An expanded repertoire of intensity-dependent exercise-responsive plasma proteins tied to loci of human disease risk

Abstract

Routine endurance exercise confers numerous health benefits, and high intensity exercise may accelerate and magnify many of these benefits. To date, explanatory molecular mechanisms and the influence of exercise intensity remain poorly understood. Circulating factors are hypothesized to transduce some of the systemic effects of exercise. We sought to examine the role of exercise and exercise intensity on the human plasma proteome. We employed an aptamer-based method to examine 1,305 plasma proteins in 12 participants before and after exercise at two physiologically defined intensities (moderate and high) to determine the proteomic response. We demonstrate that the human plasma proteome is responsive to acute exercise in an intensity-dependent manner with enrichment analysis suggesting functional biological differences between the moderate and high intensity doses. Through integration of available genetic data, we estimate the effects of acute exercise on exercise-associated traits and find proteomic responses that may contribute to observed clinical effects on coronary artery disease and blood pressure regulation. In sum, we provide supportive evidence that moderate and high intensity exercise elicit different signaling responses, that exercise may act in part non-cell autonomously through circulating plasma proteins, and that plasma protein dynamics can simulate some the beneficial and adverse effects of acute exercise.

Figure 1

Discovery of exercise regulated plasma proteins at moderate and high intensity exercise. (a) Study design. Participants all underwent CPET with determination of individual peak VO₂. Participants were then randomized to two treadmill sessions consisting of a moderate intensity (5 mile/h steady state) or high intensity (maximal effort) exercise session. Participants who underwent a moderate intensity session first later underwent a high intensity session and those who underwent an initial high intensity session later completed a moderate intensity session. Blood was drawn before and immediately after each session. (b) Breath-by-breath cardiopulmonary exercise test data from a representative participant is shown. Moderate vs. high exercise intensity is defined physiologically by an inflection point observed at the ventilatory anaerobic threshold (vertical line at 182 bpm) and distinguishes moderate from high intensity exercise. (c,d) Post-exercise Cortisol kinetics at (c) moderate (p < 0.001) and (d) high intensity (p = 0.013) exercise confirm exercise intensity. Volcano plots show proteins that rise (red) and fall (green) with (e) moderate and (f) high intensity exercise, highlighting greater complexity of the dynamic proteome with high intensity exercise. 1,305 proteins examined. n = 12 participants. p < 0.05 was considered significant and values were adjusted for multiple hypothesis testing (Benjamini–Hochberg). A paired t-test was performed to examine post-exercise cortisol kinetics in panels c,d.

Figure 2

Differential intensity dependent and independent plasma protein responses to moderate and high intensity acute exercise. (a,b) The 25 proteins with the greatest positive and negative fold change at (a) moderate and (b) high intensity exercise are shown. (c) Proteins common to both moderate and high intensity exercise (n = 159) are plotted with high intensity fold change (y-axis) against moderate intensity fold change (x-axis). Relative intensity-dependence (darker blue) and intensity-independence (lighter blue) of protein species are depicted.

Figure 3

Transcriptional inference reveals multisystem tissue contributions of proteins from the exercise plasma proteome. (a) Among proteins increased in plasma at moderate (n = 120) and high (n = 250) intensity exercise, transcriptional inference suggests systemic contribution of donor protein species into the plasma. Dominant inferred sources of protein diversity include the nervous, cardiovascular, and gastrointestinal systems at both exercise intensities. (b) Inferred tissue sources of proteins increasing in plasma with exercise are compared against tissue expression of entire SomaLogic platform, revealing relative enrichment during exercise for proteins with expression in blood, cardiovascular, skeletal muscle, and nervous tissue.

Figure 4

Exercise regulates plasma proteins tied to human traits. (a) Genomic locations of pQTLs (red, cis; blue, trans). X and Y axes represent chromosomal locations of the pQTL and the associated protein, respectively. (b) Quantification of exercise-responsive plasma proteins (FDR p < 0.05) with associated pQTLs, and associated phenotypic traits (p < 5 × 10⁻⁸) permit raw effect estimation. (c,d) The pQTL-linked proteins with corresponding risk pQTLs associated with coronary artery disease (c) and blood pressure (d) are plotted on forest plots and aligned in a manner whereby higher plasma protein concentration associates with either higher or lower disease-specific risk (x-axes). Higher plasma protein concentrations move laterally away from the midline. A red arrow depicts the direction of the simulated impact of exercise-associated acute changes in protein concentration. Forrest plots depict the GWAS β point estimate and standard error.