Effect of exercise training on the FNDC5/BDNF pathway in spontaneously hypertensive rats

Abstract Increased sympathetic activity contributes to the development of cardiovascular diseases such as hypertension. Exercise training lowers sympathetic activity and is beneficial for the prevention and treatment of hypertension and associated cognitive impairment. Increased BDNF expression in skeletal muscle, heart, and brain may contribute to these actions of exercise, but the mechanisms by which this occurs are unknown. We postulated that hypertension is associated with decreased hippocampal BDNF, which can be restored by exercise‐mediated upregulation of fibronectin type‐II domain‐containing 5 (FNDC5). Spontaneously hypertensive rats (SHR) and normotensive Wistar–Kyoto rats (WKY) were subjected to 5 weeks of motorized treadmill training. BDNF and FNDC5 expressions were measured in the left ventricle (LV), quadriceps, soleus muscle, and brain areas. Exercise training reduced blood pressure (BP) in both strains. BDNF and FNDC5 protein in the LV were increased in SHR, but exercise increased only BDNF protein in both strains. BDNF mRNA, but not protein, was increased in the quadriceps of SHR, and BDNF mRNA and protein were decreased by exercise in both groups. FNDC5 protein was higher in SHR in both the quadriceps and soleus muscle, whereas exercise increased FNDC5 protein only in the quadriceps in both strains. BDNF mRNA was lower in the dentate gyrus (DG) of SHR, which was normalized by exercise. BDNF mRNA expression in the DG negatively correlated with BP. No differences in FNDC5 expression were observed in the brain, suggesting that enhanced BDNF signaling may contribute to the cardiovascular and neurological benefits of exercise training, and these processes involve peripheral, but not central, FNDC5.


| INTRODUCTION
Regular moderate-to high-intensity exercise is beneficial for the prevention and treatment of cardiovascular disease (CVD) and CVD risk factors, such as hypertension (Lavie et al., 2015;Lear et al., 2017). The anti-hypertensive benefits of exercise are governed by both peripheral and central mechanisms of blood pressure (BP) control. However, the mechanisms by which exercise induces these changes remain unclear.
One potential mechanism by which exercise may increase BDNF is through myokines such as irisin, which are secreted by skeletal muscle into the circulation following exercise and then signal to the brain (Delezie & Handschin, 2018;Fiuza-Luces et al., 2018). Irisin-the cleaved protein product of fibronectin type-II domain-containing 5 (FNDC5) expressed abundantly in heart, skeletal muscle, and brain-has been associated with beneficial effects on the hypertensive vasculature (Fu et al., 2016;Ling et al., 2018), cardiac remodeling and fibrosis (Chen et al., 2019;Liao et al., 2019;Yu et al., 2019), and metabolism (Bostrom et al., 2012;Lee et al., 2014). FNDC5 is expressed downstream of the master skeletal muscle regulator peroxisome proliferator-activated receptor coactivator 1-alpha (PGC1α), and following exercise training, the activation of the PGC1α/FNDC5 pathway leads to increased BDNF expression in the hippocampus (Wrann et al., 2013). Whether FNDC5 might be involved in the beneficial effects of exercise in hypertension remains to be determined.
Thus, we postulated that hypertension is associated with downregulation of BDNF-TrkB signaling in the hippocampus, which can be rescued by exercise training through the upregulation of FNDC5 expression in skeletal muscle and brain. In this study, we subjected SHR and normotensive control WKY to a moderate-intensity exercise treadmill program for 5 weeks. The heart (LV), skeletal muscle (quadriceps and soleus muscles), and brain were collected for the measurement of BDNF, TrkB, and FNDC5 mRNA and protein expression. Our findings demonstrate that endurance exercise training increases BDNF-TrkB-signaling in the heart and the dentate gyrus (DG) of the hippocampus and increases FNDC5 expression in the quadriceps but not the heart and brain. room maintained at constant temperature and humidity. Rats were housed in pairs under a 12:12-hr light-dark cycle and allowed standard laboratory chow and tap water ad libitum.

| Exercise training
Rats were randomly divided into four groups: sedentary WKY (WKY-Sed), exercise-trained WKY (WKY-ExT), sedentary SHR (SHR-Sed), or exercise-trained SHR (SHR-ExT). A treadmill exercise program was conducted for 5 weeks, 5 days/week in the afternoon as previously described (Lee et al., 2017;Zheng, Sharma, Liu, & Patel, 2012). Rats were acclimatized to running on a motor-driven treadmill (Columbus Instruments) at low speed (10 m/min), grade 0%, and short duration (10-15 min/day) for the first 3 days. For the remainder of week 1 and week 2, the speed, grade, and duration were gradually increased to 15-20 m/ min, 5%-10% incline, and 60 min/day, respectively, which is considered moderate intensity for rats (Zheng et al., 2012). These exercise parameters were maintained at the maximal level for weeks 3, 4, and 5. In every session from week 2 onward, a 5-min warm-up was conducted at low speed (10 m/ min) and 0% inclines which was not included in the total duration of the exercise training session. Two WKY rats refused to run and were removed from the exercise group. Sedentary groups were handled daily under the same conditions as the exercise-trained groups and exposed to the immobile treadmill for comparable times. Body weight was measured at the beginning of each week, as well as after the final exercise session of week 5 prior to hemodynamic measurements.

| Blood pressure measurements
Rats were placed into individual cages 24 hr prior to cannulation. In the morning, under anesthesia with 2.5% isoflurane in oxygen, the right femoral artery was cannulated using PE50-10 tubing, filled with heparinized saline. Cannulae were secured by a stopper pin and exteriorized to the back. Rats were then individually housed in a quiet room near the data acquisition system. In the afternoon, intra-arterial cannulae were connected to pressure transducers. After a rest period of 30 min, BP and heart rate (HR) were recorded for a period of 10 min using AcqKnowledge 3.8 software (Biopac Systems Inc.).

| Tissue collection
After sacrifice by decapitation, the brain was removed and immediately frozen in chilled methylbutane (Sigma-Aldrich) and then stored at -80°C. The heart was immediately rinsed in ice-cold 0.9% saline. Working rapidly on ice, the right ventricle (RV) was separated from the left ventricle (LV) then blotted dry, weighed, and frozen in liquid nitrogen. Soleus muscle and quadriceps from the noncannulated left hind leg were collected, blotted dry, weighed, and frozen in liquid nitrogen. Trunk blood was collected into prechilled heparinized falcon tubes and plasma was isolated by centrifugation. Top-most layer of plasma was centrifuged again and collected for BDNF analysis. The remaining plasma was aliquoted for measurement of pro-and anti-inflammatory cytokines. Plasma samples were immediately stored at −80°C.

| Real-time qPCR
Samples were homogenized in 1 ml (per 100 mg tissue) QIAzol Lysis Reagent (Qiagen Inc.) using a prechilled Polytron, and RNA was isolated as per manufacturer's instructions. To eliminate potential genomic DNA contamination, a quantity of 20 µg total RNA was treated with DNase I (Invitrogen). A quantity of 5 µg of DNase I-treated RNA was used for cDNA synthesis by incubation with 200 U/µl RevertAid H Minus reverse transcriptase and Oligo-dT as primer (Thermo Scientific Inc.) at 42°C for 60 min. For brain RNA, serial 50-µm coronal-oriented brain cryosections were obtained, and brain nuclei were collected as described . Briefly, brain tissue was collected from −0.80 to −3.14 mm and −11.80 to −13.80 mm posterior to bregma for the forebrain and hindbrain, respectively, based on the rat brain atlas of Paxinos & Watson (1998). The PVN, CA1-CA3, and DG regions of the hippocampus, and RVLM were extracted via micro-punch and homogenized immediately in ice-cold lysis buffer (ReliaPrep RNA tissue mini-prep system, Promega). Total RNA was isolated according to the manufacturer's instruction. A quantity of 300-500 ng of DNase I-treated RNA was used for cDNA synthesis by Verso cDNA kit (Thermo Scientific Inc.).
Primers For FNDC5, a 316-bp PCR fragment corresponding to position 208-523, BDNF, a 165-bp fragment at 1212-1372 position, and TrkB, a 197-bp fragment at 1502-1698, were amplified and then subcloned into pCRII-TA vector (Invitrogen) followed by restriction endonuclease analysis. The concentrations of the plasmids were determined via NanoDrop spectrophotometry at 260 nm. An external standard curve for each plasmid was created using serial 10-fold dilutions (eg. 100 pg/µl to 0.001 pg/µl) of plasmid under the same real-time qPCR conditions described above.
Real-time qPCR was performed with a Roche LightCycler LC480 using LC480 SYBR Green I (Roche Diagnostics). PCR conditions were set as follows: initial at 95°C for 10 s followed by 45 cycles of denaturation at 95°C for 10 s, annealing for 15 s at 57°C for FNDC5 and BDNF, 59°C for TrKB, and 62°C for AT 1 R, and MR, followed by extension at 72°C. Specificity of the real-time qPCR products was determined by both melting curve analysis and agarose gel electrophoresis. mRNA abundance of various genes was normalized against phosphoglycerate kinase 1 (PGK1) levels as the endogenous reference. The external standard of PGK1 was the same as previously described (Wang et al., 2010).

| Biochemical plasma assays
Total BDNF (pro-BDNF plus mature BDNF) in the plasma was measured with a commercially available ELISA kit (ChemiKine, Cat#CYT306, EMD Millipore) according to manufacturer's instructions. Plasma samples were diluted 1:1 with a sample buffer provided by the kit and run in duplicate. Total BDNF levels are expressed as pg/ml of plasma.
To measure plasma pro-and anti-inflammatory cytokine levels, a commercially available rat cytokine 12-plex Bio-Plex assay kit (Bio-Rad) was used according to manufacturer's instructions. Plasma samples were diluted 1:4 with a sample buffer provided by the kit and run in singlicate. Cytokine levels are expressed as pg/ml of plasma.

| Statistical analysis
All values are expressed as mean ± SD. Two-way ANOVA was performed to determine the effects of both exercise and hypertension on the various parameters using IBM SPSS statistics 25 software. When significant differences were found between groups, a student's Newmans-Keuls test was run for post hoc comparative analysis. Correlations analyses were done by Pearson's correlation analysis. Statistical significance was defined as p < .05.

| Effect of exercise on BP, HR, and cardiac weights
As expected, SHR had significantly higher SBP, DBP, and HR when compared to WKY rats (Table 1). Following exercise, SBP was 7 mmHg lower in WKY-ExT versus WKY-Sed and 28 mmHg lower in SHR-ExT versus SHR-Sed. DBP was 10 mmHg lower in WKY-ExT versus WKY-Sed and 7 mmHg lower in SHR-ExT versus SHR-Sed. HR was unaffected by exercise (Table 1).
SHR had significantly higher body weight compared to WKY, which was notable in SHR-Sed versus WKY-Sed (Table 1). LV weight was larger in SHR when compared to WKY, while RV weight was similar. Exercise had no significant effect on LV and RV weights (Table 1). Thus, we demonstrate that SBP and DBP in both hypertensive and normotensive rats were significantly decreased following 5 weeks of endurance exercise training.

| Effect of hypertension and exercise on plasma BDNF and cytokines
Plasma BDNF tended (p = .09) to be lower in SHR than WKY and was not affected by exercise in either strain ( Table  2). The anti-inflammatory cytokine IL-13 was significantly elevated in SHR versus WKY and tended (p = .09) to be higher following exercise in SHR-ExT versus SHR-Sed. The anti-inflammatory cytokine IL-4 was similar in SHR versus WKY, but tended (p = .08) to be higher following exercise. The proinflammatory cytokine IL-1α was similar between SHR versus WKY, but tended (p = .07) to be lower following exercise in WKY-ExT versus WKY-Sed. No other cytokines were affected by strain or exercise (Table 2).

| Effects of hypertension and exercise on BDNF, TrkB, and FNDC5 mRNA and protein expression in the heart and skeletal muscle
In the LV, BDNF mRNA expression tended (p = .06) to be lower in SHR versus WKY, but protein expression was higher in SHR versus WKY and significantly elevated by exercise training in both SHR and WKY. TrkB mRNA expression tended (p = .09) to be lower in SHR versus WKY, while protein expression was similar. FNDC5 mRNA expression was similar, but protein expression was significantly higher in SHR compared to WKY (Figure 1). TrKB and FNDC5 in the LV were not changed by exercise.
In the quadriceps, BDNF mRNA was significantly higher in SHR versus WKY, but BDNF protein was similar. BDNF protein expression was moderately reduced in both strains following exercise training. TrkB mRNA was similar in sedentary SHR and WKY and was elevated in SHR-ExT versus SHR-Sed group, while protein expression was similar among groups. FNDC5 mRNA was moderately lower while protein expression was significantly higher in SHR versus WKY. FNDC5 protein, but not mRNA, was elevated by exercise training in WKY (Figure 2a).
In the soleus muscle, BDNF expression was similar between strains, and exercise tended (p = .097) to increase BDNF mRNA but not protein. TrkB mRNA tended (p = .077) to be lower in SHR, whereas protein expression was elevated by exercise training in both strains. FNDC5 mRNA tended (p = .097) to be lower in SHR, but protein expression was higher in SHR (Figure 2b). Neither was affected by exercise.

FNDC5, AT 1 R, and MR mRNA expression in the brain of SHR versus WKY
We next examined the effects of exercise on BDNF, TrkB, and FNDC5 mRNA expression in the PVN and RVLM of SHR versus WKY. To evaluate the potential effects on angiotensinergic signaling, we also evaluated the expression of AT 1 R and MR (Table 3). In the PVN, TrkB expression was elevated in SHR versus WKY, but BDNF, FNDC5, AT 1 R, and MR expressions were similar in the two strains. Exercise had no effect on their expressions. In the RVLM, only MR was significantly lower in SHR versus WKY, and tended (p = .053) to be lowered by exercise in both strains. BDNF, TrkB, FNDC5, and AT 1 R expression were similar in the two strains with no effect of exercise.
Given that hypertension is associated with neurocognitive deficits, we examined the expression of BDNF, TrkB, FNDC5, AT 1 R, and MR in the CA1-3 and DG regions of the hippocampus. In the CA regions, TrkB expression was higher in SHR versus WKY, while BDNF, FNDC5, AT 1 R, and MR expressions were similar. Exercise did not affect BDNF, TrkB, FNDC5, AT 1 R, or MR expression. In the DG, BDNF was significantly lower in SHR-Sed compared to WKY. BDNF expression in SHR-ExT was significantly higher than SHR-Sed, and 5 weeks of exercise training normalized BDNF deficits in SHR. Of note, relative BDNF expression in the DG negatively correlated with SBP in SHR (r = −.87, p < .01; Figure  3b), but not in WKY. Relative skeletal muscle FNDC5 protein tended to correlate with relative BDNF expression in the DG in WKY (quadriceps: r = −.74, p = .055; soleus muscle: r = .68, p = .096; Figure 4), but not SHR. TrkB expression was higher in SHR versus WKY, and unaffected by exercise. FNDC5 was similar between groups. AT 1 R was lower in SHR versus WKY, and not affected by exercise. MR expression was lower in SHR versus WKY, whereas exercise lowered MR in WKY but increased MR in SHR.

| DISCUSSION
This study explored the effect of exercise on BDNF and FNDC5 in SHR versus WKY rats. Five weeks of moderate-intensity treadmill exercise training-as expected-reduced SBP and DBP in both SHR and WKY. In the LV, BDNF protein expression was elevated in SHR and increased by exercise in both strains, while FNDC5 protein was also higher in SHR but unaffected by exercise. In skeletal muscle, FNDC5 protein was higher in SHR in both the quadriceps and soleus muscle, and exercise increased FNDC5 protein in the quadriceps. In the brain, BDNF expression in the PVN and RVLM were unaffected, but the expression of its receptor TrkB was significantly higher in SHR. Furthermore, hypertension was associated with significantly reduced BDNF mRNA in the DG of the hippocampus but not the CA regions, PVN, or RVLM. The reduced BDNF in SHR was normalized by exercise training. FNDC5 expression in these brain regions was similar between strains and unaffected by exercise training.
Our finding that 5 weeks of moderate intensity treadmill exercise reduced BP in both hypertensive and normotensive rats is consistent with other studies of similar type, intensity, and duration (Hsu et al., 2011;Lee et al., 2017Lee et al., , 2018Petriz et al., 2015). A shorter (7 days) exercise training period did not elicit these changes (Monnier et al., 2017).
BDNF is highly expressed in the heart where it may serve cardioprotective functions, through, for example, modulating cardiomyocyte calcium cycling and promoting coronary angiogenesis (Kermani & Hempstead, 2019). Exercise promotes improved remodeling and function of the heart, in part by promoting the expression of BDNF (Lee et al., 2017(Lee et al., , 2018. Indeed, BDNF protein (but not mRNA) expression was elevated in the LV of SHR, and following exercise, was increased in both hypertensive and normotensive strains. Amoureux et al. (2012) reported that aortic BDNF expression is elevated in SHR beginning at the development of hypertension (6 weeks of age) and continuing into adulthood (13 weeks of age). In contrast, Prigent-Tessier et al. (2013) reported decreased mBDNF protein in the coronary endothelium and whole hearts of 10-11-week-old SHR versus WKY by immunohistochemistry and western blot respectively, and this was attenuated by 1 week of treadmill exercise training. We also observed elevated FNDC5 protein (but not mRNA) in the LV of SHR. While FNDC5 expression has not yet been examined in the hypertensive heart, Ling et al. (2018) reported decreased FNDC5 mRNA and protein in the aortas of 4 and 12-week-old SHR versus WKY, which was evident throughout the medial and adventitial layers, and FNDC5 over-expression attenuated medial thickening in SHR. Yu et al. (2019) demonstrated that transverse aortic constriction-induced hypertrophy upregulated FNDC5 protein in mice, which is similar to our finding. In our previous study, we reported decreases in FNDC5 mRNA and increases in FNDC5 protein in rats post-MI (Lee et al., 2017). All together, these findings suggest that changes in FNDC5/BDNF signaling may play a role in cardiovascular pathologies such as hypertension or MI. Exercise did not affect FNDC5 expression in either group. Previously we reported that 5 weeks of exercise training also did not affect FNDC5 expression in the LV of rats with heart failure post-MI (Lee et al., 2017). Collectively, our findings suggest that exercise-mediated benefits occur independent of The effect of hypertension on skeletal muscle BDNF expression has been largely unexplored. In this study, BDNF mRNA was elevated in the quadriceps of SHR with or T A B L E 3 mRNA expression of BDNF, TrkB, FNDC5, AT 1 R, and MR in brain areas of SHR and WKY with or without exercise training WKY SHR Sed (n = 10) ExT (n = 7) Sed (n = 6) ExT (n = 9) BDNF PVN mRNA/PGK1 (×10 -2 ) 4.3 ± 0.5 3.7 ± 0.3 4.0 ± 0.4 4.2 ± 0.3 RVLM mRNA/PGK1 (×10 -2 ) 1.6 ± 0.2 1.7 ± 0.4 2.4 ± 0.5 2.0 ± 1.2 CA mRNA/PGK1(×10 -2 ) 11.1 ± 1 11.3 ± 0.8 9.8 ± 2 10.3 ± 2 DG mRNA/PGK1(×10 -2 ) 11.5 ± 1.0 10.2 ± 0.8 7.8 ± 1.3* 10.1 ± 2.4*#
In this study, we evaluated the effect of hypertension and exercise on two key functional regions in the brain: (a) cardiovascular regulatory centres involved in central control of BP, and (b) neurotrophin-mediated cognition in the hippocampus. Hypertension and exercise had no effect on BDNF expression in the PVN or RVLM, but its receptor TrkB was upregulated in the PVN of SHR. Previous studies have demonstrated that BDNF mRNA in the PVN increases in response to salt-induced osmotic stress (Aliaga, Arancibia, Givalois, & Tapia-Arancibia, 2002) or immobilization-induced psychological stress (Smith, Makino, Kim, & Kvetnansky, 1995). Acute and chronic BDNF elevation in the PVN increases BP in rats (Erdos et al., 2015;Schaich, Wellman, Koi, & Erdos, 2016). Complementing our finding that TrkB was elevated in SHR, Schaich et al. (2018) inhibited BDNF-TrkB signaling by overexpressing a truncated-form of TrkB (TrkB.T1) and observed a decrease in BP in rats, suggesting that TrkB contributes to BDNF-TrkB signaling to raise BP. Assessing downstream angiotensinergic signaling in the PVN and RVLM, AT 1 R expression was unaffected in SHR or by exercise, while MR was decreased in the RVLM F I G U R E 3 Correlations between blood pressure and BDNF gene expression in the brain. Correlations between systolic blood pressure versus relative BDNF mRNA expression in the (A) paraventricular nucleus (PVN), and (B) the dentate gyrus (DG) of the hippocampus. Gene expression is normalized to the housekeeping gene PGK1 F I G U R E 4 Correlations between skeletal muscle FNDC5 protein expression and hippocampal BDNF gene expression. Correlations between relative quadriceps or soleus muscle FNDC5 protein expression versus relative BDNF mRNA expression in the dentate gyrus (DG) of the hippocampus. Protein expression is normalized to housekeeping protein GAPDH and expressed as fold change of WKY-Sed. Gene expression is normalized to the housekeeping gene PGK1 of SHR and unaffected by exercise. In contrast, Pietranera et al. (2012) reported higher MR mRNA expression in the PVN and hippocampus of SHR. Nakagaki et al. (2012) reported no difference in MR protein expression in the stroke-prone SHR strain. The hippocampus is a critical site of neurogenesis and cognitive function, which may be disrupted in association with hypertension. Shih et al. (2016) reported decreased hippocampal BDNF expression, worse long-term memory, and fewer BrdU-incorporated DG cells suggesting less neuronal proliferation in mice subjected to a renal artery clip-induced hypertension. This present study also observed decreased BDNF in the DG region. BDNF levels were restored by exercise training, which is consistent with previous work showing increased BDNF in the DG following exercise (Fang et al., 2013;Liu & Nusslock, 2018;Monnier et al., 2017). In contrast, TrkB was elevated in both the CA1-3 and DG regions of SHR. FNDC5 expression in the PVN, RVLM, CA1-3, or DG was unaffected in SHR or by exercise, suggesting that peripheral, not central, FNDC5 expression may contribute to FNDC5-mediated regulation of BDNF.
In this study, we observed discrepancies between mRNA and protein expressions, suggesting the involvement of post-transcriptional modulators such as microRNAs (miR-NAs). miRNAs are small noncoding RNAs which repress gene translation by binding to mRNA transcripts and may affect protein production without affecting mRNA production . Several miRNAs, such as microRNAs-1 (Ma et al., 2015), −210 (Lin et al., 2010), −322 (Yang, Song, & Lv, 2016), are known to modulate BDNF expression in both cardiovascular and brain pathophysiology. Recently, microRNA-135a-5p was shown to affect FNDC5 expression (Metwally et al., 2019). Changes to the FNDC5 protein after translation may also account for discrepancies in mRNA and protein expression. Similarly, the conversion of proBDNF to mBDNF may account for differences in BDNF protein versus mRNA expression. ProBDNF levels were not examined in this study, but posits an area of future experimental examination.
Inflammation is associated with the onset of hypertension and its subsequent comorbidities such as neurological defects. Trends toward lower levels of pro-inflammatory cytokine IL-1α and higher levels of anti-inflammatory cytokines IL-4 and IL-13 were found following exercise, although these did not reach statistical significance. IL-4 and IL-13 are known to be expressed and secreted by skeletal muscle in response to exercise training (Peake, Della Gatta, Suzuki, & Nieman, 2015;Suzuki, 2018) and contribute to the phenotypic switching of macrophages (including microglia in the brain) from a pro-inflammatory M1 state to a reparative M2 state (Littlefield & Kohman, 2017). Exercise may also enhance clearance of neurotoxic compounds. The tryptophan-kynurenine pathway produces neurotoxic metabolites which leads to N-methyl-D-aspartate (NMDA) receptor activation, free radical production, and BDNF reduction (Vécsei, Szalárdy, Fülöp, & Toldi, 2013). In addition to inducing FNDC5, exercise-induced skeletal muscle PGC-1α increases aminotransferases which catalyzes the conversion of kynurenine to kynurenic acid (Agudelo et al., 2014), a form that cannot cross the blood-brain barrier. This consequently protects the brain from neuro-inflammation (Campbell, Charych, Lee, & Moller, 2014;Fukui, Schwarcz, Rapoport, Takada, & Smith, 1991) and prevents decreases in brain BDNF (Calabrese et al., 2014).
We identified several limitations to this study and potential future directions for further study. Firstly, changes in BDNF in SHR versus WKY may be due to strain differences, rather than hypertension per se, which may be delineated by, for example, the use of inhibitors of BDNF-TrkB signaling. In addition, our assays do not discern between different cell types within the tissues of interest, and thus cannot account for heterogeneity in expression. For example, different cardiac cell types may be differentially affected by hypertension or exercise (Tao, Bei, Zhang, Xiao, & Li, 2015;Tirziu, Giordano, & Simons, 2010). Similarly, skeletal muscles differ in their proportion of oxidative and glycolytic muscle fibre types, and the brain consists of neurons and glia, all of which have been shown to have differing responses to exercise (Hyatt et al., 2015;Stevenson, Lensmire, & Swain, 2018). Further studies are required to determine cell type-specific responses to hypertension or exercise. Lastly, this study reported moderate differences in circulating cytokines and plasma BDNF. The cellular sources of these plasma factors remain to be determined. Previous studies have suggested the brain as a major contributor of plasma BDNF after exercise training Seifert et al., 2010).
In summary, 5 weeks of treadmill exercise training reduced BP in both SHR and WKY. BDNF mRNA was increased in the quadriceps and BDNF protein was increased in the LV of SHR, while BDNF mRNA in the DG was lower in SHR, suggesting that hypertension increased BDNF expression in peripheral tissues while decreasing hippocampal BDNF. Exercise normalized BDNF mRNA in the DG of SHR, while BDNF protein was increased in the LV and decreased in the quadriceps of both strains. FNDC5 mRNA was lower in the LV while FNDC5 protein was higher in the LV, quadriceps, and soleus muscle of SHR. Exercise increased FNDC5 protein (but not mRNA) only in the quadriceps of both strains. FNDC5 in the heart and skeletal muscle appears involved in the response to hypertension, while only skeletal muscle is involved in the exercise response. No differences were observed in brain FNDC5, indicating that central FNDC5 is not involved in the response to hypertension or exercise. Thus, targeting the skeletal muscle FNDC5/BDNF pathway through endurance exercise may benefit the heart, skeletal muscle, and brain and lead to novel therapeutic approaches against hypertension and its associated comorbidities.