Treatment of Myocardial Infarction with Gene-modified Mesenchymal Stem Cells in a Small Molecular Hydrogel

The effect of transplanted rat mesenchymal stem cells (MSCs) can be reduced by extracellular microenvironment in myocardial infarction (MI). We tested a novel small-molecular hydrogel (SMH) on whether it could provide a scaffold for hepatocyte growth factor (HGF)-modified MSCs and alleviate ventricular remodeling while preserving cardiac function after MI. Overexpression of HGF in MSCs increased Bcl-2 and reduced Bax and caspase-3 levels in response to hypoxia in vitro. Immunocytochemistry demonstrated that cardiac troponin (cTnT), desmin and connexin 43 expression were significantly enhanced in the 5-azacytidine (5-aza) with SMH group compared with the 5-aza only group in vitro and in vivo. Bioluminescent imaging indicated that retention and survival of transplanted cells was highest when MSCs transfected with adenovirus (ad-HGF) were injected with SMH. Heart function and structure improvement were confirmed by echocardiography and histology in the Ad-HGF-SMHs-MSCs group compared to other groups. Our study showed that: HGF alleviated cell apoptosis and promoted MSC growth. SMHs improved stem cell adhesion, survival and myocardial cell differentiation after MSC transplantation. SMHs combined with modified MSCs significantly decreased the scar area and improved cardiac function.


Results
The effects of recombinant adenovirus and Ad-HGF on MSCs. The HGF level in MSCs was detected by Western blot (Figures S3). Annexin V/PI flow cytometry analysis was carried out 12 hours after hypoxia or normoxia. The percentage of apoptotic cells was lower in the Ad-HGF-MSCs group (37.83 ± 3.67%) compared to the Ad-EGFP-MSCs (52.87 ± 5.09%) or MSCs (53.07 ± 4.58%) group in hypoxia (P < 0.05). We observed a consistent trend in the 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay among the Ad-EGFP-MSCs, MSCs, and Ad-HGF-MSCs groups in hypoxia. Thus, Ad-HGF overexpression remarkably suppressed MSC apoptosis induced by hypoxia (Fig. 1A,B). HGF in the cell supernatant was detected by ELISA ( Figure S4). In future research, we will study in depth the mechanism by which HGF mediates apoptosis in MSCs.
MTT analysis showed that MSC growth and proliferation were normal in the normoxic group. However, the optical density (OD) value of the Ad-HGF-MSCs group was significantly increased compared with the Ad-GFP-MSCs and MSCs groups in hypoxia at 15 days (P < 0.05). In addition, no statistical difference was observed between the Ad-GFP-MSCs and MSC groups in hypoxia. The results indicate that HGF overexpression had a beneficial effect on cell proliferation under hypoxic conditions (Fig. 1C).
The significant increase in Bcl-2 level was observed in the Ad-HGF-MSCs group compared to the Ad-EGFP-MSCs and MSCs groups 12 hours after hypoxia or normoxia, and there was no significant difference between Ad-HGF-MSCs in hypoxia and normoxic MSCs. In contrast, Bax and caspase-3 levels were markedly increased in the hypoxic Ad-GFP-MSCs and MSCs groups compared to the normoxic MSCs group. We also found no significant difference in Bcl-2, Bax and caspase-3 expression between the Ad-GFP-MSCs and MSCs groups in hypoxia (Fig. 1D).
Proliferation, survival, and differentiation of MSCs in the presence SMHs. Cell apoptosis and necrosis were studied on days 5 and 7 in SMHs in vitro. AO/EB staining revealed that the MSCs cultured in SMHs were still highly proliferative as evidenced by a healthy nucleus with green fluorescence compared with red fluorescence under normal culture conditions. This result was confirmed by flow cytometry with Annexin V/PI staining. The number of apoptotic cells increased in the MSCs only group compared with the SMHs-MSCs group 12 hours after hypoxia ( Fig. 2A,B). The total Bcl-2 level in MSCs and SMHs-MSCs groups after hypoxia was significantly higher than that in control (MSCs hypoxia) group. The level of Bcl-2 in the hypoxic SMHs-MSCs group was higher than that in the hypoxic MSCs group (P < 0.05). There was an opposite trend with regards to the levels of Bax and caspase-3, such that protein expression decreased in the SMHs-MSCs group whereas their levels were increased in the MSCs group (P < 0.05). Therefore, SMHs can can induce changes in protein expression to minimize the effects of apoptosis (Fig. 2C).
Gene transcript expression levels of NKx2.5 and GATA-4 were significantly elevated in the SMHs-5-aza-MSCs group compared with the 5-aza-MSCs group after 2 weeks of incubation, as demonstrated by (Real-time Quantitative PCR Detecting System) qPCR (Fig. 2D). The expression of cardiac troponin T (cTnT) in the 5-aza + SMHs group increased significantly compared to the 5-aza only group. cTnT was not expressed in the PBS control group (without 5-aza). The MSCs induced by 5-aza in the SMHs formed clusters and generated cell junctions, which could be seen under phase contrast microscopy. cTnT and desmin were visible in the SMHs-5-aza-MSCs and 5-aza-MSCs groups, however, higher fluorescence intensity was observed in the SMHs-5-aza-MSCs group (Fig. 2D,E).
Image Pro Plus (IPP) 6.0 software analysis showed that higher blue (DAPI) and green fluorescence were found in the Ad-HGF-SMHs-MSCs and SMHs-MSCs groups compared to the Ad-HGF-MSC and other groups (P < 0.01, data not shown). SMHs significantly improved the retention and survival of Ad-HGF/MSCs after 16 days. The infarcted and surrounding area injected with Ad-HGF-SMHs-MSCs exhibited higher density of cardiac biomarkers including cTnT and Cx43 at 8 and 16 days (Fig. 4A). Some MSCs formed connections to the myocardium.
Non-invasive bioluminescent imaging was used to monitor the retention and survival of transplanted cells. Results indicated that the bioluminescent intensity of the Ad-HGF-SMHs-MSCs group was significantly increased compared with the Ad-HGF-MSCs group at day 2 (3.25 ± 0.26 × 10 6 vs. 0.84 ± 0.13 × 10 6 p/sec/cm 2 /sr; P < 0.05, n = 6) and at day 4 (2.72 ± 0.15 × 10 6 vs. 0.34 ± 0.08 × 10 6 p/sec/cm 2 /sr; P < 0.01, n = 6). Bioluminescent signal was also detected in the Ad-HGF-SMHs-MSCs and SMHs-MSCs groups from day 8 to day 12. In contrast, there was no signal observed in the Ad-HGF-MSCs group (Fig. 4B). Therefore, SMHs can promote stem cell retention and survival. MSCs-luc2+ bioluminescent in vitro imaging, correlation analysis, and quantitative data of photon emission for each animal and time-point are shown in Supplementary Material ( Figure S8).

Discussion
The novelty of this study was the use of SMHs in stem cell transplantation for MI treatment. Specifically, we demonstrated that SMHs provided scaffold materials for cardiac tissue engineering. We also showed that Ad-HGF transfected MSCs reduce apoptosis and improve resistance to oxidative stress in vivo. We further demonstrated   hydrogels may be less accommodating for cells embedded in these hydrogels to stretch compared with SMHs, and these cells have relatively lower viability and proliferative capacity 30,31 . Thus, challenges and limitations persist in the application of biomaterials.
Extracellular matrix (ECM) hydrogels are not eligible for tissue engineering as the result of low range stiffness relative to extremely high stiffness of myocardium. ECM hydrogels also have cytotoxic effects caused by glutaraldehyde. Therefore, ECM hydrogels may not be the best choice for cell-based application 32,33 . We designed SMHs and modified the properties, including elasticity, crystallinity, cross-linking density, and porosity of short peptide SMHs. This resulted in impactful results on MSC differentiation and engraftment capacity. Song et al. indicated that enhancement of cell therapy was attributed to mechanical stimulation that activated cell-surface receptors and subsequent signaling pathways 34 .
The SMHs used in this study was formed by the peptide D FEFK D FEFKYRGD. As our previous reported 35 , the peptide D FEFK D FEFKYRGD can form hydrogels which possess excellent histocompatibility and elasticity could be applied for the production of cell colony of HeLa cells. However, we found that the peptide D FEFK D FEFK (without RGD) could not form a clear solution in pure water at pH of 7.4. Consequently, it can not form hydrogels. Thus, we were unable to encapsulate MSCs. The peptide D FEFK D FEFKYRGD in the present study possesses an Arg-Gly-Asp (RGD) sequence. The RGD sequence is a motif in cell-surface integrins that facilitates ligand binding and can be used to modify the behavior of scaffold to promote cell adhesion and differentiation. Currently, an RGD-modified sequence is being used to treat inflammation, cardiovascular disease and metastatic tumors 36,37 . Previous studies have demonstrated that the RGD structural domain regulates integrin αvβ3 expression in the early stage of heart development, and promotes the differentiation of MSCs into cardiac muscle cells 38 . On the other hand, SMHs provide the chemical microenvironment to reduce toxicity of 5-aza. Our hydrogel SMHs was prepared without any chemical triggers and could be gelled at physiological conditions with thixotropic (time-dependent shear thinning effect) properties, allowing it to be considered an injectable. Histocompatibility and homeostasis tests showed that the injection of 1% SMHs did not lead to a severe inflammatory response. The in vivo biocompatibility and biodegradation tests of the hydrogel indicated that our hydrogel was suitable as a scaffold material for cardiac tissue engineering and regenerative medicine.

Ad-HGF transfected MSCs reduce apoptosis and improve oxygen resistance ability in SMHs.
HGF stimulates angiogenesis, limits scar formation, and reduces fibrosis [39][40][41] . HGF can reduce apoptosis of myocardial cells in vivo and in vitro 42 . In this study, adenovirus was chosen as the vehicle to to increase forced expression of HGF in MSCs. Consistent with previous studies 43 , we found that after 48 hours of Ad-HGF infection, HGF expression was increased at both mRNA and protein levels in MSCs, and reached a maximum value of 121.4 ng/ mL at 3 weeks. Madonna et al. reported that HGF could protect against cell apoptosis caused by many stimuli 44 . It is also well known that Bcl-2 and Bcl-xL play essential roles in anti-apoptosis. Bax apoptotic sub-family (Bax and Bak) contribute to their oligomer conformation, which responsible for permeabilizing membrane of the mitochondrion and resulting in apoptosis by activating executioner caspases. Thus, Bcl-2 and Bax are crucial in modulating apoptosis through upstream regulation of caspase-3 activation. Hypoxic-ischemic-induced apoptosis contributed to decreased expression of Bcl-2 and increased expression of Bax 45 . In our experiment, MSCs transfected with Ad-HGF after 12 hours of hypoxia showed overexpression of Bcl-2 and decreased expression of Bax and caspase-3. This result is consistent with previous reports 46 . In contrast, HGF-mediated c-MET activation increased angiogenesis and reduced fibrosis. The colon cancer 1 (MACC1) gene is an essential pro-metastatic factor in human colon cancer. MACC1 has been demonstrated to activate the HGF/c-MET pathway, which results in multiple cellular responses regulating proliferation, cell migration and ECM breakdown 47 . Notwithstanding, the mechanism by which HGF in MSCs ameliorates apoptosis and fibrosis, as well as increase angiogenesis, is still unclear and needs to be studied in the future.
Previous research has showed that endogenous HGF is blocked in occlusion-reperfusion studies, and, consequently, the infarct size and mortality rate are increased significantly. As a pharmacological delivery system, hydrogel or gelatin can improve the effect of the HGF released by stem cells 28,29 . The results of our study also indicate that SMHs work with HGF to exert excellent therapeutic effects on systemic disease requiring the participation of stem cells. SMHs have great potential for tissue engineering 48 and drug delivery 49 . Prior research reported that self-assembling peptide hydrogels encapsulated with curcumin could serve as injectable delivery vehicles 50 . To some extent, injectable SMHs provide sustained and controlled HGF release in the MI zone, as well as retention and efficacy, promoting reduced apoptosis and improved oxygen resistance ability.

SMHs carrying gene-modified MSCs improve MI.
We observed that SMHs mediated expression of apoptosis-related proteins. The MSC group showed an overexpression of Bcl-2, Bax and caspases-3 following exposure to hypoxia for 12 hours,indicating the presence of cell apoptosis [51][52][53] . In the SMHs-MSCs group, Bax and caspases-3 protein levels were downregulated, accompanied by an overexpression of Bcl-2. Thus, SMHs can improve cell survival and resistance to a hypoxic microenvironment. Notably, SMHs served as a 3D platform for cell culture, which regulated apoptosis as evidenced by Western blot, CCK-8 and MTT via following mechanisms: (1) Mechanical stimulation of 3D microenvironment activated intracellular signaling pathways to directly improve MSC survival; (2) binding of specific ligands in SMHs with surface receptors on MSCs to regulate downstream signaling pathways, and (3) exertion of paracrine effects on cell survival mechanisms under anoxic conditions. NKx2.5, GATA-4 and desmin are the early differentiation markers in cardiac progenitor cells 54,55 . Cx43, a gap junction protein, ensures electrical and metabolic coupling between cardiomyocytes and coordinates their contractility 56 . Yang et al. found that genes indicative of MSC differentiation into cardiomyocytes induced by 5-aza was verified by qPCR, Western blot and immunohistology 57,58 . In our study, the expression of NKx2.5 and GATA-4 was up-regulated in SHMs-5-aza-MSCs compared to the 5-aza MSCs control. Moreover, specific myocardial proteins, such as cTnT and Cx43, were expressed at higher levels in SHMs-5-aza-MSCs compared with the 5-aza MSCs control in vitro, and in Ad-HGF-SMHs-MSCs compared with others in vivo. The TUNEL results demonstrated that SMHs protected against cell apoptosis induced by MI. As evidenced by immunostaining, we found that SMHs improved cell retention and survival. DAPI and GFP fluorescence were stronger in Ad-HGF-SMHs-MSCs compared to PBS-Ad-HGF-MSCs. Additionally, we hypothesized that MSCs in PBS were vulnerable in the heart but SMHs provided a support to protect the cells.
Echocardiographic analysis showed that the Ad-HGF-SMHs-MSCs group had increased LV wall thickness, systolic function, and LV chamber dimension compared with the PBS and Ad-HGF-MSCs control groups. Most of the angiogenesis and microvessels were found in the peri-infarcted and adjacent areas in the Ad-HGF-SMHs-MSCs group 4 weeks after transplantation. We propose that this result is due to the following mechanisms: (1) SMHs serving as cell scaffold to provide the microenvironment that is similar to in vivo conditions; (2) enhanced tenacity of SMHs affecting the elasticity of scar tissue by increasing the thickness of infarcted tissue to improve tissue compliance and heart function 59 ; (3) SMHs ameliorating angiogenesis to increase cell survival; (4) RGD in SMHs interacting with MSCs to protect the cell from anoikis 60 ; (5) SMHs stimulating heart development-related genes in the early phase mediated by RGD; and (6) SMHs controlling Ad-HGF release, therefore prolonging the effect of HGF to improve MSCs survival, differentiation and angiogenesis.
Limitations and future research. Although our study confirmed the apparent value of SMH therapy, This study has a number of limitations with Ad-HGF-MSCs in experimental MI, the pathways by which SMHs affect MSCs proliferation, survival, and differentiation are still unknown. In the current literature, most studies have focused on studying the properties of biomaterial and its apparent effects, but there is little focus on the underlying mechanisms. In order to determine the mechanisms of SMHs' systematic effect on different cells, we will devote efforts to study the relative effect of peptide reactive group biomaterial on cell surface protein kinases or antigenic determinants in our future research.

Summary
HGF can effectively reduce cell apoptosis and promote growth and proliferation of MSCs. Chemical and physical factors provided by the SMHs promote cell growth, proliferation and differentiation. Delivery of MSCs by SMHs improved the microenvironment of transplanted stem cells. A multitude of mechanisms associated with SMHs favors stem cell adhersion, survival and direction differentiation of myocardial cells, leading to reduced scar area, enhanced angiogenesis and an improvement in myocardial function.

Materials and Methods
Preparation of SMHs. Hydrogels were produced through non-covalent interactions (π -π, hydrogen bond, ionic bond, etc.). At Nankai University, we previously designed a novel small molecule, D FEFK D FEFKYRGD, which can form hydrogels in physiological environments 35 . The hydrogel characterization is shown in the Supplement (Figs S2-4). The peptide was prepared by solid phase peptide synthesis (SPPS). 10 mg SPPS was dissolved in 0.4 mL H 2 O 2 , then the pH was adjusted to 7.4 by sodium carbonate, and mixed with 0.5 mL phosphate buffered saline (PBS). The vessel was placed upside down to promote formation of a jellylike colloid. The examination of hydrogel biocompatibility were conducted in previous work 31 . MSC origin and maintenance. MSCs were obtained from Cyagen Biosciences Inc, (USA) and cultured in DMEM supplemented with 10% FBS, streptomycin (100 U/mL), and penicillin (100 U/mL). Detection of markers CD34, CD45, CD11b/c was done by flow cytometry analysis and multi-potential differentiation. All MSCs used in this study were used by passage 5 (P5). MSC culture and passage were performed according to instructions from Cyagen Biosciences Inc. MSCs in P5 were cultured for 7 days before injection. 5-aza (4 μmol/L) was added to DMEM on day 4 and incubated with MSCs for 24 hours. All MSCs were then cultured in serum-free DMEM. The respective groups were cultured in anoxic conditions for 12 hours: 94% N 2 , 1% O 2 and 5% CO 2 . All groups were fasted for 12 hours and given water before surgery. Rats were anesthetized using 10% chloral hydrate (0.3 mL/100 g) and prior to tracheal intubation. ' After thoracotomy, muscle layers were separated, pericardial membrane was removed, and the left auricle was exposed. The left coronary artery, as tied off about 2 mm below the left auricle using a 6-0 silk suture at a depth of 0.5 mm to form the experimental MI. The same procedure was followed in sham group except the artery was not tied. ST elevation was detected by RM6240BD-ECG. Penicillin was administered for three days post-surgery. Seven days post-surgery, thoracotomy and tracheal intubation were carried out again. MSCs were incubated with DAPI for 15 min before injection. The MSCs (5 × 10 6 ) were suspended in SMHs and injected around the area of MI with a 29-gauge needle. Bioluminescent imaging, histological and immunohistochemical staining. Non-invasive bioluminescent imaging was used to measure MSC retention. The rats in the MSC group and therapy groups were anaesthetized by 10% chloral hydrate, underwent coronary artery ligation, and were injected with MSCs or Ad-HGF-MSCs infected with lentiviral-luciferase2 (luc2) with or without SMHs near the MI zone (5 × 10 6 , 150 µl, n = 8). The rats were administrated 150 µl D-luciferin (15 mg/mL) by intraperitoneal injection. The signal was captured by imaging technology (Night OWL II LB983 NC100, Berthold) and analyzed after 1 to 14 days.

Transfection of MSCs and cell proliferation and viability analyses. Lentivirus infection and reporter
To assess the degree of apoptosis in the peri-infarct areas, the left ventricle (LV) tissue was sampled 2 weeks post-surgery. LV samples were fixed in 4% paraformaldehyde, embedded with liquid paraffin, and cut into 5 µm thick sections. The sections were treated according to instructions in the One Step TUNEL Apoptosis Assay Kit (Takara). Samples were imaged using a fluorescence microscope. Enumeration was performed counting the number of positive cells in five views per sample and five slides per group.
LV samples were fixed in 10% neutral buffered formalin, embedded with paraffin, and sectioned. Myocyte cross-sectional area was used to evaluate the proportion of MI in the LV. Briefly, sections were deparaffinized with xylene and stained with H&E. Sections were stained with a Masson trichrome kit (Baso, BA4079) according to the manufacturer's instructions. The stained sections were examined under a light microscope (Nikon ECLIPSE Ti-U). The area of the MI, cross sectional area of the LV chamber, and the thickness of the wall were measured by Image Pro Plus 6.0 (IPP, Media Cybernetics, Carlsbad, CA).
Frozen sections of rat heart (n = 6) were fixed in acetone and perfused with PBS twice, followed by permeabilization with 0.1% Triton X-100 and blocked with normal goats serum for 20 min. Cardiac troponin (cTnT) and connexin 43 (Cx43) antibodies (SantaCruz, USA) were diluted in blocking solution (l:500). Cy5-labeled goat anti-mouse IgG (l: 500, Chemicon, USA) was used for incubation in the dark. We observed positive cells by determining the green and red fluorescence using fluorescence microscopy.
Evaluation of cardiac function. At 4 weeks post-surgery, rats were weighed and anesthetized with chloral hydrate before examination. Transthoracic two-dimensional guided M-mode echocardiography was conducted using an IE33 echocardiographic system (Philips Medical Systems) outfitted with the 15-MHz (s12) transducer, as we previously described 61 . The parameters of the systolic function fractional shortening (FS) and LV ejection fraction (EF) were obtained by the improved Simpson method. Hemodynamic parameters were measured according to the following protocol. The animals were anesthetized with 10% chloral hydrate solution on the same day that the cardiac ultrasonography examination was performed. The right common carotid was separated and the distal end was ligated with thread while the proximal end was occluded. The multichannel animal physiological signal acquisition processing system was applied to record the LVP curve LVEDP, and +dp/dtmax was also recorded by the PowerLab system. Quantitative real-time qPCR and Western blotting. Cell RNA was extracted by Trizol. The relative expression level of NKx2.5 and GATA-4 was analyzed by real-time PCR. We used SYBR Green combined with DNA emitted fluoresce. Data were analyzed using the 2 −ΔΔCT to calculate the CT value. Relative protein expression level of Bcl-2, Bax and caspase-3 was analyzed by Western blot. We chose the Bio-Rad Dc Protein Assay Reagent to determine protein concentration. Briefly, protein samples were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidene difluoride (PVDF) membranes. PVDF membranes were probed with cTnT antibody (Abcam, 1:300), Bcl-2 antibody (Invitrogen, 1:500), caspase-3 (Invitrogen, 1:500), and Bax (Invitrogen, 1:500). The antibody-antigen complexes were detected using the ECL PLUS kit (Bio-Rad). The protein bands were analyzed with Image Pro Plus 6.0.

Statistical analysis.
Values are expressed as means ± standard error of the mean (SEM). One-way analysis of variance was used to analyze variations between the means of groups; SPSS 17.0 (SPSS Inc., Chicago, IL, USA) was utilized for analysis. Significant values are defined as P < 0.05.
A post hoc t-test (Newman-Keuls) was conducted in the method of S-N-K following the one-way ANOVA in SPSS.