Exercise reduces circulating biomarkers of cellular senescence in humans

Abstract Cellular senescence has emerged as a significant and potentially tractable mechanism of aging and multiple aging‐related conditions. Biomarkers of senescent cell burden, including molecular signals in circulating immune cells and the abundance of circulating senescence‐related proteins, have been associated with chronological age and clinical parameters of biological age in humans. The extent to which senescence biomarkers are affected by interventions that enhance health and function has not yet been examined. Here, we report that a 12‐week structured exercise program drives significant improvements in several performance‐based and self‐reported measures of physical function in older adults. Impressively, the expression of key markers of the senescence program, including p16, p21, cGAS, and TNFα, were significantly lowered in CD3+ T cells in response to the intervention, as were the circulating concentrations of multiple senescence‐related proteins. Moreover, partial least squares discriminant analysis showed levels of senescence‐related proteins at baseline were predictive of changes in physical function in response to the exercise intervention. Our study provides first‐in‐human evidence that biomarkers of senescent cell burden are significantly lowered by a structured exercise program and predictive of the adaptive response to exercise.

The preclinical promise of senescent cell clearance as a translatable intervention for human health has highlighted the need for accessible and informative biomarkers of senescent cell burden. Two promising candidates, based on their cross-sectional associations with chronological age and clinical indices of biological age (e.g., disease burden and frailty) in humans, include the expression of the CDKI p16 in peripheral blood T cells and the circulating abundance of SASP proteins (Liu et al., 2009;Schafer et al., 2020). The responsiveness of these biomarkers of cellular senescence in older adults to intervention, however, has not yet been examined.
Exercise remains the most promising intervention to improve physical function in older adults, and prior observations in preclinical models and humans suggest it may influence senescent cell burden (Chen et al., 2021). Correspondingly, we studied the effects of a 12-week structured exercise intervention comprised of progressive strength and endurance training 2 days/weeks on the molecular phenotype of isolated CD3 + T cells and the circulating concentrations of SASP proteins in older adults ( Figure 1a). We hypothesized that structured exercise would drive significant reductions in biomarkers of cellular senescence.
Following the intervention, participants exhibited significant improvements in body composition, evidenced by significant reductions in body weight, BMI, waist circumference, and fat mass To assess the impact of exercise on biomarkers of cellular senescence, we first profiled peripheral blood CD3 + T cells by qPCR before and after the intervention. Impressively, exercise significantly and consistently reduced the expression of the CDKIs p16 and p21 as well as mediators and constituents of the SASP. Components of the cGAS-STING pathway (e.g., cGAS, IFNγ, and TNFα), which triggers inflammation and reinforces the senescence program (Li & Chen, 2018), were also reduced ( Figure 2a). Next, we examined plasma concentrations of prototypical SASP components, including cytokines, chemokines, matrix remodeling proteins, and growth factors, using a multiplex assay (Schafer et al., 2020). The intervention modestly but again consistently lowered the circulating abundance of multiple senescence-related proteins, including the inflammatory factor, myeloperoxidase (MPO), and thrombotic, fibrinolytic, and inflammatory factor, serpin E1 (PAI1) ( Figure 2b). Based on these findings and the diversity of our SASP panel, we utilized partial least squares discriminant analysis (PLSDA) to integrate values for baseline protein concentrations and calculate a multi-dimensional SASP index for each participant.
Notably, the SASP index differentiated participants who responded favorably to the intervention (responders) from those that did not (non-responders), based on the TUG measure, an integrative measure of physical function (Figure 2c). Variable importance projection (VIP) calculations, which rank individual SASP factors based on their overall importance for separating responder and non-responder data clouds, showed that SASP factors had a range of importance for distinguishing responders from non-responders (Figure 2d). Moreover, the SASP index was predictive of the degree to which participants' performance changed in response to the exercise intervention ( Figure 2e).
Our data demonstrate the responsiveness of the molecular phenotype of CD3 + T cells and components of the SASP to intervention, which provides a critical level of support for their use as biomarkers of senescence. This criterion adds to prior work that has established other key attributes of these candidates, including relevance to aging, measurement feasibility and reliability, and associations with clinical outcomes (Justice et al., 2018;Liu et al., 2009;Schafer et al., 2020).
Biomarkers offer significant utility to clinical trials as they can help determine indications and even individuals that may be most responsive to intervention. Further, our data suggest senescence transcripts in CD3 + T cells and a diverse panel of SASP proteins in plasma can potentially serve as surrogate endpoints that, in combination with predictive analytics like PLSDA, forecast clinical outcomes. This is of particular value given the time course for meaningful changes in clinical endpoints, such as physical, cardiovascular, and cognitive function measures, disease progression, and, of course, mortality (Biomarkers Definitions Working & G, 2001). As core properties of the senescence program, alterations in CDKI expression and SASP abundance may also reflect target engagement and the bioactivity of emerging interventions. Additional research is needed to understand the extent to which these circulating biomarkers reflect senescent cell burden in specific tissues and respond to senotherapeutic interventions in the context of age-related diseases and geriatric syndromes.
Our study highlights the impact of exercise on the fundamental biology of aging and, in turn, human health. Twelve weeks of structured exercise did not alter habitual physical activity, a finding that increases our confidence that the changes in senescence biomarkers can be ascribed to the exercise program itself, and not extrinsic alterations in participant behavior. We do note that including a control group in the study design would have further reinforced these findings. Additional research and advocacy for lifestyle interventions as a means to counter aging and extend human healthspan is warranted.

CO N FLI C T O F I NTE R E S T
The authors declare no conflicts of interest. ClinicalTrials.gov Identifier: NCT04897373.

AUTH O R CO NTR I B UTI O N S
NL provided the study design. MY involved in the study design. CF, BK, DL, and MS involved in subject recruitment and human data collection. DE, AS, AH, TW, EA and NL acquired, analyzed, or interpreted the data. DE and AS drafted the manuscript. All authors reviewed the manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.