MicroRNA-377 Regulates Mesenchymal Stem Cell-Induced Angiogenesis in Ischemic Hearts by Targeting VEGF

MicroRNAs have been appreciated in various cellular functions, including the regulation of angiogenesis. Mesenchymal-stem-cells (MSCs) transplanted to the MI heart improve cardiac function through paracrine-mediated angiogenesis. However, whether microRNAs regulate MSC induced angiogenesis remains to be clarified. Using microRNA microarray analysis, we identified a microRNA expression profile in hypoxia-treated MSCs and observed that among all dysregulated microRNAs, microRNA-377 was decreased the most significantly. We also validated that vascular endothelial growth factor (VEGF) is a target of microRNA-377 using dual-luciferase reporter assay and Western-blotting. Knockdown of endogenous microRNA-377 promoted tube formation in human umbilical vein endothelial cells. We then engineered rat MSCs with lentiviral vectors to either overexpress microRNA-377 (MSCmiR-377) or knockdown microRNA-377 (MSCAnti-377) to investigate whether microRNA-377 regulated MSC-induced myocardial angiogenesis, using MSCs infected with lentiviral empty vector to serve as controls (MSCNull). Four weeks after implantation of the microRNA-engineered MSCs into the infarcted rat hearts, the vessel density was significantly increased in MSCAnti-377-hearts, and this was accompanied by reduced fibrosis and improved myocardial function as compared to controls. Adverse effects were observed in MSCmiR-377-treated hearts, including reduced vessel density, impaired myocardial function, and increased fibrosis in comparison with MSCNull-group. These findings indicate that hypoxia-responsive microRNA-377 directly targets VEGF in MSCs, and knockdown of endogenous microRNA-377 promotes MSC-induced angiogenesis in the infarcted myocardium. Thus, microRNA-377 may serve as a novel therapeutic target for stem cell-based treatment of ischemic heart disease.


Introduction
The formation of new blood vessels is critical for the repair of ischemic myocardium, and VEGF is one of the most extensively characterized angiogenic factors [1]. While direct administration of VEGF into the ischemic myocardium has been used successfully to stimulate therapeutic angiogenesis in animal models, clinical trials of VEGF have been largely unsuccessful [2], [3]. These results underscore our incomplete knowledge of myocardial angiogenesis under ischemic conditions.
During the past decade, it has been demonstrated that MSCs can facilitate new blood vessel growth by secretion of proangiogenic factors (e.g. VEGF, IGF-1a, HGF, etc.) that contribute to cardiac repair and enhance the reparative process [4][5][6]. MSCs are, however, highly sensitive to ischemic conditions, and the majority of injected MSCs die within several hours of delivery in vivo [7]. In this regard, multiple approaches (e.g. hypoxic treatment, genetic modification, and pre-conditioning) have been applied to MSCs in an effort to improve their survival and proangiogenic capacity both in vivo and in vitro [8]. Although hypoxia is well recognized to promote MSC-mediated myocardial angiogenesis by induction of VEGF expression [9], [10], the underlying mechanisms underlying these effects have not been delineated.
In this study, we first sought to determine the alterations of miRs in MSCs under hypoxic conditions in rats. We then identified which hypoxia-inducible miRs could directly regulate VEGF expression and tested whether manipulation of such miR in MSCs could affect MSC-induced angiogenesis in ischemic myocardium. Our results show for the first time that miR-377   In vitro tube formation assay indicated that knockdown of miR-377 enhanced the formation of capillary-like structures, but this effect was limited by miR-377 overexpression. Scale bars = 500 mm. (C): Total capillary tube lengths and tube branch points were measured by analytical software Image-Pro Plus 6.0 (IPP). All values were expressed as means 6 SE; n = 6 independent experiments for each group; *P,0.05. doi:10.1371/journal.pone.0104666.g002 was strongly down-regulated in hypoxia-treated MSCs, which was a major factor contributing to the increased VEGF levels. Thus, injection of miR-377-knockdown MSCs into an infarcted myocardium reduced overall infarction size and improved contractile function by promoting angiogenesis. These findings suggest that miR-377 may serve as a novel potential therapeutic target for treatment of ischemic heart diseases.

Animal Experiments
All research protocols conformed to the Guidelines for the Care and Use of Laboratory Animals published by the National Institutes of Health (National Academies Press, 8 th edition, 2011). All animal use protocols and methods of euthanasia (pentobarbital overdose followed by thoracotomy) used in this study were approved by the University of Cincinnati Animal Care and Use Committee. The Institutional Biosafety Committee (IBC) conducted an independent review and approval of our cell and virus methods.

In Vitro Studies
In order to culture MSCs, they were extracted from Sprague-Dawley (SD; 8-wk-old male) rats following previously published procedures from our laboratory [9]. MSCs were then cultured in Dulbecco's Modified Eagle Medium(DMEM) supplemented with 10% (v/v) fetal bovine serum (FBS) and antibiotics (100 U/mL penicillin and 100 mg/mL streptomycin). The cells were kept in a humidified 5% CO 2 incubator at 37uC and culture medium was changed after 3 days. Non-adherent cells were removed by changing the medium and the remaining adherent cells were primary MSCs. Passage 2-4 MSCs were used in this study.
Hypoxic MSCs were cultured in DMEM without glucose and with 1% FBS under hypoxic conditions of 1% O 2 , 5% CO 2 and 94% N 2 at 37uC in a hypoxic incubator (O 2 /CO 2 incubator-

RNA Extraction and RT-PCR
Total RNA from the MSCs was extracted using the Trizol reagent (Invitrogen, Carlsbad, Calif., United States), as recommended by the manufacturer. Total RNA concentrations were determined by NanoVue plus (GE Healthcore, Piscataway, New Jersey, USA). The mRNA levels of VEGF and miRs were examined by reverse transcription-polymerase chain reaction (RT-PCR) or quantitative real-time PCR (qPCR), and b-Actin or U6 was used as an internal reference. The primers for VEGF and b-Actin were designed as follows: VEGF forward: 59-GCAACAC-CAAGTCCGAATGCAGAT-39, reverse: 59-TCTGGCTTCA-CAGCACTCTCCTTT-39; b-Actin forward: 59-TGTGATGGTGGGAAT GGGTCAGAA-39, reverse: 59-TGTGGTGCCAGATCTTCTCCATGT-39. The primers for miRNA and U6 were purchased from QIAGEN. The primers for miRs consisted of a specific primer (Rn_miR-377_2 miScript Primer Assay) and a universal primer (106 miScript Universal Primer). The amplification profiles for PCR: 94uC 5 min., followed by 30 cycles of 94uC 30 sec., 55uC 30 sec., 72uC 45 sec., and a final 5 min. extend; and for qPCR: 95uC 15 min., followed by 40 cycles of 94uC 15 sec., 55uC 30 sec., 72uC 30 sec. with 0.5uC/15 sec. in 55uC,95uC. PCR products were analyzed with 1.5% agarose gel. The qPCR expression of VEGF mRNA relative to b-Actin under experimental and control conditions was calculated based on the threshold cycle (Ct) as n = 2 2D (DCt) , where DCt = Ct VEGF 2 Ct b-Actin and D (DCt) = DCt experimental 2 DCt control. Individual experiments were repeated at least 3 times, and the n-mean value was calculated.

Western-Blotting Analysis
Protein samples were collected from MSCs treated under different conditions, and 60 mg of protein was loaded and subjected to SDS-PAGE, as described previously [6]. PageRuler TM Plus Prestained Protein Ladder (Thermo Scientific Inc., MA, USA) was loaded as a protein marker to estimate molecular Western blot assay showed protein level changes of VEGF in MSCs induced by miR-377 mimic (miR-377) and miR-377 inhibitor (Anti-377) as well as its quantitative data. All values were expressed as means 6 SE; n = 8 for each group; *P,0.05 was considered statistically significant. (C): qPCR analysis after normalization against b-actin showed significant up-regulation of VEGF in hypoxia-treated MSCs in comparison with normoxia-treated MSCs, which was further increased in MSC Anti-377 , when compared with NC group and miR-377 groups. All values were expressed as means 6 SE; n = 8 for each group. *P,0.05 was considered statistically significant. doi:10.1371/journal.pone.0104666.g004 MicroRNA-377 Regulates Angiogenesis in Ischemic Hearts PLOS ONE | www.plosone.org weight of samples. A VEGF antibody (Rabbit, 1:500) was purchased from Santa Cruz Biotechnology. b-Actin (mouse 1:1000) was purchased from Santa Cruz Biotechnology.

miRNA Array Analysis and Target Prediction
Total RNA samples obtained from MSCs under normoxia or hypoxia were sent to LC Sciences (Houston, TX) for miRNA microarray profiling. Data was analyzed by LC Sciences with inhouse developed computer programs. Intensity values were transformed into log2 scale, and fold changes were given in log2 scale. A t-test was performed between normoxic MSCs and hypoxic MSCs, and statistical significance was considered at P, 0.01. The microarray data were confirmed using an miRNA detection protocol with RT 2 miRNA First Strand Kit (SA biosciences). Computational miRNA target prediction analysis was performed with TargetScan (version 6.2) and miRDB to predict potential binding between VEGF 39UTR and miRNA.

Dual-Luciferase Reporter Assay
Rat VEGF 39-UTR (nt1660-3545) was inserted into the Dual-Luciferase reporter vector (pEZX-MT01, Genecopoeia Corp. MD, USA) downstream from the Firefly luciferase (hLuc) reporter gene, and was driven by SV40 Enhancer promoter. In addition, Renilla luciferase (hRLuc) reporter driven by a CMV promoter was cloned into the same vector, serving as the tracking gene and internal control. The dual-reporter vector system enabled transfection-normalization for accurate across-sample comparison. The 293TN cells were assigned into three groups to be transfected with A. pEZX-MT01 vector; B. pEZX-MT01 vector+NC-miR-377 mimic; C. pEZX-MT01 vector+miR-377 mimic. Cell lysates were collected and assayed 48 hours after transfection. Firefly and Renilla luciferase activities were measured using a Dual Luciferase Reporter Assay System kit (Promega Corp. WI, USA) and each transfected well was assayed in triplicate as described [27]. The mutated pEZX-MT01 plasmid containing the mutated VEGF-39UTR with mutation in the seed region was synthesized using Phusion TM site-directed mutagenesis kit (New England Biolabs. MA, USA) with the following primer, mutated VEGF 39UTR forward primer 59-AAGGATAAAATAGACATTGC-TATTCTG-39; reverse primer 59-AGACTATATACATAAACA-TATATATATATATATACAC-39.

In Vitro Tube Formation Assay
HUVECs were purchased from American Type Culture Collection (ATCC) and cultured in endothelial cell growth medium (Cell Application). HUVECs were transiently transfected with A. negative control (NC miR /NC Anti ); B. miR-377 mimic; C. miR-377 inhibitor; D. miR-377 inhibitor+VEGF siRNA. After 48 h, in vitro tube formation assay was performed with a tube formation assay kit (Chemicon), per the manufacturer's instructions. Briefly, ECMatrix Solution was thawed on ice for 1,2 hours, then was mixed with 106ECMatrixdilutent (v: v = 9:1). The mix was added to a 96-well tissue culture plate (50 ml/well) and was placed at 37uC for 1 hour to allow the matrix solution to solidify. HUVECs were digested by 0.125% trypsin and were placed (1610 4 cells/well) on top of the solidified matrix solution and incubated at 37uC for 18 hours. Cellular network structures were fully developed and photos were taken using an inverted light microscope at 406 magnification. Total capillary tube length and tube branch points were measured using analytical software Image-Pro Plus 6.0 (IPP, Media Cybernetics, Carlsbad, CA). Tube formation was defined as a structure exhibiting a length four times its width. Five independent fields were assessed for each well, and the average number of tubes was calculated [28].

Surgical Procedures for the LAD Occlusion and MSC Implantation
An MI model was developed in SD rats (200-250 g), as described previously [29]. Briefly, isofluorane anesthesia was induced by spontaneous inhalation. The animals were mechanically ventilated with room air supplemented with oxygen (1.5 L/ min) using a rodent ventilator (Model 683; Harvard Apparatus, South Natick, MA). Body temperature was carefully monitored with a probe (Cole Parmer Instrument, Vernon Hill, IL) and was maintained at 37uC throughout the surgical procedure. The heart was exposed by left side limited thoracotomy, and the left anterior descending coronary artery (LAD) was ligated with a 6-0 polyester suture 1 mm from the tip of the normally positioned left auricle. MSCs (30 ml, 2610 6 ) were injected into border area of the left ventricle (LV) wall at 10 minutes after LAD ligation. The chest was closed with 5-0 silk sutures. Approximately 10% of rats succumbed during surgical procedures.

Immunohistochemical Analysis
Immunohistochemical studies were performed on heart tissue at 4 weeks after cell implantation. Heart tissue sections were harvested, fixed in 10% Formalin, and sectioned at 5-mm thickness. The cardiac troponin T (cTnT) antibody (Thermo Scientific) was used to identify cardiomyocytes, while 49, 6diamino-2-phenyindole (DAPI, Sigma) was used to identify nuclei. Von Willebrand Factor (vWF, DAKO, Agilent Technologies) rabbit polyclonal antibody (Santa Cruz Biotechnology) and asmooth muscle actin (SMA) mouse monoclonal antibody (Sigma) were used to assess capillary and vascular density. Fluorescence labeled secondary antibodies (Jackson Immuno Research Laboratories or Molecular Probes) were used following these primary antibodies. Fluorescent imaging was performed with an Olympus BX41 microscope (Olympus America Inc., Melville, NY, U.S.A.) equipped with epiflourescence attachment, and images were recorded using a digital camera with MagnaFire 2.1 software.

Measurement of Infarct Size
Fixed hearts were embedded in paraffin, and sections from apex, mid-LV, and base were stained with Masson's Trichrome. An Olympus BX41 camera was used to obtain images of LV area on each slide using MagnaFire (Olympus) software. Fibrosis and total LV area of each image were measured using Image J software, and the percentage of the fibrotic area was calculated as shown: (fibrosis area/total LV area) 6100, as previously described [9].

Assessment of Heart Function
Heart function was assessed by transthoracic echocardiography, which was performed at 4 weeks after MI using iE33 Ultrasound System (Phillips) with a 15-MHz probe. After rats were anesthetized with pentobarbital sodium (40 mg/kg) by intraperitoneal injection, hearts were imaged two-dimensionally in longaxis view at the level of the greatest LV diameter. This view was used to position the M-mode cursor perpendicular to the LV anterior and posterior walls. The LV end-diastolic diameters (LVDd) and LV end-systolic diameters (LVDs) were measured from M-mode recordings according to the leading-edge method. LV parameters were obtained from two-dimensional images. LV ejection fraction (EF) was calculated as: EF (%) = [(LVDd) 3 -(LVDs) 3 ]/(LVDd) 3 6100. Fractional shortening (FS) was measured using the equation FS (%) = [(LVDd 2 LVDs)/LVDd] 6100. All echocardiographic measurements were averaged from at least three separate cardiac cycles.

Statistical Analysis
Experiments were performed in quadruplicate and repeated at least three times. Data are expressed as means 6 SE. Statistical significance was assessed by one-way ANOVA followed by Bonferroni/Dunn testing. P,0.05 was considered statistically significant.

Reduced Expression of miR-377 in HUVECs Promotes Angiogenesis
HUVECs were transfected with either miR mimic to overexpress miR-377 or miR inhibitor to specifically knockdown miR-377 ( Fig. 2A) in order to determine the significance of miR-377 in angiogenesis, followed by an in vitro tube-formation assay using Matrigel-precoated wells. HUVECs transfected with either negative control miR mimic (NC miR ) or negative control miR inhibitor (NC Anti ) was used as negative control groups. The negative control (NC miR and NC Anti ) groups exhibited some tubelike shapes and half-full cellular networks, while the miR-377 mimic group (miR-377 group) revealed less tube-like structures and hardly formed cellular networks (Fig. 2B). However, the miR-377 inhibitor group (Anti-377 group) displayed the formation of full and dense cellular networks (Fig. 2B). The cumulative capillary tube length, measured using IPP software, was reduced by 4066% in miR-377-HUVECs, whereas it was increased by 5465% in Anti-377 cells when compared to negative controls (NC Anti ) (Fig. 2C). No significance was observed between NC miR group and NC Anti group. In addition, the number of tube branch point in miR-377 group (19.562.2) was less than that of negative controls (NC miR 36.062.9; NC Anti 38.063.0 respectively), and was significantly increased in Anti-377 group (55.063.5) (Fig. 2C).

MiR-377 Acts Directly at the 39UTR of VEGF
Computational miRNA target prediction analysis was performed to elucidate the potential mechanism of miR-377 in the regulation of angiogenesis, using TargetScan and miRDB. VEGF-A (usually referred to as VEGF) is listed among the top of assumed targets for rno-miR-377 and the seed sequence of VEGF 39UTR interacting with rno-miR-377 is highly conserved among the species of rat, human, chimpanzee, rhesus, bushbaby, treeshrew, and mouse (Fig. 3A). HEK293TN cells were transfected with a Dual-Luciferase reporter vector containing the 39-UTR of VEGF or mutated 39-UTR of VEGF fused downstream to the Luciferase coding sequence (Fig. 3B) along with miR-377 mimic or a NC (NC miR ) to validate whether miR-377 directly recognizes the 39UTR of VEGF. Luciferase activity was repressed by 67% when miR-377 was co-expressed with the VEGF-39-UTR luciferase reporter vector (Fig. 3C), whereas luciferase activity of mutated VEGF-39-UTR was not affected. In contrast, transfection with NC did not affect the activity of luciferase.
Negative Effects of miR-377 in Angiogenesis are Largely Dependent on VEGF It is important to determine whether miR-377 reduction-caused angiogenesis is dependent on VEGF given that VEGF is a target for miR-377. The expression of VEGF was knocked down in miR-377-reduced HUVECs by siRNA, followed by in vitro tube formation assay. Similar to previous findings (Fig. 2), it was observed that miR-377-inhibitor-transfected HUVECs had welldeveloped networks of capillary-like tubes, evidenced by a 1.5-fold increase of cumulative tube length and 1.45-fold increase of tube branch points, compared with NC miR cells. In contrast, HUVECs co-transfected with miR-377 inhibitor+VEGF siRNA exhibited sparse capillary-like structures in which the tube length and the number of tube branch points are similar to NC group ( Fig. 5A and 5B). These results indicate that enhanced effects in the formation of tube-like structures induced by knockdown of miR-377 were abolished by inhibition of VEGF expression.

Reduced Expression of MiR-377 in MSCs Limits Fibrosis and Improves Contractile Function in Infarcted Hearts
Fibrosis area was evaluated using Masson's Trichrome staining in which normal myocardium was colored red while fiberized myocardium was blue in color due to its inner collagen. The percentage of fibrosis in the left ventricle wall was significantly reduced in the MSC Anti-377 group, but was increased in MSC miR-377 -implanted hearts, as compared with MSC Null (Fig. 8A-B).  (Fig. 8C-D). However, no significant differences were noted in these parameters between MI and MI+PBS groups.

Discussion
Angiogenesis is a key regenerative event in ischemic injured hearts after MI, and VEGF plays an important role in MSCinduced angiogenesis [1]. Given that miRNAs are endogenous regulators of gene expression, it is reasonable to hypothesize that miRNAs may be involved in the regulation of VEGF expression in MSCs. We therefore employed hypoxia, a well-established VEGF inducer, to pre-treat MSCs and determined the miRNA expression profile. The results showed that miR-377 expression was decreased by more than 2-fold in hypoxia-treated MSCs as compared with normoxic condition. By computational miRNA target prediction analysis, we identified VEGF as a potential target of miR-377. Furthermore, both dual-luciferase reporter assay and Western-blotting verified that miR-377 can directly bind with VEGF 39UTR leading to negatively regulation of its expression. Accordingly, in vivo by transplanting MSCs with genetic overexpression or knockdown of miR-377 in the rat MI hearts, we observed that myocardial angiogenesis was significantly improved in MSC Anti-377 -treated hearts, whereas it was poor in MSC miR-377 -treated hearts when comparable to MSC Null -injected hearts. It is important to note here, while the degree of fibrosis was less in MSC Anti-377 -treated myocardium than MSC Null -injected group, there were no significant changes in MSC miR-377 -treated group when compared to controls. Consistent with the alteration of myocardial fibrosis, cardiac function was significantly improved in the MSC Anti-377 -treatedgroup, but there were no obvious changes in MSC miR-377 -implanted hearts as compared with MSC Null -injected hearts. This may be interpreted that myocardial angiogenesis is reduced in MSC miR-377 -treated hearts, but not enough to affect MSC-induced beneficial effects on the reduction of fibrosis and improvement of function in infarcted hearts. However, numerous studies have indicated that an increase in myocardial angiogenesis improves contractile function in the infarcted myocardium [10], [32], [33].

Conclusion
In conclusion, our study indicates thathypoxia-reducedmiR-377 directly targets VEGF, and knockdown of endogenous miR-377 promotes MSC transplantation-induced angiogenesis and subsequent heart function improvement post MI. These data may suggest a new therapeutic strategy for ischemic heart disease treatment in the future.