To test the hypothesis that host age at the time of infection is critical to the course of SIV disease, the pace of host aging, and comorbidity risks, we longitudinally assessed its impact on the course of SIV infection. We included 6 aged RMs (15–22 years) to model elderly people living with HIV, and 5 young RMs (4–6 years) to model young adults (Table 1). RMs received 300 tissue culture ID50 SIVmac239 (34), and tissues critical for HIV/SIV pathogenesis were sampled at baseline and throughout infection until progression to AIDS (Table 2). We show the dynamics of each biomarker in Figures 1, 2, 3, 4 for individual values and in Supplemental Figures 1–7 for averages (supplemental material available online with this article; https://doi.org/10.1172/JCI189574DS1).

Viral RNA and DNA burdens. Dynamics of plasma viral loads (vRNA copies/mL) (A) and cell-associated viral RNA (CA-vRNA) (copies/106 cells) in the superficial lymph nodes (SLNs) (B) and duodenum (C) in old (red lines) and young (black lines) rhesus macaques. CA-vRNA (D) and cell-associated vDNA (E) in necropsy tissues (cerebellum, heart, spleen, liver, peritoneal fat, transverse colon) from the advanced stage of infection. Significance was determined using a 2-tailed, unpaired t test.

Longitudinal changes in CD4+ and CD8+ T cell profiles. Dynamics of CD4+ T cell counts in circulation (A), superficial lymph nodes (SLNs) (B), and gut (C). Changes in T lymphocyte proliferation assessed by measuring the expression of the fraction of circulating CD4+ T cells (D) and CD8+ T cells (G) expressing Ki-67. Dynamics of the expression of HLA-DR and CD38 on circulating CD4+ T cells (E) and CD8+ T cells (H). Dynamics of the expression of the early activation biomarker CD69 assessed on circulating CD4+ T cells (F) and CD8+ T cells (I). Old RMs are shown as red lines; the young ones as black lines.

Changes in the plasma levels of inflammatory molecules in SIVmac239-infected old (red) and young (black) rhesus macaques: IL-1B (A), IL-1RA (B), CXCL-9 (MIG) (C), IL-15 (D), IL-12 (E), CXCL-11 (I-TAC) (F), CXCL-8 (IL-8) (G), and CXCL-10 (IP-10) (H).

Probability of survival after SIV infection in old (red lines) versus young (black lines) rhesus macaques. (A). Changes in the levels of biomarkers that were reported to be predictive for HIV disease progression and death in SIVmac-infected old (red) and young (black) rhesus macaques: D-dimer (B); C-reactive protein (CRP) (C); and sCD14 (D).

Demographic characteristics of the rhesus macaques included in the 2 study groups

Terminal tissue samples used for DNAme analysis

For the DNAme studies, we first examined a dataset of 40 PBMCs collected from 10 RMs throughout the follow-up: at the baseline and during acute, early chronic, and late chronic infection (Table 3). We also analyzed the DNAme profiles in 56 tissue samples from the gut (transverse colon), cardiometabolic tissues (liver, heart, and peritoneal fat), and immune tissues (spleen) obtained from 10 RMs during the late chronic infection (except for fat, which was only acquired from 6 RMs) (Table 3).

PBMC samples used for the longitudinal DNAme analysis

Young SIV-infected RMs yield higher plasma viral loads than aged RMs due to a greater abundance of target cells. Gradual depletion of CD4+ T cells and increasing plasma viral loads are major hallmarks of HIV/SIV disease progression (1, 35). CD4+ T cells decreased throughout infection both in the periphery (P < 2 × 10–16) and in the superficial lymph nodes (SLNs) (P = 1.5 × 10–8) (Supplemental Table 1). The young SIVmac-infected RMs exhibited slightly higher plasma viral loads than the aged ones (Figure 1A and Supplemental Figure 1A) and, to a lesser extent, in the SLNs and gut (Figure 1, B and C, and Supplemental Figure 1, B and C). The CD4+ T cells were significantly higher in the young compared with the old RMs in the periphery at the baseline (P = 0.0087) and during the early chronic infection (P = 0.043) (Supplemental Table 1).

The higher plasma viral loads observed in the young versus aged RMs were associated with higher baseline CD4+ T cell counts in the young RMs (Figure 2A and Supplemental Figure 2A). This trend was maintained throughout the early chronic infection due to an increased capacity to restore the CD4+ T cells by young RMs (Figure 2A and Supplemental Figure 2A). More robust plasma viral loads contributed to an accelerated CD4+ T cell loss in young RMs, and virtually complete depletion during chronic infection, similar in both groups. In the lymph nodes, both groups experienced CD4+ T cell depletions upon infection, slightly faster in aged RMs (Figure 2B and Supplemental Figure 2B). Finally, mucosal CD4+ T cell losses were similarly massive and swift in both groups, resulting in nearly complete depletion around 28 days after infection (dpi) (Figure 2C and Supplemental Figure 2C).

There was no clear association between age and the viral RNA/viral DNA content in most tissues (Figure 1, D and E), yet the aged RMs yielded a higher amount of SIV DNA in the heart (P = 0.0004) (Figure 1E) and a trend for higher virus replication in the cerebellum (P = 0.0978) (Figure 1E) compared with young RMs.

Chronic T cell activation and proliferation increase in young RMs up to or above the levels observed in aged RMs. Chronic activation drives CD4+ T cell loss and HIV/SIV disease progression (4). Meanwhile, aging-associated immune activation contributes to age-related diseases and increases the risk of uncontrolled inflammatory responses to infections (4). We monitored multiple CD4+ and CD8+ T cell activation markers: Ki-67, CD69, HLA-DR, and CD38 (Figure 2, D–I, and Supplemental Figure 2, D and I), and CD25 (Supplemental Figure 3, A–D). Aged RMs expressed higher baseline levels of activation markers: Ki-67 on CD4+ (P = 0.055) and CD8+ T cells (P = 0.03), CD69 on CD4+ (P = 0.03) and CD8+ T cells (P = 0.03), and CD25 on CD4+ and CD8+ T cells, yet without statistical significance (Figure 2, D, F, G, and I; Supplemental Figure 2, D, F, G, I; and Supplemental Figure 3, A–D). Conversely, CD38 and HLA-DR were similarly expressed by CD4+ and CD8+ T cells in the 2 groups (Figure 2, E and H, and Supplemental Figure 2, E and H).

Upon infection, Ki-67 and CD69 expression on CD4+ and CD8+ T cells and the CD8+ T cell fraction expressing CD38 and HLA-DR steeply increased in young RMs, reaching the levels observed in the aged RMs (Ki-67+, shown in Figure 2D and Supplemental Figure 2D, and HLA-DR+CD38+CD4+ T cells, shown in Figure 2E and Supplemental Figure 2E), or even exceeding them (Ki-67+, shown in Figure 2G and Supplemental Figure 2G; HLA-DR+CD38+, shown in Figure 2H and Supplemental Figure 2H; and CD69+CD8+ T cells, shown in Figure 2I and Supplemental Figure 2I; but not CD69+CD4+ T cells, for which increases were less discernible, as shown in Figure 2F and Supplemental Figure 2F). CD25 dynamics on CD4+ and CD8+ T cells were similar in young and old RMs (Supplemental Figure 3). Taken together, our data suggest that despite the initial advantage, young RMs catch up with or even surpass the old RMs with respect to CD4+ and CD8+ T cell activation.

The markers with higher levels in the old versus young RMs in early chronic infection, that is, Ki67+ on CD4+ (P = 0.078) and CD8+ (P = 0.038) T cells and CD69+CD8+ T cells (P = 0.0079) also showed a faster increase throughout the chronic infection stage in the young versus old RMs, that is, Ki67+ on CD4+ (P = 0.07) and CD8+ (P = 0.00056) T cells and CD69+CD8+ T cells (P = 8.4 × 10–9) (Supplemental Table 1). Overall, higher activation levels were observed in the old RMs for CD69+ on CD4+ (P = 0.0043) and CD8+ (P = 0.082) T cells.

Greater increases of inflammatory cytokines occur in young SIV-infected RMs. Inflammation occurs early in SIV/HIV infection and its persistence during chronic infection is associated with disease progression, even in people with ART-suppressed virus (4). We assessed the dynamics of inflammatory responses to SIVmac infection (Figure 3 and Supplemental Figure 4), and we report that both their acute and chronic increases tended to be more robust in young versus aged RMs (Figure 3, A, C, and F, and Supplemental Figure 4, A, C, and F), showing significant or near statistically significant results for IL-1B (P = 0.082), CXCL-9 (P = 0.068), and CXLC-11 (P = 0.027) in acute infection and significant results for IL-1B (P = 0.046) and CXCL-9 (P = 0.046) in early chronic infection (Supplemental Table 1). IL-8 did not follow this pattern, but yielded higher levels in the old RMs in acute and early chronic infection, trending toward significance (P = 0.064 and P = 0.065) (Figure 3G and Supplemental Figure 4G). However, some markers did not adhere to this pattern, for example, IL-1RA, IL-15, IL-12 or CXCL-10 (Figure 3, B, D, E, and H, and Supplemental Figure 4, A, B, D, E, and H).

Accelerated progression to AIDS occurs in aged versus young SIV-infected RMs. During the follow-up period, 4 of 6 old RMs progressed to AIDS (median survival: 269 days), as opposed to only 1 of 5 young RMs. We found a significantly lower probability of survival for aged RMs compared with the young ones (P = 0.0487) (Figure 4A). The main cause of death in the old RMs was interstitial pneumonia.

Excessive coagulation and inflammation occur preferentially in young SIV-infected RMs, reaching levels similar to those observed in old RMs with advanced infection. Systemic levels of coagulation, inflammation, and T cell immune activation predict SIV/HIV disease progression (4, 15, 16, 36, 37). We assessed multiple biomarkers: D-dimer, a biomarker of the hypercoagulation state (15) (Figure 4B and Supplemental Figure 5A); CRP, an acute-phase inflammatory marker (15) (Figure 4C and Supplemental Figure 5B); and sCD14, a biomarker of monocyte activation in response to microbial translocation (16) (Figure 4D and Supplemental Figure 5C). These markers also increase with age (38–40), so their baseline levels were higher in old versus young RMs, albeit they did not reach significance for D-dimer (P = 0.067) and CRP (P = 0.052) (Supplemental Table 1). All 3 markers significantly increased throughout infection: D-dimer (P = 4.3 × 10–11), CRP (P = 0.015), and sCD14 (P = 0.0032) (Supplemental Table 1). The timing and magnitude of D-dimer and CRP changes were different in young versus aged RMs, albeit the overall dynamics were similar, with acute peaks, partial control during early chronic infection, and gradual increases with disease progression (Figure 4, B–D). Early chronic sCD14 levels were significantly higher in old versus young RMs (P = 0.031). Although both groups largely managed to achieve a partial postacute control of D-dimer, CRP, and sCD14, the young RMs exhibited a more substantial control to nearly baseline levels.

Monocyte subsets are similar in young versus old RMs with advanced SIV infection. Monocytes are key innate immune effectors, monocyte/macrophage activation being a major determinant of inflammation/immune activation in HIV/SIV infection (41). Monocytes from people with HIV exhibit reduced counts and increased activation (41). Nonclassical monocytes contribute to an inflammatory phenotype in elderly people with HIV (41), and an increased proportion of intermediate monocytes is associated with cognitive impairments in people with HIV (42). We monitored the impact of infection on classical (CD14++CD16–); intermediate (CD14+CD16+); and nonclassical (CD14dimCD16+) monocytes (43). Classical monocytes decreased slowly during the late chronic infection stage in both aged and young RMs (Supplemental Figure 6A). Conversely, intermediate and nonclassical monocytes tended to rise with disease progression, more slowly in young RMs, eventually reaching similar levels (Supplemental Figure 6, B and C). Monocyte activation status, assessed by measuring sCD163 plasma levels (44), increased more robustly in young RMs, being partly controlled during chronic infection in both groups (Supplemental Figure 7). Meanwhile, the early chronic sCD14 levels were higher in old versus young RMs (Figure 4D and Supplemental Figure 5C).

Taken together, the young RMs initially had either advantageous baselines or greater capacity to control the initial impact of SIV infection on many (yet not all) pathogenesis biomarkers compared with old RMs. Thus, in the initial stages of SIV infection, the young RMs better preserved CD4+ T cells, had lower expression of CD4+ T cell (CD69) and CD8+ T cell (Ki67, CD69, CD38) activation biomarkers, and kept at bay hypercoagulation (D-dimer) and inflammation (CRP). Then, young RMs gradually lost these abilities during chronic infection, achieving an aged activated phenotype similar to that of old RMs by the end of the follow-up period.

SIV-induced changes in DNAme. HIV infection widely influences DNAme profiles across the human genome (45), yet the impact of SIV infection in the reference RM model remains unknown. We characterized the dynamics and tissue specificity of the host DNAme responses to SIV infection in the tissues involved in the pathogenesis of HIV-associated non-AIDS conditions in young and old RMs.

Epigenome-wide association studies of the stages of infection in PBMCs. It is rarely possible to obtain samples prior to and during the acute HIV infection; therefore, we longitudinally assessed the effect of SIV infection on DNAme in RMs including these critical time points. We performed epigenome-wide association studies (EWAS) using the entire content of the methylation chip to analyze how SIVmac infection affects the global methylation profiles in the PBMCs in our aggregated group of young and aged RMs (Figure 5A and Supporting Data Values file).

Longitudinal DNAme analysis of PBMCs. (A) Epigenome-wide association study of dpi in PBMCs based on the entire methylation array. The volcano plots display the −log10 (P values) and the directionality of association between CpG sites and infection stages (A, EC, and LC) compared with B: A versus B (left panel), EC versus B (center panel), and LC versus B (right panel). Each dot represents a specific DNAme site. Shown are significantly associated CpG sites (q < 0.05) with hypomethylation (blue), hypermethylation (red), and nonsignificant (gray). The horizontal axis represents the mean methylation change (i.e., the difference between group means), and the vertical axis represents −log10 (P values). (B) Changes in EA during each infection stage (A, EC, and LC) relative to B. Biological age analysis was performed based on subsets of clock CpGs. EA at the 3 infection time points was compared with B using mixed-effects linear regression modeling of longitudinal EA changes in PBMCs based on 10 epigenetic clocks. The results are shown separately for young (right) and old (left) RMs. Epigenetic age changes in young (blue) and old (red) RMs are shown. Saturated colors indicate statistically significant changes (P < 0.05); pale colors indicate nonsignificant changes (P > 0.05). A statistically significant increase in EA was observed only in young RMs. B–H, Benjamini–Hochberg correction; DMP, differentially methylated positions; dpi, days after infection; RMs, rhesus macaques; B, baseline; A, acute; EC, early chronic; LC, late chronic; EA, epigenetic age.

Compared with baseline, deregulation of methylation of as many as 1,943 CpGs (associated with 890 genes) emerged during acute infection; the vast majority of CpG sites returned to the baseline during early chronic infection, and only 187 CpGs (associated with 126 genes) remained differentially methylated (Figure 5A). During late chronic infection, 2,150 CpGs (associated with 911 genes) underwent a new round of widespread aberrant methylation coincident with disease progression. There was a marked bias toward CpG hypomethylation: 83% during acute, 55% during early chronic, and 71% during late chronic SIV infection. Notably, these discernible shifts in DNAme occurred within 1 year of pathogenic infection, which is the timeframe in which the majority of SIVmac-infected RMs progress to AIDS (46). The most significant longitudinal changes in PBMC DNAme throughout infection were at the CpG sites associated with the SPRED2 (hypomethylation) and ZBTB7B, also known as ThPOK (hypermethylation) genes.

We then used ingenuity pathway analysis (IPA) to longitudinally assess how methylation changes across biological pathways throughout SIV infection (47). The top canonical pathways associated with the acute infection included “aryl hydrocarbon receptor signaling,” the “role of BRCA1 in DNA damage response” pathway, and the “activin-inhibin signaling” pathway (–log10[P] > 1.3, Benjamini-Hochberg–corrected) (Supplemental Figure 8A). DNAme changes in acute SIV infection were enriched in a cluster of conditions related to liver tumorigenesis and congenital heart disease, whereas early chronic SIV infection was associated with dilated cardiomyopathy (Supplemental Figure 8B). DNAme shifts in acute and late chronic infection were enriched among various oncological categories and T cell malignancies (Supplemental Figure 9A); terms related to transcription regulation; lymphocyte homeostasis and proliferation; and differentiation, development, and homeostasis of T cells (Supplemental Figure 9B).

Given that SIV infection alters blood cell composition, mainly through the massive depletion of CD4+ T cells, we reran EWAS, accounting for the absolute CD4+ T cell counts in the blood. We observed considerably fewer significant EWAS hits (132 for acute vs. baseline, 47 for early chronic vs. baseline, and 76 for late chronic vs. baseline) (Supplemental Figure 10) in comparison to the results of EWAS without adjustment for CD4+ T cell population size, suggesting that CD4+ T cell depletion significantly contributes to the observed changes in global DNAme patterns.

DNAme studies of biological age. Epigenetic clocks predicting biological age allow assessment of the impact of various factors, including HIV, on aging (22, 23). To understand the impact of the host age on virus-driven aging, we determined the EA for each DNAme sample by using 10 epigenetic clocks previously developed based on the diverse tissue set and/or blood of primates and other mammals (48). EAA was determined as the difference between EA and chronological age.

Links between EA, stage of infection, and biomarkers of pathogenesis in PBMCs. We hypothesized that SIV infection of RMs, similar to HIV infection in humans, may affect EA. To test this hypothesis, we modeled the difference in the EA at each experimental time point compared with baseline in young and aged RMs by using a mixed-effects model that included the time point, sex, and age. A significant increase in the EA was observed during late chronic infection compared with baseline in young RMs but not in the old ones, based on 8 of 10 mammalian clocks (adjusted P < 0.05, Figure 5B). The concordance of predictions with 8 clocks implies that late-stage SIV disproportionately increases EA in the PBMCs of young RMs.

Multiple factors can affect biological aging, including infection stage, host age and sex, and ongoing pathogenic processes. We used variance analysis to evaluate the associations between EA, predicted based on the Universal Blood Clock 3 developed for blood and the key markers of SIV/HIV pathogenesis, including disease progression (CD4+ T cell counts, plasma viral loads, immune activation, and inflammation), microbial translocation, and hypercoagulation. We considered EA as a dependent variable and modeled it as a function of infection stage, age, sex, and individual. We tested whether including each biomarker in the model significantly improved the EA predictions.

In young RMs, the EA of the circulating PBMCs was negatively associated with the abundance of CD4+ T cells in SLNs (P = 0.016) and positively associated with systemic T cell activation, that is, CD25+CD4+ T cells (P = 0.027), HLA-DR+CD38+CD4+ T cells (P = 0.0082), and CD8+CD25+CD8+ T cells (P = 0.039) (Figure 6A). In old RMs, EA was positively associated with the plasma levels of CRP (P = 0.027) and sCD163 (P = 0.014) (Figure 6B).

Biomarkers associated with epigenetic age in PBMCs across infection stages. P values above bars were obtained from a 2-sided ANOVA analysis comparing a full mixed-effects model, which included biomarker levels (measured at the time point closest to the clock measurements) as independent variables to a null model. The full model accounted for time point (B, A, EC, and LC), age, sex, and given biomarker; the null model excluded biomarker levels. For significantly associated markers, P values are shown on the plot, along with an indication of the directionality of the association (positive or negative). The results for young (A, blue) and old (B, red) RMs are shown. Saturated colors indicate statistically significant changes (P < 0.05); pale colors indicate nonsignificant changes (P > 0.05). Biomarker levels were measured during infection stage corresponding to clock measurements: 0 dpi, B; 9–21 dpi, A stage; 42–49 dpi, EC stage (except the immune activation markers in the gut, measured at 70 dpi); and more than 130 dpi, LC stage. B, baseline; A, acute; EC, early chronic; LC, late chronic; dpi, days after infection; RMs, rhesus macaques.

Tissue-specific differences in EAA associated with SIV infection. To assess the impact of SIV infection on tissue aging, we used the Wilcoxon test and compared EAA between young and old RMs in 6 solid tissues during advanced-stage infection (Figure 7). EAA estimates based on the third universal pan-mammalian epigenetic clock (Universal Clock 3), the most robust multitissue pan-mammalian clock, indicated greater EAA in the young RMs versus aged RMs, particularly in the cerebellum, heart, and spleen, whereas the colon exhibited greater EAA in the aged versus young RMs. When applying 9 other epigenetic clocks, EAA increased more prominently in the young RMs in the cerebellum (per all 10 clocks), heart (per 9 of 10 clocks), spleen (per 5 of 10 clocks), liver (per 2 of 10 clocks), and colon (per the Primate Relative Clock) (Wilcoxon, P < 0.05) (Supplemental Figure 11). We did not observe EAA in the adipose tissue of RMs, but a smaller number of samples of this tissue was analyzed.

Epigenetic age acceleration in young (blue) and old (red) rhesus macaques in late chronic SIVmac infection in 6 tissue types based on Universal Clock 3. The differences in epigenetic age acceleration between the young and old rhesus macaques were assessed using a 2-sided Wilcoxon test.

We next assessed the effect of age status on EAA in solid tissues with a subset of 4 epigenetic clocks most representative of internal tissues (Figure 8). The clocks concordantly pointed to a positive effect of age status on EAA in the colon, cerebellum, spleen, fat, and heart in the young RMs and in the colon, cerebellum, and fat in the aged RMs. These findings suggest that susceptibility to virus-driven EAA is tissue-specific and that the host’s age is a risk factor differentially influencing different tissues. Although the colon has a high cell turnover, the heart and cerebellum comprise extremely long-lived cells, which in the long-term may be particularly vulnerable to HIV-associated non-AIDS comorbidities.

Effects of age (young or old) on epigenetic age acceleration in tissues collected at necropsy from young and old rhesus macaques during late chronic SIV infection. The results for young (blue) and old (red) rhesus macaques are shown. Saturated colors indicate statistically significant differences (P < 0.05); pale colors indicate nonsignificant differences (P ≥ 0.05). The effects were assessed using linear regression analysis, where epigenetic age acceleration was modeled as a function of age status (young or old). The clock measurements were obtained from tissues collected during necropsies in the late chronic infection stage (225–360 dpi for old and 366–380 dpi for young). Dpi, days after infection.

Correlation of biomarkers of SIV pathogenesis and EAA in tissues during the late chronic infection stage. Given that HIV pathogenesis biomarkers are associated with EAA in the circulation and the gut (49), we calculated Spearman’s rank correlation between EAA measured using Universal Clock 3 in the solid tissues and circulating biomarkers of SIV pathogenesis, including CD4+ T cells, viral loads, T cell activation and proliferation status, inflammation, and hypercoagulation. We also analyzed the relationship between the estimated SIV DNA and RNA and EAA for each tissue, but did not find any significant associations.

High levels of circulating markers of immune activation were positively correlated with EAA in tissues (Figure 9). In the young RMs, HLA-DR+CD38+CD4+ T cells correlated with EAA in the spleen (P = 0.013), and HLA-DR+CD38+CD8+ T cells correlated with EAA in the spleen (P = 0.031) and colon (P = 0.038) (Figure 9A). Conversely, in aged RMs, Ki-67+CD8+ T cells positively correlated with EAA in the spleen (P = 0.012) and displayed a positive trend with EAA in the cerebellum (P = 0.05). Taken together, these results imply a relationship between EAA and CD4+ and CD8+ T cell activation in young RMs and CD8+ T cell expansion in aged RMs (Figure 9B).

Association between tissue epigenetic age and peripheral pathogenesis markers. Correlation between biomarkers of immune activation, inflammation, coagulation, and cell-associated viral DNA and epigenetic age acceleration in tissues collected from young (A) and old (B) rhesus macaques during late SIV infection. Tissues for clock estimates were obtained during necropsy. Measurements of biomarker levels were obtained either from the necropsy or the nearest available time point (between 200 dpi and necropsy). Pearson’s correlation was used to assess these relationships. Statistically significant correlations (P < 0.05) are shown in blue; nonsignificant correlations (P > 0.05) are shown in yellow. Dpi, days after infection.

We also found that several inflammatory markers correlated with the rate of EAA. In old RMs, EAA in the spleen was directly related to the plasma levels of IL-1B, a key mediator of inflammatory response (P = 0.038), and IL-12, a biomarker of cardiovascular disease (50) and promoter of spleen hematopoiesis (P = 0.032) (51). Also, D-dimer exhibited a positive correlation trend with EAA in the cerebellum in the old RMs (P = 0.051).

In young RMs, colon EAA correlated with plasma levels of IL-8 (a biomarker of chronic inflammation acting as a neutrophil chemoattractant that stimulates neutrophil phagocytic activity; ref. 52; P = 0.02), whereas EAA in the heart exhibited a positive correlation trend with plasma IL-12 levels (P = 0.055). IL-15 exhibited the strongest correlation among the plasma biomarkers, and was the only biomarker negatively correlated with tissue EAA. Taken together, our results indicate the role of systemic inflammation (increased IL-1 and IL-12 levels) in the aging process and potential splenic dysfunction in aged RMs. In young RMs, aging of tissues with low cell turnover (i.e., heart and cerebellum) was related to IL-15 and IL-12 deregulation.