Evaluation of microRNA expression in plasma and skeletal muscle of thoroughbred racehorses in training

BACKGROUND Circulating miRNAs (ci-miRNAs) are endogenous, non-coding RNAs emerging as potential diagnostic biomarkers. Equine miRNAs have been previously identified including subsets of tissue-specific miRNAs. In order to investigate ci-miRNAs as diagnostic tools, normal patterns of expression for different scenarios including responses to exercise need to be identified. Human studies have demonstrated that many ci-miRNAs are up-regulated following exercise with changes in expression patterns in skeletal muscle. However, technical challenges such as haemolysis impact on accurate plasma ci-miRNA quantification, with haemolysis often occurring naturally in horses following moderate-to-intense exercise. The objectives of this study were to identify plasma ci-miRNA profiles and skeletal muscle miRNAs before and after exercise in Thoroughbreds (Tb), and to evaluate for the presence and effect of haemolysis on plasma ci-miRNA determination. Resting and post-exercise plasma ci-miRNA profiles and haemolysis were evaluated in twenty 3 year-old Tbs in sprint training. Resting and post-exercise skeletal muscle miRNA abundance was evaluated in a second cohort of eleven 2 year-old Tbs just entering sprint training. Haemolysis was further quantified in resting blood samples from twelve Tbs in sprint training. A human plasma panel containing 179 miRNAs was used for profiling, with haemolysis assessed spectrophotometrically. Data was analysed using a paired Student's t-test and Pearson's rank correlation. RESULTS Plasma ci-miRNA data for 13/20 horses and all skeletal muscle miRNA data passed quality control. From plasma, 52/179 miRNAs were detected at both time-points. Haemolysis levels were greater than the threshold for accurate quantification of ci-miRNAs in 18/25 resting and all post-exercise plasma samples. Positive correlations (P < 0.05) between haemolysis and miRNA abundance were detected for all but 4 miRNAs, so exercise-induced changes in plasma ci-miRNA expression could not be quantified. In skeletal muscle samples, 97/179 miRNAs were detected with 5 miRNAs (miR-21-5p, let-7d-3p, let-7d-5p, miR-30b-5p, miR-30e-5p) differentially expressed (DE, P < 0.05) between time-points. CONCLUSIONS The degree of haemolysis needs to be determined prior to quantifying plasma ci-miRNA expression from horses in high-intensity exercise training. Identification of DE miRNAs in skeletal muscle indicates modification of miRNA expression may contribute to adaptive training responses in Tbs. Using a human plasma panel likely limited detection of equine-specific miRNAs.