Systematic physical exercise is an effective clinical option to improve quality of life in many types of patients, and the beneficial effects of exercise on cardiac function and exercise capacity have been well documented. It is well established that regular physical exercise promotes the improvement of life quality and lengthens lifespan (Moore et al., 2012; Arem et al., 2015). Furthermore, exercise has been reported to effectively reduce cardiovascular morbidity and mortality in patients with heart diseases (Gayda et al., 2016; Ribeiro et al., 2017). The augment of exercise capacity needs cumulative adaptation of the circulatory, pulmonary, and musculoskeletal systems, and exercise capacity has become an important indicator of cardiovascular health (Bassett and Howley, 2000; Wilson et al., 2016).

Exercise training could impact multiple signaling pathways, which influence energy metabolism, inflammation, regeneration and remodeling of myocardial and skeletal muscle (Egan and Zierath, 2013; Tao et al., 2015). However, whether exercise causes benefits depends on the intensity, training time, exercise type, environmental factors and the personal health status. Increases of circulating biomarkers of destruction parameters [Creatine kinase (CK), Lactic dehydrogenase (LDH)], cardiac markers [Creatine kinase-MB isoenzyme (CKMB), Troponin T and NH2-terminal prohormone of brain natriuretic peptide (NT-proBNP)], inflammatory marker [high-sensitivity C-reactive protein (Hs-CRP)] have been well documented (Clyne and Olshaker, 1999; Rifai et al., 1999; Bhalla et al., 2004; Hekimsoy and Oktem, 2005). However, these biomarkers provide limited insight into the exercise-induced adaptive processes. Thus, new biomarkers capable of monitoring cardiac changes and evaluating exercise physiology are critically needed.

MicroRNAs (miRNAs, miRs) are endogenous small noncoding RNAs and range from 18 to 25 nucleotides in length. miRNAs regulate gene expression at post-transcriptional levels through inhibition of translation or mRNA degradation in sequence-dependent manner (Guo et al., 2010). miRNAs have been widely detected in various human tissues, playing pivotal roles in a wide range of physiological and pathological processes. Besides tissues, miRNAs can also be detected in body fluids such as serum, plasma, urine, and saliva (Gautam et al., 2016). Circulating miRNAs are stable, easily detectable, and sensitive response to health changes, which endow the plasma and /or serum miRNAs with the potential use as biomarkers.

To date, increasing evidence has reported circulating miRNAs as potential therapeutic biomarkers for cardiovascular diseases, and numerous circulating miRNAs are found to be associated with exercise both in healthy human and patients. Numerous cardiac and/or skeletal muscle enriched miRNAs related with muscle damage, inflammation, cardiac adaptation, and angiogenesis have been shown altered after acute and/or sustained exercise (Baggish et al., 2011; Xu et al., 2016), indicating that circulating miRNAs can strongly monitor exercise influence on pathological and physiological processes of cardiovascular and muscles. Profiles of circulating miRNAs are varying under different exercise type, intensity and duration time. Dynamic regulation of circulating miRNAs response to acute and sustained exercise has been deciphered both in healthy participants and athletes (Baggish et al., 2011; Nielsen et al., 2014; Denham and Prestes, 2016). However, the intensity of long-term exercise training performed in previous studies is moderate, and regarding the energy metabolism system, nearly all the exercise type in previous studies is aerobic exercise (running, cycling, ski, and rowing etc.) (Baggish et al., 2011; Aoi et al., 2013; Tonevitsky et al., 2013; Nielsen et al., 2014; Uhlemann et al., 2014; Schmitz et al., 2017) or anaerobic exercise (sprint interval cycling; Cui et al., 2015). It is still unknown whether and how the mixed anaerobic-aerobic exercise with high intensity (basketball, football, volleyball, and handball etc.) influences circulating miRNAs and cardiac adaptation.

Here, we inspect how acute and long-term exercise training affects specific circulating miRNAs with well-known roles in adaptive process in basketball athletes. Specifically, we determined the change of circulating miRNA levels, general echocardiographic indexes, biochemical parameters and cardiorespiratory fitness at baseline and after a 3-month basketball season. Besides, the circulating miRNA levels and biochemical parameters are also assessed during acute exercise testing. We found that cardiorespiratory fitness and inflammation markers are significantly increased without the change of general echocardiographic indexes after 3-month basketball matches. Circulating miR-208b was decreased and miR-221 was increased after long-term exercise, while circulating miR-221, miR-21, miR-146a, and miR-210 were decreased response to acute exercise. Moreover, we correlated the circulating miRNA changes with biochemical parameters and cardiorespiratory fitness, showing the potential of circulating miRNA as biomarkers of initial stress induced by acute exercise and adaptation response to long-term basketball exercise.

The intent of this study is to determine how circulating miRNAs are regulated by an acute exhaustive exercise and long-term exercise training in basketball athletes. Although these athletes were trained regularly before enrolled in our study, their exercise capacities were still enhanced after 3-month exercise training. We reported for the first time, to our best knowledge, that 3 month period of basketball exercise changed miR-208b and miR-221 expression in male athletes. Given that miRNA changes were correlated with biochemical indexes and cardiorespiratory fitness, the present study also provides potential biomarkers for evaluating exercise effects.

To date, the circulating miRNA changes during exercise adaptation in basketball athletes have not been previously published. Competitive basketball is characterized by intermittent exercise with multiple and repeated episodes of high intensity and short duration exertion. The energy demands of basketball were contributed from both aerobic and anaerobic energy systems. These characteristics of basketball make the signature of circulating miRNAs be different from other aerobic and anaerobic exercise types. Our study, for the first time, showed that miR-146a, miR-21, miR-221, and miR-210 were decreased in response to acute exercise in basketball athletes. In addition, our study demonstrated that miR-208b was decreased while miR-221 was increased after the 3-month basketball match in 10 basketball athletes. And other miRNAs (miR-1, miR-21, miR-222, miR-486, miR-146a, and miR-126, etc.) have no changes after 3-month basketball match, which have been shown dynamically altered after different programs of long-term aerobic exercise training (Baggish et al., 2011; Aoi et al., 2013; Nielsen et al., 2014; Denham and Prestes, 2016; Schmitz et al., 2017). Considering most of the previous studies focused on aerobic or anaerobic exercise, our results from mixed aerobic-anaerobic exercise provided novel insights into exercise physiology.

miR-21, miR-146a, and miR-210 regulate multiple functions relevant to cardiovascular diseases and exercise. It can be speculated that acute exercise might cause local short-term hypoxic conditions in skeletal muscle and cardiovascular systems. miR-21 plays anti-apoptosis roles in endothelial cells (Weber et al., 2010) and cardiomyocytes under ischemia (Dong et al., 2009). miR-210 has been found induced by hypoxia in a variety of ischemic conditions and tissues (Bostjancic et al., 2009; Biswas et al., 2010; Hu et al., 2010). By targeting several transcripts, miR-210 regulates multiple aspects of cellular response to hypoxia such as inhibiting apoptosis (Fasanaro et al., 2008; Zaccagnini et al., 2014), promoting the shift from mitochondria respiration to glycolysis (Chan et al., 2009) and inducing angiogenesis (Fasanaro et al., 2008; Alaiti et al., 2012). Thus, the decrease of miR-21 and miR-210 in present study may reflect the initial cellular response to hypoxia after acute exercise, such as increased cell apoptosis and enhanced mitochondrial metabolism. In addition, inflammatory response is also another process accompanied with exercise, especially acute exercise with high intensity. Both miR-21 (Sheedy, 2015), miR-221 (Corsten et al., 2015) and miR-146a (Yang et al., 2012; Cheng et al., 2013) have been identified as vital brake in inflammatory response. It can be speculated that the decrease of miR-21, miR-221 and miR-146a in present study may reflect the initial pro-inflammatory process that occur after acute exercise. Among them, time course of plasma miR-21 level has showed that miR-21 was downregulated immediately after acute exercise while increased during recovery in healthy young men (Cui et al., 2017). Thus, in the future, time course study of circulating miR-21, miR-146a and miR-221 levels will demonstrate the potential roles of circulating miRNAs in real-timely reflecting the inflammatory responses after acute exercise in basketball athletes or other similar types of athletes.

Long-term exercise training causes various adaptations including angiogenesis, mitochondrial metabolism, skeletal muscle regeneration and myocardial remodeling, which will directly interfere with the exercise capacity and health outcomes. miR-208b is expressed in skeletal muscle and embryonic hearts, and its ectopic expression caused multiple cardiovascular diseases (Huang and Li, 2015). Aerobic exercise reduces miR-208b expression and modulates the expression of miR-208b downstream genes such as cardiac myosin heavy chain (MHC) isoforms, PPARβ, and histone deacetylase (HDAC) class I and II, which are associated with cardiac metabolic and contractile adaptation (Soci et al., 2016). The decrease of miR-208b may contribute to the maintenance of physiological cardiac hypertrophy and improvement of cardiac function after exercise training (Soci et al., 2016). miR-221 shows vital roles in vasculature network by inhibiting the proliferation of endothelial cells (Poliseno et al., 2006) and supporting the vascular smooth muscle cell dedifferentiation, proliferation and migration (Liu et al., 2009; Zampetaki and Mayr, 2012). miR-221 has also been shown promoting skeletal muscle regeneration through modulating muscle cell differentiation and maturation (Cardinali et al., 2009; Togliatto et al., 2013). Angiogenesis and muscle regeneration are all key aspects of exercise adaptation. Therefore, miR-221 elevation after long-term exercise training might represent the enhanced muscle regeneration and controlled angiogenesis. Considering miR-221 was the only circulating miRNA in response to both acute and long-term exercise training, findings from this study set the stage for identifying miR-221 as a promising biomarker capable of evaluating the magnitude of stress and adaptation induced by exercise.

In our study, after 3-month exercise training, echocardiography showed that EF% value and other cardiac indexes were not changed, whereas their exercise capacities are enhanced. These results indicate that echocardiography and biochemical measurements are unable to comprehensively evaluate the benefits of long-term exercise training. The biochemical measurements showed that the contributions of both acute and 3-month basketball match to tissue damage does not appear to be significant in the present study. To further determine the uniqueness of circulating miRNA changes, we next compared circulating miRNA profiles to those of conventional biomarkers such as EF%, AT VO₂, peak VO₂ and biochemical indexes. With the improvement of exercise capacities, we showed the changes of miR-208b after 3-month exercise training moderately correlated with AT VO₂ at baseline and the changes of miR-221 correlated with AT VO₂, peak work load and serum CK levels after 3-month exercise training. Thus, these results indicated the possibility of circulating miR-221 and miR-208b as potential biomarkers for assessing or revealing the muscle damage and adaptations of exercise capacity induced by long-term basketball exercise.

As an initial study, several limitations of this study were still existent. Firstly, considering the variations of exercise intensity and duration time of every subject during the basketball game, the number of enrolled athletes should be expanded. Secondly, in view of the difference of exercise abilities between male and female, it would be interesting to see the miRNA changes in response to long-term exercise training using female athletes as subjects. Thirdly, the current study was restricted to basketball athletes. Future study is required to determine whether the changes of circulating miRNAs are applicable to football or other similar exercise types. Fourthly, our exercise training program involved team-based basketball match. Thus, the volume or intensity of the exercise training could not be controlled and calculated. However, quantitative assessment was taken to document playing time and fitness time of every athlete. Finally, quantitative analysis was restricted to a subset of relevant miRNAs, and high-throughput screening is needed to obtain a more complete profile of circulating miRNA regulation during long-term basketball exercise training.

In conclusion, serum circulating miRNA levels are dynamically regulated in response to acute cycling exercise and long-term basketball exercise training in basketball athletes. Future studies are highly needed to define these miRNAs as potential and useful biomarkers of long-term exercise training, and disclose their direct biological roles in exercise mode-specific training adaptations.

YL: undertook the qRT-PCR experiments and contributed to the drawing up of figures and writing of the paper; MY: contributed to the drawing up of figures and ELISA experiments; QZ: undertook the qRT-PCR experiments and cardiopulmonary exercise test; YC, LC, and JiX: provided technical support in echocardiographic diagnosis and serum sample collections; JuX: was involved in discussion on experimental design and contributed to the writing of the paper; ZS and YB: designed the experiments and contributed to the writing of the paper.