Transcriptomic analyses of peripheral blood mononuclear cells reveal age-specific basal and acute exercise responsiveness differences in humans

Abstract

Aging is associated with alterations in immune cell function, contributing to age-related diseases and frailty. As peripheral blood mononuclear cells (PBMCs) are key drivers of the immune response, we investigated their transcriptome using RNA-sequencing before and immediately after a single bout of high-intensity knee-extension exercise in young (Young; n=7, 23 ± 4 years) and older individuals (Old; n=8, 65 ± 7 years). We used bioinformatics analyses to identify the biological processes and pathways that may be altered with age and in response to acute exercise. At baseline, 665 genes differed between Young and Old with notable differences in pathways involved in DNA Damage/Telomere Stress Induced Senescence, NAD Signaling Pathway, and Oxidative Stress Induced Senescence. After the exercise bout, 53 genes were differentially expressed in Young, while 1,026 genes changed in Old. In Young, the enriched processes and predicted pathways were linked to natural killer cells, while in Old these pathways were associated with cell signaling immune responses. Lastly, 26 genes exhibited similar responses to exercise between groups, enriching the biological process of natural killer cell-mediated immunity regulation. Our findings indicate that PBMC gene expression and the response to acute exercise are altered with aging, where exercise induced more pronounced PBMC transcriptomic adaptations in the Old. Additionally, while aging is associated with increased expression of genes linked to cellular dysfunction and suppressed immune function, acute exercise attenuated these age-related differences by downregulating the genes related to those pathways. Finally, acute exercise activated similar immune-related pathways in both age groups.

New and Noteworthy:

This study demonstrates that aging alters the transcriptional landscape of PBMCs at rest and in response to acute high-intensity exercise. Older adults exhibited greater transcriptomic responsiveness to exercise, particularly in pathways related to immune signaling and cellular stress. Notably, exercise elicited shared activation of NK cell-mediated processes across age groups, suggesting a conserved immunomodulatory effect. These findings provide molecular insight into how aging and exercise interact to shape immune cell function.

Graphical Abstract

Introduction

Aging is a complex process characterized by the functional decline of multiple systems in the body as a result of cell specific alterations in molecular processes (1). Critically, there are hallmarks of aging in the immune system, including a phenomenon known as immunosenescence, where both the innate and adaptive immune responses become less effective. Immunosenescence is associated with chronic inflammation and a decline in the production of naive T cells, leading to a reduced immune response to infections (2). Together, these alterations contribute to age-related conditions such as sarcopenia and frailty (3). Exercise training combats these hallmarks of aging as it has been shown to reduce inflammation and enhance immune function. Specifically, exercise in older individuals increases T cell proliferation, cytokine production, and natural killer (NK) cell cytotoxicity (4, 5), all of which are normally diminished with an aging immune system (6).

Peripheral blood mononuclear cells (PBMCs), which include lymphocytes (T cells, B cells, NK cells) and monocytes, play a vital role in immune responses, helping to repair damage induced by oxidative stress and inflammation (7, 8). Importantly, these cells provide insight into age-related immune deficiencies critically mediating stress and inflammation via the production of cytokines, chemokines, and growth factors that support immune responses (9–11). While PBMCs are the primary immune responders, they are also modulated by multiple systemic exercise-induced signals, including the release of exerkines (e.g., IL-6) from other tissues. These circulating factors can influence PBMC gene expression and contribute to systemic adaptations to exercise (12). Given PBMCs crucial role in the immune system, their responsiveness to acute stressors has been suggested to provide a better assessment of functional differences between populations (13). A single bout of exercise is a potent stimulus for PBMCs, and the subsequent RNA changes reveal the molecular adaptation processes. Notably, previous findings show that 30-minutes of exercise significantly alters global gene expression in PBMCs, triggering an immediate proinflammatory response followed by rapid anti-inflammatory signaling that prevents chronic inflammation (14). These gene expression changes are largely transient, returning to or dropping below baseline levels within an hour of recovery (15). In addition, the effects of exercise extend beyond immune pathways to include genes involved in energy metabolism, growth, and tissue repair (14–16). Thus, these cells provide key mechanistic insight into how aging influences immune regulation, helping to identify the molecular mechanisms underlying age-related immune system changes.

Several important studies have independently assessed the effect of age on resting PBMC gene expression (17–19) or the effect of different modalities of exercise on PBMC gene responsiveness (16, 20). However, the direct effect of age is difficult to determine in these studies due to varied exercise stimuli between groups. For example, studies investigating PBMC responses vary widely in exercise intensity (16) and duration, ranging from 30 minutes (21) to 240 minutes (22). Moreover, studies also differ substantially with regard to sample collection timing as the first post-exercise samples have been taken as early as immediately after exercise (0 minutes, (14)) or as late as 240 minutes post-exercise (22). While the trend of each study is an inflammatory response in these cells, it becomes difficult to generalize these findings or disentangle specific pathways from the effects of variables such as age due to methodological inconsistencies across studies. Therefore, to understand the effect of age on the exercise-induced transcriptomic response in PBMCs, we utilized RNA-sequencing before and immediately after an acute bout of aerobic exercise in healthy young and older individuals. To complement transcriptomic profiling and better characterize the immune response of the exercise protocol, we also measured circulating levels of interleukin-6 (IL-6) and tumor necrosis factor-alpha (TNF-α). These cytokines were selected because they are well-established markers of acute inflammation in response to exercise (12). We used bioinformatics analyses to identify the biological processes and pathways that may be altered with age and in response to acute exercise. We hypothesized that i) age differentially modulates basal gene expression in PBMCs and this altered gene expression relates to changes in other pathways known to be affected by aging (i.e., autophagy, inflammation, and cellular senescence) and ii) younger individuals would exhibit a greater gene expression response to exercise and the altered genes would be specific to redox regulation and inflammatory pathways compared to the older individuals.

Methods

Ethical approval

Experimental procedures were reviewed and approved by the University of Utah and the Salt Lake City Veterans Affairs Medical Center Institutional Review Board and were conducted according to the standards set by the latest revisions of the Declaration of Helsinki.

Participants

Eligible participants were between the ages of 18-30 years old (Young) or 55-80 years old (Old), and free of overt cardiovascular and metabolic diseases. Participants were required to be relatively healthy, with no history of chronic diseases that could impact exercise performance or confound the study results. Eight old adults (Sex = 4M/4F, 65 ± 7 years, 80.6 ± 16.5 kg, 26.7 ± 3.5 kg/m²; mean ± SD) and seven young adults (Sex = 2M/5F, 23 ± 4 years, 79.3 ± 15.4 kg, 26.9 ± 5.1 kg/m²) provided verbal and written consent to participate prior to the data collection procedures outlined below.

Experimental Design

During visit 1, participants completed the informed consent and health history questionnaire and were given an accelerometer to wear for a minimum of seven days prior to returning to the laboratory for visit 2. During visit 2, participants completed a graded single-leg knee extension (KE) exercise protocol to determine their KE maximal work rate on a custom knee extension ergometer (23, 24). Lastly, for visit 3, participants reported to the laboratory around 8:00 AM following an overnight fast, with all testing completed within two hours. All testing was performed using the same leg for each participant. Height and weight were recorded, and prior to and immediately following a 60-minute bout of aerobic KE, blood was obtained via an indwelling venous catheter in the antecubital space. Baseline blood glucose, total cholesterol, low-density lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL) cholesterol, and triglycerides were analyzed by a diagnostic laboratory (Roche Cobas Pro C503 and E801). Inflammatory markers IL-6 (Cat No: HS600C R&D systems) and TNFα (Cat No: DTA00D, R&D systems), were measured from blood samples collected immediately before and after the exercise.

Max Test

After familiarization with KE, participants performed consecutive 1-min stages at 60 rpm. The resistance started unloaded and increased five watts every stage until volitional exhaustion. At the end of each stage, blood pressure, heart rate, and ratings of perceived exertion (RPE) were recorded. Maximal work rate was determined as the highest work rate maintained for greater than 30 seconds. Participants then rested for 10 min before completing three minutes of KE at 80% of their maximal work rate to verify they could sustain this work rate during visit 3.

Testing Protocol

KE exercise started at 20% of maximal work rate and increased by 20% every three minutes. Following this initial progression, 3-minute intervals were performed at 80% of maximal work rate with three minutes of passive recovery separating each exercise interval. A total of nine intervals were performed and the total duration of the exercise bout was 1 hour. This small muscle mass exercise paradigm isolates the quadriceps and minimizes central limitations to exercise that may be present when comparing young and older adults.

Collection of Peripheral Blood Mononuclear Cells

Venous blood was collected before exercise and within one minute of stopping exercise into two 10 ml BD K2EDTA vacutainers (Cat No: BD-366643; VWR Avantor, Radnor PA) for fresh PBMC isolation. PBMCs were separated by density-gradient centrifugation using SepMate^™ −50 (Cat No: 85450; STEMCELL technologies, Vancouver, Canada) tubes with 15 ml of a density gradient medium, Histopaque (Cat No: 10771; Sigma-Aldrich, St. Louis, MO). Next, EDTA anti-coagulated blood was diluted 1:1 with phosphate-buffered saline (PBS) by mixing 14-16 ml of whole blood with 14-16 ml of PBS and centrifuged at 1200g for 10 minutes at room temperature (with the brake on). The PBMC layer was carefully transferred into a 50 ml conical tube with 25 ml PBS and centrifuged at 700g for 10 minutes. PBMCs were then washed by discarding the supernatant, resuspending the pellet in 1 ml PBS before bringing the total volume to 45 ml and centrifuging again at 700g for 10 minutes. The final PBMC pellet was resuspended in 5 ml of freezing media (fetal bovine serum, supplemented with 10% dimethyl sulfoxide, Sigma-Aldrich, St. Louis, MO and EBM-2, Lonza, Basel Switzerland) to be stored at −80 °C for 24 hours before the vials were transferred and stored at −150 °C until RNA isolation described below. Cell count and viability were determined from a hemocytometer from a 20-fold dilution using 90 μl of freezing media, 10 μl of PBMCs and 10 μl of trypan blue (Cat No: T10282; Thermo Fisher).

RNA Isolation and RNA-Sequencing

Total RNA extraction for PBMC samples was achieved using RNeasy Plus Mini Kit (Cat No: 74134; Qiagen), with DNase treatment, with RNA isolated from 1 to 5 × 10^6 cells. Total RNA samples (5-500 ng) were hybridized with NEBNext rRNA Depletion Kit v2 (Human, Mouse, Rat) (E7400) to diminish rRNA from the samples. A total amount of 200 ng RNA per sample was used as input material for the RNA-seq sample preparations. Stranded RNA sequencing libraries were prepared as described using the NEBNext Ultra II Directional RNA Library Prep Kit for Illumina (E7760L). Purified libraries were qualified on an Agilent Technologies 4150 TapeStation using a D1000 ScreenTape assay (25 5067-5582 and 5067-5583). The molarity of adapter-modified molecules was defined by quantitative PCR using the Kapa Biosystems Kapa Library Quant Kit (25KK4824). Individual libraries were normalized and pooled in preparation for Illumina sequence analysis. NovaSeq X Series 10B Reagent Kit 150x150 bp Sequencing: Sequencing libraries were chemically denatured in preparation for sequencing. Following transfer of the denatured samples to an Illumina NovaSeq X instrument, a 151 x 151 cycle paired end sequence run was performed using a NovaSeq X Series 10B Reagent Kit (20085594).

The human GRCh38 genome and gene annotation files were downloaded from Ensembl release 110 and a reference database was created using STAR version 2.7.9a with splice junctions optimized for 150 base pair reads (26). Optical duplicates were removed from the paired end FASTQ files using clumpify v38.34 and reads were trimmed of adapters using cutadapt 1.16 (27). The trimmed reads were aligned to the reference database using STAR in two pass mode to output a BAM file sorted by coordinates. Mapped reads were assigned to annotated genes using featureCounts version 1.6.3 (28). The output files from cutadapt, FastQC, FastQ Screen, Picard CollectRnaSeqMetrics, STAR and featureCounts were summarized using MultiQC to check for any sample outliers (29). Prior to differential expression analysis, genes were filtered to remove 1) genes with zero counts across all samples and 2) genes with fewer than 5 reads in every sample. Differentially expressed genes were identified using a 5% false discovery rate with DESeq2 version 1.40.2 (30). Log2 fold changes in DESeq2 were shrunk using the apeglm shrinkage estimator.

Statistical analysis

All statistical analyses were performed using SPSS Version 25 (IBM SPSS Statistics Software). Prior to analysis, we tested for normality on all baseline dependent variables using the Shapiro-Wilk test, and all variables were found to be normally distributed. Baseline comparisons from Old and Young were made using independent samples t-tests. Two-way (age × time) repeated measure ANOVA were used to determine changes in IL-6 and TNFα. Percent changes in IL-6 and TNFα were calculated ([[Post−Pre]/Pre] × 100) and analyzed using independent samples t-tests to compare between age groups. Statistical significance was established at p < 0.05. Differentially Expressed Genes (DEGs) were identified using a threshold of p < 0.01, an adjusted p-value (False Discovery Rate (FDR)) of p < 0.01. Among these significant genes, we separated up-regulated and downregulated DEGs into two groups. For each group, we calculated the mean Log2 fold change, and only genes with a Log2 fold change greater than the group specific mean were used for further analysis. These DEGs, including both up- and down-regulated genes, were analyzed separately and entered into g:Profiler (31) to detect statistically significant enriched biological processes (Gene Ontology), using an FDR threshold of p < 0.05 (32). To analyze pathways and networks from the RNA-seq data, we used the Ingenuity Pathway Analysis (IPA) bioinformatics software with the Qiagen Knowledge Base (33). The DEG data were uploaded into IPA, and canonical pathway analysis was performed to identify the significantly enriched biological pathways among the DEGs based on the ratio, p-value, and z-score. The ratio was calculated as the number of DEGs in the dataset divided by the total number of genes in the canonical pathway database. The z-score represents the deviation from the expected activation state of a canonical pathway, based on the expression of the genes in the data set. The p-value for canonical pathway analysis was calculated using Fisher’s exact test. All figures were constructed using GraphPad Prism v9.2.0 (San Diego, CA, United States).

Results

Baseline Characteristics

Besides age, cholesterol and LDL cholesterol, the participants were well matched for baseline characteristics (Table 1). Out of the 19,762 total genes, RNA-Seq revealed 665 genes were significantly different between groups at baseline (Old compared to Young: 446 up, 199 down, Figure 1a). The genes that were expressed more in older individuals were associated with the biological process of “multicellular organismal processes” (GO:0032501). Numerous canonical pathways were significantly activated by these genes, including DNA Damage/Telomere Stress Induced Senescence (z score = 2.0), NAD Signaling Pathway (z score = 1.9), and Oxidative Stress Induced Senescence (z score = 1.3), with additional pathways presented in Table 2.

Table 1.

Data are presented as means ± standard deviation values.

Variable	Young	Old	p-value
Age (years old)	23 ± 4	65 ± 7*	<0.001
Sex (Male / Female)	2 / 5	4 / 4
Body Mass (kg)	79.3 ± 15.4	80.6 ± 16.6	0.872
BMI (kg/m2)	26.9 ± 5.1	26.7 ± 3.5	0.932
Maximal Power (Watts)	35.0 ± 6.5	29 ± 18.1	0.422
Daily Step Count	3980 ± 1845	6185 ± 3805	0.187
Sedentary Time (min/day)	1268 ± 60	1267 ± 57	0.984
Light Activity Time (min/day)	104 ± 45	135± 47	0.227
Moderate Activity Time (min/day)	25 ± 14	35 ± 27	0.417
Vigorous Activity Time (min/day)	1 ± 2	1 ± 2	0.880
Glucose (mg/dL)	96.0 ± 24.7	88.4 ± 6.6	0.415
Total Cholesterol (mg/dL)	153.6 ± 15.1	195.4 ± 29.4*	0.005
Triglycerides (mg/dL)	101.5 ± 76.8	90.4 ± 31.9	0.713
HDL Cholesterol (mg/dL)	49.2 ± 7.4	58.0 ± 11.7	0.113
LDL Cholesterol (mg/dL)	90.5 ± 19.0	126.73 ± 23.4*	0.006
Total Cholesterol/HDL Ratio	3.2 ± 0.7	3.4 ± 0.5	0.465

Abbreviation: HDL, high-density lipoprotein; LDL, low-density lipoprotein.

^*Indicates statistically significant from Young (p < 0.05).

Figure 1.

Differentially expressed genes (DEGs) viewed as a volcano plot where colored DEGs indicate Old have a >mean Log2 fold-change (p < 0.01) from Young (a), Young post-exercise have a >mean Log2 fold-change (p < 0.01) from pre-exercise (b) and Old post-exercise have a >mean Log2 fold-change (p < 0.01) from pre-exercise (c). DEGs biological processes (gene ontology [GO] terms) associated with up- or down-regulated DEGs in Young PBMCs (d) and Old PBMCs (e) after exercise (False Discovery Rate p < 0.05).

Table 2.

Data presented are from Ingenuity Pathway Analysis for predicted pathways in peripheral blood mononuclear cells for Old vs Young. A −log(p-value) greater than 2 indicates significant pathways, which is calculated using the right-tailed Fisher’s Exact Test. Ratio is determined by dividing the number of significant differentially expressed genes by the total number of genes within the pathway. The z-score indicates the activation or inhibition state of the pathway. A z-score greater than or equal to 2 indicates prediction of activation, while a z-score less than or equal to -2 indicates prediction of inhibition. Not A Number (NaN) indicates that pathway is not eligible for z-score prediction because there were fewer than four genes in the pathway.

Ingenuity Canonical Pathways	−log(p-value)	Ratio	z-score
Pre-NOTCH Expression and Processing	2.53	0.0658	2.236
NGF-stimulated transcription	2.82	0.103	2
DNA Damage/Telomere Stress Induced Senescence	2.24	0.0714	2
NAD Signaling Pathway	2.47	0.0464	1.89
ABRA Signaling Pathway	2.92	0.0652	1.633
G-Protein Coupled Receptor Signaling	3.5	0.0299	1.528
Oxidative Stress Induced Senescence	2.3	0.0581	1.342
DNA methylation	2.36	0.115	NaN
SIRT1 negatively regulates rRNA expression	2.23	0.103	NaN
PRC2 methylates histones and DNA	2	0.0857	NaN

Exercise Induced Changes in Gene Expression in Peripheral Blood Mononuclear Cells

PBMC RNA-Seq in the Young revealed 53 out of 13,695 genes were significantly altered in response to the exercise bout (52 up-regulated, 1 down-regulated, Figure 1b). These 52 up-regulated genes significantly enriched three biological processes including “natural killer cell mediated cytotoxicity” (GO:0042267), “regulation of natural killer cell mediated immunity” (GO:0002715), and “natural killer cell inhibitory signaling pathway” ((GO: 0002769), Figure 1d). The one down-regulated gene, SH2D3A, was not significantly linked to an enriched biological process, but the 53 genes were significantly related to numerous canonical pathways (significant relevant pathways presented in Table 3). In the Old, 1,026 out of 19,101 genes were significantly altered (359 up-regulated, 572 down-regulated, Figure 1c). These 359 up-regulated genes significantly enriched one biological process, “regulation of transcription by RNA polymerase II” ((GO:0006357), Figure 1e). The 572 down-regulated genes significantly enriched nine biological processes (Figure 1e). Both up- and down-regulated genes were run through IPA, which indicated significant predicted inhibition of numerous canonical pathways (significant relevant pathways presented in Table 3).

Table 3.

Data presented are from Ingenuity Pathway Analysis for predicted pathways in peripheral blood mononuclear cells changes with exercise in both Young and Old. A −log(p-value) greater than 2 indicates significant pathways, which is calculated using the right-tailed Fisher’s Exact Test. Ratio is determined by dividing the number of significant differentially expressed genes by the total number of genes within the pathway. The z-score indicates the activation or inhibition state of the pathway. A z-score greater than or equal to 2 indicates prediction of activation, while a z-score less than or equal to -2 indicates prediction of inhibition. Not A Number (NaN) indicates that the pathway is not eligible for z-score prediction because there were fewer than four genes in the pathway.

Ingenuity Canonical Pathways	−log(p-value)	Ratio	z-score
Peripheral Blood Mononuclear Cells Young Pre vs Young Post
Immunoregulatory interactions between a Lymphoid and a non-Lymphoid cell	8.85	0.0446	3
Th1 Pathway	3.71	0.0342	2
FAK Signaling	2.43	0.00853	1.89
Pathogen Induced Cytokine Storm Signaling Pathway	2.84	0.0147	0.447
Natural Killer Cell Signaling	9.02	0.0466	0.333
Crosstalk between Dendritic Cells and Natural Killer Cells	7.03	0.0659	NaN
Granzyme B Signaling	3.16	0.125	NaN
Th1 and Th2 Activation Pathway	3.12	0.024	NaN
Immunogenic Cell Death Signaling Pathway	2.94	0.0357	NaN
DAP12 interactions	2.32	0.0476	NaN
Peripheral Blood Mononuclear Cells Old Pre vs Old Post
Platelet homeostasis	2.89	0.0814	−2.646
GP6 Signaling Pathway	2.03	0.0569	−2.646
HGF Signaling	3.03	0.0687	−2.236
α-Adrenergic Signaling	2.38	0.066	−2.236
Collagen biosynthesis and modifying enzymes	2.04	0.0746	−2.236
Cargo concentration in the ER	2.42	0.118	−2
PI3K Cascade	2.01	0.0909	−2
Gαi Signaling	2.26	0.0571	−1.89
Actin Cytoskeleton Signaling	2.18	0.0458	−1
Protein Kinase A Signaling	2.04	0.0381	−0.832

Between Young and Old, 26 genes had a similar response to exercise (Figure 2a). These 26 genes only significantly enriched one biological process, “regulation of natural killer cell mediated immunity” ((GO:0002715), Figure 2b), and no similar pathways were found through IPA. Interestingly, the 665 genes that were significantly different at baseline between Young and Old was reduced to only six genes after the exercise (Old compared to Young: one up, five down, not shown).

Figure 2.

Similarities between groups in gene response to exercise in peripheral blood mononuclear cells (PBMCs). (a) Heatmap showing the 26 significantly increased genes shared between Old and Young in response to acute exercise. (b) The similar biological processes (gene ontology [GO] terms) from the up-regulated DEGs in both Old and Young with acute exercise in PBMCs (False Discovery Rate p < 0.05). (c) Venn diagram illustrating similar and different DEGs in response to acute exercise in Old and Young in PBMCs.

Systemic Cytokine Response to Exercise

To evaluate systemic inflammatory responses to acute exercise, plasma concentrations of IL-6 and TNF-α were measured before and after exercise (Figure 3). IL-6 revealed a significant main effect of exercise (p < 0.001), with no significant main effect of age (p = 0.598) or age × time interaction (p = 0.339, Figure 3a), indicating that IL-6 increased post-exercise across both age groups. However, the percent change in IL-6 differed significantly between groups, with a greater increase in Young (223.39 ± 200.83%) compared to Old (88.57 ± 75.47%, p = 0.050). For TNF-α, there was also a significant main effect of exercise (p = 0.004), with no significant main effect of age (p = 0.064) or age × time interaction (p = 0.066, Figure 3b). Percent change analysis showed a significantly greater reduction in TNF-α levels in Old (−13.39 ± 9.33%) compared to Young (−4.17 ± 10.42%, p = 0.047).

Figure 3.

Circulating levels of (a) interleukin-6 (IL-6) and (b) tumor necrosis factor alpha (TNF-α) were measured before (white bars) and after (gray bars) acute single-leg knee extension exercise.

Discussion

While numerous studies have explored the effects of age and acute exercise on PBMCs, to our knowledge, this study is the first study to investigate the transcriptome of PBMCs between different age groups before and after acute exercise. Our key findings reveal that, at baseline, older individuals exhibit a higher number of up-regulated genes in PBMCs compared to their younger counterparts, and these genes are linked to other pathways known to be affected by aging. After exercise, younger individuals show significant activation of immune-related processes, including cytokine release and immune cell crosstalk, suggesting enhanced immune function and/or better regulation of exercise induced inflammation. In contrast, older individuals display a dramatic reduction in the expression of RNA after exercise, which leads to the down-regulation of genes involved in pathways of cellular dysfunction, impaired tissue repair, and immunosuppression. Notably, young and old participants shared some similar gene up-regulation in response to exercise, which were related to the biological process regulation of NK cell immunity. Lastly, while there were substantial differences in gene expression at baseline between young and old, these differences were diminished after exercise, driven by attenuated gene expression in older adults. Together, these findings suggest that while aging is associated with increased baseline expression of genes linked to cellular dysfunction and suppressed immune function, acute exercise appears to normalize age-related differences in the PBMC transcriptome, partially through the activation of specific immune-related processes in both age groups, with exercise-induced immune suppression observed in older adults.

Baseline differences in peripheral blood mononuclear cells with age

At baseline, significant gene expression differences were observed in PBMCs between the two age groups. Specifically, older individuals had a marked up-regulation of genes associated with age related pathways, including increased methylation, DNA damage, and senescence. Although this study did not directly investigate these biomarkers, existing research provides evidence that aligns with the predictive pathways identified here (17, 34–37). For example, DNA methylation provides a reliable biomarker for predicting donor age in PBMCs (17). Similarly, Steegenga et al. demonstrated that age-related DNA methylation of genes involved in developmental processes occurs in PBMCs, which may explain the observed multicellular organismal process associated with aging in our dataset. In addition to methylation, senescence is a key age-related process that has been well-documented in PBMCs (34). In our dataset, senescence was accompanied by the terms of oxidative stress and DNA damage/telomere stress contributing to senescence. Age-related oxidative stress is recognized to shorten telomere length in PBMCs, which has been correlated with broader markers of aging, such as vascular aging or sarcopenia (35, 36, 38). Another notable age-related change in PBMCs is a decline in energy metabolism (37). Energy metabolism is crucial for the activation, proliferation, and overall function of PBMCs (39, 40). In our current study, we found that the NAD signaling pathway, a key regulator of cellular energy homeostasis, was significantly enriched with age, further supporting the notion of altered metabolism in aging PBMCs. When combined with increased senescence and methylation, these findings further emphasize the importance of exploring PBMC functionality as a biomarker in the context of aging and the impact on the immune system.

Peripheral blood mononuclear cells with acute exercise

PBMCs are commonly studied due to their accessibility, but most research has been conducted using cross-sectional studies or a single age group, leaving a definitive gap in the information about acute exercise response with aging. The current findings reveal considerable age-related differences in gene expression in PBMCs, but these differences were diminished following exercise. These changes occurred due to the down-regulation of genes that were highly expressed in Old at baseline, rather than the up-regulation of genes in Young. One possible explanation for this dramatic drop in gene expression could be that the exercise causes a transient suppression of the immune system (41–43). This has been described previously as part of the body’s adaptive response, where the immune system temporarily becomes less active before recovering and potentially becomes more efficient over time (44, 45). Old showed a greater relative reduction in TNF-α following exercise, consistent with transcriptional patterns indicating immune suppression and reduced inflammatory signaling. In our current dataset, the predicted pathway inhibition of hepatocyte growth factor (HGF) signaling or α-adrenergic signaling suggests an impaired migration and activation of immune cells such as T cells and macrophages, potentially leading to reduced tissue repair, and diminished ability to combat inflammation after exercise (46–48). This may have been the response in the older individuals for several reasons. First, older individuals already have weaker immune systems and combining this with exercise may have exacerbated the response (49, 50). Second, the more pronounced transcriptomic response observed in older participants may indicate that the relative physiological stress imposed by the exercise bout was greater in this group. In younger individuals, exercise often evokes a less extensive transcriptomic shift unless the activity is prolonged or extremely strenuous (44, 45). However, this does not imply a blunted or dysregulated immune response in the Young. The significantly higher increase in IL-6 post-exercise, a well-known exerkine, is often associated with beneficial immune activation and metabolic signaling during recovery (51, 52). Third, the analyses were conducted separately for each age group, resulting in different gene sets being compared, which introduces a potential limitation in that the differing gene sets may impact interpretation of the comparing results across the groups. Last, while the changes observed following an acute bout of exercise may suggest immune suppression, these effects are transient, with studies showing that PBMC transcriptional alterations return to baseline within 30 minutes post-exercise (15) and improved with chronic exercise (47, 48). While acute exercise may temporarily suppress immune function in older individuals, these effects are likely short-lived and may contribute to long-term improvements in immune system efficiency with regular, chronic exercise.

Similar gene expression in Old and Young with exercise

The exercise response across age groups showed remarkably different responses in gene expression yet there was a small overlap of 26 genes. Regardless of age, these genes up-regulated with exercise, and their combination activity enriched a single biological process, regulation of NK cell mediated immunity. This suggests that exercise preferentially influences the activity of NK cells independent of age. This alone is not unique as NK cells are a key component of the first line of defense in the innate immune response (53). In fact, it is well-documented that acute exercise increases NK cell numbers (54–56). While this connection to the regulation of natural killer cells is clear, another important aspect of NK cell function during exercise is their cytotoxic response to stressors induced by exercise. However, the response of NK cell cytotoxicity to exercise remains inconsistent in the literature (54). This inconsistency can be attributed to variations in exercise modalities and methodologies used to assess cytotoxicity. Nevertheless, the prevailing trend suggests an increase in NK cell cytotoxicity immediately following exercise (53). Data on age-related differences in NK cell cytotoxicity is limited, previous work demonstrated that while both young and older individuals experienced increases in NK cell count after exercise, only the young group showed an increase in NK cell cytotoxicity (57). While the current study did not measure NK cell numbers, it is reasonable to conclude that an increase in NK cell numbers may have occurred in both young and older individuals, based on previous literature and the similar biological processes following exercise (54–57). However, it is an intriguing possibility that NK cell cytotoxicity may only be up-regulated in the young (shown with the enriched biological processes natural killer cell mediated cytotoxicity), while remaining unchanged in the elderly following exercise. Since cytotoxicity data was not collected, future studies are needed to test this hypothesis.

Conclusion

This study provides novel insights into the transcriptome response of PBMCs with aging in conjunction with acute exercise. Our findings reveal that significant age-related differences are evident in PBMCs. Specifically, there is an up-regulation of genes associated with aging (i.e., DNA methylation, oxidative stress induced senescence, DNA damage/telomere stress induced senescence) in PBMCs of older individuals at baseline. While the mitigation of these age-related differences following exercise were diminished, the exercise-induced down-regulation of differentially expressed genes in the older participants were primarily related to pathways of the immune response (i.e., HGF Signaling, Platelet homeostasis, α-Adrenergic Signaling). Collectively, these findings support the divergent effect of exercise to modulate the immune system in different cell signaling pathways across the lifespan. However, the small sample size restricted our ability to evaluate sex differences, including potential effects of menstrual cycle phases on immune and inflammatory responses to exercise (58). Future studies with larger, sex-balanced cohorts are warranted to better understand these dynamics. In addition, longitudinal research is needed to confirm these findings by examining the acute and long-term effects of various exercise regimens on gene expression, cellular stress responses, and functional outcomes in these heterogeneous immune cell populations.

Data availability

All RNA-Seq data generated in this study are publicly available and can be accessed through the Gene Expression Omnibus (GEO) database. The data can be accessed at the following URL: www.ncbi.nlm.nih.gov/geo/, under the GEO accession number GSE293163.