Aerobic exercise training induced LVH involves regulatory microRNAs, decreased ACE-ANG II, and synergistic regulation of ACE2-ANG (1–7)

Abstract

Aerobic exercise training leads to a physiological, non pathological left ventricular hypertrophy (LVH); however, the underlying biochemical and molecular mechanisms of physiological LVH are unknown. The role of microRNAs regulating the classic and the novel cardiac renin angiotensin system (RAS) was studied in trained rats assigned to three groups: sedentary, swimming trained with protocol 1 (T1: moderate volume training) and protocol 2 (T2: high volume training). Cardiac Ang I levels, ACE activity and protein expression, as well as Ang II levels were lower in T1 and T2, however AT1 mRNA levels (69% in T1 and 99% in T2) and protein expression (240% in T1 and 300% in T2) increased after training. AT2 receptor mRNA levels (220%) and protein expression (332%) were shown to be increased in T2. In addition, T1 and T2 were shown to increase ACE2 activity and protein expression, and Ang (1–7) levels in the heart. Exercise increased microRNA-27a and 27b, targeting ACE and decreasing microRNA-143 targeting ACE2 in the heart. LVH induced by aerobic training involves microRNAs regulation and an increase in cardiac AT1 receptor without the participation of Ang II. Parallel to this, increase in ACE2, Ang (1–7) and AT2 receptor in the heart by exercise suggests that this non classic cardiac RAS counteracts the classic cardiac RAS. These findings are consistent with a model in which exercise may induce LVH, at least in part, altering the expression of specific microRNAs targeting RAS genes. Together these effects might provide the additional aerobic capacity required by the exercised heart.

INTRODUCTION

Left ventricular hypertrophy (LVH) induced by aerobic exercise training is an important physiological compensatory mechanism in response to chronic increases in hemodynamic overload. This phenotype is associated with sarcomeres added in series to lengthen the cardiac cell, as well as in parallel. The increased cross sectional area contributes to increased ventricular stroke volume and cardiac output which improves aerobic capacity. In contrast, pathological LVH in cardiovascular diseases (CVD) is associated with increased fibrosis, and lowered aerobic capacity leading to high mortality (1–3).

Several studies have reported that the renin angiotensin system (RAS) plays an important role in the progression of LVH (4–6). However, there are only limited data about the mechanisms of exercise training involved in RAS and LVH. The aim of this study was to elucidate these mechanisms of exercise training on physiological LVH.

Pathological LVH occurs with arterial hypertension (7, 8), myocardial infarction (9, 10) and heart failure (11). These disease states are also associated with increased local cardiac RAS levels, represented by augmented angiotensinogen, angiotensin-converting enzyme (ACE) and angiotensin II (Ang II). Blockade of the classical RAS promotes therapeutic benefits to patients with essential hypertension and CVD (8). Cardiac Ang II is implicated in the induction of fibrosis, but is not required for LVH (12). LVH is produced without the participation of Ang II in transgenic animal models for RAS components (13–15) and in AT1 receptor activation by mechanical stress (16–18). A novel cardiac RAS includes angiotensin-converting enzyme 2 (ACE2) which is an essential regulator of heart function (19) and plays a pivotal role in angiotensin (1–7) formation. This novel RAS is implicated in vasodilatation and control of fibrosis (20–23). Previous findings suggest that ACE2 maintains the important balance between the Ang II and Ang (1–7), favoring cardiovascular homeostasis. However, the role of exercise training in the cardiac ACE2-Ang (1–7) axis is unknown. We have shown that AT1R blockade prevents physiological LVH induced by resistance training (18) and by aerobic exercise training (24). Moreover, exercise training promoted LVH by cardiac RAS stimulation independent of the systemic RAS (18,24).

Several genes are regulated by microRNAs (miRNAs). MiRNAs are endogenous, small and non-coding RNA, which are targeted to specific genes and function as negative regulators of gene expression by inhibiting translation or promoting degradation of target mRNAs. Recent studies have shown the roles played by miRNAs in different forms of CVD and pathological LVH (25,26). However, miRNAs may be important for normal development and in physiological cardiac hypertrophy induced by aerobic exercise training. In the present study, it was hypothesized that exercise training alters specific miRNAs that regulate their target cardiac RAS genes and tip the balance of classic RAS genes in favor of the novel RAS genes to contribute to physiological LVH.

MATERIAL AND METHODS

Animal Care

All protocols and surgical procedures used were in accordance with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals and were approved by the Ethics Committee of the School of Physical Education and Sport of the University of Sao Paulo. Female normotensive Wistar rats (190 to 220 g, n=42) were used. The animals were housed 3–5 per cage at a controlled room temperature (22°C) with a 12-hour dark-light cycle and fed standard rat chow having access to water ad libitum.

The rats were randomly divided into three experimental groups, each with 14 rats: sedentary (S, n=7), swimming trained with protocol 1 (T1, n=7) and swimming trained with protocol 2 (T2, n=7). Each group was subdivided into 2 groups, one for hemodynamic, biochemical, and molecular studies, and the other for morphological and histological studies. For detailed Material and Methods, please see the online data supplement, available at http://hyper.ahajournals.org.

RESULTS

Hemodynamic Parameters

Table 1 summarizes systolic blood pressure (SBP), diastolic blood pressure (DBP), mean blood pressure (MBP) and heart rate (HR) results of the groups S, T1 and T2. There was no difference in blood pressure among the three groups. However, HR decreased significantly after 10 weeks of swimming training in T1 (301.2±15.3 beats/min) and T2 (309±14 beats/min) groups when compared with S group (344.8±12.1 beats/min, P<0.05).

Table 1.

Hemodynamic Parameters.

Groups	SBP, mmHg	DBP, mmHg	MBP, mmHg	HR, beats/min
S	127.6 ± 3.9	97.6 ± 10.3	113.5 ± 7.3	344.8 ± 12.1
T1	123.3 ± 8.5	96.4 ± 5.2	110.7 ± 6.7	301.2 ± 15.3*
T2	123.0 ± 8.4	94.3 ±8.9	108.3 ± 9.0	309.0 ± 14.0*

Values are means ± SD; SBP, systolic blood pressure; DBP, diastolic blood pressure; MBP, mean blood pressure; HR, heart rate. Significant difference vs. *S, P < 0.05.

Cardiac Hypertrophy

Body weight (BW) before and after swimming training was similar among all the studied groups. Left ventricle (LV) and right ventricle (RV)/BW ratio was used as an index of hypertrophy. The absolute values referring to BW, LV/BW and RV/BW in all groups of rats are summarized in the Table 2. LVH obtained by T1 and T2 was 13% (2.8±0.14 mg/g; P<0.05) and 27% (3.2±0.12 mg/g; P<0.01), respectively, in comparison with S group (2.5±0.06 mg/g). Right ventricle hypertrophy obtained by T1 and T2 was 15% (0.68±0.06 mg/g; P<0.05) and 35% (0.80±0.08 mg/g; P<0.01), respectively, in comparison with S group (0.59±0.04 mg/g). The increase in LV/BW ratio observed with swimming training was further confirmed by the increase in LV myocyte diameter in T1 (13.2±1.3 μg) and T2 (14.4±1.3 μg) groups when compared with S group (11±1.1 μg), P<0.05 (Table 2). Supplementary Figure 1 shows representative histological sections of increase in LV myocyte diameter in T1 and T2 in comparison with S group.

Table 2.

Cardiac Hypertrophy.

Groups	BW, g	LV/BW, mg/g	RV/BW, mg/g	Myocyte Diameter, μg
S	245.8 ± 9.4	2.5 ± 0.06	0.59 ± 0.04	11 ± 1.1
T1	232.2 ± 6.4	2.8 ± 0.14*	0.68 ± 0.06*	13.2 ± 1.3*
T2	234.7 ± 17.3	3.2 ±0.12†,‡	0.80 ± 0.08†,‡	14.4 ± 1.3*

Values are means ± SD; BW, body weight; LV, left ventricular weight; RW, right ventricular weight. Significant difference vs. *S, ^†T1, P < 0.05; ^‡S, P < 0.01.

Molecular Markers of Pathological LVH

Pathological cardiac hypertrophy is characterized by the induction of genes normally expressed during fetal development, such as, atrial natriuretic factor (ANF), skeletal muscle α-actin and a decrease in the ratio of α/β-myosin heavy chain (MHC). The mRNA levels of these 4 genes were assessed in the LV of S and trained groups (T1 and T2) by real-time PCR. The results of this study showed that swimming training did not modify ANF gene expression. Similarly, T1 training did not change the gene levels of skeletal α-actin and α/β-MHC, although T2 exercise significantly reduced the LV levels of skeletal α-actin by 53% (P<0.05), and increased the LV levels of α/β-MHC by 98% (P<0.05), when compared with S group.

RAS Biochemical Analysis

To evaluate the role of swimming exercise training in systemic RAS, the serum ACE activity and plasma renin activity were measured. Figure 1A shows that there was an increase in serum ACE activity of 3.8% in T1 (P=NS) and 23.5% in T2 (P<0.05) in comparison with the S group. Plasma renin activity also was increased by 20% in T1 (P=NS) and 126% in T2 (P<0.01) when compared with S group (Figure 1A). In contrast, there was a reduction in local cardiac ACE activity of 11% (P=NS) and 15% (P=NS) in the RV and LV, respectively, in the T1 group. When the T2 group was compared with S group, there was a decrease of 40% (P<0.05) and 32% (P<0.05) in RV and LV, respectively (Figure 1B). Interestingly, Figure 1B also shows that LV ACE2 activity was increased by 12% in T1 (1,708±354 uF/min/mg, P=NS) and 41% in T2 (2,160±218 uF/min/mg, P<0.01) when compared with S group (1,531±174 uF/min/mg).

Figure 1.

Effect of exercise training on systemic and cardiac RAS activity. Serum ACE activity and Plasma Renin activity (A). Right Ventricle (RV) ACE activity, Left Ventricle (LV) ACE activity and Left Ventricle (LV) ACE2 activity (B). Groups: S- sedentary, T1- swimming training protocol 1 and T2- swimming training protocol 2. Data are reported as means ± SD. Significant difference vs. ^*S, ^†T1, P < 0.05, ^**S, P < 0.01.

RAS Molecular Analysis

In order, to test whether swimming exercise training modulates cardiac RAS gene expression, real-time PCR was used to assess ACE, ACE2, AT1 and AT2 receptor gene expression in the heart. mRNA levels of ACE showed a small decrease and ACE2 mRNA a small increase in both trained groups but without significance (data not shown). AT1 receptor gene expression increased in T1 (69%, P<0.05) and T2 (99%, P<0.01) when compared with S group (Figure 4B). In addition, AT2 receptor gene expression increased by 26% (P=NS) in T1 and 332% in T2 (P<0.001); T1 differed from T2 (P<0.001) (Figure 4D).

Figure 4.

Effect of exercise training on AT1 and AT2 receptor gene and protein expression in the heart. Data are presented as means ± SD. Cardiac AT1 receptor (A) and AT2 receptor (C) protein expression analyzed by western blotting accompanied by their representative blots from sedentary (S) and trained groups (T1 and T2). Targeted bands were normalized for cardiac α-tubulin. Cardiac AT1 receptor (B) and AT2 receptor (D) gene expression analyzed by real-time PCR. Targeted genes were normalized by cyclophilin mRNA. Significant difference vs. ^*S, P < 0.05; ^**S, P < 0.01 and ^***S, ^‡T1 P < 0.001.

Similar results were obtained for RAS proteins and peptide levels determined by western blotting and high performance liquid chromatography (HPLC), respectively. Figure 2A shows that swimming exercise training decreased cardiac angiotensinogen levels by 26% (P<0.05) in T1 and 44% in T2 (P<0.05) when compared with S group. As angiotensinogen is a substrate for Ang I production, this reduction was accompanied by a decrease of 25.6% (P<0.05) in cardiac Ang I levels in T1 and 44% in T2 (P<0.001) when compared with S group (Figure 2B). The next step was to measure the levels of ACE, since it is responsible for converting Ang I into Ang II. Accordingly, Figure 2C shows cardiac ACE levels were decreased by 22% (P=NS) in T1 and 31% in T2 (P<0.05) in comparison with S group. This reduction was accompanied by a decrease of 23% (P<0.001) in Ang II levels in T1, and 20% in T2 (P<0.001) in comparison with S group (Figure 2D), indicating an attenuation of the ACE-Ang II axis induced by swimming exercise training.

Figure 2.

Effect of exercise training on classical ACE-Ang II axis in the heart. Data are presented as means ± SD. Cardiac angiotensinogen (A) and ACE (C) protein expression analyzed by western blotting accompanied by their representative blots from sedentary (S) and trained groups (T1 and T2). Targeted bands were normalized for cardiac α-tubulin. Cardiac Ang I (B) and Ang II (D) peptide concentration analyzed by HPLC. Significant difference vs. ^*S, P < 0.05; ^***S, P<0.001.

Swimming exercise training also had an effect on the protein and peptide levels of novel RAS, ACE2 and Ang (1–7) in the heart. As shown in Figure 3A, swimming training increased ACE2 protein expression in both trained groups (68% in T1 and 91% in T2, P<0.05), when compared with the S group. There was increased Ang (1–7) formation (160% in T1 and 120% in T2, P<0.01; Figure 3B) in comparison with S group. Figure 3C shows an increase in the Ang (1–7)/Ang II ratio in both trained groups (180% in T1, P<0.05 and 160% T2, P<0.001) in comparison with S group, suggesting an aerobic training-mediated increase in Ang (1–7) formation from Ang II by ACE2.

Figure 3.

Effect of exercise training on novel ACE2- Ang (1–7) axis in the heart. Data are presented as means ± SD. Cardiac ACE2 (A) protein expression analyzed by western blotting accompanied by its representative blot from sedentary (S) and trained groups (T1 and T2). Targeted band was normalized for cardiac α-tubulin. Cardiac Ang (1–7) peptide concentration analyzed by HPLC (B). Angiotensin (1–7) generation from Angiotensin II represented by Ang (1–7)/Ang II ratio (C). Significant difference vs. ^*S, P < 0.05; ^**S, P < 0.01; ^***S, P<0.001.

Protein expression of AT1 receptor, in concert with the increase in AT1 receptor mRNA levels, was 2.4 fold greater in T1 (P<0.05) and 3.0 fold greater in T2 (P<0.05) when compared with the S group (Figure 4A). In addition, AT2 receptor protein expression was increased 1.6 fold in T1 (P=NS) and 2.2 fold (P<0.05) in T2 (Figure 4C).

MiRNAs Analysis by Microarray

Microarray analysis of miRNA was restricted to those miRNAs that underwent a significant change from baseline. Figure 5A shows miRNAs targeting ACE: in the S group the relative expression value of miRNA-27a was 1,760±108 a.u. In T1 group the value was 2,225±78 a.u., (26% increase in comparison with S; P<0.05) and in T2 group the expression was 3,218±30 a.u., (83% increase in comparison with S; P<0.01). In addition, T1 differed from T2 (P<0.01). Similarly, in the S group the relative expression value of miRNA-27b was 3,409±89 a.u. In T1 group the value was 4,341±124 a.u., (27% increase in comparison with S; P<0.05) and in T2 group the expression was 4,939±59 a.u., (45% increase in comparison with S; P<0.01). In addition, T1 differed from T2 (P<0.01). Figure 5A also shows miRNA targeting ACE2: in the S group the relative expression value of miRNA-143 was 6,556± 157 a.u., T1= 6,095±83 a.u. was not significant when compared with S, however in T2 group the expression was 4,249±32 a.u., (35% decrease in comparison with S; P<0.01). T1 also differed from T2 (P<0.01).

Figure 5.

Effect of exercise training on specific miRNAs targeting RAS genes. MiRNAs associated with ACE (miRNA-27a and 27b) and ACE2 (miRNA-143) analyzed by microarray (A). Confirmation of miRNAs-27a, 27b and 143 by real-time PCR (B). Targeted miRNAs were normalized by U6 expression. MiRNAs were isolated using the mirVana^™ qRT-PCR- miRNA. Significant difference vs. ^*S, ^†T1, P < 0.05 and ^**S, ^§T1, P < 0.01.

MiRNAs Analysis by Real-Time PCR

To confirm the miRNAs that targeted RAS genes in physiological LVH, the miRNAs-27a, 27b and 143 were quantified by real-time PCR. MiRNAs- 27a (S: 1.0±0.08, T1: 1.52±0.06, T2: 2.04±0.13 a.u.) and 27b (S: 1.0±0.08, T1: 1.30±0.05, T2: 1.59±0.08 a.u.) were upregulated in T1 and T2 in comparison with S, while miRNA-143 (S: 1.0±0.11, T1: 0.82±0.02, T2: 0.58±0.05 a.u.) was downregulated in T2 in comparison with S. The miRNAs expression in T1 and T2 in comparison with S group confirmed the microarray results.

DISCUSSION

The results of the present study show that swimming exercise training: 1) induced physiological LVH, 2) is not correlated with pathological cardiac hypertrophy markers, 3) increased AT1 and AT2 receptor expression, 4) decreased cardiac ACE and Ang II levels and increased ACE2 and Ang (1–7) levels, 5) altered the expression of specific miRNAs that target RAS genes.

The LV/BW ratio, myocyte diameter and the resting bradycardia confirmed the exercise-associated adaptations of physiological LVH (24, 27). In contrast to markers for pathological hypertrophy (28), the physiological LVH reported here was not associated with activation of fetal genes, such as, ANF, skeletal muscle α-actin and β-MHC. There were no pathological hypertrophy markers in the exercised trained T1 or T2 groups.

The development of LVH after training did not appear to involve Ang II, which was decreased, supporting recent evidence from several transgenic animal models that increased formation of local Ang II in the heart does not directly develop hypertrophy, except when excess cardiac Ang II enters the circulation and causes an increase in blood pressure (12–14). Xiao et al. (15) reported that in mice expressing ACE only in the heart, increased cardiac Ang II was not associated with cardiac hypertrophy. Our group demonstrated that in transgenic mice harboring one, two, three or four ACE gene copies, the magnitude of physiological LVH induced by swimming training was not correlated with ACE levels (17).

Increased AT1 receptor was also found with resistance training (18). Losartan treatment blocked LVH in the same two swimming training protocols (24). The mechanism for overexpressing AT1 receptor may be related to an independent action of the AT1 receptor. Zou et al. (16) showed in vitro and in vivo that AT1 receptor is a mechanical sensor and converts mechanical stress into a biochemical signal inducing LVH without involvement of Ang II. Moreover, Yasuda et al. (29) showed that mechanical stress activates an anticlockwise rotation of transmembrane 7 domain of AT1 receptor causing a conformational change of the receptor, independently of Ang II. AT1 receptors have no direct cell signaling pathway to tyrosine kinase and the MAPK pathways for cell growth (30, 31). The AT1 receptors have an indirect, membrane transactivating step to stimulate epidermal growth factor receptor (EGFR) (30, 31). Inhibition of the EFGR directly prevents Ang II induced LVH in rats (31). The mechanical activity of exercise could therefore activate the EGFR pathway via mechanical activation of AT1 even in the absence of Ang II.

Aerobic exercise training also increased AT2 receptor genes and protein expression in the left ventricle. Studies suggest that AT1 and AT2 receptor may serve opposing functions in the heart, although they exhibit the same ligand binding affinity (4, 12). The role of AT2 receptors in cardiac regulation is not fully understood. AT2 receptor has been associated with dephosphorylation and inactivation of growth factor-activated MAPK and inactivation of ERK 1/2 providing a protective role in the heart (32). Furthermore, AT2 receptor activates nitric oxide and bradykinin, inducing vasodilation (33). Yang et al. (34) demonstrated that in transgenic animals, the overexpression of AT2 receptors preserved left ventricle function after myocardial infarction. The results of the present study suggest that the AT2 receptor plays a cardioprotective role in opposing deleterious cardiac remodeling in CVD. In contrast to the pathological condition, an increase in AT2 receptor expression in the heart could aid vasodilatation induced by aerobic exercise training. This modulation might increase blood and oxygen transport to the exercising cardiac muscle to facilitate high performance.

The present study demonstrated the effect of aerobic exercise training on ACE2 and Ang (1–7) in the heart of rats. The discovery of ACE2 revealed that classical RAS has a reciprocal side, the novel RAS. The results of this study show a reversal of balance in favor of the novel RAS with exercise training. When compared with sedentary animals the trained groups had increased ACE2 and Ang (1–7) activity and protein expression in the left ventricle. ACE2 cleaves Ang I to generate the inactive Ang (1–9) peptide, but the preferred pathway with 500 fold greater efficiency is Ang II to generate the vasodilator Ang (1–7) (6, 35, 36). In hypertension, a decrease in ACE2 mRNA and protein expression, leads to increases in local Ang II levels and might reduce myocardial blood flow preferentially via coronary vasoconstriction or microcirculatory dysfunction. However, with the use of ACE inhibitor or AT1 receptor blocker, the ACE2 level is enhanced (35–37). Transgenic animal models overexpressing cardiac ACE2 by systemic lentiviral delivery resulted in a regression of pathological LVH in hypertensive rats (21). In fact, studies have suggested that Ang (1–7) can reduce hypertension-induced cardiac remodeling through a direct effect on the heart and raise the possibility that pathologies associated with ACE2 inactivation are partly mediated by a decrease in Ang (1–7) production (38, 39).

AT1 receptor blockade augmented the plasma Ang (1–7)/Ang II ratio suggesting increased generation of Ang (1–7) from Ang II (20). In addition, Crackower et al. (19) showed that deletion of ACE2 in mice resulted in elevated cardiac and plasma Ang II, together with impaired cardiac contractility and exhibited left ventricle dilatation. Therefore, ACE2 might protect against pathological LVH by reducing Ang II concentration and increasing Ang (1–7) generation (6, 20–23, 38, 39).

Thus the mechanism of aerobic exercise training to prevent LVH could occur by diminished vascular resistance leading to increased cardiac flow, due to the reduction in ACE and Ang II levels and the vasodilator effects of the ACE2-Ang (1–7) expression, mediating the release of different vasoactive factors such as nitric oxide, prostaglandins and bradykinin (40,41).

Another aspect of this report is the correlation of miRNAs with angiotensin related genes. The implication of specific miRNA regulating RAS genes in cardiac hypertrophy induced by exercise training has not previously been reported. The target predictions of miRNAs are all based on 3′UTR of mRNA of RAS components in web based bioinformatics TargetScan 4.2 and 5.1, MiRanda and PicTar. Confirming these predictions, the literature provides more details of angiotensin gene regulation through qPCR and angiotensin gene measurements. The ACE gene has recently been shown to be regulated by miRNA-27a and 27b (42). It has been demonstrated that the ACE2 gene is regulated by miRNA-143 (43).

This study reveals potential molecular mechanisms for the results. MiRNAs target multiple genes but targeted genes are controlled by specific miRNAs (25, 26). Increased expression of miRNA indicates inhibition of the target gene. This appears to be the case with miRNA-27a and 27b as ACE decreased by 22% (T1) and 31% (T2) in comparison with the control, while miRNA-27a increased by 26% (T1) and 88% (T2) when compared with the control and miRNA-27b increased by 27% (T1) and 44% (T2). By the same principle, decreased expression of miRNAs reflects increased expression of target genes. Whereas, in the T2 group in which the expression of ACE2 was highest, the miRNA that target the ACE2 gene, miRNA-143, were at their lowest level of expression when compared with the control or T1. Thus aerobic exercise training exerts an effect on the expression of miRNAs and thereby might regulate their specific target genes.

PERSPECTIVES

Exercise is widely recognized as an important lifestyle factor in lowering hypertension and improving cardiac health. This study reveals some of the biochemical and molecular mechanisms of aerobic exercise training involved in physiological, non pathological cardiac hypertrophy. The results clearly indicate that in aerobic exercise trained animals; LVH is physiological and associated with decreased ACE and Ang II versus increased ACE2 and Ang (1–7), and increased AT1 and AT2 receptors. In addition there was a reciprocal differential expression of specific miRNAs and these genes. These findings are consistent with a model in which exercise may influence these changes, at least in part, altering the expression of specific miRNAs targeting RAS genes. Together, these effects might provide the additional aerobic capacity required by the exercised heart. The results imply that a decrease in miRNA-143 could upregulate cardioprotective genes in the heart and an increase of miRNA-27 expression inhibits ACE levels. These results suggests that a basis for treatment to prevent of the development of pathological LVH might be to inhibit specific miRNAs, probably with antisense or siRNA, to inhibit ACE and Ang II and increase ACE2 and Ang (1–7).