Pterostilbene Attenuates Fructose-Induced Myocardial Fibrosis by Inhibiting ROS-Driven Pitx2c/miR-15b Pathway

Excessive fructose consumption induces oxidative stress and myocardial fibrosis. Antioxidant compound pterostilbene has cardioprotective effect in experimental animals. This study is aimed at investigating how fructose drove fibrotic responses via oxidative stress in cardiomyocytes and explored the attenuation mechanisms of pterostilbene. We observed fructose-induced myocardial hypertrophy and fibrosis with ROS overproduction in rats. Paired-like homeodomain 2 (Pitx2c) increase, microRNA-15b (miR-15b) low expression, and p53 phosphorylation (p-p53) upregulation, as well as activation of transforming growth factor-β1 (TGF-β1)/drosophila mothers against DPP homolog (Smads) signaling and connective tissue growth factor (CTGF) induction, were also detected in fructose-fed rat hearts and fructose-exposed rat myocardial cell line H9c2 cells. The results from p53 siRNA or TGF-β1 siRNA transfection showed that TGF-β1-induced upregulation of CTGF expression and p-p53 activated TGF-β1/Smads signaling in fructose-exposed H9c2 cells. Of note, Pitx2c negatively modulated miR-15b expression via binding to the upstream of the miR-15b genetic loci by chromatin immunoprecipitation and transfection analysis with pEX1-Pitx2c plasmid and Pitx2c siRNA, respectively. In H9c2 cells pretreated with ROS scavenger N-acetylcysteine, or transfected with miR-15b mimic and inhibitor, fructose-induced cardiac ROS overload could drive Pitx2c-mediated miR-15b low expression, then cause p-p53-activated TGF-β1/Smads signaling and CTGF induction in myocardial fibrosis. We also found that pterostilbene significantly improved myocardial hypertrophy and fibrosis in fructose-fed rats and fructose-exposed H9c2 cells. Pterostilbene reduced cardiac ROS to block Pitx2c-mediated miR-15b low expression and p-p53-dependent TGF-β1/Smads signaling activation and CTGF induction in high fructose-induced myocardial fibrosis. These results firstly demonstrated that the ROS-driven Pitx2c/miR-15b pathway was required for p-p53-dependent TGF-β1/Smads signaling activation in fructose-induced myocardial fibrosis. Pterostilbene protected against high fructose-induced myocardial fibrosis through the inhibition of Pitx2c/miR-15b pathway to suppress p-p53-activated TGF-β1/Smads signaling, warranting the consideration of Pitx2c/miR-15b pathway as a therapeutic target in myocardial fibrosis.


Introduction
Fructose overconsumption increases oxidative stress, inflammation, and cardiomyocyte hypertrophy, causing myocardial fibrosis [1,2]. Transforming growth factor-β1/(small) mothers against decapentaplegic homologs (TGF-β1/Smads) signaling is known to mediate the pathological process of fibrosis. Its activation is observed in left ventricle tissues of Western diet-fed mice with myocardial fibrosis [3]. TGF-β1 activates the promoter of connective tissue growth factor (CTGF) to induce its expression in rat primary cardiac myocytes, in parallel with myocardial infarction in rats and cardiac ischemia patients [4]. Activation of Smad3/4 is essential for TGF-β1-induced CTGF transcription in rat proximal tubular epithelial cells with the progression of tubulointerstitial fibrosis [5]. Consistently, high levels of hyaluronic acid, hydroxyproline, and collagen volume fraction (important indicators in clinical diagnosis of myocardial fibrosis) are observed in heart tissues from hypertrophic cardiomyopathy patients [6][7][8]. Furthermore, mRNA expression levels of cardiac hypertrophy-related genes such as atrial natriuretic peptide (ANP), brain natriuretic peptide (BNP), and beta myosin heavy chain (β-MHC) and fibrosis-related genes such as collagen I, collagen III, CTGF, and TGF-β1 are also increased in aortic banding-induced experimental cardiac hypertrophy and fibrosis [9]. Of note, TGF-β1, alpha smooth muscle-actin (α-SMA), and fibroblast specific-1 (FSP-1) are increased in mouse hearts and the primary cardiomyocytes during fructose-induced myocardial fibrosis [10]. How high fructose intake causes myocardial fibrosis and its possible pathological mechanism are still unknown.
Recent study shows that the microRNA-15 (miR-15) family acts as a novel regulator of cardiac hypertrophy and fibrosis by inhibiting TGF-β pathway [11]. Early downregulation of miR-15b precedes the activation of profibrogenic mediators and then accelerates fibrotic remodeling in the hearts of type-2 diabetic mice [12]. Moreover, a bioinformatics approach predicts that osteoblastic specific miR-15b targets 16 genes in the tumor suppressor p53 signaling pathway [13]. Interestingly, p53 as a highly labile transcription factor is increased in myocardial biopsies of patients with heart disease [14]. Heart overexpression of p53 and TGF-β1 is also detected in high oxygen-exposed rats with cardiomyocyte hypertrophy and enhanced fibrosis [15]. A specific p53 inducer, indoxyl sulfate, can enhance p53-TGF-β1/Smad3 pathway in kidney fibrosis of rats [16]. However, it is unclear whether miR-15b regulates p-p53 to activate TGF-β1/Smads signaling in fructose-induced myocardial fibrosis.
Oxidative stress and reactive oxygen species (ROS) overproduction cause myocardial damage during the progression of myocardial fibrosis [17]. p53 acts as a finely tuned regulator of redox-dependent physiological processes [18]. Redox regulation by paired-like homeodomain transcription factor 2 (Pitx2) affects cardiac structure and function [19,20]. Pitx2 promotes heart repair by activating antioxidant response after cardiac injury [20]. Surprisingly, Pitx2c is reported to negatively regulate miR-15b expression in cell proliferation of myoblasts [21]. In fact, fructose induces oxidative stress in myocardial fibrosis of rats [17]. The possible molecular mechanism by which fructose affects Pitx2c via oxidative stress to dysregulate miR-15b in myocardial fibrosis needs to be explored.
In this study, we investigated whether fructose induced Pitx2c to negatively regulate miR-15b in myocardial fibrosis and examined what the molecular basis could be. Our findings demonstrated that fructose-induced ROS overload increased Pitx2c to downregulate miR-15b expression, then activated p-p53-dependent TGF-β1/Smads signaling, causing CTGF-mediated myocardial fibrosis. Pterostilbene with antioxidation downregulated Pitx2c to upregulate miR-15b and reduced p-p53 to suppress TGF-β1/Smads signaling activation and CTGF expression in the attenuation of fructose-induced myocardial fibrosis.
Pterostilbene (25,50, and 100 mg/kg) is reported to ameliorate cardiac oxidative stress, hypertrophy, and right ventricle systolic dysfunction in monocrotaline-induced pulmonary hypertension of rats [26]. It also alleviates myocardial ischemia/reperfusion injury of rats at 10 mg/kg [24]. Allopurinol, with antioxidative activity, is clinically used to treat some cardiovascular diseases [29,30]. Our previous studies showed that pterostilbene (10, 20, and 40 mg/kg) or allopurinol (5 mg/kg) reduced fructose-induced oxidative stress and inflammation in the heart, liver, or kidney of fructose-fed rats [17,[31][32][33]. Additionally, allopurinol restores a high-fat and high-fructose diet-induced myocardial oxidative stress, cardiomyocyte hypertrophy, interstitial fibrosis, and left ventricular diastolic dysfunction in mice [34] as well as ventricular relaxation impairment and cardiac ischemia in rats [35]. Thus, 10, 20, and 40 mg/kg were selected as the dosages of pterostilbene administration, and 5 mg/kg allopurinol was the positive control in this study. Rat body weight was measured weekly. Animal welfare and experimental procedures were carried out in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals prepared by the National Academy of Sciences and published by the National Institutes of Health (NIH publication 86-23 revised 1985).

2.2.
Collection of Blood and Tissue Samples. During the last feeding week, animals were anesthetized according to a previously described protocol [17]. Serum samples were collected by centrifugation (4000 × g, 4°C) for 10 min and stored at -80°C for biochemical assays. Heart tissue samples were rapidly dissected on ice and stored at -80°C for microarray, biochemical, qRT-PCR, and Western blot assay, respectively, while some of which were fixed for histological study.
2.3. miRNA Microarray Analysis. A microarray-based approach was used to identify miRNA expression difference in plasma samples between normal and fructose-fed rats [36]. The microarray analysis for miRNA profiling using the miRCURY LNA Array system (Exiqon, Vedbaek, Denmark) was conducted by the KangChen Bio-tech Inc. (Shanghai, China). The threshold value for significance used to define upregulation or downregulation of miRNAs was a fold change > 1:5 or <0.6. Here, circulating levels of miR-15b showed a relatively obvious downtrend with fold change of 0.5641 (fructose vehicle vs. normal control P value = 0.0505) (Table S1).
2.6. Determination of Hydroxyproline and Hyaluronic Acid Levels in Rat Hearts. Rat heart tissues were homogenized in sodium chloride (10 wt/vol) on ice and then centrifuged (10000 × g, 4°C) for 15 min to obtain the supernatants. Hydroxyproline and hyaluronic acid levels were measured by standard diagnostic kits (Jiancheng Biotechnology Co., Ltd., Nanjing, China), respectively.

Fluorescence In Situ Hybridization (FISH) for miR-15b
Detection in Rat Hearts. miR-15b-FISH detection was performed according to a previously described protocol with some modifications as listed below [11]. miR-15b probe was synthesized (Wuhan Servicebio Technology Co., Ltd., Wuhan, China), and then, the FISH Tag RNA Green Kit with Alexa Fluor 488 was used (Invitrogen, Burlington, ON, Canada). The sequence of Rat-miR-15b probe for in situ hybridization was 5′-TGTAA ACCAT GATGTGCTGC TA-3′.
Nuclear and cytoplasmic staining for detecting miRNA precursors and mature miRNA was carried out in cardiomyocytes, the images of which were obtained using an upright microscope (Nikon ECLIPSE CI, Nikon, Japan).
H9c2 cells were incubated in serum-free DMEM for 12 h. These cells were pretreated with ROS scavenger Nacetylcysteine (NAC, 1 mM, Amresco, Solon, USA) for 1 h and then coincubated with 5 mM fructose in the presence or absence of pterostilbene (10 μM) or allopurinol (30 μM) for further 24 or 48 h.
H9c2 cells were transfected with pEX1-Pitx2c plasmid (1 μg/mL) or Pitx2c siRNA (50 nM) (GenePharma) for 6 h and then used to determine miR-15b expression for another 24 h. These sequences were listed in Tables S2 and S3. The transfection efficiency ( Figures S6 and S7) in H9c2 cells was detected by qRT-PCR, respectively.
Pterostilbene and allopurinol were dissolved in DMSO, while NAC was dissolved in ultrapure water at the respective stock concentrations. The concentrations of DMSO in all cell cultures were less than 0.1%. Cell culture supernatants and lysates were collected separately. Total RNAs or proteins were extracted and stored at -80°C for further biochemical, qRT-PCR, or Western blot analysis, respectively.
2.9. Determination of Oxidative Stress. Determination of oxidative stress was performed according to the previously described protocol [17,40]. Total ROS levels were measured using a commercial kit (Beyotime Institute of Biotechnology, Haimei, China). Activity of NADPH oxidase was represented as the rate of NADPH consumption as previously described [40]. Malondialdehyde (MDA) levels were determined by a standard diagnostic kit (Jiancheng Biotechnology Co., Ltd., Nanjing, China).
2.10. Immunofluorescence (IF) Analysis. Heart samples were snap frozen in prechilled isopentane (fill a glass beaker with isopentane and keep on dry ice) and stored at -80°C. The frozen samples were cut into 8 μm thick sections. The IF assay for rat heart tissue and H9c2 cell samples was performed according to the previously described protocols with minor modifications [41]. Briefly, except ANP used paraffinembedded sections of the heart herein, frozen sections of the heart and cells were fixed using 4% formaldehyde, then incubated overnight at 4°C with a polyclonal anti-rabbit ANP (ab76743, dilution 1 : 200), a polyclonal anti-rabbit α-SMA (ab5694, dilution 1 : 200), a monoclonal anti-mouse SPG3A (namely, FSP-1, ab58273, dilution 1 : 100), a polyclonal anti-rabbit CTGF (ab125943, dilution 1 : 200) (Abcam, Cambridge, MA, USA), and a polyclonal anti-rabbit TGF-β1 (sc-146, 1 : 500; Santa Cruz, CA, USA), respectively. The secondary antibodies conjugated to an Alexa Fluor 555 goat anti-rabbit IgG, 555 goat anti-mouse IgG, and 488 goat anti-rabbit IgG (A-21428; A-21424; A-11008; Life Technologies, Oregon, USA) were added and incubated for 0.5-1 h further at room temperature. After rinsing, the slides were incubated with DAPI for 8 min for nucleus staining. Images were acquired by a Multiphoton Confocal Microscope (Lei TCS SP8-MaiTai MP, State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University, Nanjing, China) and processed with Photoshop software (Adobe, San Jose, CA, USA).
2.11. Chromatin Immunoprecipitation (ChIP) Assay. ChIP assay was performed as previously described using an Epi-Quik™ ChIP Kit (EpiGentek, New York, USA) with the following modifications [42]. H9c2 cells were transfected with 10 μg pEX1-Pitx2c plasmid using Lipofectamine 2000 in a 75 cm 2 culture flask. After 6 h, cells were swapped by Opti-MEM® I Reduced-Serum Medium (Invitrogen) for another 24 h and then cross-linked with 1% formaldehyde for 10 min at room temperature on a rocking platform, lysed, and sonicated. 1% of the sheared DNA-protein complex was used for an input DNA sample. Soluble chromatin was incubated for 90 min at room temperature with anti-Pitx2 (ab192495; Abcam, MA, USA) and anti-RNA polymerase II as a positive control or normal mouse IgG as a negative control (supplied by the kit), respectively. All PCRs were performed at an annealing temperature of 55°C. Different primers were used to amplify the DNA regions containing the Pitx2c binding site 6 kb upstream of the coding sequences for miR-15b (Table S4). Normal mouse IgG exhibited nonspecific immunoprecipitation with chromatin. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) used as a positive control showed the efficacy of the kit reagents and protocols ( Figure S8); its primers were illustrated in Table S4. Three parallel qRT-PCRs were performed in triplicate with dilutions of input DNA to determine the linear range of amplification.
2.12. RNA Isolation and qRT-PCR Analysis. Total RNA was isolated from rat heart tissues and H9c2 cells using TRIzol reagent (Invitrogen). The primers were synthesized by Shanghai Generay Biotech Co., Ltd. (Shanghai, China), and the sequences were listed in Table S2.
For mRNA qRT-PCR analysis, single-stranded cDNA was reverse transcribed from RNA using oligo-dT primers. For miRNA qRT-PCR analysis, miRNA was extracted with specific stem loop from RNA by reverse transcription. Reverse transcription was conducted using a HiScript® II Q RT SuperMix (Vazyme, Nanjing, China) and iQ™ SYBR® Green Supermix (Bio-Rad Laboratories, Hercules, CA, USA), respectively.
The qRT-PCR analysis was performed according to our previously described protocol [17]. The relative expression of mRNAs was normalized to GAPDH, while that of miR-NAs was normalized to U6, respectively.
2.14. Statistical Analysis. All data were analyzed with oneway analysis of variance (ANOVA) and followed by Tukey's Multiple Comparisons Test. Data were presented as the mean ± S:E:M:, and P < 0:05 was considered as statistically significant.

Pterostilbene Reduces Fructose-Induced Cardiac ROS to Block Pitx2c-Mediated miR-15b Low Expression in
Cardiomyocytes with the Suppression of p-p53-Activated TGF-β1/Smads Signaling. Pitx2c mRNA and protein levels were significantly increased in fructose-fed rat hearts (Figures 6(a) and 6(b)) and fructose-exposed H9c2 cells compared with the control group (Figures 6(c) and 6(d)). As shown in Figures 6(e) and 6(f), low miR-15b expression displayed significantly after Pitx2c overexpression, whereas Pitx2c siRNA caused miR-15b high expression in H9c2 cells compared with the negative control cell group, supporting negative regulation of miR-15b expression by Pitx2c.
Next, we screened potential conserved Pitx2c binding sites upstream of the miR-15b genetic loci. Target genes in the most enriched bins were further analyzed for the presence and the evolutionary conservation of Pitx2c consensus binding sequence, TAATCY (namely, TAATCC or TAAT CT), on the -20 kb, intronic, and coding gene sequences [43]. Three conserved Pitx2c binding sites were identified -6 kb upstream of the miR-15b genetic loci. The ChIP assay showed that exogenous Pitx2c bounds to the all-putative binding sites upstream of the miR-15b genetic loci compared with the IgG-negative control group in H9c2 cells (Figure 6(g)). Additionally, miR-15b mimic or miR-15b inhibitor failed to alter fructose-induced Pitx2c overexpression in H9c2 cells compared with the fructose-vehicle cell group (Figures 6(h) and 6(i)), suggesting the involvement of Pitx2c/miR-15b pathway in myocardial cells under high fructose induction.
With respect to myocardial injury, cardiac biomarkers CK, CK-MB, MB, and cTn-T levels in serum were significantly increased in fructose-fed rats, being consistent with the previous reports [45,46]. Fructose induced-cardiomyocyte hypertrophy and fibrosis were also observed in rats. Importantly, this study showed that p-p53 upregulation drove TGF-β1/Smads signaling activation in the hearts and cardiomyocytes under fructose stimulation. It is noteworthy that overexpression of p53 and TGF-β1 is detected in the heart of high oxygen-exposed rats with cardiomyocyte hypertrophy and fibrosis [15]. These results indicate that p-p53 upregula-tion in myocardial cells may induce myocardial fibrosis by activating TGF-β1/Smads signaling.
Osteoblastic specific miR-15b is predicted to target 16 genes in p53 signaling pathway identified by the bioinformatics approach [13]. In this study, fructose was found to induce a significant downregulation of miR-15b expression in plasma and heart of rats as well as in myocardial cells, which were consistent with the recent reports in plasma and heart of diabetic patients and mice [12]. In case of fructose-induced myocardial fibrosis, there was evidence of miR-15b low expression-mediated p-p53 to activate TGF-β1/Smads signaling. High fructose triggered miR-15b low expression, which may be a risk factor in myocardial fibrosis. Continuous monitoring of miR-15b levels in plasma may be considered as a noninvasive biomarker for cardiac fibrotic remodeling identification in patients.
Recent study has revealed the negative regulation of miR-15b by Pitx2c on the transcription level [21]. Pitx2c is significantly reactivated in the left ventricular myocardium of patients with systolic heart failure [47] and myocardial injury In situ hybridization showed miR-15b downregulation ((b), green) in the heart of fructose-fed rats (scale bar 50 μm). DAPI was used for staining nuclei, and merged views were shown in the right panels. miR-15b expression was measured in rat hearts (c) and H9c2 cells (d) by qRT-PCR analysis (n = 6). The relative miR-15b expression levels were normalized to U6. Data are expressed as the mean ± S:E:M: ## P < 0:01, ### P < 0:001 vs. normal animal control group or normal cell control group; * P < 0:05, * * P < 0:01, and * * * P < 0:001 vs. fructose-vehicle animal group or fructose-vehicle cell group.   Figure 5: Pterostilbene and allopurinol increase miR-15b expression to suppress p-p53-dependent TGF-β1/Smads signaling activation in fructose-exposed H9c2 cells. Cellular levels of p-p53 protein (a) and miR-15b expression (b) were determined in p53 siRNA-transfected H9c2 cells coincubated with 5 mM fructose, 10 μM pterostilbene, and 30 μM allopurinol (n = 6). The relative miR-15b expression levels were normalized to U6. Cellular p-p53 protein levels were determined in miR-15b mimic (c) or miR-15b inhibitor (d)-transfected H9c2 cells coincubated with 5 mM fructose, 10 μM pterostilbene, and 30 μM allopurinol, respectively. Protein levels of TGF-β1, p-Smad2/3, and Smad4 were determined in these cells ((e-h) miR-15b mimic; (i-l) miR-15b inhibitor) (n = 6), respectively. Relative protein levels of p-p53 were normalized to p53, of TGF-β1 and Smad4 were normalized to GAPDH, and of p-Smad2/3 were normalized to Smad2/3, respectively. Data are expressed as the mean ± S:E:M: # P < 0:05, ## P < 0:01, and ### P < 0:001 vs. normal cell control group; * P < 0:05, * * P < 0:01, and * * * P < 0:01 vs. fructose-vehicle cell group or fructose-vehicle+p53 siRNA or miR-15b mimic or miR-15b inhibitor control cell group.  Figure 6: Continued. [20]. High Pitx2c expression is detected in atrial myocytes or atrial appendages in chronic atrial fibrillation patients [48,49]. Here, we showed significant Pitx2 overexpression in fructose-induced myocardial fibrosis of rats and cells. We hypothesized that targeting Pitx2c and miR-15b might alleviate myocardial fibrosis under fructose induction. We further supported Pitx2c-mediated negative regulation of miR-15b expression in fructose-stimulated myocardial cells. More importantly, exogenous Pitx2c bounds to the all-putative binding sites upstream of the miR-15b genetic loci. These results provide an evidence that Pitx2c/miR-15b pathway possibly mediates fructose-induced myocardial fibrosis. Pitx2 has been demonstrated to promote heart repair by activating antioxidant response [20]. High-fat and highfructose diet increases myocardial ROS production and oxidative stress, cardiomyocyte hypertrophy, interstitial fibrosis, and left ventricular diastolic dysfunction in mice [34]. Our previous study showed that fructose-induced ROS was a major initiator of myocardial damage in rats with myocardial fibrosis [17]. In this study, ROS inhibitor prevented fructoseinduced Pitx2c upregulation and miR-15b low expression, as well as p-p53-activated TGF-β1/Smads signaling in myocardial cells. These observations indicate that fructose-triggered cardiac ROS may be the primary step driving Pitx2c upregulation to reduce miR-15b expression, providing a novel mechanistic insight into the link between oxidative stress and myocardial fibrosis. Thus, ROS-driven Pitx2c/miR-15b pathway was required for p-p53-dependent TGF-β1/Smads signaling activation in myocardial fibrosis. Pitx2c may be a novel prominent target in the development of fructoseinduced myocardial fibrosis for further studies.
fructose. The abnormal expression of CTGF could be employed as a diagnostic marker for fructose-induced myocardial hypertrophy and fibrosis.
The safety of pterostilbene has been confirmed in humans (up to 250 mg/day), without statistically significant major adverse drug reactions [55]. Based on our findings, the inclusion of blueberries and grape vines in the diet, especially offering a broader range of available pharmacological compound pterostilbene in subjects, may reduce the risk factors associated with myocardial diseases, comprising a more efficacious therapeutic option. On the other hand, the novel findings of this study may be of importance contributing to understanding the potential beneficial effects of allopurinol treatment in myocardial fibrosis.

Conclusion
In conclusion, the results from this study demonstrate that Pitx2c may be a novel factor participating in fructoseinduced myocardial fibrosis. High fructose-triggered cardiac ROS may be the primary step accounting for Pitx2c upregulation to reduce miR-15b expression and then activating TGF-β1/Smads signaling in CTGF-mediated myocardial fibrosis ( Figure S13). Pterostilbene and allopurinol with antioxidant activity downregulate Pitx2c and upregulate miR-15b and then suppress p-p53-dependent TGF-β1/Smads signaling activation to reduce CTGF, resulting in the alleviation of fructose-induced myocardial fibrosis. Thus, this study suggests that inhibition of Pitx2c-mediated miR-15b pathway by pterostilbene and allopurinol may provide a novel therapeutic strategy for myocardial fibrosis associated with excess fructose consumption.