeks of age, mice were infected with either 106 pfu HSV-1 (strain 17, as described (17)) or 105 pfu MCMV (strain Smith) intraperitoneally. For mice infected with both HSV-1 and MCMV, mice were first infected with HSV-1 as above, rested for 7 weeks, then infected with MCMV. 23 month-old C57BL/6 mice were purchased from the National Institute of Aging Aged Mouse Colony (Charles River Laboratories), rested in-house for 1 week, then infected with HSV-1 or MCMV as above. All mice were maintained under specific pathogen-free conditions in the animal facility at the University of Arizona and experiments were conducted under guidelines set by the University of Arizona Institutional Animal Care and Use Committee.
To confirm life-long infection with herpesviruses and antiviral memory CD8 T-cell inflation, a subset of animals were monitored at 3-month intervals throughout their life. Peripheral blood lymphocytes were stained with fluorochrome-conjugated antibodies specific for CD4 (GK1.5), CD8 (53-6.7), CD44 (IM7), CD62L (MEL-14), and KLRG1 (2F1), as well as gB498-505:Kb (HSV gB inflating CD8 epitope) or m139419-426:Kb (MCMV inflating CD8 epitope) tetramers (National Institutes of Health Tetramer Core Facility, Emory University, Atlanta, GA), then evaluated by flow cytometry.
Mice were infected with HSV-1, MCMV, or both viruses at 9 weeks of age, then monitored daily for all-cause mortality from 11 months of age, through 23 months of age. Survival up to 23 months of age (prior to their challenge with L. monocytogenes) was compared to age-matched uninfected control animals housed in the same room throughout the experiment by log-rank test (Mantel Cox) using GraphPad Prism software (GraphPad Software Inc, San Diego, CA).
At 21 months following herpesvirus infection (23 months-old), mice were systemically infected by intravenous injection in the lateral tail vein with 1-2×105 colony forming units (CFU) of recombinant L. monocytogenes expressing ovalbumin (Lm-OVA; (18)) in 100 μl sterile PBS. 10 week-old, naïve young C57BL/6 mice (Jackson Labs) as well as 23 month-old C57BL/6 mice (NIA) were included as additional control groups for analyses of the response to Lm-OVA challenge. The dose of inoculated bacteria was determined retrospectively by plating serial dilutions of the injected bacterial suspension onto BHI agar and counting colonies the next day.
Seven days following Lm-OVA challenge, splenocytes were collected and passed through a 40 μm mesh screen to prepare single-cell suspensions. Cells were incubated for 6 hrs at 37°C in a total volume of 100 μl RPMI 1640 + 5% FCS containing 0.1 μg/well Brefeldin A (eBioscience) plus either 10-6 M OVA257-264 peptide (SIINFEKL), 10-5 M LLO189-200 peptide (WNEKYAQAYPNV), or no peptide. Alternatively, cells were stimulated for 6 hours in the presence of PMA/ionomycin (BD Leukocyte Activation Cocktail). Cells were washed, stained overnight at 4°C with fluorochrome-conjugated antibodies specific for the surface markers CD4 (GK1.5), and CD8α (53-6.7), washed, fixed and permeabilized, then stained for intracellular IFNγ (XMG1.2), TNFα (MP6-XT22), Granzyme B (GB11), IL-2 (JES6-5H4), and IL-17A (eBio17B7) using the eBioscience FoxP3 Fix/Perm buffer kit according to the manufacturer's directions. Data acquisition was performed on a custom-made, four-laser BD Fortessa flow cytometer (Becton Dickinson, Sunnyvale, CA), and was analyzed using FlowJo software (Tree Star, Inc., Ashland, OR). A minimum of 10,000 CD8+ events within the lymphocyte gate was collected for all files.
The tetramer-enrichment protocol was slightly modified (19). The spleen, inguinal, cervical and axillary lymph nodes were pooled from individual mice in attempt to collect the majority of the naïve T cell population. Cells were resuspended in 1 ml of isolation buffer (PBS with 0.2% NaN3, 0.5% BSA, and 2mM EDTA) and stained with anti-CD8α (53-6.7), PE- and APC-labeled OVA257-264-Kb or B8R20-27-Kb tetramers (NIH Tetramer Core Facility) and Fc block for 1 hr at room temperature. Cells were washed, resuspended in 500 μl isolation buffer plus 50 μl anti-APC and anti-PE microbeads (Miltenyi Biotec), and slowly rocked for 30 minutes at 4°C. Cells were washed, resuspended in 500 μl isolation buffer, and passed over a LS magnetic column (Miltenyi Biotec) according to the manufacturer's instructions. The columns were removed from the magnetic field, and bound cells eluted by pushing 5 ml of isolation buffer through the column with a plunger. The resulting tetramer-enriched fractions were stained with a cocktail of fluorochrome-labeled antibodies for 30 minutes at 4°C that served as a “dump” gate: anti-CD19 (eBio1D3), anti-CD4 (GK1.5), anti-MHC class II (M5/114.15.2), anti-F4/80 (BM8). Cells were washed and the entire sample analyzed by flow cytometry. Precursors were also isolated from young adult animals as a control for every experiment, and served as quality control demonstrating that we can isolate the expected precursor numbers in young adult animals, and that therefore the lower precursor numbers in old animals were not caused by technical difficulties.
Splenocytes collected 7 days following Lm-OVA challenge were processed into single cell suspension as above, then negatively enriched for CD8+ cells using immunomagnetic beads (Miltenyi Biotec). Enriched CD8+ cells were stained with fluorochrome-conjugated anti-CD8α (53-6.7), anti-CD4 (GK1.5), anti-CD44 (IM7), anti-Vβ5.1/5.2 (MR9-4), and OVA257-264-Kb tetramers (NIH Tetramer Core Facility) for 45 minutes on ice, then washed twice. CD8+ CD4- CD44+ OVA-Kb+ Vβ5+ lymphocytes were sorted as single cells into 96-well plates with a FACSAria cell sorter (BD Biosciences). Control wells without sorted cells were included on every plate to control for contamination.
Our RT-PCR protocol was adapted from previous studies (20, 21). Single CD4- CD8+ OVA257-264-Kb+ Vβ5+ CD44+ cells were sorted directly into 96-well PCR plates containing 5 μl cDNA reaction mix: 0.25 μl Sensiscript reverse transcriptase (Qiagen), 1x cDNA buffer (Qiagen), 0.5 mM 2’-deoxynucleoside 5’-triphosphate (Qiagen), 100 μg/ml tRNA (Invitrogen), 50 ng oligo-dT12-18 (Invitrogen), 20 U RNAse Out (Invitrogen), and 0.1% TritonX-100 (Sigma-Aldrich). cDNA synthesis was performed immediately after sorting by incubating plates at 37°C for 90 min, followed by 5 min at 95°C. Plates were immediately stored at -80°C.
Vβ5 transcripts were amplified by nested PCR with the entire 5 μl cDNA reaction used for the first PCR reaction in a final 25 μl volume containing 1.25 U DreamTaq polymerase (Fisher Scientific) in the manufacturer's 1x Buffer with 200 μM each 2’-deoxynucleoside 5’-triphosphate (Fisher Scientific), and 100 nM external degenerate sense Vβ5 primer (5’-GGGGTTGTCCAGTCTCC-3’) and external antisense Vβ5 primer (5’-CCAGAAGGTAGCAGAGACCC-3’). The PCR cycling program utilized 5 min at 95°C; 40 cycles of 20 sec at 95°C, 20 sec at 56°C, and 45 sec at 72°C; ending with 5 min at 72°C. A 4 μl aliquot of the first PCR product was used for the second PCR reaction with the internal degenerate Vβ5 sense primer (5’-CCAGCAGATTCTCAGTCC-3’) and the internal antisense Vβ5 primer (5’-GGGTAGCCTTTTGTTTGTTTG-3’). The second PCR program was the same as the first, with 35 rounds of amplification. PCR products were purified with MinElute 96 UF PCR purification kits (Qiagen), and sequenced with 12 pmol of the internal degenerate Vβ5 sense primer, using an Applied Biosystems 3730KL DNA Analyzer at the University of Arizona Genomics Core (Tucson, AZ).
OVA257-264-specific CD8+ TCRβ clonotypes were characterized by sequentially aligning each TCRβ sequence with the Vβ5.1 or 5.2 (TRBV12 in IMGT nomenclature) gene and then the best-match Jβ gene, using the IMGT reference alleles for the Mus musculus TRB genes (22). The CDR3β sequence was then identified between, and inclusive of, the conserved cysteine in the Vβ-region and the conserved phenylalanine in the Jβ-region.
All analyses were performed using GraphPad Prism software (GraphPad Software Inc, San Diego, CA). Survival comparisons were performed by log-rank (Mantel-Cox) test. Other analyses were performed by 1-way ANOVA with Dunnett's post-test, using herpesvirus-free aged mice as the comparison control. Probability values of p < 0.05 were considered to be significant. The following notations have been used to denote p values in all figures: *p < 0.05; **p < 0.01; ***p < 0.001, ****p <0.0001.
The goal of this longitudinal study was to evaluate how life-long persistent systemic infections with the herpesvirus family members HSV-1 and MCMV might impact the ability of the adaptive immune system to mount a productive response to a new pathogen late in life. Specifically, we were interested in whether a lifetime of interactions between the immune system and a persistent virus might alter the function and/or repertoire diversity of the naïve CD8 T-cells remaining in old animals. To that end, 4 cohorts of 9-week old C57BL/6 mice were infected systemically with either HSV (produces MCMV-like memory inflation (10)), MCMV, both viruses, or neither virus. Mice were then monitored for survival and allowed to age with their persistent virus(es) until they reached 23 months of age. The establishment of productive herpesvirus infection was monitored by memory CD8 T-cell inflation to either the gB498-505 (HSV infection) or m139419-426 (MCMV infection) determinants over the first 15 months following infection (Fig. 1) (10, 11).
Our general strategy was to monitor animals for mortality up to the time they were challenged at 21 months post-herpesvirus infection (23 months old) with recombinant L. monocytogenes expressing the ovalbumin protein. Seven days following Lm-OVA challenge we evaluated 1) bacterial burden in the liver, 2) effector properties of the OVA-specific CD8 T-cell response in the spleen by polyfunctional intracellular cytokine staining, and 3) CD8 effector repertoire diversity by TCR Vβ chain sequencing of individually sorted OVA257-264-Kb tetramer+ cells.
It has been observed in human subjects that the accumulation of memory CD8 T-cells specific for cytomegalovirus correlates with increased mortality in late life (3, 9, 23, 24). To determine whether such an effect is causal to CMV infection or whether, perhaps, genetic or lifestyle factors that predispose to early mortality may also predispose to CMV infection, we used a mouse model in which other potential contributors such as genetic background, lifetime pathogen exposure, diet and environment were controlled for. We monitored mortality in our life-long herpesvirus-infected cohorts to 23 months of age, just prior to challenge with Lm-OVA. Survival of each group (prior to Lm-OVA infection) was compared with age-matched herpesvirus-free control mice that were housed in the same animal room throughout the experiment. Comparisons of survival curves between these groups (in the absence of any deliberate pathogen challenge, and prior to all reaching natural deaths) found that animals that were dually infected with both HSV-1 and MCMV had just approached the threshold of significance for mortality relative to age-matched controls (Fig. 2, p= 0.05). Single infection with HSV-1 or MCMV did not significantly shorten lifespan by this measure, although it is important to emphasize that these animals were not allowed to live until natural death occurred, but rather were monitored until their challenge with Lm-OVA at 23 months of age. Whether increased mortality would be observed in single herpesvirus-infected mice (HSV-1 or MCMV) beyond 23 months of age is unknown, although there was a trend for higher mortality with infection with HSV-1 alone (p=0.0586).
We have previously characterized pronounced functional defects in the CD8 T-cell responses of aged mice following acute infection with West Nile virus or Listeria monocytogenes. In these studies, aged mice showed decreased proliferation and up-regulation of activation markers within the pathogen-specific CD8 T-cell population, resulting in a numerically reduced effector CD8 T-cell pool at the peak of the response, lower capacity to produce multiple effector molecules on a per-cell basis (including cytokines and lytic proteins, termed “polyfunctionality”), reduced quantities of the individual effector molecule(s)/cell, and a diminished ability to lyse targets bearing cognate antigen (25, 26). Similar age-associated defects in CD8 expansion and/or effector function have been observed in response to influenza, LCMV, and E. cuniculi infections as well (27-32). Collectively, these reports indicate that aging alone has severe consequences on the ability of the immune system to mount a robust CD8 T-cell response to pathogens of varying classes, pathogenesis, and host-cell tropism.
We evaluated whether further impairment above and beyond age-associated CD8 T-cell defects would be evident in animals that had experienced life-long infections with herpesviruses. To that end, 23 month-old mice that had been infected with HSV, MCMV, or both viruses in early life were challenged with Lm-OVA, and the splenic OVA257-264 specific CD8 T-cell population was evaluated 7 days later. As expected, there was a marked decrease in the magnitude of the OVA-specific CD8 T-cell response in aged mice, as measured by IFNγ production following brief in vitro stimulation (Fig. 3) (26). This reduction was not significantly worsened in aged mice that had been infected with HSV-1, MCMV or both viruses since 9 weeks of age, compared to their age-related uninfected littermates, suggesting that life-long viral infection does not significantly impact the magnitude or expansion of a primary CD8 T-cell response to an intracellular bacterium in late life.
Polyfunctional cytokine production – that is, the ability of individual cells to produce multiple different effector molecules in response to antigen recognition – is believed to be a hallmark of highly functional CD8 T-cell populations that successfully control pathogens. As a measurement of the “robustness” of the effector CD8 T-cells generated in our life-long infected cohorts, we determined whether the OVA257-264-specific CD8 T-cells elicited following Lm-OVA challenge were functionally altered. Following in vitro stimulation with OVA257-264 peptide, the ability of IFNγ-producing cells to additionally make TNFα, Granzyme B, IL-17A and/or IL-2 was measured. Very little production of IL-2 or IL-17A was seen in the OVA-specific CD8 T-cell population, so these data have not been stratified out into separate populations for simplicity. Further, IFNγ-producing cells were the largest population seen in all groups, indicating this cytokine maintains its’ position at the top of the effector molecule hierarchy for all cohorts evaluated.
As shown in Fig. 4A, relative to the distribution of polyfunctional CD8 T-cells seen in young adults, aged mice showed (as expected) a significant loss of most functional effector populations: those able to simultaneously produce IFNγ, TNFα, and Granzyme B (black bars, I+T+G+) through those producing only IFNγ (white bars, I+T-G-), with conservation of the small population able to produce both IFNγ and TNFα (hatched bars, I+T+G-). In aged (herpesvirus-free) mice, the largest effector population included cells producing only IFNγ (~0.7% of the CD8 T-cell population, white bars, I+T-G-), with all other functional subgroups at less than 0.3% of the total CD8 T-cell pool.
In animals that had experienced life-long persistent infection(s) with HSV-1, MCMV, or both viruses, there was significant further erosion in two subpopulations of functional T-cells beyond that attributable to aging alone: both the IFNγ+/TNFα+ (hatched bars, I+T+G-) and IFNγ-only (white bars, I+T-G-) subsets were significantly reduced in life-long herpesvirus infected cohorts relative to the response seen in virus-free aged mice (Fig. 4A). Thus, in addition to the effector molecule defects seen in primary CD8 T-cell responses that result as a consequence of aging alone, these data show for the first time that life-long infection with herpesviruses further erodes the functional capacity of effectors stimulated in response to a new infection in late life.
When we performed the same polyfunctional evaluation of the Lm LLO189-200 specific CD4 T-cell response, no such impact was seen (Fig. 4B). The distribution of polyfunctional Lm-specific CD4 T-cells into 4 functional categories defined by IFNγ, ± TNFα and ± Granzyme B production was comparable in aged mice with and without life-long persistent herpesvirus infection. Although significant loss of all polyfunctional populations was observed as a consequence of aging (Adult vs. Old), further attrition was not seen in any functional CD4 subsets in the life-long-infected cohorts (Old vs. HSV, MCMV, or both). This suggests that the detrimental impact of life-long persistent herpesvirus infection on T-cell functionality is restricted to the CD8 T-cell pool.
We next asked whether this impact on CD8 polyfunctionality might reflect a larger, global impairment of CD8 T-cell function in general. To address this possibility we measured the polyfunctional responses of both the CD8 and CD4 subsets in the same animals following non-specific stimulation with phorbol 12-myristate 13-acetate (PMA) and ionomycin, bypassing the TCR signal transduction machinery. Although age-related decreases in some functional subsets were observed (comparing Adult vs. Old animals), for the most part these were not further depressed in the life-long-infected cohorts (comparing Old vs. HSV, MCMV, or both; Supp. Fig. 1), although MCMV infection (both with and without co-infection with HSV) significantly decreased the CD8 effector population capable of producing both IFNγ and TNFα (hatched bars, I+T+G-, Supp. Fig. 1A). Thus, persistent infection with herpesviruses appears to primarily impact the functional properties of CD8 T-cells recruited to a new infection, and does not reflect an overall dampening of T-cell responsiveness.
Persistent viral infection might influence anti-listerial immunity as a side-effect of pro-inflammatory cytokine production due to the ongoing anti-herpesvirus response (33, 34). Control experiments were performed to assess this issue. First, we determined whether short-term herpesvirus infection of aged mice would have a similar dampening effect on the CD8 T-cell response to Lm-OVA challenge as life-long infection did in Fig. 4. To that end, 23 month-old (naïve) mice were infected with HSV or MCMV, and then rested for 1 month. Animals were subsequently challenged with Lm-OVA, and the polyfunctional analysis of OVA257-264-specific CD8 T-cells in the spleen was determined 7 days later. Under this short-term viral infection condition, we found no unfavorable influence of either HSV or MCMV infection on the polyfunctional CD8 T-cell response to Lm-OVA (Fig. 5). In fact, in HSV-infected animals, the IFNγ+ Granzyme B+ subset (Grey bars, I+T-G+) of OVA257-264 specific effectors was significantly improved. Collectively, these data suggest that long-term, persistent infection with HSV or MCMV specifically impairs the ability to mount highly functional CD8 T-cell responses to a new pathogen in late life, and that life-long interactions between the immune system and the persistent virus are required for this erosion of CD8 T-cell function.
Due to its unique intracellular lifecycle and ability to spread from cell to cell, pathogen clearance and protective immunity to Listeria is primarily mediated by CD8 T-cells (reviewed in 35). As such, we measured whether the functional decay in the OVA257-264 specific CD8 T-cell population in life-long herpesvirus-infected animals impaired bacterial clearance after Listeria challenge. On day 7 after Lm-OVA infection, livers were homogenized and the bacterial burden was determined. Although there was a high degree of variability in the bacterial loads, animals harboring life-long persistent MCMV infection had a significantly higher number of bacteria than the herpesvirus-free, age-matched controls (Fig. 6, Old vs. MCMV). This suggests that there are biologic consequences to the impaired polyfunctional CD8 T-cell responses in mice with life-long herpesvirus infections (Fig. 4).
Aging, in the absence of persistent infections, results in decreased naïve CD8 T-cell precursor numbers and diversity (36, 37) leaving holes in the repertoire that increase susceptibility to new pathogens (15).
To evaluate how life-long persistent herpesvirus infection influences the naïve CD8 T-cell pool over time, the number of naïve CD8 T-cells/spleen in HSV, MCMV, dually infected (HSV+MCMV) mice, and uninfected age-matched controls was determined at various post-infection times. We found no evidence that persistent herpesvirus infection further significantly eroded the global naïve CD8 T-cell pool in the spleen above and beyond the effect of aging alone (Fig. 7A). We also found that the number of naïve CD8 T-cell splenocytes held steady throughout most of the lifespan, with a precipitous drop in late life (Fig. 7A).
We next evaluated the influence of aging alone on the number of naïve CD8 T-cell precursors for two different Kb-restricted antigens: OVA257-267 used for our polyfunctional analysis above, and the B8R20-27 determinant shared by numerous poxvirus family members (including vaccinia, variola, ectromelia, and cowpox; (38)). Pooled spleens and lymph nodes from individual adult and 18 month old naïve C57BL/6 mice were negatively enriched for CD8 T-cells, then naïve OVA- or B8R-specific CD8 T-cells were magnetically isolated by the tetramer pull-down technique (representative FACS plots are shown in Supp. Fig. 2) (19, 39). We found that both antigen-specific precursor populations were significantly reduced in aged naïve (herpesvirus-free) mice (Fig. 7B-C). In adult mice, the average number of OVA-and B8R-specific naïve CD8 precursors was 142 and 760 cells, respectively. In aged mice, those precursor populations had been reduced to 45 (OVA) and 140 (B8R) naïve cells, relative losses of 67% and 80%.
As the OVA-specific naïve precursor pool was already reduced in aged mice to near the limits of detection, we focused on the B8R-specific naïve CD8 precursors to evaluate whether life-long persistent herpesvirus infection would further diminish the naive CD8 precursor pool. Our rationale was that additional naïve precursor loss as a consequence of life-long viral infection would be difficult to conclusively measure in the OVA precursor population due to the small pool present in aged mice (in the absence of life-long infection). In contrast, the B8R-specific precursor population was still relatively large in aged mice (mean of 140 cells/aged mouse). Of interest, in Figure 7D, mice that had experienced life-long infection with HSV had reduction in naïve B8R-specific CD8 precursors (mean = 83 cells) that was not significant, whereas both MCMV and dually infected (HSV+MCMV) mice showed a significant reduction in their B8R precursor pool, as compared with age-matched controls (74 and 79 cells, respectively). These data indicate that life-long persistent herpesvirus infection can further erode the naïve CD8 precursor pool, exacerbating the pronounced loss of precursors that occurs due to aging alone.
Our experiments to determine the number of naïve CD8 T-cell precursors specific for the B8R peptide suggested that life-long MCMV infection markedly changed the naïve pool that remains in late life (Fig. 7D). Further, our polyfunctional cytokine response suggested that the OVA257-264-specific CD8 T-cells recruited into the response to Lm-OVA were functionally compromised in animals with life-long persistent herpesvirus infection. To directly assess the impact of persistent viral infection on the T-cell repertoire, TCR Vβ CDR3 sequencing was performed. Life-long MCMV-infected mice (+/- HSV co-infection) and uninfected controls were challenged with Lm-OVA at 23 months of age. On day 7 after Lm-OVA challenge, Vβ5+ OVA257-264-Kb tetramer+ CD8 T-cells were individually sorted for single-cell TCR Vβ sequencing analysis. The results from this analysis were striking. Across the herpesvirus-free, old mice, there were 4 TCR Vβ clonotypes that were each shared between 2 of 3 mice. One of these 4 clonotypes was dominant within at least one of the mice and all 4 were found to constitute ~65-70% of the OVA257-264 specific population in all uninfected animals (Fig 8, top row, clonotypes in red-orange). However, in old mice with life-long MCMV infection, none of the clonotypes observed in the herpesvirus-free, old mice, including these 4 main shared clonotypes, were found. Rather, 5 different shared clonotypes were found in at least 2 out of the 4 animals with life-long MCMV infection (some also with HSV infection), and these clonotypes were not recovered from the herpesvirus-free, old mice (middle and bottom rows, dominant clones in blue-green). The dominance of the shared sequences was much lower in the MCMV-infected mice; in 1 animal these 4 sequences made up 94% of the entire response, while two other mice predominantly used “unshared sequences” that did not appear in any other mouse, MCMV-infected or not (represented in white). The list of recovered TCR Vβ sequences and their distributions in animals is provided in Supplemental Table 1. Although performed on a limited sample, these data nonetheless suggest that substantial repertoire changes occur in mice that have experience life-long persistent infection. In herpesvirus-free old mice, the diversity of the naïve repertoire pool appears to become narrowed to a few shared dominant clonotypes. In contrast, these clonotypes are absent in animals with persistent, life-long MCMV infection, and others emerge in their place.
In this study, we have deliberately infected mice with different herpesviruses and longitudinally explored the effects of such infection upon T-cell homeostasis and function. We report three main and novel observations: (i) an increase of all-cause mortality in herpesvirus-infected animals that just nearly reached significance (p=0.05) at the time of late-life challenge with a novel bacterial pathogen; (ii) reduced bacterial clearance and depressed and altered CD8 (but not CD4) T-cell functional reactivity; and (iii) repertoire disturbances that were evident both in the unimmunized naïve and microbial pathogen-mobilized CD8 T-cell repertoires. These findings, for the first time, causally connect life-long persistent herpesvirus infection to manifestations associated with impaired naïve T-cell maintenance and function in aging.
CMV was associated with reduced residual life span in octo- and nona-generians in a Swedish study (23) and with increased cardiovascular mortality in Hispanic-Americans (40). While our study was not designed to assess full-course mortality at the end of life, it was nonetheless interesting to find that all-cause mortality was increased to just the point of significance at 23 months of age in the CMV+HSV infected group, compared to controls. Of note, our systemic HSV infection model is, from the standpoint of infection and establishment of latency much more like CMV in humans than HSV in humans (10). Proper longevity and mortality studies are currently in progress; however, given that in the Hadrup et al. study (23), T-cell repertoire loss correlated to increased mortality, this raises a highly intriguing parallel between that study and our data here, and mandates tracking the two in individual mice.
The second fundamental discovery was that herpesvirus+ mice exhibited perturbations in naïve repertoire more profound than those seen in aging alone. A highly diverse T-cell population is critical for protection against pathogens. The diversity of the T-cell response to infection (i.e. the number of different clonotypes participating) is a better correlate of protection than the magnitude of the response (13, 41-43). As little as a 2 to 3-fold reduction in TCR repertoire diversity dramatically impairs antigen-specific responses (44, 45), and it is the T-cell defects in the primary immune response that were identified as a major contributor to immune senescence. Thymic involution and the reduced generation of new naïve T-cells, followed by an initially successful, but eventually failing, peripheral maintenance of a naïve T-cell pool, lead to the loss of numbers and diversity within the naïve T-cell pool over a lifetime, contributing to impaired immunity in the elderly. In terms of absolute cell numbers, we found a dramatic loss of splenic naïve CD8 T-cells between ages 18 and 22 months in mice (16-20 months post-infection, Figure 7A), that was not further exacerbated by life-long persistent herpesvirus infection. To our knowledge, this is the first formal demonstration of maintenance and loss of naïve T-cell numbers in a longitudinal mouse study.
While CMV in our experiments did not lead to global numerical loss of phenotypically-defined naïve T-cells, it led to reduction of precursors specific for the poxvirus B8R epitope, whose numbers were large enough to allow analysis using pMHC tetramer pull-down. Of note, tetramer pull-down does not discriminate between different categories of cells that are antigen inexperienced, for example the truly phenotypically naïve CD44low CD62Lhi and the phenotypically not fully “naïve” CD44hiCD62Lhi cells (46). It is becoming increasingly clear that the use of such phenotypic markers in aged mice is confounded by significant homeostatic/virtual memory conversion of antigen inexperienced cells into expression of cell surface markers incongruous with the prototypical “naïve” phenotype (36, 37). Rather, our data present a picture of total epitope-specific precursor loss. The implication of these findings is that not all specificities will be affected by this process equally, however, at the present, we do not understand the rules that govern the survival of naïve lymphocytes in aging, and, likewise, the changes in the old repertoire under the influence of CMV.
Still, one should consider the following: whereas TCR rearrangement and pairing can theoretically produce >1015 different TCR αβ chain combinations in the mouse (47), current measurements and estimates place the number of different TCR clonotypes at ~2×106-7 per mouse (48), and ~108 in humans (49, 50). Given this massive disparity between the theoretical maximum of 1015 and estimated upper actual size of 2×107, it would seem that, even in genetically identical mice, the possibility that a given clonotype (defined by its’ unique TCR sequence) will be found in separate animals highly unlikely. Yet numerous studies of antigen-specific responses to infection have found identical TCRβ chain sequences in the responding CD8 T-cell populations in different animals, suggesting the presence of shared naïve clonotypes in the T-cell pool. Such “public” clonotypes stand in contrast to “private” clonotypes that are unique to individual animals (51, 52). In our evaluation of the CD8 TCR Vβ repertoire elicited in old mice challenged with Lm-OVA, we found a broad degree of sharing between aged mice, as well as the expected clonally limited response. However, the finding that no single sequences found in MCMV-infected mice were recovered in herpesvirus-free old controls – and vice versa – was unexpected. This suggests that the life-long interactions between persistent viral infection shapes the naïve CD8 T-cell repertoire to such an aggressive degree, that the entire naïve pool shared no overlap with the repertoire that exists in the absence of persistent herpesvirus infection. Clearly, additional studies, including deep sequencing of the total naïve and tetramer+ repertoire from unimmunized mice will be necessary to provide a better quantitative handle on this issue.
Importantly, the observed repertoire alterations were accompanied in herpesvirus+ animals by altered functional responses of the remaining CD8, but not CD4, T-cells, and with reduced ability to clear the bacterial infection, which was over and above the one seen in the absence of herpesvirus infection. While CMV and/or HSV-1 did not lead to significant numerical or percent reductions to the OVA-specific CD8 T-cell population, the responses were already highly reduced by aging, and therefore the overall low numbers of the responding cells may have precluded detection of this effect.
Subtle differences were seen when comparing life-long persistent infection with HSV vs. MCMV (Summarized in Table 1). In general, defects in T-cell function, pathogen clearance, and T cell repertoire were predominant in mice infected with MCMV (with or without co-infection with HSV). In contrast, increased all-cause mortality prior to late-life challenge with Lm-OVA, as well as the lack of effect on Lm-OVA clearance was only observed in mice with life-long HSV infection. Of consideration is that the levels of viral reactivation with persistent HSV vs. MCMV infection may be quite different, and thus have diffent impact(s) on the inflammatory environment generated in response to reactivation (both quantitatively in the frequency of viral reactivation, and qualitatively in the anatomical locations of reactivation and the immune response triggered at those sites). To the best of our knowledge, these interesting questions remain unanswered. One possibility is that the background inflammatory environment in old mice harboring persistent HSV may be somewhat more resistant to Listeria infection, as previously demonstrated for both γHV68 and MCMV infection, although was NOT seen for HSV-1 infection in young mice (33). Further, although this effect (with other herpesvirus family members) was shown to be transient in young animals (34), it is unclear how lifelong infection may alter the inflammatory environment over a lifetime and into old age.
It is of further note that such dramatic influences on the immune system were observed in our persistently-infected cohorts that showed relatively lower levels of antiviral CD8 memory inflation compared to those we and others have previously reported (9, 11). As we only monitored one inflating MCMV epitope throughout our longitudinal analysis, we cannot comment on whether the entire population of known inflating MCMV-specific CD8 epitopes was similarly dampened in these animals. Memory inflation is a complex and incompletely understood process that is likely driven by multiple factors, including microbial flora variations within animal facilities, and, based on the microbial barrier control in the facilities we have experienced at the Oregon Health & Science University (2001-2008) and the University of Arizona (2008-present) likely explains differences in memory inflation seen here and in prior work (9,11). However, what is clear from our results, is that even relatively modest levels of antiviral memory CD8 inflation can significantly alter immune function in aging.
Overall, we believe that our data present a comprehensive picture of accumulating hits to the CD8 T-cell compartment in mice experiencing a life-long herpesvirus infection. Over time, age-associated loss of naïve CD8 T-cell precursors is further exacerbated in these animals. In addition to the loss of precursors, those that remain are of entirely different clonotypic composition compared to those in herpesvirus-free controls. The repertoire that remains in herpesvirus+ mice is functionally compromised, resulting in the priming of an effector CD8 T-cell population with decreased polyfunctional capacity in response to Lm-OVA infection, and a reduced ability to control the infection. It has been proposed that peripheral turnover is most likely the primary mechanisms for maintenance of a diverse memory T-cell pool throughout life in humans, whereas thymic output is likely the primary mechanism in mice (53). If this holds true, then the impact of life-long persistent herpesvirus infection on the naïve CD8 T-cell repertoire would be expected to be even more pronounced in humans, an issue that requires immediate attention.
We thank members of the Nikolich-Zugich laboratory for help and stimulating discussion. Expert cell sorting assistance was provided by Paula Campbell at the University of Arizona Cancer Center/ARL-Division of Biotechnology Cytometry Core Facility. TCR Vβ sequencing was performed by the University of Arizona Genomics Core. Tetramers were provided by the NIH Tetramer Core Facility (Emory University).
1This work was supported by USPHS grants U51 AI081680 (Pacific Northwest Research Center of Excellence in Biodefense and Emerging Diseases), HHSN27220110017C from the National Institute of Allergy and Infectious Diseases, N01-AI-00017 from the National Institute of Allergy and Infectious Diseases, and AG20719 from the National Institute on Aging, NIH. MPD is an NHMRC Senior Research Fellow and VV is an ARC Future Fellow. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.