Patterns of Circulating piRNAs in the Context of a Single Bout of Exercise: Potential Biomarkers of Exercise-Induced Adaptation?

Abstract

Background: Physical activity induces a range of physiological and molecular adaptations, particularly affecting skeletal muscle and the cardiovascular system, regulating both tissue architecture and metabolic pathways. Emerging evidence suggests that PIWI-interacting RNAs (piRNAs) may serve as potential biomarkers for these adaptations. Here, we analyzed piRNA patterns in the context of exercise.

Methods: This study selected eight participants of the iReAct study (DRKS00017446) for piRNA analysis. Baseline assessments included demographic profiling and fitness evaluation, particularly maximal oxygen uptake (V̇O2max) assessment. In addition, blood samples were collected pre- and (for six of the eight participants) post- standard reference training sessions. Subsequently, subjects underwent 6-week training protocols, employing standardized high-intensity interval training (HIIT) and moderate-intensity continuous training (MICT) regimens. Next, RNA sequencing was conducted to identify differentially expressed piRNAs, and correlation analyses were performed between piRNA expression patterns and training-associated changes in V̇O2max. Finally, to identify piRNAs potentially of interest in the context of exercise, different screening procedures were applied.

Results: There were unique and specific changes in individual piRNA expression levels in response to exercise. In addition, we could define correlations of piRNA expression patterns, namely of piR-32886, piR-33151, piR-12547, and piR-33074, with changes in V̇O2max. These correlations did not reach significance in the small sample size of this pilot study, but might be verified in larger, confirming studies.

Conclusions: This hypothesis-generating study identifies characteristic piRNA patterns in the context of exercise. Their significance as biomarkers is yet to be determined.

Keywords: endurance training; epigenetics; exercise biomarkers; piRNA; training adaptation.

Figure 1

Differentially expressed piRNAs filtered by five distinct criteria. The number of piRNAs identified by five selection criteria: ‘relative standard deviation’, ‘log2(fc)’, ‘zero expression’, ‘increase/decrease’, and ‘fold change’. The bar graph indicates the number of piRNAs meeting each individual criterion and the different criteria combinations (labelled in black). The ‘fold change’ criterion identified the highest number of piRNAs (212), with significant overlap observed with the ‘increase/decrease’ (128) and ‘zero expression’ (70) criteria. The ‘relative standard deviation’ criterion showed minimal overlap with others, identifying 24 piRNAs. Only three piRNAs satisfied four criteria (‘log2(fc)’, ‘fold change’, ‘zero expression’, and ‘increase/decrease’). An upset plot was used for visualizing these data, highlighting the degree of complexity and distinctiveness of each criterion.

Figure 2

Venn diagrams illustrating the relationships between the five selection criteria—‘log2(fc)’, ‘fold change’, ‘zero expression’, ‘increase/decrease’, and ‘relative standard deviation’. (A) Three piRNAs (hsa-piR-9137, hsa-piR-14091, hsa-piR-16886) satisfied the ‘log2(fc)’, ‘fold change’, ‘zero expression’, and ‘increase/decrease’ criteria. (B) Five piRNAs (hsa-piR-20111, hsa-piR-33055, hsa-piR-14091, hsa-piR-9137, hsa-piR-16886) met both the ‘log2(fc)’ and ‘zero expression’ criteria. (C) Minimal overlap was observed between the ‘fold change’ and ‘relative standard deviation’ criteria, with only two piRNAs (hsa-piR-25111, hsa-piR-32002) identified. (D) Five piRNAs (hsa-piR-32857, hsa-piR-32836, hsa-piR-32002, hsa-piR-11121, hsa-piR-33046) met both the ‘relative standard deviation’ and ‘increase/decrease’ criteria. Numbers indicate total piRNAs selected for each criterion, highlighting overlaps across the different requirements.

Figure 3

Kinetics of normalized expression of the n = 13 selected piRNAs. (‘at rest’: n = 8; ‘immediately’ and ‘+3 h’: n = 6).

Figure 4

Correlation of piRNA patterns with ΔV̇O2max between baseline and FU1. Diagrams illustrate correlations with ‘at rest’ expression (n = 8) and with changes between ‘at rest’ and ‘+3 h’ (n = 6) for which correlation coefficients (Spearman) of either >0.5 or <−0.5 were found.

Figure 5

Standardized reference training. The reference training on a bicycle ergometer consisted of a 10-min warm-up at 90% of the intensity corresponding to the first lactate threshold, followed by 50 min cycling at the intensity halfway between the first and the second lactate thresholds. Blood samples were collected at three time points: 3 h before training (‘at rest’), directly after training (‘immediately’), and 3 h after training (‘+3 h’).