Hepatocyte Growth Factor or Vascular Endothelial Growth Factor Gene Transfer Maximizes Mesenchymal Stem Cell–Based Myocardial Salvage After Acute Myocardial Infarction
Background— Mesenchymal stem cell (MSC)-based regenerative strategies were investigated to treat acute myocardial infarction and improve left ventricular function.
Methods and Results— Murine AMI was induced by coronary ligation with subsequent injection of MSCs, hepatocyte growth factor (HGF), vascular endothelial growth factor (VEGF), or MSCs +HGF/VEGF into the border zone. Left ventricular ejection fraction was calculated using micro–computed tomography imaging after 6 months. HGF and VEGF protein injection (with or without concomitant MSC injection) significantly and similarly improved the left ventricular ejection fraction and reduced scar size compared with the MSC group, suggesting that myocardial recovery was due to the cytokines rather than myocardial regeneration. To provide sustained paracrine effects, HGF or VEGF overexpressing MSCs were generated (MSC-HGF, MSC-VEGF). MSC-HGF and MSC-VEGF showed significantly increased in vitro proliferation and increased in vivo proliferation within the border zone. Cytokine production correlated with MSC survival. MSC-HGF– and MSC-VEGF–treated animals showed smaller scar sizes, increased peri-infarct vessel densities, and better preserved left ventricular function when compared with MSCs transfected with empty vector. Murine cardiomyocytes were exposed to hypoxic in vitro conditions. The LDH release was reduced, fewer cardiomyocytes were apoptotic, and Akt activity was increased if cardiomyocytes were maintained in conditioned medium obtained from MSC-HGF or MSC-VEGF cultures.
Conclusions— This study showed that (1) elevating the tissue levels of HGF and VEGF after acute myocardial infarction seems to be a promising reparative therapeutic approach, (2) HGF and VEGF are cardioprotective by increasing the tolerance of cardiomyocytes to ischemia, reducing cardiomyocyte apoptosis and increasing prosurvival Akt activation, and (3) MSC-HGF and MSC-VEGF are a valuable source for increased cytokine production and maximize the beneficial effect of MSC-based repair strategies.
Mesenchymal stem cell (MSC)-based regenerative strategies for treatment of acute myocardial infarction (AMI) are currently under investigation. Two clinical studies using autologous MSC therapy have been published,1,2 and both demonstrated the safety of intracoronary MSC infusion and an improvement in overall left ventricular (LV) function. However, the optimal MSC source (bone marrow versus adipose tissue), the ideal application method (intracoronary versus intramyocardial), and the underlying mechanisms are unclear. Although initially aimed at regeneration, MSC therapy may limit maladaptive remodeling and improve heart function mainly through paracrine mechanisms in the absence of regeneration. Due to the stressful procedure of cell harvest causing morphological changes with loss of spindle-shaped fibroblast-like shape and intercellular connections and shear stress during needle aspiration and cell injection, combined with a harsh milieu in the ischemic heart of the host, substantial cell losses are common early after cell transplantation.3 Despite these significant losses, beneficial effects on cardiac remodeling have been consistently observed, and these data suggest that even a small cell mass could have a profound impact on their local environment if they secrete key bioactive mediators.4
In this study, we examined the paracrine effects of MSC therapy to define mechanisms of benefit and optimize the recovery of function with this emerging therapy. We previously identified 2 cytokines that are produced by acutely ischemic myocardium and exert strong proproliferatory and promigratory effects on MSCs5: hepatocyte growth factor (HGF) and vascular endothelial growth factor (VEGF). Prolonged MSC survival using these factors could amplify the proposed regenerative effect of these cells. We tested different cytokine-based or MSC-based strategies for their potential to improve myocardial recovery.
MSC Isolation, Purification, and Labeling
Balb/c mouse MSCs were isolated from long bones as described previously.6 MSCs were characterized using flow cytometry as CD34−, CD45−, CD90+, sca-1+, c-kit−, CD14−, CD44+, and CD105+.6 Cells were transfected to express firefly luciferase (fLuc) for in vivo IVIS imaging.7,8
Myocardial Infarction Model
Mice weighing 20 to 25 g were purchased from Harlan (Indianapolis, Ind) and underwent myocardial infarction through left anterior descending (LAD) artery ligation as previously described.9 All study groups comprised of 6 animals.
Mice received 60 μL of cell suspension with 0.5×106 MSCs, 60 μL of HGF, (40 ng/mL) of VEGF (8 ng/mL), or MSC with HGF or VEGF immediately after the LAD ligation. Using thoracotomy access, multiple injections were performed within the presumed infarct and the border zone.
Bioluminescence In Vivo Imaging
For bioluminescence imaging (BLI), the reporter probe D-Luciferin (375 mg/kg) was injected intraperitoneally, and animals were imaged using the IVIS 200 system (Xenogen, Hopkinton, Mass). Bioluminescence was quantified in units of maximum photons per second per centimeter squared per steradian (p/s/cm2/sr), as described earlier.10
Small-Animal X-Ray Computed Tomography
Micro–computed tomography (CT) was performed for postinfarction LV imaging and to calculate the left ventricular ejection fraction (LVEF). For details, please see the online-only Data Supplement.
Generation of Murine HGF and VEGF Overexpressing MSCs (MSC-HGF and MSC-VEGF)
The technique for HGF/VEGF gene transfer is described in the Data Supplement. HGF and VEGF production was assessed in the cell culture supernatant of MSC-HGF, MSC-VEGF, or MSCs transfected with an empty vector (MSC-EV), after 24 hours of cell culture. Rabbit–anti-mouse antibodies against HGF (Santa Cruz Biotechnology, Santa Cruz, Calif) and VEGF (Calbiochem, San Diego, Calif) were used with an ECL Western Blot System (Amersham, Piscataway, NJ). Integrated densities were calculated using National Institutes of Health imageJ 1.41 (Bethesda, Md).
MSC Proliferation Assay
In vitro proliferation of MSCs grown with or without HGF 40 ng/mL or VEGF 8 ng/mL and of MSC-HGF and MSC-VEGF was determined. MSCs were seeded in 24-well plates (3000 cells per well) and grown in medium with or without HGF 40 ng/mL or VEGF 8 ng/mL. MSC-HGF and MSC-VEGF were cultured without additional cytokines. Daily cell counts were determined using the CellTiter 96 Cell Aqueous One Solution Proliferation Assay (Promega, Madison, Wis). Absorbance at 490 nm was measured with the Magellan ELISA Reader and Software (Tecan Systems Inc, San Jose, Calif). Triplets were performed and expressed as percent cell proliferation compared with 100% on day 0.
Histological Determination of Infarct Size
Histological slides were stained with Masson trichrome, and computerized morphometry was used to measure scar size. Hearts were fixed in 4% paraformaldehyde solution and embedded in paraffin an. Masson trichrome was performed according to the manufacturer’s protocol (DBS, Pleasanton, Calif). Serial sections were performed to identify the midinfarct region. Computerized morphometry was used to measure the length of the scar as well as of the entire circumferences. The scar extent was calculated as the ratio of scar and ventricle sizes and is presented in percentage.
Determination of Capillary Density
The capillary density was determined as described previously.11 Tissue sections were stained using CD31 antibody (Santa Cruz Biotechnology). Vessels in the peri-infarct zone were counted in randomly chosen high-power fields (HPFs, magnification ×400). The results are expressed as capillaries per HPF.
Assessment of Intramyocardial Cytokine Levels
The infarcted areas were excised after 5 days and homogenized using cell lysis buffer. Protein concentrations were measured using the BCA Protein Assay Kit (Thermo Scientific, Rockford, Ill). Self-designed Cytokine Antibody Arrays (Raybiotech, Norcross, Ga) were used to identify HGF and VEGF concentrations. Therefore, membranes were covered with 100 μg protein from tissue lysates according to the manufacturer’s protocol. Integrated densities were calculated using ImageJ 1.41.
Isolation of Cardiomyocytes and Hypoxic Cell Culture
Cardiomyocytes were isolated by fractionated DNase/Trypsin digestion as described earlier.12 For details, please see the Data Supplement. Cardiomyocytes were then cultured in conditioned glucose-free medium obtained after 24 hours from MSC, MSC-HGF, or MSC-VEGF cultures. After 5 days of culture, cardiomyocytes were placed into a hypoxic incubator for 24 hours in an atmosphere of a N2 (95%) and CO2 (5%). The lactate dehydrogenase (LDH) release, as a marker for cell injury, was quantified, the percentage of apoptotic cells was assessed using the TUNEL assay, and the Akt activation, a prosurvival marker, was determined (see Data Supplement).
Data are presented as mean±standard deviation. Comparisons between groups were done by ANOVA between groups with least significant differences post hoc tests. The critical α-level for these analyses was set at P<0.05.
Micro-CT imaging was used to follow LV function of mice after LAD ligation for a period of 6 months. LVEF continuously declined from 68±4% before surgery to 18±1% in the control group that did not receive any treatment (Figure 1A).
MSC survival after intramyocardial injection was monitored with bioluminescence imaging (BLI). We found that MSC signals without cytokine enhancement steadily decreased and became undetectable after 7 days. In the study groups that received concomitant MSC and cytokine injection, BLI showed that early MSC losses could be effectively alleviated. Initially, MSC signals increased and peaked after 3 days at 167±43% of the day-0 value with HGF and after 4 days at 144±55% with VEGF, respectively (Figure 1B). Signals in the study groups remained significantly higher than those of the control group until day 6 (P<0.05). Longer-term MSC survival, however, could not be significantly prolonged by cytokine treatment because BLI signals vanished by day 8 in all groups.
In the control group, the reduced LVEF of 18±1% corresponded to a scar size of 30±4% of the LV, and LVEF (Figure 1C) and scar size (Figure 1D) were not significantly improved by sole MSC injection (MSC group). In contrast, HGF (P=0.007 and P=0.006) and VEGF injection (P=0.003 and P=0.003) as well as concomitant MSC with HGF (P=0.001 and P=0.011) or VEGF injection (P<0.001 and P=0.039) significantly improved the LVEF over 6 months and reduced the scar size compared with the sole MSC group. Interestingly, no significant differences were observed between the groups receiving either only cytokines or concomitant cytokine and MSC injections, contributing the observed benefits to the paracrine effects of the injected cytokines rather than a regenerative response from the transplanted cells.
Transplantation of Gene-Enhanced MSCs
To provide a sustained local delivery of targeted cytokines, HGF or VEGF gene transfer was performed in MSCs. Western blot analyses of cell culture supernatants after 24 hours of in vitro culture confirmed the markedly increased cytokine production of genetically modified MSCs compared with MSCs transfected with an empty vector (MSC-EV, Figure 2A). The HGF release was increased 3.7-fold, and the VEGF release was 2.0-fold. Although HGF and VEGF are among the cytokines that are naturally released by native MSCs and also MSC-EV, their production could be markedly stimulated with our gene enhancement technique.
To assess the autocrine proliferatory effects of MSC-produced HGF and VEGF on the same MSCs, in vitro proliferation assays were performed (Figure 2B). After 4 days, proliferation of MSC-HGF and MSC-VEGF was significantly better than that of nontransfected MSCs (P<0.001) but was still significantly less than that of MSCs cultured in medium with constantly high cytokine concentrations of HGF at 40 ng/mL (P=0.004 versus MSC-HGF) or VEGF at 8 ng/mL (P=0.020 versus MSC-VEGF).
Gene-enhanced MSCs were then injected intramyocardially after LAD ligation, and cell survival was followed using BLI (Figure 2C). MSC-EV did not show any survival benefit compared with nontransfected MSCs. MSC-HGF and MSC-VEGF, however, demonstrated initial proliferation with a peak after 5 and 4 days, respectively, and maximum signals of 269±115% and 245±79% (Figure 2D). For 7 days, BLI signals in both cytokine gene enhancement groups were significantly higher than in the MSC-EV group (P<0.05). Thereafter, the cell count rapidly decreased, and all signals vanished after 9 days.
To confirm increased cytokine production by gene-enhanced MSCs after transplantation, cytokine arrays of the infarct tissue and peri-infarct LV areas were performed after 5 days, around the time of peak BLI signals (Figure 2E). Compared with the tissue cytokine concentrations after MSC-EV transplantation, we could detect 3.2-fold and 2.7-fold elevated HGF and VEGF tissue concentrations after MSC-HGF and MSC-VEGF transplantation, respectively. Increased cytokine production by gene-enhanced MSCs, which was shown earlier in vitro (Figure 2A), could thus been shown to persist after cell transplantation.
LV function and histology was assessed after 6 months. MSC-HGF– and MSC-VEGF–treated animals showed smaller average infarct areas (Figure 3A) and significantly smaller scar sizes (P=0.004 and P=0.001) when compared with MSC-EV (Figure 3B). Longitudinal and transversal micro-CT images suggested mildly decreased LV anterior wall aneurysm formation in the MSC-HGF and MSC-VEGF groups (Figure 3C). Accordingly, LV function was better preserved with MSC-HGF (P=0.003) and MSC-VEGF (P=0.002, Figure 3D). Both gene enhancement strategies improved the LVEF by approximately 10% points. Vessel densities within the peri-infarct zone were assessed with immunohistochemistry. The number of capillaries per HPF was elevated in the groups receiving cytokine-producing MSCs (P=0.022 for MSC-HGF and P=0.013 for MSC-VEGF versus MSC-EV; Figure 3E and 3F).
Effects of MSC-HGF and MSC-VEGF on Cardiomyocytes
To explore a potential molecular cardiomyocyte-protective mechanism underlying the functional benefits observed, rat cardiomyocytes were isolated, cultured (Figure 4A), and then exposed to hypoxic conditions for 24 hours. The normalized LDH release, as a marker for cell damage, was reduced 25±17% and 31±18% if cardiomyocytes were maintained in conditioned medium obtained from MSC-HGF (P=0.010 versus cardiomyocytes) or MSC-VEGF (P=0.002 versus cardiomyocyte) cultures, respectively (Figure 4B). Likewise, 36±39% and 23±32% fewer cardiomyocytes were apoptotic if kept in MSC-HGF (P=0.003) or MSC-VEGF medium (P=0.0032, Figure 4C). Cardiomyocyte Akt activity was increased 2.0-fold and 1.9-fold with MSC-HGF or MSC-VEGF medium compared with cardiomyocytes cultured in normal medium (Figure 4D). Conditioned medium from nontransfected MSCs was not effective.
Using stem cell–enriching and/or cytokine-enriching strategies after AMI, we found the effect of the injected cytokines HGF and VEGF to be more important to improve postinfarction myocardial function than the transplanted MSCs. Our data therefore support the emerging findings that stem cell therapy may enhance cardiac function primarily via paracrine mechanisms as opposed to a regeneration of new cardiomyocytes.
To date, this study provides the longest follow-up period after murine LAD ligation in the literature. We show a slow LV remodeling over time with development of anterior wall aneurysms. Although the biggest drop in LV function occurred within 1 day of the ischemic event, the LVEF continued to decline between 2 weeks and 6 months. We found a good correlation between micro-CT imaging and histology for LV morphology. Animals with extended anterior wall scars in histological slides showed typical wall thinning with aneurysm formation in micro-CT scans. ECG gating ensured low-artifact images to assess systolic and diastolic LV volumes and calculate an accurate LVEF. Micro-CT imaging proved to be very valuable for noninvasive longitudinal follow-up monitoring.
Although MSCs are capable of producing a great variety of cytokines, including HGF and VEGF, their posttransplantation state may not support high metabolic activity. HGF and VEGF gene-enhanced MSCs showed increased cytokine secretion in vitro and in vivo and maximized the beneficial paracrine effects of MSC transplantation. Importantly, we demonstrated that cytokine release continued for several days after intramyocardial MSC injection and generated sustained elevated tissue cytokine levels, although shear stress during cell transplantation, ischemia, and an inflammatory environment that is inadequate for donor cell retention are factors that may impair MSC metabolism and cytokine production.
For noninvasive in vivo tracking of transplanted stem cells, our group previously developed and validated BLI techniques.13 Because the fLuc reporter gene is integrated into the DNA and expressed only by living cells, BLI is a highly accurate tool to follow in vivo cell survival.
MSC transplantation usually results in huge early cell losses, which could also be significantly alleviated with the use of gene-enhanced MSCs. Pons et al14 previously demonstrated MSC Akt upregulation by these cytokines, which increased their resistance to culture stress and hypoxia. This autocrine cytokine effect may explain the significantly improved early post-transplantation cell survival of cytokine-overexpressing compared with native MSCs in this study. Because both HGF and VEGF are potent promigratory and proproliferatory factors for MSCs, we could also show that MSC-HGF and MSC-VEGF establish a strong enough autocrine stimulus to allow significantly increased proliferation in vitro. Cytokine production after gene transfer does not obey physiological feedback mechanisms and ensures persistently elevated cytokine production. The fact that MSC proliferation in high HGF or VEGF media was even more effective supports the assumption that MSC proliferation is concentration-dependent.5
Possible strategies to treat AMI involve cardiomyocyte Akt overexpression, cardiomyocyte HGF/VEGF gene transfer, and transplantation of HGF/VEGF gene-enhanced MSCs. It has been demonstrated that adenoviral gene transfer of cardiomyocytes with constitutively active Akt protects the cells from apoptosis after ischemia-reperfusion injury.15–17 However, constitutively active Akt mutants (such as myristolated or phosphomimetic mutants) accumulate predominantly throughout the cytoplasm and have been associated with the development of hypertrophic cardiac remodeling. This contrasts sharply with the physiological nuclear redistribution of its activated wild-type form, which was found to be antihypertrophic.18
HGF and VEGF, besides more physiological Akt activation, provide additional cytokine effects. They promote angiogenesis and may contribute to revascularization of the ischemic area and to salvage hibernating native cardiomyocytes.4 HGF has also been shown to activate the endogenous cardiac repair mechanism by inducing cardiac stem cell activation and migration19 and to markedly reduce the collagen content.20 HGF also upregulates Bcl-xL and Bcl-2, both of which have further antiapoptotic actions.21 Direct intramyocardial adenoviral VEGF or HGF gene transfer after AMI has been investigated by other groups.21–23 Jayasankar et al21 performed LAD ligations in rats and subsequently injected adenoviral vectors containing the rat HGF transgene into the infarct border zone. After 6 weeks, they observed significantly better preserved LV geometry, increased LV wall thickness, and enhanced angiogenesis as compared with injections of empty null virus. Using VEGF-encoding adenovirus injections in 1-week-old AMIs, Hao et al23 found increased arteriolar density and improved LV function 4 weeks later. Ruixing et al22 injected VEGF165 cDNA into myocardial infarcts and reported a decreased apoptotic index with inhibition of the proapoptotic proteins p53, Fas, and Bax, and an increase of Bcl-2 expression. Thus, paracrine effects arising from direct intramyocardial VEGF or HGF gene transfer already appear to somewhat improve cardiac function and induce angiogenesis. That is, although Leor et al24 found that gene transfer into infarcted myocardium, although feasible, was very limited by low transfection efficiency. They performed direct injections of a β-galactosidase–encoding adenovirus vector into healthy myocardium or myocardium subjected to 60 minutes of coronary artery occlusion followed by sustained reperfusion. The areas of transgene expression in infarcted tissues were only 5% to 10% of the areas of noninfarcted hearts. Accordingly, Hao et al23 reported that VEGF expression by cardiomyocytes after adenoviral gene transfer was considerably low in ischemic myocardium, only 7% of that expressed in normal myocardium. Furthermore, increased cardiomyocyte apoptosis within the peri-infarct area was found with adenoviral gene transfer, which was believed to be a direct cytotoxic effect of adenovirus carriers.
Transplantation of HGF/VEGF-overexpressing MSCs has 2 major advantages. First, in vitro transduction of cultured MSCs is highly effective as opposed to the variable and much less reliable in vivo transduction of cardiomyocytes. Second, donor MSCs may also participate in the formation of new vessels by incorporating into the newly forming vessel wall.4 Recent studies demonstrated the integration of transplanted bone marrow stem cells into neocapillaries in myocardial infarct areas.25,26 As a limitation of our study, we did not assess a structural integration of transplanted MSCs into new capillaries.
Earlier studies using HGF or VEGF overexpressing MSCs after AMI20,27 may have overestimated the transdifferentiation potential of MSCs with incorporation into the ischemic myocardium, which is not supported by our current knowledge.28
It is unknown for how long the elevated cytokine concentrations are needed in the healing heart and for how long they remain beneficial. Too widespread and prolonged overexpression of growth and angiogenesis factors may increase the risk for hemangioma growth, and the observed fate of MSCs after 8 to 9 days may be advantageous.
In conclusion, MSC gene enhancement strongly upregulated cytokine production and augmented both autocrine and paracrine mechanisms involved in cell survival and myocardial recovery. Our data suggest that the strong paracrine properties are the most important mechanism of MSC transplantation and that gene-enhanced MSCs maximize the beneficial cardioprotective effect. HGF- and VEGF-overexpressing MSCs were similarly potent in initiating angiogenesis, increasing the tolerance of cardiomyocytes to ischemia, reducing cardiomyocyte apoptosis, and resulting in decreased scar size and improved LV function. The beneficial value of gene-enhanced over native MSCs warrants their future clinical investigation.
We thank Christiane Pahrmann and Beverly Bonfert for their great help.
Sources of Funding
Dr Schrepfer received a research grant from the German Research Foundation (SCHR992/2-1). Dr Deuse received a research grant from the “Deutsche Herzstiftung”. The present study was supported by the Falk Research Fund for the Department of Cardiothoracic Surgery at Stanford University.
Presented in part at American Heart Association Scientific Sessions 2008, November 8–12, 2008, New Orleans, La.
The online-only Data Supplement is available with this article at http://circ.ahajournals.org/cgi/content/full/120/11_suppl_1/S247/DC1.
Chen SL, Fang WW, Ye F, Liu YH, Qian J, Shan SJ, Zhang JJ, Chunhua RZ, Liao LM, Lin S, Sun JP. Effect on left ventricular function of intracoronary transplantation of autologous bone marrow mesenchymal stem cell in patients with acute myocardial infarction. Am J Cardiol. 2004; 94: 92–95.
Katritsis DG, Sotiropoulou PA, Karvouni E, Karabinos I, Korovesis S, Perez SA, Voridis EM, Papamichail M. Transcoronary transplantation of autologous mesenchymal stem cells and endothelial progenitors into infarcted human myocardium. Catheter Cardiovasc Interv. 2005; 65: 321–329.
Schrepfer S, Deuse T, Reichenspurner H, Robbins R, Pelletier M. Stem cell transplantation for the ischemic myocardium: answering the basic questions of bone marrow stromal cell migration, homing, and survival [abstract]. J Heart Lung Transplant. 2008; 27: S321.
Fujio Y, Nguyen T, Wencker D, Kitsis RN, Walsh K. Akt promotes survival of cardiomyocytes in vitro and protects against ischemia-reperfusion injury in mouse heart. Circulation. 2000; 101: 660–667.
Matsui T, Tao J, del Monte F, Lee KH, Li L, Picard M, Force TL, Franke TF, Hajjar RJ, Rosenzweig A. Akt activation preserves cardiac function and prevents injury after transient cardiac ischemia in vivo. Circulation. 2001; 104: 330–335.
Urbanek K, Rota M, Cascapera S, Bearzi C, Nascimbene A, De Angelis A, Hosoda T, Chimenti S, Baker M, Limana F, Nurzynska D, Torella D, Rotatori F, Rastaldo R, Musso E, Quaini F, Leri A, Kajstura J, Anversa P. Cardiac stem cells possess growth factor-receptor systems that after activation regenerate the infarcted myocardium, improving ventricular function and long-term survival. Circ Res. 2005; 97: 663–673.
Jayasankar V, Woo YJ, Bish LT, Pirolli TJ, Chatterjee S, Berry MF, Burdick J, Gardner TJ, Sweeney HL. Gene transfer of hepatocyte growth factor attenuates postinfarction heart failure. Circulation. 2003; 108 (suppl 1): II-230–II-236.
Hao X, Mansson-Broberg A, Grinnemo K-H, Siddiqui AJ, Dellgren G, Brodin LA, Sylven C. Myocardial angiogenesis after plasmid or adenoviral VEGF-A(165) gene transfer in rat myocardial infarction model. Cardiovasc Res. 2007; 73: 481–487.
Kamihata H, Matsubara H, Nishiue T, Fujiyama S, Amano K, Iba O, Imada T, Iwasaka T. Improvement of collateral perfusion and regional function by implantation of peripheral blood mononuclear cells into ischemic hibernating myocardium. Arterioscler Thromb Vasc Biol. 2002; 22: 1804–1810.