Aerobic Exercise Upregulates DNA Methylation of Agtr1a and Mas1 Genes to Improve Mesenteric Arterial Function in Spontaneously Hypertensive Rats

The imbalance between vasoconstrictive axis and vasodilative axis of the renin-angiotensin system (RAS) is involved in the pathogenesis of hypertension. Exercise modulates components of the RAS and inuences vascular function. This study aimed to investigate the balance of RAS axes and the mechanism of DNA methylation of the Agtr1a (AT 1a R) and Mas1 (MasR) genes in aerobic exercise-induced improvement of the function of mesenteric arteries (MAs) in hypertension. Spontaneously hypertensive rats (SHRs) and Wistar-Kyoto (WKY) rats were subjected to exercise training or kept sedentary. Plasma RAS peptides, vascular function, and molecular properties were assessed. Aerobic exercise signicantly decreased blood pressure in SHR. Plasma levels of RAS peptides increased in SHR, and the level of Ang II was much higher than Ang-(1–7), whereas exercise eciently inhibited this inappropriate increase. In addition, Ang II-induced maximal contraction of MAs is largely through Type 1 angiotensin receptors (AT 1 R), while Mas receptor (MasR) inhibits this contribution. Exercise effectively suppressed hypertension-associated mRNA and protein expression upregulation of AT 1 R and MasR and increased MasR/AT 1 R ratio in SHR by triggering hypermethylation of Agtr1a and Mas1 genes, with increasing DNMT1 and DNMT3b protein expression and ratio of SAM/SAH. These ndings suggest that aerobic exercise alleviates vascular tone by upregulating the methylation status of the Agtr1a and Mas1 genes and inhibiting the imbalanced increase in the vasoconstrictive and vasodilative axes during hypertension. PCR amplication on the promoter and nearby regions of the Agtr1a (at chromosome 17 from position 35957898 to 35958307, 410 bp in length, containing 21 CpG sites) or Mas1 (at chromosome 1 from position 48076585 to 48077039, 455 bp in length, containing 7 CpG sites) gene was performed. The products were separated with 2.0% agarose gels, and the bands were resolved using the TIAN gel Midi Purication Kit (Tiangen, Beijing, China). The puried samples were cloned into a pEASY-T1 Cloning vector (TRAN). Sample with 10 clones were sequenced. The primers used for modied BSP were 5′-ATGAGGGAGTTAGGATTAGTTGAG-3′ (forward) and 5′-CACTCCRAACTCTAATCACCAC-3′ (reverse) for Agtr1a; The primers used for modied BSP were 5′-TTTAAGAGTAGAGGGGGTTTGG-3′ (forward) and 5′-TACCCTACTTAATACATAACCCCTT-3′ (reverse) for Mas1. The percentage of methylation level of each CpG site was calculated as the ratio of methylated cytosines/total tested cytosines. The average methylation level was calculated using methylation levels of all measured CpG sites within the gene.


Introduction
Hypertension is a large and growing public health problem worldwide and is related to approximately 1/2 of all strokes and heart failure [1]. It is one of the leading causes of morbidity and mortality globally. In clinical and animal models, hypertension is associated with vascular pathological changes, increased vasoconstriction, and arterial wall remodeling. Resistant arteries are responsible for blood pressure control and regional distribution of blood ow, which have an enhanced arterial tone in hypertension [2]. Renin-angiotensin system (RAS) has been shown to be a key factor in the development of hypertension [3]. RAS, as an endocrine system, mainly consists of a vasoconstrictor axis where circulating renin regulates cardiovascular function through angiotensin-converting enzyme (ACE), angiotensin II (Ang II), and angiotensin II receptor type 1 (AT 1 R) and AT 2 R on target tissues and a vasodilator axis, including ACE2, angiotensin-(1-7) [Ang-(1-7)], and its receptor, Mas (MasR). The enhanced vasoconstrictor axismediated vascular smooth muscle cell (VSMC) dysfunction plays a major role in hypertension [4].
The vascular and systemic RAS are involved in increased vasoconstrictor tone [5]. Exercise training has been shown to improve vasodilatory function and decrease vasoconstrictor tone in critical tissues for cardiovascular control [6,7]. A previous study has shown that exercise training represses the mRNA and protein expression of AT 1 R [5]. Ang-(1-7) participates in enhanced insulin-induced vasorelaxation through MasR after exercise [8]. Recent studies have indicated that exercise contributes to the restoration of the vasoconstrictor/vasodilator balance of the RAS axes to improve vascular function [9,10].
Epigenetics refers to stable changes in gene function, without changes in the DNA sequence per se. Many factors such as environment can in uence epigenetic modi cations. DNA methylation is one of the pivotal epigenetic mechanisms in mammals and plays a crucial role in the biological processes during hypertension. Studies have suggested that DNA methylation of RAS components may be associated with hypertension. For instance, hypomethylation of the Agt gene promoter in rat visceral adipose tissue with a high-salt diet results in hypertension [11]. Human studies show that hypomethylation of the AT 1 R gene is likely associated with the risk of hypertension [12,13].
Studies that have focused on the RAS axes in hypertension are controversial, particularly at the level of the vasodilator axis. Here, we sought to investigate the effects of aerobic exercise on RAS axes, and the DNA methylation mechanism of exercise to characterize AT 1 R and MasR gene pro les on SHRs. Our data suggest that both RAS axes are elevated in hypertension, and the vasoconstrictor axis plays a more important role in spontaneously hypertensive rats (SHRs), whereas exercise training signi cantly suppressed RAS axes, with more inhibition on vasoconstrictor axis than vasodilator axis. Moreover, exercise-induced hypermethylation of Agtr1a and Mas1 reprograms AT 1a R and MasR transcription and protein expression in mesenteric arteries (MAs) during hypertension. The change in gene expression is attributed to an epigenetic program with DNMTs.

Animals and exercise protocol
Twelve-week-old male normotensive Wistar-Kyoto rats (WKYs) and SHRs were randomly assigned into a sedentary (WKY-C and SHR-C) and an exercise (WKY-E and SHR-E) group (n = 24 in each group). After one-week habituation, rats in the exercise group were subjected to a motor treadmill (20 m/min, 0% grade, 60 min, 5 days/week, about 55-65% of maximal aerobic velocity) for 12 weeks. The maximal exercise capacity of rats was tested as described elsewhere [14]. Rats in the sedentary group were exposed to the same environments as their trained counterparts. All the rats were housed in a 12:12-h light-dark cycle environment with a temperature of 23-25°C and 40-60% humidity. The rats were provided fresh water and standard rodent chow ad libitum.
Body weight (BW) was measured before the rats were anesthetized, and the blood pressure of conscious WKY rats and SHRs was tested using an indirect tail-cuff method (BP-2010A, Softron Biotechnology, Beijing, China) every two weeks.

Ultra-performance liquid chromatography (UPLC)
Plasma for UPLC analysis was collected by centrifuging blood at 1,500g for 15 min (RT). Then, 500 µL of 1% TFA was added to 500 µL of the plasma sample. The samples were vortexed and centrifuged at 10,400g for 20 min. The pellet was discarded, and the column was prebalanced with 60% acetonitrile, 1% tri uoroacetic acid (TFA), and 39% distilled water. The column was washed with 1 mL of 1% tri uoroacetic acid (TFA), then the samples were added to the column with 1 mL 1% TFA and washed twice. The eluent was collected in a tube and evaporated to dryness. AGT, Ang I, Ang II, and Ang-(1-7) levels were determined using the Dionex Ultimate 3000 UPLC system coupled to a TSQ Quantiva Ultra triple-quadrupole mass spectrometer (Thermo Fisher, CA, USA) and equipped with a heated electrospray ionization (HESI) probe. Samples were separated by a synergi Hydro-RP column (2.0 × 100 mm, 2.5 µm, Phenomenex, CA, USA). The mobile phase contained 10 mM tributylamine with 15 mM acetic acid in water as solvent A, and methanol as solvent B. A gradient of 10-98% of B was run for a total run time of 10 min. The gradient started with 10% B, which was changed progressively to 50% B in 1.5-2.5 min, increased in 2.5-6.5 min at 98% B, then decreased in 6.5-8 min at 10% B again, and run for the next 8.1-10 min at the same composition. Column chamber and sample tray were held at 40°C and 10°C, respectively. SAM and SAH in the plasma sample were identi ed according to their retention times and transitions of 399.2/250.1 and 385.2/136.2 in the negative ion mode. The resolution for precursor and fragment ion were both 0.7 FWHM. The source parameters were as follows: spray voltage, 3,000 V; ion transfer tube temperature, 350°C; vaporizer temperature, 300°C; sheath gas ow rate, 35 Arb; auxiliary gas ow rate, 12 Arb. CID gas, 1.5 mTorr. Data analysis and quantitation were performed using the software Xcalibur 3.0.63 (Thermo Fisher, CA).

Isometric contraction studies
Rats were euthanized by intraperitoneal injection with sodium pentobarbital (50 mg/kg) at the age of 25 weeks. The MAs and its branches were removed and the segments of A3 were isolated in Krebs' solution containing (mM) 131.5 NaCl, 5 KCl, 1.2 NaH 2 PO 4 , 1.2 MgCl 2 , 2.5 CaCl 2 , 11.2 glucose, 13.5 NaHCO 3 and 0.025 EDTA (pH 7.4), mounted on a Multi Myograph System (620M; DMT, Aarhus, Denmark), gassing with 95% O 2 and 5% CO 2 . The maximum contractile response of artery rings was elicited using 60 mM KCl (K max ) and arterial contractile response was evaluated by measuring the maximum peak height and expressed as a percentage of contraction to K max . The non-selective NOS inhibitor N ω -nitro-L-arginine methylester (L-NAME, 10 -4 M) was added in the experiments. To assess contribution of AT 1 R, AT 2 R and MasR function to the vascular tone regulation, the arterial responses to NE (10 − 5 M), Ang II (10 − 5 M), losartan (AT 1 R blocker, 10 − 6 M), PD123319 (AT 2 R blocker, 10 − 5 M), and A779 (MasR blocker, 10 − 5 M) were examined, respectively. Each artery was used once, signals were recorded by Power-Lab system with Chart-5 software (AD Instruments, Bella Vista, Australia).

RT-PCR assay
To analyze the induction of the target gene Agtr1a (AT 1a R), Agtr1b (AT 1b R) and Mas1 (MasR) transcript in MAs, total RNA was extracted with TRIzol reagent (Invitrogen). 500 ng of RNA was reverse transcribed to cDNA using Maxima First Strand cDNA Synthesis Kit after DNase treatment. Real-time PCR was performed with TaqMan Fast Advanced Master Mix and inventoried TaqMan expression assays using an ABI PRISM 7500 Sequence Detection System (Applied Biosystems, Foster City, CA, USA). The target gene used in RT-PCR assays were as follows: Agtr1a (Rn02758772_s1, amplicon length = 112 bp), Agtr1b (Rn02132799_s1, amplicon length = 150 bp), Mas1 (Rn00562673_s1, amplicon length = 67 bp), and the normalization gene Actb (β-actin, Rn00667869_m1, amplicon length = 91 bp). All reagents were from Applied Biosystems. Messenger RNA abundance was calculated using the 2-ΔΔCT method and normalized to Actb mRNA and expressed as a percentage of WKY-C.

DNA bisul te sequencing PCR (BSP)
Genomic DNA from MAs were extracted using the Pure Link Genomic DNA Mini Kit (Invitrogen) and treated with sodium bisulfate with EZ DNA Methylation-GOLD Kit (Zymo Research) according to manufacturer's protocols. Unmethylated cytosine residues were converted to thymines, whereas methylcytosines remain unmodi ed. PCR ampli cation on the promoter and nearby regions of the Agtr1a (at chromosome 17 from position 35957898 to 35958307, 410 bp in length, containing 21 CpG sites) or Mas1 (at chromosome 1 from position 48076585 to 48077039, 455 bp in length, containing 7 CpG sites) gene was performed. The products were separated with 2.0% agarose gels, and the bands were resolved using the TIAN gel Midi Puri cation Kit (Tiangen, Beijing, China). The puri ed samples were cloned into a pEASY-T1 Cloning vector (TRAN). Sample with 10 clones were sequenced. The primers used for modi ed BSP were 5′-ATGAGGGAGTTAGGATTAGTTGAG-3′ (forward) and 5′-CACTCCRAACTCTAATCACCAC-3′ (reverse) for Agtr1a; The primers used for modi ed BSP were 5′-TTTAAGAGTAGAGGGGGTTTGG-3′ (forward) and 5′-TACCCTACTTAATACATAACCCCTT-3′ (reverse) for Mas1. The percentage of methylation level of each CpG site was calculated as the ratio of methylated cytosines/total tested cytosines. The average methylation level was calculated using methylation levels of all measured CpG sites within the gene.

Chemicals and Statistical Analyses
All chemical reagents were purchased from Sigma-Aldrich (www.sigma-aldrich.com) unless otherwise stated. Data were expressed as mean ± SEM. SPSS 17.0 software were used for statistical analyses. Where appropriate, differences were evaluated for statistical signi cance (P < 0.05) by a two-way ANOVA (hypertension × exercise).

Body weight and blood pressure
There were no signi cant differences in body weight (BW) between WKY rats and SHRs at baseline (12 weeks). However, BW was signi cantly lower in both the WKY-E and SHR-E groups when compared with their counterparts at the end of training (both P < 0.05, Table 1). In addition, the BW of SHR-C was signi cantly lower than that in the WKY-C group (P < 0.05). Basic systolic blood pressure (SBP), mean arterial pressure (MAP), and diastolic blood pressure (DBP) in the SHR-C were signi cantly higher than those in WKY-C. After 12 weeks of exercise training, SBP, DBP, and MAP signi cantly decreased in SHR-E compared to SHR-C (all P < 0.05).

Aerobic exercise decreases the plasma levels of RAS components
The plasma levels of AGT, Ang I, Ang II, and Ang-(1-7) were examined by UPLC. The plasma levels of these four peptides markedly increased in SHR-C, while decreased after exercise training in SHR-E (all P < 0.05, Fig. 1a-e). Figure 1f shows that the plasma Ang-(1-7)/Ang II ratio signi cantly decreased in SHR-C compared with WKY-C (P < 0.05), and exercise signi cantly increased the ratio in SHR-E (P < 0.05). These results suggest that the elevated RAS component levels from hypertension are abrogated by chronic exercise. In addition, the decrease in the amplitude of Ang II was larger than Ang-(1-7) in hypertension after exercise training.
Aerobic exercise inhibits the AT 1 R and MasR-mediated responses in contribution to vascular tone regulation in SHR MAs To assess the function of MAs from all groups, NE was added in the presence of L-NAME. The maximal contraction in SHR-C was higher compared with that in WKY-C (P < 0.05), whereas MAs from SHR-E responded with less constriction than SHR-C (P < 0.05, Fig. 2a). Then, Ang was applied to measure its effect on vasoconstriction. Figure 2b shows that in contrast to WKY-C, the increase in Ang -induced tension was more pronounced in the SHR-C arteries (P < 0.05). However, exercise training markedly inhibited this increase (P < 0.05). To examine the contribution of AT 1 R, AT 2 R, and MasR, inhibitors were used on the MAs before adding Ang . Losartan, an AT 1 R inhibitor, almost completely suppressed the Ang -induced contraction (Fig. 2c). However, there were no differences in vasoconstriction between the MAs subjected to Ang and PD123319 (AT 2 R antagonist) co-treatment when compared to the MAs treated with Ang alone (Fig. 2d). By contrast, suppression of MasR with A779 increased the Ang -induced tension in all groups, especially in hypertensive rats (Fig. 2e). Taken together, these results indicate that the tension induced by Ang is largely through AT 1 R, not AT 2 R, while MasR inhibits the contribution of Ang in vascular regulation and plays a pivotal role in resting tone.
Aerobic exercise suppresses increased AT 1a R and MasR protein expression and mRNA levels in hypertension Protein expression levels of AT 1 R and MasR were measured to assess exercise-associated vasodilatation in SHR. Figure 3a and b show the results of western blot analysis, which revealed signi cant increases in both AT 1 R and MasR protein levels of MAs in SHR-C (vs. WKY-C). Moreover, the expression of AT 1 R and MasR both decreased in SHR after exercise, but increased in WKY under the same training conditions (all P < 0.05). To further determine the contribution of AT 1 R and MasR in exercise, MasR/AT 1 R was analyzed.
The expression ratio of MasR/AT 1 R in MAs was signi cantly downregulated in SHR-C (0.60 ± 0.04) and upregulated in SHR-E (0.69 ± 0.02, both P < 0.05, Fig. 3c). These ndings suggest that AT 1 R contributes more to the vessel tone in hypertension. However, MasR has a greater contribution in exercise-induced effects.
Then, the mRNA of AT 1a R, AT 1b R, and MasR were assessed to identify whether the differences in protein expression of AT 1 R and MasR among groups were due to transcription (Fig. 3d). Compared to WKY-C, the mRNA of AT 1a R (P < 0.05), not AT 1b R (P > 0.05), signi cantly increased in SHR-C. Exercise training decreased the mRNA level of AT 1a R in the MAs of hypertensive rats, whereas increased it in normotensive rats after exercise (both P < 0.05). There were no effects on the expression of AT 1b R in either WKY or SHR after exercise training (P > 0.05). Similar to AT 1a R, the level of MasR in SHR-C was higher than that in WKY-C, and exercise decreased mRNA of MasR in SHR-E (both P < 0.05). These results on mRNA expression levels coincided with the ndings of protein expression.

Aerobic exercise increases CpG methylation of the Agtra1 and Mas1 gene promoters in SHR MAs
To further assess the DNA methylation effect on AT 1a R and MasR, we rst evaluated DNMT functions in MAs. Figure 4a and b show that hypertension is associated with a signi cant decrease in the protein expression of DNMT1 (P < 0.05), not DNMT3b (P > 0.05), in SHR-C compared with WKY-C. In normotensive rats, exercise training reduced DNMT1 and DNMT3b expression, but in hypertensive rats, DNMT1 and DNMT3b expression increased with chronic exercise (P < 0.05). DNMT3a levels did not change in any experimental group (P > 0.05). Then, the methylation capacity index (SAM/SAH ratio) was assessed. The SAM/SAH ratio was markedly reduced in SHR-C compared with WKY-C (P < 0.05). In contrast, the SAM/SAH ratio in plasma was signi cantly elevated in SHR after exercise training, but decreased in WKY-E (both P < 0.05, Fig. 4c). The DNA methylation status of the AT 1a R (Agtr1a) and MasR (Mas1) genes in MAs was assessed to con rm the correlation between DNA methylation and transcription. The CpG sites detected are shown in Fig. 4d and e. In normotensive rats, there was a signi cant decrease in promoter methylation of the Agtr1a and Mas1 genes in MAs of WKY-E compared to WKY-C. However, in hypertensive rats, exercise training signi cantly increased both Agtr1a and Mas1 gene methylation status compared with their sedentary counterparts (all P < 0.05, Fig. 4f and i). These ndings show that the changes in the methylation of the Agtr1a and Mas1 genes may partially account for the alteration in DNMT expression in MAs observed in the hypertension and exercise training group.

Discussion
In the present study, 12 weeks of aerobic exercise repressed the plasma levels of bioactive peptides including AGT, Ang I, and Ang and reversed the pathological compensation of Ang-(1-7) in hypertension, inhibiting both Ang receptor AT 1a R and Ang-(1-7) receptor MasR-mediated vascular function in the MAs of SHRs, thereby downregulating the RAS axes. In addition, exercise-induced hypermethylation of the Agtr1a and Mas1 genes via increased DNMT1 and DNMT3b expression reduced the transcription of AT 1a R and MasR (Fig. 5). The novelty of this study is that activation of DNA methylation plays an essential role in the exercise-mediated decrease of AT 1a R and MasR in hypertension, thereby maintaining the vasoconstriction/vasodilatation balance at the physiological level.
RAS is known as a major blood pressure regulator and is involved in the pathogenesis of hypertension. Evidence shows that in hypertension, the vasodilator axis is highly activated [15]. Reduced ACE2 and consequently decreased Ang-(1-7) production is associated with vascular injury in hypertension [16]. Therefore, an imbalance between the vasoconstrictor axis and the vasodilator axis in the RAS has emerged as a common denominator in cardiovascular disorders [17]. The present study demonstrated that levels of plasma peptides [AGT, Ang I, Ang , Ang-(1-7)] all increased in hypertension (Fig. 1a-e). AGT is one of the candidate genes in essential hypertension. Increased plasmatic levels of AGT may induce AGT metabolism, which elevates Ang I. Next, Ang I is converted to Ang , which is a biologically active vasopeptide that increases vascular tone. However, the level of Ang- (1-7), a vasodilator peptide, is also increased during hypertension. To determine the contribution of the vasodilator and vasoconstrictor axes, the ratio of Ang-(1-7)/Ang II was assessed (Fig. 1f). The results showed a lower ratio in hypertensive rats, which suggests that the compensatory increase of Ang-(1-7) is nevertheless insu cient in the counterbalance of Ang II during hypertension. This is accordance with research in upregulated plasma levels of both Ang-(1-7) and Ang in SHR conducted by Kohara [18] and Zhou [19].
The discrepancy in the results on Ang-(1-7) between Shaltout [16] and our study may be related to the use of different experimental animals (sheep vs. rat).
RAS regulates vascular tone and plays an important role in vascular remodeling, which is attributed to a complex interplay of alterations in vascular tone and structure. In this study, isometric contraction studies were used to examine vessel function in relation to RAS. The contractile response to Ang was higher in hypertensive vessels (Fig. 2b). Ang mediates effects via complex intracellular signaling pathways by binding to two major G protein-coupled receptors AT 1 R and AT 2 R. AT 1 R mediates most of the pathophysiological effects of Ang , including vasoconstriction, in ammation growth, and brosis, whereas AT 2 R is thought to oppose the effects of AT 1 R [20]. By preincubating with an AT 1 R inhibitor (losartan), Ang -induced vasoconstriction was almost repressed in all groups. However, the AT 2 R antagonist (PD123319) did not impart any effects on Ang II-mediated vasoconstriction (Fig. 2c and d). Using a MasR blocker (A779), the vascular tension induced by Ang increased, suggesting the role of MasR in the regulation of vascular tone. These results are further supported at the molecular level by the demonstration of increased MasR and AT 1a R, but not AT 1b R transcript (Fig. 3d) and protein expression ( Fig. 3a and b). The MasR/AT 1 R ratio also decreased in SHR-C (Fig. 3c). The ratio of the two receptors was in line with upstream Ang-(1-7)/Ang . Ang is the main active molecule of the classical RAS; upon Ang binding, AT 1 R facilitates a variety of cytoplasmic signaling pathways that mediate VSMCs remodeling, including hypertrophy and migration. The inappropriate overactivity of the Ang /AT 1 R interaction is involved in the genesis and progression of hypertension [21]. A sustained activation of AT 1 R by Ang , one of the main effectors of RAS, has been proposed to contribute to increased peripheral vascular resistance in obesity-associated hypertension [22]. Tirapelli et al. [23] has observed that Ang-(1-7) is able to relax carotid rings through activation of MasR on VSMCs, as well as the aorta [24].
In the present study, DNA methylation status of the Agtra1 and Mas1 gene promoters was well correlated with changes in transcriptional levels of AT 1a R and MasR, revealing a higher methylation status of the Mas1/Agtra1 gene promoter in MAs in the development of hypertension. DNA hypermethylation is a hallmark of gene silencing, while DNA hypomethylation promotes active transcription. Recent studies have reported that methylation of Mas1/Agtra1 plays a key role in the regulation of vascular tone [25,26]. For instance, perinatal nicotine exposure enhances vascular contractility that is associated with decreased DNA methylation of Agtra1 in the aorta in adult offspring [27]. In addition, hypomethylation of the Agtra1 gene promoter is correlated with the expression of AT 1a R in the aorta and MAs of SHRs [28]. In this study, the methylation status of the Agtra1 gene promoter was similar to Pei in MAs from SHRs. The hypomethylation of Mas1 was exhibited in the SHR-C as well. DNA methylation is catalyzed by DNMT. DNMT3a and DNMT3b are the de novo DNA methyltransferases that act on non-methylated DNA. However, DNMT1 is essential to the maintenance of methylated DNA. The present study found that the expression of DNMT1, not DNMT3a or DNMT3b, signi cantly decreased in hypertensive MAs ( Fig. 4a and  b). A pilot study on in vitro fertilization-embryo transfer on Ang -mediated vasoconstrictions in umbilical cord vein suggested that hypomethylation of Agtr1 is caused by decreased DNMT3a expression [26]. This difference may be attributed to different vascular beds and experimental models. Recently, research works have demonstrated that DNMT1 mediates de novo methylation in several cell types [29,30]. To further con rm the contribution of DNMT1 in de novo Agtra1 and Mas1 methylation, the ratio of SAM/SAH was measured by UPLC, which showed a decreased ratio in SHR-C. This nding, in combination with reduced DNMT1 expression, underlies Agtra1 and Mas1 gene promoter hypomethylation in MAs and contributes to pathological/compensatory upregulation of AT 1a R/MasR during hypertension. In addition, it reveals that DNMT1 is not purely a maintenance methyltransferase but can also participate in de novo methylation.
Aerobic exercise is often considered the cornerstone of nonpharmacological treatment of hypertension. However, the exact mechanism by which exercise improves vascular dysfunction in hypertension remains unclear. In this study, we have proven a disproportional reduction in RAS peptides in hypertension, especially the major bioactive content of Ang and Ang-(1-7) after 12 weeks of exercise. In addition, the increase in the Ang-(1-7)/Ang ratio is a more important contributor to enhancing the function of hypertensive MAs after exercise, indicating that exercise restores the abnormal balance of vascular RAS and adjusts this to a lower level. The transcript and protein expression of receptors (AT 1 R, MasR) suppressed in SHR-E group. Studies suggest that epigenetics is in uenced by the environment, and exercise is associated with DNA methylation [31]. In support of this notion, DNA-modifying enzymes in MAs and intermediate metabolites (SAM, SAH) were assessed. Exercise upregulated the expression of DNMT1 and DNMT3b in MAs and increased SAM/SAH, thus in uencing Agtra1 and Mas1 gene promoter methylation, which led to the downregulation of AT 1a R and MasR. By contrast, in normotensive rats, exercise-mediated enhancement of AT 1a R and MasR expression was correlated with decreased methylation status of Agtra1 and Mas1 gene promoters, which were attributed to reduced DNMT1 and DNMT3b expression and decreased SAM/SAH ratio. The difference in the results between physiological and pathological conditions may be multifactorial. We hypothesize that exercise training mainly restores the inappropriate overactivity of RAS to a lower level in hypertension, whereas exercise enhances vasoconstriction by increasing AT 1a R and MasR expression at the physiological level.

Conclusions
In summary, the present set of data shows that exercise reduces RAS and its counter-regulatory axes to improve vascular function in hypertensive rats. Most important, our results demonstrate that aerobic exercise decrease AT 1a R and MasR via hypermethylation of Agtra1 and Mas1 gene, counterbalancing the pathological increase. These results provide mechanistic evidence that unlike currently available pharmacological anti-hypertensive therapies, aerobic exercise plays a favorably role in treatment for hypertension. Our nding reveals crucial insight into the epigenetic mechanism by which exercise exerts bene cial effects in hypertension. Figure 1 The levels of RAS components in plasma. Plasma of AGT, Ang I, Ang II, Ang-(1-7) levels, respectively (ad), comparison of plasma levels of AGT, Ang I, Ang II and Ang-(1-7) (e), Ang-(1-7)/ Ang II in WKY-C, WKY-E, SHR-C, SHR-E plasma samples (f). n = 5-6. * P < 0.05 vs WKY-C; # P < 0.05 vs SHR-C. In each experiment, the arteries were preincubated with nonselective nitric oxide synthase inhibitor Nωnitro-l-arginine methyl ester (L-NAME, 100 μM) for 20 min (dotted arrows). n = 6 in each group. * P < 0.05 vs WKY-C; # P < 0.05 vs SHR-C.