Activation of Cannabinoid Receptor Type II by AM1241 Ameliorates Myocardial Fibrosis via Nrf2-Mediated Inhibition of TGF-β1/Smad3 Pathway in Myocardial Infarction Mice

Myocardial infarction (MI) is still a common fatal disease despite advances in drug treatment and percutaneous coronary intervention. According to the report from American Heart Association, there will be approximately one American suffering from MI within every 42 seconds [1]. Cardiac fibrosis is the key part of cardiac remodeling after MI which was characterized by excessive collagen deposition resulting in ventricular stiffening, diastolic and systolic dysfunction and eventually leading to heart failure [2,3]. It not only involves the infarct zone, but the non-infarct region is also affected. The cardiac fibrosis after MI to date is still a thorny issue for clinical doctors due to few available effective strategies [4].

The cannabinoid receptor-type II (CB2 receptor), a seven-transmembrane and G protein-coupled receptor, is widespread in peripheral tissues including liver, skin, heart and immune cells [5]. In the past several years, the CB2 receptor activation has emerged as a promising therapeutic target to protect against fibrosis in various organs, for instance, CB2 receptor agonist was reported to significantly reduce hepatic collagen content in cirrhotic rats and exert antifibrotic effects in experimental dermal fibrosis mice [6,7]. Moreover, CB2 receptor activation has exhibited great potential in anti-oxidative stress and anti-inflammation in various disease models [8,9,10]. In our previous study, we also found that CB2 receptor agonist AM1241 activated PI3K/Akt/Nrf2 signaling to reduce excess oxidative stress and inflammation in ischemic heart to promote endogenous myocardial regeneration [11]. Incidentally, it was well known that excess oxidative stress and inflammation in heart aggravated cardiac fibrosis [12,13]. Thus, we were curious to know whether CB2 receptor activation could ameliorate cardiac fibrosis post MI.

Transforming growth factor-β1 (TGF-β1) has long been demonstrated to play a crucial role in regulation of extracelluar matrix (ECM) and homeostasis of tissue. TGF-β1 predominantly transmits cell signaling through a downstream mediator protein named Smad3. Numerous evidences have proved that the TGF-β1/Smad3 pathway was a canonical pathway in the development of organ fibrosis [14,15,16]. On the other hand, nuclear factor-erythroid-2- (NF-E2-) related factor 2 (Nrf2), a major regulator of the antioxidant response elements, has long been known for resisting oxygen-free radicals and oxidative stress [17,18]. Previously, we and others have reported the involvement of Nrf2-reduced oxidative stress in alleviation of cardiac fibrosis [19,20,21]. So our current study was designed to investigate whether Nrf2 and TGF-β1/Smad3 signaling were involved in CB2 receptor activation-regulated cardiac fibrosis post MI.

In the present research, both in vivo and in vitro models were conducted to imitate cardiac fibrosis after MI. We attempted to evaluate the impact of CB2 receptor agonist AM1241 on cardiac fibrosis post MI and investigate the underlying mechanisms with a focus on Nrf2 and TGF-β1/Smad3 signaling pathways.

Eight-week-old C57BL/6 male mice between 20-25 g were purchased and housed in the Animal Center of Fourth Military Medical University. Mice were randomly divided into 4 experiment groups: 1) sham group (n=20), 2) sham+AM1241 group (n=20), 3) MI group (n=20), 4) MI+AM1241 group (n=20). All animal experiment protocols were performed in conformity with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and approved by the Fourth Military Medical University Ethic Committee on Animal Care.

The MI mice model were conducted as described previously [22], In brief, the left anterior descending (LAD) artery was ligated tightly with a 6-0 suture until the anterior wall of the left ventricle (LV) turned pale. The sham group were performed with similar procedure except LAD ligation. After operation, AM1241 (20mg/kg/d, Selleck, USA) was intraperitoneally injected for 7 consecutive days in 2) and 4) groups, while the (1) and (3) groups were given equal volume DMSO for 7 consecutive days.

To quantify fibrosis area in histological staining, we randomly selected 10 fields in remote region of heart section in "MI" group (n=5) and ‘‘MI+AM 1241'' group (n=5) under polarized light microscope. The percent of the positive staining areas were quantified with Image J software by 2 independent experienced investigators.

Primary cardiac fibroblast cells (CFs) were isolated from the left ventricular of neonatal C57BL/6 mice using the enzymatic digestion and selective plating methods as previously described [23]. Briefly, the hearts were removed from the sterilized mouse with 75% ethanol and then washed with cold PBS twice, the left ventricular were then minced before incubated with 0.1% collagenase at 37°C. The digested cell suspension was mixed with DMEM containing 20% fetal bovine serum to stop digestion and then been centrifuged (800r, 5min) and suspended again. At last, the fibroblasts were separated from myocytes and other cells by selective plating for 1 hour. The isolated CFs were cultured in DMEM containing 10% fetal bovine serum, streptomycin(100U⁄ml) and penicillin(100U⁄ml). Cultured CFs of second to third passage were used for further experiments.

The identity and purity of CFs was assessed by morphology examination and immunostaining methods as described previously [24]. CFs should have a thin and triangular profile under light microscope and be positive for fibroblast marker vimentin, and negative for endothelial marker von Willebrand factor(vWF).

The hypoxia/serum deprivation injury cell model was used to mimic MI as we have previously described [25]. In brief, after indicated treatment, CFs were exposed to hypoxia (94% N₂ ‚ 5% CO₂ and 1% O₂) in an anaerobic system (Thermo Forma, USA) at 37 °C for indicated period of time in Hanks buffer.

The small interfering RNA targeted to Nrf2 (Nrf2−siRNA) and control siRNA (Scramble siRNA) were purchased from Gene-Pharma Company (Shanghai, China). The mouse Nrf2 siRNA sequence is 5'-UCCCGUUUGUAGAUGACA-3'(Nrf2, NCBI Reference Sequence: NM_010902.3). Cardiac fibroblasts were transfected with 100nM Nrf2-siRNA or Control siRNA with Lipofectamine 2000 (Invitrogen) for 24 hours according to the manufacturer' protocol. The silencing effects were confirmed by Western blot.

Transthoracic echocardiography (VEVO2100, VisualSonic, Canada) was performed by 3 independent investigators to evaluate cardiac function post operation as previously described [26]. In brief, mice were anesthetized with inhalation of isoflurane (2%). Images were captured with M-mode echocardiogaphy. Left ventricular end-diastolic volume (LVEDV), Left ventricular end-systolic volume (LVESV), LV ejection fraction (LVEF) and LV fraction shortening (LVFS) were measured by the use of computer algorithm.

Subcellular protein fractions of cardiac LV tissues and fibroblasts were extracted with nuclear and cytoplasmic protein extraction kit (Beyotime, Guangzhou, China) following manufacturer's instructions. Protein concentration quantitation was determined by the Bradford method. Proteins were separated via SDS-PAGE (12%) (Beyontime, Guangzhou, China) and then transferred onto NC membrane. After incubated with the following primary antibodies: anti-collagen I(ab34710, abcam, 1:5000), anti-collagen III(ab7778, abcam, 1:5000), anti-α-SMA(ab7817, abcam, 1:300), anti-PAI-1(ab66705, abcam, 1:1000), anti-TIMP-1(ab86482, abcam, 1:1000), anti-fibronectin (ab2413, abcam, 1:1000), anti-TGF-β1 (ab92486, abcam, 1:1000), anti-Smad3(ab40854, abcam, 1:1000), anti-P-Smad3(ab52903, Ser423/425, abcam, 1:1000), anti-Nrf2(ab31163, abcam, 1:1000), anti-HO-1(abl3243, abcam, 1:1000), overnight at 4°C, the membrane was then incubated secondary antibodies(PR-0255, ZhongshanJinqiao Biotechnology Co., 1:2000) for 1 hour, protein levels were detected using chemiluminescene system (Amersham Biosciense, Buchinghamshire, UK) and analyzed with Image J software.

Confocal dishes with CFs were washed with PBS three times and fixed in 4% paraformaldehyde for 20 min. After being permeabilized in 0.03% Triton X-100 for 10 min, the dishes were then blocked with goat serum for 1 h at room temperature. After that, CFs were incubated with primary antibodies at 4°C overnight. The next morning, secondary antibodies were incubated for 1 h and DAPI were counterstained for 10 min according to the manufacture's instructions. All images were captured with a digital confocal microscopy (FV-1000, Olympus, Japan).

The sensitive ROS fluorescent probe- DCFH-DA (KGT010-1; KeyGEN BioTECH, Nanjing, China) was used to detect the ROS production following the manufacturer's instructions. In brief, cardiac fibroblasts incubated with 5 µM DCFH-DA at 37°C for 30 min. The fluorescence was then detected on flow cytometer (Becton Dickinson Biosciences, Franklin Lakes, NJ).

The level of MDA, GSH, SOD activity in the supernatant of CFs were determined by commercially available assay kit (Nanjing Jiancheng Bioengineering Institute, China) in accordance with the manufacturer's protocols. All samples were stored at -80°C before assay and the amount of MDA, GSH, SOD activity in the samples were then measured by comparing the OD450 to the standard curves.

All data were analyzed using the GraphPad Prism 6.02 and expressed as mean ± standard deviation (SD). The differences between groups were subjected to one-way ANOVA or Student's two-tailed unpaired t test. The statistical significance was accepted at P < 0.05.

Western blot analysis of cardiac tissue in remote region revealed that collagen I, the major cardiac extracellular agents in fibrosis formation, was remarkably augmented after MI and approached the peak at 4 weeks. The content of collagen I at 4 weeks was approximately fivefold compared with that in sham-operated tissues (P< 0.05, Fig. 1 A, B). Moreover, content of collagen reflected by picrosirius red staining were notably higher in remote region of MI group as compared to sham group, which was attenuated by AM1241 treatment (fibrosis area: 18.37±1.12% in MI group vs. 11.40±0.66% in MI+AM1241 group, P< 0.05, Fig. 1 C, D). However, collagen content of sham group and sham+AM1241 group, infarct region of MI group and MI+AM1241 group did not differ. These results demonstrated that activation of CB2 receptor by AM1241 mainly ameliorated cardiac fibrosis in remote region after MI in mice.

In order to confirm the anti-fibrosis effect of AM1241 after MI at protein level, we used Western blot to examine the fibrosis associated proteins collagen I, collagen III, fibronectin, Plasminogen activator inhibitor(PAI)-1 and tissue inhibitor of metalloprotease(TIMP)-1. As shown in Fig. 1F-H, the expressions of cardiac collagen I, collagen III and fibronectin, the main components constituting cardiac extracellular matrix, were not significantly altered by AM1241 treatment in sham operated mice. However, significantly decreased expressions of collagen I, collagen III and fibronectin were observed in MI+AM1241group (P < 0.05 vs. MI). Meanwhile, PAI-1 and TIMP-1 which were markers of extracellular matrix degradation inhibitors, were also markedly increased in MI group as compared to sham group, whereas, this increase was partially reversed by AM1241 treatment (P < 0.05, Fig. 1 I, J). Taken together, these findings demonstrated that AM1241 inhibited synthesis and accelerated the degradation of extracellular matrix in infarcted hearts.

To evaluate the cardiac functional improvement of CB2 receptor agonist AM1241 in MI mice, we performed echocardiograph 4 weeks after MI. As shown in Fig. 2, AM1241 significantly increased LVEF, LVFS in MI+AM1241 group compared with that of MI group (EF: 60.04±5.93% vs. 42.87±3.16%, P < 0.05; FS: 26.17±1.88% vs. 21.64±1.03%, P < 0.05). Meanwhile, AM1241 treatment also decreased LVEDV, LVESV as compared to MI group (LVEDV: 0.32±0.01 ml vs. 0.38±0.02 ml, P < 0.05; LVESV: 0.12±0.01 ml vs. 0.19±0.01 ml, P < 0.05) These results indicated that CB2 receptor agonist AM1241 ameliorated cardiac fibrosis, which thereby partly prevented the loss of cardiac systolic and diastolic function.

To confirm the anti-fibrosis effect of AM 12 41 in vitro, we isolated cardiac fibroblasts(CFs) as described in methods section and used in vitro hypoxia/serum deprivation(H/SD) cell model to mimic in vivo MI as described in methods section. As shown in Fig. 3A, the majority of the cells of third passage had a thin and triangular profile under light microscope in accordance with description in previous reports. Next, we identified cultured CFs by immunofluorescent staining for CFs marker vimentin, and endothelial marker von willebrand factor (VWF). The results showed that >98% cultured cells were positive for vimentin and negative for VWF (Fig. 3B, C). To identify the proper dosage of AM1241 in vitro, α-smooth muscle actin (α-SMA), a fibroblast-myofibroblast transformation marker, was examined by Western blot in CFs subjected to H/SD injury. As shown in Fig. 3D, F, α-SMA expression elevated with the duration of H/SD injury within 12 hours and reached the peak at 12 hours. Next, the expression levels of α-SMA with different concentrations of AM1241 (0, 0.1, 1, 5, l0µM) treatment in CFs subjected to 12 hours' H/SD injury were detected. The results indicated that 0.1, 1, 5, l0µM AM1241 all decreased α-SMA expression level, of which 5µM AM1241 was the most effective one (P < 0.05, Fig. 3. E, G). Thus, the AM1241 dosage of 5 µM (mass concentration, 2.52 mg/L) and H/SD injury for 12 hours were chosen in the following experiments.

Previously, we and others have demonstrated that CB2 receptor activation could activate PI3K/Akt pathway [11,27], activation of Akt was shown to activate Nrf2, a well-known anti-oxidative protein which is frequently related to anti-fibrotic actions [20,28]. To achieve a better understanding of the possible signaling pathways and mechanisms implicated in the action of CB2 receptor activation, we detected the expression of CB2 receptor, Akt, Nrf2 and Nrf2 downstream heme oxygenase-1(HO-1) in CFs. Our data showed that H/SD could induce a mild elevation of CB2 receptor, phosphorylation-Akt(p-Akt), nucleus Nrf2, and HO-1(P < 0.05 vs. Control), the effect of which was further augmented by AM1241 treatment (P < 0.05 vs. H/SD, Fig. 3 H-J). However, treatment with a CB2 receptor selective antagonist AM630(1µM, Selleck, USA)reversed these effects of AM1241 (P < 0.05 vs. H/SD+AM1241, Fig. 3 H-M). These data suggested a pivotal role of CB2 receptor/Akt/Nrf2/HO-1 signaling pathway in the anti-fibrotic effects of AM1241.

A Nrf2 targeted small interfering RNA(siRNA) were used to determine the role of Nrf2 in the effects of AM1241 on H/SD induced cardiac fibroblast-to-myoblast transformation. The interfering efficiency of siRNA was determined in Fig. 4A, B, which indicated more than 60% down-regulation of Nrf2 protein level. As expected, AM1241 significantly decreased H/SD induced up-regulation of α-SMA, a common marker for fibroblast-to-myoblast transformation, which was partially reversed by Nrf2 siRNA (P < 0.05, Fig. 4 C, D). A similar trend was confirmed in immunofluorescent staining for α-SMA under confocal microscope (Fig. 4 E). Collectively, these data revealed that AM1241 suppressed the transformation of CFs under H/SD condition, in which Nrf-2 was strongly involved.

The cardiac fibroblasts were attracted to the site of infarct region to secrete collagen which helps scar formation by various chemotactic factors [29]. Therefore, we evaluated the effects of AM1241 on the secretory properties of CFs. The results illustrated that the up-regulated levels of collagen I and collagen III induced by H/SD injury were significantly decreased by AM1241, and Nrf2 siRNA partially abolished this trend (P < 0.05, Fig. 5 A-C). Next, to further confirm the result, we used the immunofluorescence staining to detect collagen I secretion in CFs. Unsurprisingly, AM1241 reduced the secretion of collagen I in H/SD+AM1241 group compared with that of the H/SD group, and Nrf2 siRNA partially abrogated this reduction (Fig. 5 D). These data demonstrated that AM1241 inhibited the collagen production properties of CFs under H/SD injury in an Nrf2 dependent manner.

Oxidative stress levels has been demonstrated to be closely related to fibrosis [30]. So we next sought to explore the effects of AM1241 on oxidative stress levels in CFs subjected to H/SD injury. Fluorescence intensity of ROS probe DCFH-DA was detected by flow cytometer in response to the levels of ROS. Fluorescence intensity enhanced by H/SD was significantly reduced by AM1241(P < 0.05, Fig. 6 A, B). N-acetylcysteine (NAC) was selected for its common use as a well-known antioxidant of ROS measurement. Our data showed that pre-treatment with NAC (5 mM) for 2 h could remarkably alleviate the ROS triggered by H/SD injury. Furthermore, Malondialdehyde (MDA), recognized as a marker for oxidative stress, is the product of oxidative free radical. Superoxide dismutase (SOD) and Glutathione (GSH) were ROS scavengers which protect cells from ROS damage. As shown in Fig. 6 C-E AM1241 treatment significantly decreased H/SD-induced MDA up-regulation (2.45±0.29 vs. 3.42±0.27 nmol/mg protein, P < 0.05) and increased H/SD-reduced GSH level and SOD activity (GSH: 17.96±l.90 vs. 9.62±1.84 nmol/mg protein, P < 0.05; SOD: 126.10±4.90 vs. 77.85±2.61 U/mg protein, P < 0.05). These findings suggested that the anti-fibrotic effect of AM1241 might be related to reduction in oxidative stress level, and promotion of anti-oxidative GSH and SOD activity.

It is well known that the TGF-β1/Smad3 pathway is a classic cell signaling pathway involved in the development of fibrosis. So we next sought to explore the effects of AM1241 on TGF-β1/Smad3 signaling pathway in CFs subjected to H/SD injury. Western blot assay revealed that the levels of TGF-β1, Smad3 phosphorylation significantly increased under H/SD conditions as compared to Control group (Fig. 7 A-C, P < 0.05). However, these increases were inhibited by AM1241. In contrast, silencing the Nrf2 by siRNA partially abolished the AM1241's effect. These data indicated that AM1241 inhibited TGF-β1/Smad3 pathway activation in CFs subjected to H/SD injury in an Nrf-2 dependent manner, Fig. 8.

In the present study, we demonstrated that cannabinoid II (CB2 receptor) receptor agonist AM1241 ameliorated cardiac fibrosis and enhanced cardiac functional recovery in myocardial infarcted mice. In cardiac fibroblasts subjected to H/SD injury, AM1241 inhibited cardiac fibroblast-to-myoblast transformation, collagen production including collagen I and collagen III in an Nrf2 dependent manner. Moreover, AM1241 reduced ROS generation and MDA production, and meanwhile increased anti-oxidative GSH level and SOD activity. Furthermore, AM1241 not only activated and accelerated the translocation of Nrf2 to nucleus, but also inhibited TGF-β1/Smad3 Pathway in an Nrf2 dependent manner.

Cardiac fibrosis and remodeling is one of the major pathological process during the post myocardial infarction phase (post-MI) which is characterized by the imbalance between extracellular matrix(ECM) synthesis and degradation. The overproduction of collagen due to dysregulation of repair inevitably deposited not only in infarct region but also in remote region, resulting in cardiac fibrosis [29]. Cardiac fibrosis curbed cardiac diastolic and systolic function severely which ultimately resulted in heart failure and death [2,31,32,33]. In our present study, we demonstrated that CB2 receptor agonist AM1241 ameliorated cardiac fibrosis, which involved in decreased synthesis and increased degradation of extracellular matrix. By retarding cardiac fibrosis, CB2 receptor agonist AM1241 provides incremental benefits for cardiac diastolic and systolic functional recovery. However, relieved cardiac fibrosis may not be the only reason for AM124's improvement of cardiac function post MI. In our previous study, we proved that CB2 receptor agonist AM1241 could activate cardiac progenitor cells and facilitate endogenous myocardial regeneration [11]. Furthermore, the direct cardio-protection of CB2 receptor agonist in ischemic heart has been well elucidated by a number of studies [34,35]. Thus, multiple mechanisms may be involved in CB2 receptor agonist mediated cardiac diastolic and systolic functional recovery in ischemic heart diseases.

Previous evidences have indicated the feasibility of CB2 receptor agonist in the treatment of tissue fibrosis. Munoz-Luque and colleagues reported that CB2 receptor activation reduced liver fibrosis in cirrhotic rats [6]. Moreover, CB2 receptor knockout mice was reported to exhibit increased dermal fibrosis in bleomycin-induced dermal fibrosis [7]. Not surprisingly, CB2 receptor-deficiency in mice reportedly leaded to interstitial fibrosis after repetitive cardiac ischemia/reperfusion injury [35]. In present study, CB2 receptor agonist AM1241 is proved to possess a significant anti-fibrotic effect in MI mice, which may represent a novel strategy to retard cardiac fibrosis in ischemic heart disease and other fibrosis associated heart disorders in a much broader context.

Cardiac fibroblasts (CFs) were thought to play a pivotal role in the pathogenesis of cardiac fibrosis. Under normal circumstances, cardiac fibroblasts (CFs) were protected by the ECM framework and produced minimal amounts of ECM. Once injury happened, fibroblasts were activated and differentiated into contractile myofibroblast characteristicly expressing α-smooth muscle actin (α-SMA) and exhibiting secretory properties [36,37]. Previous reports have proved that hypoxic stimuli increased the fibroblast-to-myofibroblast transformation and collagen synthesis of CFs [38,39]. In addition, another previous study has shown that exposure the primary fibroblasts from wild-type and CB2 receptor knockout mice to H₂O₂ for 15 min, the expression of α-SMA in CB2 receptor knockout fibroblasts were significantly higher than that of wildtype fibroblasts. However, JWH133, a CB2 receptor agonist, could inhibit α-SMA up-regulation [40]. Consistent with previous findings that certain environmental stimuli could promote the differentiation of CFs into myofibroblasts that drive aggressive remodeling of the ECM, we observed that simulated MI in vitro—H/SD stimuli remarkably increased the expression of α-SMA and enhanced the secretion of collagens that aggravate ECM remodeling. Moreover, we also found that CB2 receptor agonist AM1241 retarded cardiac fibrosis in MI mice by reversing the fibroblast-to-myofibroblast transformation and collagen synthesis of CFs. After MI, CFs quickly differentiated to myofibroblasts which exhibited increased fibrillar collagen synthesis, αSMA expression and capacity to aggravate cardiac fibrosis [37]. By reversing this process, CB2 receptor agonist provided incremental benefits to infarcted hearts—alleviating cardiac fibrosis as well as facilitating cardiac functional recovery.

Nrf2, a well-known anti-oxidative gene, has been reported to be closely related to cardiac fibrosis. Nrf2 reportedly inhibit cardiac fibrosis directly. Nrf2 knockout mice exhibited myocardial fibrosis significantly after transverse aortic constriction surgery, while the overexpression of Nrf2 dramatically inhibited myocardial fibrosis [41]. Moreover, as a crucial regulator of oxidative stress and reactive oxygen species(ROS) level, Nrf2 could reduce cardiac fibrosis through regulating ROS levels. Accumulating evidences already demonstrated that ROS generation promoted the development of cardiac interstitial fibrosis and reducing the overproduction of ROS contributed to ameliorated cardiac fibrosis [42,43,44]. Nrf2 was dissociated with Keapl following ROS generation and translocated into nucleus to bind the antioxidant response elements (AREs), then, initiated the transcription of antioxidant genes such as Heme oxygenase-1 (HO-1), superoxide dismutase (SOD), glutathione (GSH) to defense the ROS productions [17,45,46]. In line with these reports, we observed the down-regulated levels of ROS and MDA and up-regulated levels of GSH and SOD activity en route to the beneficial actions of CB2 receptor agonist AM1241 against cardiac fibrosis. Furthermore, by using an Nrf2 siRNA, we confirmed the anti-fibrotic effects of AM1241 on CFs in vitro were Nrf2 dependent. Thus, CB2 receptor activated Nrf2 signaling might reduce cardiac fibrosis through direct anti-fibrotic effects of Nrf2 and indirect down-regulation of ROS levels. In addition, we also observed that CB2 receptor, its downstream molecular Akt, nucleus Nrf2 and HO-1 slightly increased under H/SD condition. AM1241 treatment further enhanced the CB2 receptor/Akt/Nrf2 activation, the effect of which was inhibited by CB2 receptor antagonist AM630. These data suggested a pivotal role of CB2 receptor/Akt/Nrf2/HO-1 signaling pathway in the anti-fibrotic effects of AM1241.

TGF-β1/Smad3 Pathway, a canonical signaling pathway, has been proved by numerous reports to be involved in maintaining normal cardiac structure extracellular matrix (ECM) homeostasis and regulating cardiac fibrosis [14,39,47]. Once TGF-β1 was activated, the signals transmitted through phosphorylating a cytoplasmic protein called Smad3 into nucleus, thus regulating the fibrosis related gene expression, such as collagen I, collagen III, a-SMA, PAI-1 and TIMP-1 [12,14,15,39,48,49]. Interestingly, Nrf2 has been reported to attenuate dystrophic muscle fibrosis through inhibition of TGF-β1/Smad3 Pathway [50]. In our present study, we used siRNA to inhibit Nrf2 expression, and demonstrated that Nrf2 silencing significantly increased the expression of TGF-β1 and phosphorylation of Smad3. ROS have been shown to promote the development of fibrosis through activating latent TGF-β. ROS upregulated TGF-β1 expression in various ways, such as oxidizing latency-associated protein(LAP), activating MMPs and up-regulating Nox/Duox family, which in turn cleaved LAP to release more TGF-β1, thus forming a positive feedback cycle [51,52]. On the other hand, TGF-β1 was reported to directly promote reactive oxygen species (ROS) generation, while deletion of Smad3 abolished this effect of TGF-β, suggesting a pivotal role of TGF-β-Smad3 signaling in ROS generation [53]. Thus, there may be a crosstalk between TGF-β signaling and ROS generation is cardiac fibrogenesis. In our research, we also observed the co-activation of TGF-β signaling and ROS generation in CFs subjected to H/SD injury, while AM1241 treatment leaded to a concurrent decrease of both TGF-β signaling and ROS generation.

Despite the clinical relevance of our present study, there are still some limitations in our study. Firstly, the detailed mechanisms for how Nrf2 interacted with and inhibited TGF- β1/Smad3 pathway was not yet elucidated in the present study, which need further investigations. Secondly, in our current research, we observed that 10 µM AM1241 was not as effective as 5 µM AM1241. Yao B and colleagues have also reported that AM1241 exhibited partial agonist activity in the activation of p42/p44 ERK [54]. While some reports have denoted that activation of ERK inhibited progression of fibrosis, other reports indicated that ERK pathway was implicated in pro-fibrotic actions [55] and diabetic mouse heart [56]. In addition, TGF-β1 reportedly activated Erk in lung fibroblasts [57]. Thus, the overall effects of AM1241 on cardiac fibrosis might be ascribed to the complex interplay among ERK, Nrf2 and TGF-β1 signals. Whether higher concentrations of AM1241 in CFs might make a reversion of the anti-fibrotic actions and the underlying mechanisms remained to be investigated.

In conclusion, our work demonstrated a beneficial role of CB2 receptor activation in ameliorating cardiac fibrosis after myocardial infarction possibly through Nrf2 mediated inhibition of TGF-β1/Smad3 pathway. It is suggested that therapeutic strategy targeting CB2 receptor might be a potential approach to attenuate myocardial fibrosis induced by MI and other fibrosis associated heart disorders in a clinical perspective application.

This work was supported by the National Key Research Program of China (2016YFA0100903), National Funds for Distinguished Young Scientists of China (81325009) and National Nature Science Foundation of China (No. No.81270168, 81530058, 81570272), (FCao BWS12J037), Beijing Nature Science Foundation (No.7152131), Innovation Team Grant of Shaanxi Province (No.2014KCT-20).

None.

X. Li and D. Han contributed equally to this work and should be considered as co-first authors.