Author + information
- Received December 4, 2009
- Revision received November 3, 2010
- Accepted November 23, 2010
- Published online June 14, 2011.
- Jérôme Roncalli, MD, PhD⁎,†,
- Marie-Ange Renault, PhD⁎,
- Jörn Tongers, MD⁎,‡,
- Sol Misener, AAS⁎,
- Tina Thorne, MS⁎,
- Christine Kamide, BS⁎,
- Kentaro Jujo, MD⁎,
- Toshikazu Tanaka, MD⁎,
- Masaaki Ii, MD, PhD§,
- Ekaterina Klyachko, PhD⁎ and
- Douglas W. Losordo, MD⁎,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Douglas W. Losordo, Feinberg Cardiovascular Research Institute, Northwestern University Feinberg School of Medicine Tarry 14-725, 303 East Chicago Avenue, Chicago, Illinois 60611
Objectives This study was designed to compare the effectiveness of Sonic hedgehog (Shh) gene transfer, AMD3100-induced progenitor-cell mobilization, and Shh-AMD3100 combination therapy for treatment of surgically induced myocardial infarction (MI) in mice.
Background Shh gene transfer improves myocardial recovery by up-regulating angiogenic genes and enhancing the incorporation of bone marrow–derived progenitor cells (BMPCs) in infarcted myocardium. Here, we investigated whether the effectiveness of Shh gene therapy could be improved with AMD3100-induced progenitor-cell mobilization.
Methods Gene expression and cell function were evaluated in cells cultured with medium collected from fibroblasts transfected with plasmids encoding human Shh (phShh). MI was induced in wild-type mice, in matrix metalloproteinase (MMP)-9 knockout mice, and in mice transplanted with bone marrow that expressed green-fluorescent protein. Mice were treated with 100 μg of phShh (administered intramyocardially), 5 mg/kg of AMD3100 (administered subcutaneously), or both; cardiac function was evaluated echocardiographically, and fibrosis, capillary density, and BMPC incorporation were evaluated immunohistochemically.
Results phShh increased vascular endothelial growth factor and stromal cell–derived factor 1 expression in fibroblasts; the medium from phShh-transfected fibroblasts increased endothelial-cell migration and the migration, proliferation, and tube formation of BMPCs. Combination therapy enhanced cardiac functional recovery (i.e., left ventricular ejection fraction) in wild-type mice, but not in MMP-9 knockout mice, and was associated with less fibrosis, greater capillary density and smooth muscle–containing vessel density, and enhanced BMPC incorporation.
Conclusions Combination therapy consisting of intramyocardial Shh gene transfer and AMD3100-induced progenitor-cell mobilization improves cardiac functional recovery after MI and is superior to either individual treatment for promoting therapeutic neovascularization.
Recent investigations suggest that the effectiveness of angiogenic gene therapy may be increased by combining it with treatments that mobilize progenitor cells from the bone marrow. Kawamoto et al. (1) combined gene transfer of vascular endothelial growth factor (VEGF) with progenitor-cell mobilization in animal models of acute and chronic myocardial infarction, revealing that the combination therapy was more effective than either treatment alone. However, these promising results were not reproduced in a phase I clinical study that combined intramyocardial injection of VEGF-A plasmid and progenitor-cell mobilization in patients with severe chronic ischemic heart disease (2); both of these investigations mobilized stem cells via the administration of granulocyte colony-stimulating factor, which has been associated with an increased risk of restenosis in 1 report (3).
Previously we demonstrated that myocardial infarction (MI) reactivates the Hedgehog signaling pathway in adult mammals and that Sonic hedgehog (Shh) up-regulates the expression of several angiogenic and cardioprotective genes, including VEGF. Consequently, the vascular benefits of Shh gene therapy may be more robust and extensive than those observed with VEGF alone. Shh gene transfer also preserved left ventricular function, reduced fibrosis, increased angiogenesis, and enhanced the contribution of bone marrow–derived progenitor cells (BMPCs) to myocardial neovascularization in animal models of acute and chronic MI (4).
Because Shh gene therapy appears to improve recovery from MI, at least in part, by increasing the incorporation of BMPCs, it may be particularly effective when administered with treatments that increase progenitor-cell mobilization. We tested this hypothesis in a murine model of MI by combining myocardial injection of a plasmid encoding the human Shh gene (phShh) with subcutaneous injections of AMD3100, a pharmacological agent that mobilizes progenitor cells by disrupting interactions between CXC-chemokine receptor 4 (CXCR4) and stromal cell–derived factor (SDF)-1. Our findings indicate that combined therapy improves cardiac functional recovery after MI and more effectively promotes neovascularization than either individual treatment.
AMD3100 was purchased from Sigma-Aldrich Co. (St. Louis, Missouri), and Shh protein was provided by William Munger, PhD, of Curis Inc. (Lexington, Massachusetts). The phShh was prepared as described previously (4) and as summarized in the Online Appendix.
Cell lines and media
In vitro experiments were performed with National Institutes of Health 3T3 fibroblasts (ATCC, Manassas, Virginia), bovine aortic endothelial cells (ATCC), and BMPCs. BMPCs were isolated from the total mononuclear cell population (5,6); mononuclear cells were obtained from mouse tibias and femurs by density gradient selection, plated on culture dishes coated with rat vitronectin, and cultured in endothelial-growth medium; nonadherent cells were discarded. Procedures for cell maintenance and transfection, collection of the Shh-conditioned and control media, and BMPC isolation and expansion are summarized in the Online Appendix.
Cell functional assays
Migration, proliferation, and tube formation were assessed via standardized protocols as summarized in the Online Appendix. Assessments were performed by personnel who were blinded to treatment.
All experimental procedures were approved by the Institutional Animal Care and Use Committee of Northwestern University. MI was surgically induced in 8- to 10-week-old wild-type C57BL/6J mice (The Jackson Laboratory, Bar Harbor, Maine), wild-type (Charles River Laboratories International, Inc., Wilmington, Massachusetts) and matrix metalloproteinase (MMP)-9 knockout (The Jackson Laboratory) FVB mice, and in wild-type C57BL/6J mice transplanted with bone marrow from age- and sex-matched green-fluorescent protein (GFP)-expressing C57BL/6-Tg(ACTB-EGFP)1Osb/J mice (The Jackson Laboratory). Bone marrow transplantation was performed as described previously (1) and as summarized in the Online Appendix.
Surgical induction of MI and treatment
MI was induced as described in the Online Appendix. Ten minutes later, 0.02 ml of sterile saline containing 100 μg of either phShh or the empty plasmid was injected into 2 sites within the region perfused by the left anterior descending artery; the injection sites were located at the border zone of the infarct, as indicated by changes in the color of the myocardium after left anterior descending artery ligation. AMD3100 (5 mg/kg) was administered subcutaneously approximately 1 h after MI surgery. Subsequent analyses were performed by personnel who were blinded to treatment.
Echocardiographic assessment of cardiac function
Left ventricular ejection fractions (LVEFs) were calculated from echocardiographic measurements performed before MI and 7 ± 1 days, 14 ± 1 days, and 28 ± 2 days afterward. The echocardiographic protocol is summarized in the Online Appendix.
Histology and immunology
Histological and immunologic assessments are summarized in the Online Appendix. Fibrosis was reported as the ratio of the area of fibrosis to the left-ventricular area. Capillaries were identified by either: 1) injecting mice with fluorescein isothiocyanate–labeled BS1-lectin (Vector Laboratories, Burlingame, California) 10 min before sacrifice; or 2) staining tissue sections with fluorescein isothiocyanate–labeled isolectin B4. Smooth muscle–containing vessels were identified as tubular structures stained positively for smooth-muscle α actin and isolectin B4 or smooth-muscle α actin and BS1-lectin. MMP-9–expressing cells were identified by staining sections with anti–MMP-9 antibodies. Bone marrow–derived cells were identified by staining tissue sections with anti-GFP antibodies rather than via intrinsic GFP expression, because GFP fluorescence is low after bone marrow transplantation, and the infarcted tissue autofluoresces.
Quantitative reverse transcriptase polymerase chain reaction
All values were expressed as mean ± SEM, and a p value <0.05 was considered statistically significant. For comparisons between 2 groups, statistical significance was evaluated with the unpaired t test. Comparisons among 3 or 4 groups were assessed via 1-factor analysis of variance followed by the Fisher protected least significant difference post-hoc test when analysis of variance p < 0.05. The relationship between LVEF and capillary density was tested via linear regression. In vitro conditions were evaluated in triplicate (or higher replicates), and data points were used as the unit of analysis; each experiment was performed at least twice. Additional details are provided in the Online Appendix.
Human Shh activates the hedgehog pathway in fibroblasts and induces secretion of VEGF and SDF-1
The mRNA expression of hShh, the Hedghog receptor Patched-1, and Gli1 (a transcription factor targeted by Hedgehog signaling) was significantly greater (p < 0.05) in mouse 3T3 fibroblasts transfected with phShh than in fibroblasts transfected with the empty vector for up to 5 days after transfection (Online Figs. 1A to 1C); maximum hShh expression was observed 48 h after transfection, whereas Patched-1 and Gli1 expression peaked at 72 h. Plasmid hShh gene therapy was also associated with a 5-fold up-regulation of VEGF-A mRNA expression (p < 0.01) and a 2-fold up-regulation of SDF-1α mRNA expression (p < 0.01) at 48 h (Online Figs. 1D and 1E). Supernatant levels of VEGF and SDF-1α protein were quantified 72 h after transfection; VEGF (p < 0.0001) and SDF-1α (p < 0.01) protein levels were significantly higher in supernatants collected from phShh-transfected cells than from cells transfected with the empty vector (Online Figs. 1F and 1G). Incubation with hShh protein (1 μg/ml) for 24 h also led to higher supernatant levels of VEGF (p < 0.0001) and SDF-1α (p < 0.0001) (Online Figs. 1H and 1I).
Shh-conditioned medium promotes capillary morphogenesis, proliferation, and migration in BMPCs and endothelial cells
The migration of both BMPCs (p < 0.001) and bovine aortic endothelial cells (p < 0.01) was significantly greater in cells treated with Shh-conditioned medium (i.e., supernatants collected from phShh-transfected fibroblasts) than with unconditioned medium (i.e., supernatants collected from fibroblasts transfected with an empty vector) (Figs. 1A and 1B), and Shh-conditioned medium also enhanced proliferation (p < 0.01) (Figs. 1C and 1D) and tube formation (p < 0.05) (Figs. 1E and 1F) in BMPCs.
Combination therapy enhances the recovery of cardiac function after MI
One day after surgically induced MI, Shh expression was approximately 200-fold greater in mice administered phShh gene therapy alone (100 mg/mouse) or combination therapy consisting of both phShh administration and AMD3100 (5 mg/kg) than in mice treated with AMD3100 alone or in untreated and control mice (p < 0.0001); the differences between groups remained statistically significant on day 5 (p < 0.05) (Fig. 2A). SDF-1α mRNA expression was similar in all groups on day 1 (data not shown), but on day 5, expression in the infarcted region and in the border zone of the infarct was significantly greater in animals administered combination therapy than in any other treatment group. In noninfarcted tissue, SDF-1α mRNA expression on day 5 was significantly greater in the combination therapy group than in animals treated with phShh alone or in untreated and control animals (p < 0.01) (Fig. 2B).
Cardiac function was evaluated echocardiographically before MI and 7, 14, and 28 days afterward (Figs. 2C and 2D); the interaction between time periods and treatment groups was significant (p < 0.0001). Before MI, the mean LVEF for all animals was 77 ± 1% (Fig. 2E). LVEF increased between days 7 and 14 in the combination therapy group and was significantly greater in animals administered combination therapy than in untreated animals on day 28 (p < 0.05). Improvement was delayed until after day 14 in animals administered phShh alone.
Combination therapy reduces cardiac fibrosis and enhances the development of capillaries and smooth muscle-containing vessels after MI
In mice sacrificed 28 days after MI, the extent of fibrosis was significantly lower, and capillary density was significantly greater, in all treated groups than in untreated animals (fibrosis, p < 0.01; capillary density, p < 0.05) and in the combination therapy group than in animals administered either individual treatment (fibrosis, p < 0.05; capillary density, p < 0.05) (Figs. 3A to 3D). When the infarcted region and border zone were analyzed independently, capillary density was significantly greater in the infarcted zone of the Shh-treatment group (p < 0.05), in the border zone of the AMD3100-treatment group (p < 0.05), and in both regions after treatment with combination therapy (p < 0.05) than in untreated animals (Figs. 3E and 3F). The difference between the combination therapy and AMD3100 treatment groups approached significance in the infarcted region (p = 0.051), and a post-hoc analysis indicated that border-zone capillary density and LVEF on day 28 were correlated (p < 0.05) (Online Fig. 2). Smooth muscle–containing vessels were significantly more prevalent in animals administered combination therapy or phShh alone than in untreated animals (p < 0.05) and in the combination therapy group than in the AMD3100 treatment group (p < 0.05) (Figs. 3G and 3H).
Combination therapy enhances the incorporation of BMPCs
Because SDF-1 is a chemoattractant for BMPCs, and SDF-1α mRNA expression after MI was significantly elevated in mice administered combination therapy, we investigated whether combination therapy increased BMPC incorporation after MI in wild-type mice transplanted with bone marrow from GFP-expressing mice. One week after MI, significantly more bone marrow–derived cells were identified in the myocardial endothelium of mice from all 3 treatment groups than in untreated mice (p < 0.01) and in the combination therapy group than in mice administered either individual treatment (p < 0.001) (Figs. 4A and 4B). The bone marrow–derived cells were primarily incorporated in the infarcted region and the border zone, and cross-sectional views of individual vessels also identified vascular GFP-expressing cells surrounding the endothelium, which suggests that the neovasculature had matured (Fig. 4C).
MMP-9 contributes to the beneficial effects associated with combination therapy
One day after MI, MMP-9 mRNA expression was significantly lower in mice administered AMD3100, both alone and in combination therapy, than in untreated mice (Fig. 5A); however, expression on day 5 was significantly higher in the infarcted and border zone regions of myocardial tissue from mice treated with combination therapy than in the corresponding regions of tissues harvested from untreated mice or from mice administered either individual treatment (p < 0.05) (Fig. 5B). Seven days after MI, MMP-9–positive cells were significantly more prevalent in sections from mice in the combination therapy group than in sections from untreated mice or from mice administered either individual therapy (Fig. 5C).
To determine whether MMP-9 expression is required for the enhanced cardiac functional recovery observed in mice administered combination therapy, we monitored LVEFs in wild-type and MMP-9 knockout mice before MI and 7, 14, and 28 days afterward (Fig. 5D). Twenty-eight days after MI, LVEFs were significantly lower in MMP-9 knockout mice treated with combination therapy than in wild-type mice treated with combination therapy. Collectively, these observations indicate that MMP-9 expression contributes to the beneficial effects associated with combination therapy after MI.
To date, studies of cardiovascular regeneration have primarily investigated gene therapy (7) and cell mobilization (8) independently. However, the promising pre-clinical results achieved with each individual treatment have not been adequately reproduced in clinical studies. Although this discrepancy may be partly caused by issues in trial design, we considered whether a combination of approaches could be superior to monotherapy. Because gene therapy is more practical for treating patients who develop myocardial dysfunction after MI than for preventing myocardial reperfusion injury in patients experiencing an acute cardiac event, we induced MI by permanently ligating the coronary artery, rather than by temporarily restricting blood flow via coronary banding. Furthermore, we mimicked one form of clinical gene delivery (e.g., via the NOGA system, Biosense Webster, Diamond Bar, California) (9) by administering phShh directly to the myocardium. Our results demonstrate that Shh gene therapy and AMD3100-stimulated progenitor-cell mobilization reduced cardiac fibrosis and promoted the development of capillaries and smooth muscle–containing vessels after MI more effectively when combined than when either treatment was administered individually. Furthermore, LVEF measurements were significantly greater in the combination therapy group than in untreated mice, and cardiac functional recovery was more rapid after combination therapy than gene therapy alone, which may improve efficacy and alleviate some of the safety concerns associated with gene therapy (10).
Previously, members of our lab have investigated therapy that combined local VEGF signaling with the cell-mobilizing agent granulocyte colony-stimulating factor (G-CSF) (1). The VEGF–G-CSF combination, like the phShh-AMD3100 combination investigated here, was more effective than either individual treatment, but the benefits of combined VEGF–G-CSF therapy were not observed in a subsequent clinical trial, despite significant increases in circulating progenitor-cell counts (2). Although this relatively small clinical study may have lacked the statistical power or endpoints needed to establish efficacy, the authors suggested that adding local SDF-1 administration to the VEGF–G-CSF combination may increase the recruitment of mobilized cells to the ischemic tissue and improve patient outcomes. Both VEGF and SDF-1 are up-regulated by Shh (4), so the phShh-AMD3100 combination likely harnesses many of the same signaling pathways that would be triggered by a 3-component combination of VEGF, SDF-1, and G-CSF, as well as the angiogenic and cytoprotective mechanisms activated by the Shh-induced up-regulation of angiopoietin and insulin-like growth factor-1 (4). Furthermore, AMD3100 rapidly mobilizes progenitor cells after a single administration and, consequently, may be more suitable for clinical use than G-CSF, which must be administered repeatedly over several days. Use of G-CSF may also increase the risk of restenosis (3), and a recent meta-analysis found no evidence of functional improvement with G-CSF–induced cell mobilization alone (8).
Because Shh regulates the expression of numerous angiogenic factors, the beneficial effects of Shh-AMD3100 combination therapy likely evolve through several mechanisms. BMPC recruitment was higher in animals treated with combination therapy than in other treatment groups, and our in vitro results suggest that this enhancement may have been caused by increased expression of SDF-1 from cardiac fibroblasts, which are the most common cells in infarcted myocardium. Local SDF-1 expression is required for progenitor-cell recruitment after MI (11,12), and SDF-1 overexpression improves progenitor-cell homing (13); however, SDF-1 administration alone may not sufficiently enhance cell incorporation (13), so the Shh-induced upregulation of other factors, including VEGF, may also contribute to progenitor-cell incorporation. Furthermore, Shh-conditioned media enhanced migration, proliferation, and tube formation in BMPCs, and migration in endothelial cells, so the angiogenic potential of incorporated BMPCs and locally resident endothelial cells may also be increased by combination therapy.
Investigations into the role of the SDF-1/CXCR4 axis during recovery from myocardial injury have yielded apparently conflicting results. Proulx et al. (14) have linked AMD3100 administration to reductions in infarct size and improved systolic function, whereas reports by Askari et al. (11) and Yamaguchi et al. (12) suggest that the interaction between SDF-1 and CXCR4 contributes to ischemic tissue repair. Notably, the plasma half-life of AMD3100 (2 to 3 h) indicates that the direct effects of AMD3100-induced CXCR4 blockade will dissipate within a day of administration, whereas the enhanced BMPC mobilization triggered by AMD3100 endures for days or weeks (15), and combination therapy enhanced SDF-1 expression in the infarcted and border-zone regions on day 5 after infarction. Furthermore, AMD3100 has been linked to up-regulated VEGF and MMP-9 expression in BMPCs (15), which suggests that AMD3100 is not simply a CXCR4 antagonist; rather, it appears to possess certain agonist properties, and consequently, the benefits associated with AMD3100 administration cannot be exclusively attributed to declines in SDF-1 activity. Thus the acute effects of AMD3100 could increase the number of BMPCs released from the bone marrow by disrupting the SDF-1/CXCR4 axis, whereas the long-term effects of AMD3100 and Shh could increase the incorporation of the mobilized cells by elevating the expression of angiogenic cytokines in BMPCs and enhancing SDF-1 expression in ischemic tissue.
Recent experiments performed by Jujo et al. (15) indicate that AMD3100 does not enhance BMPC mobilization in MMP-9 knockout mice and that the benefit of treatment with AMD3100 alone after MI requires MMP-9 expression in bone marrow cells, but not in the ischemic region. Furthermore, MMPs are believed to promote cell migration (16) by degrading type IV collagen in the basement membranes of vessels and the extracellular matrix (17), and Xiang et al. (18) have linked improvements in vascularity and cardiac recovery after surgical MI in rats to increases in BMPC trans-endocardial migration. Thus MMP-9 likely mediates the benefit of combination therapy both by facilitating AMD3100-induced BMPC mobilization and by improving BMPC migration. The potential relationship between MMP-9 and SDF-1α expression also warrants continued investigation.
Our findings demonstrate that combination therapy consisting of intramyocardial Shh gene transfer and AMD3100-induced progenitor-cell mobilization improves cardiac functional recovery after MI and is superior to either individual treatment for promoting therapeutic neovascularization. Thus this therapeutic combination may prove to be a viable treatment for a variety of ischemic diseases.
The authors thank Joelle Fourcade, MD, MS, and Alfred Rademaker, PhD, for assistance with statistical analyses; W. Kevin Meisner, PhD, ELS, for editorial support; Ashley Peterson for administrative support; and William Munger for providing the Shh protein.
For an expanded Methods section and supplementary figures and table, please see the online version of this article.
Sonic Hedgehog-Induced Functional Recovery After Myocardial Infarction Is Enhanced by AMD3100-Mediated Progenitor-Cell Mobilization
This work was supported in part by National Institutes of Health grants HL-53354, HL-57516, HL-77428, HL-95874, HL-80137, and PO1HL-66957. Dr. Roncalli was supported by the French Federation of Cardiology and Sanofi-Synthelabo Inc, Paris, France. The authors have reported that they have no relationships to disclose.
- Abbreviations and Acronyms
- bone marrow–derived progenitor cell
- CXC-chemokine receptor 4
- granulocyte colony-stimulating factor
- left ventricular ejection fraction
- myocardial infarction
- matrix metalloproteinase
- plasmid encoding the human Shh gene
- stromal cell–derived factor
- Sonic hedgehog
- vascular endothelial growth factor
- Received December 4, 2009.
- Revision received November 3, 2010.
- Accepted November 23, 2010.
- American College of Cardiology Foundation
- Kawamoto A.,
- Murayama T.,
- Kusano K.,
- et al.
- Ripa R.S.,
- Wang Y.,
- Jorgensen E.,
- Johnsen H.E.,
- Hesse B.,
- Kastrup J.
- Kang H.J.,
- Kim H.S.,
- Zhang S.Y.,
- et al.
- Asahara T.,
- Murohara T.,
- Sullivan A.,
- et al.
- Ii M.,
- Nishimura H.,
- Iwakura A.,
- et al.
- Tongers J.,
- Roncalli J.G.,
- Losordo D.W.
- Zohlnhofer D.,
- Dibra A.,
- Koppara T.,
- et al.
- Vale P.R.,
- Losordo D.W.,
- Milliken C.E.,
- et al.
- Yamaguchi J.,
- Kusano K.F.,
- Masuo O.,
- et al.
- Abbott J.D.,
- Huang Y.,
- Liu D.,
- Hickey R.,
- Krause D.S.,
- Giordano F.J.
- Jujo K.,
- Hamada H.,
- Iwakura A.,
- et al.
- Van Den Steen P.E.,
- Wuyts A.,
- Husson S.J.,
- Proost P.,
- Van Damme J.,
- Opdenakker G.
- Creemers E.E.,
- Cleutjens J.P.,
- Smits J.F.,
- Daemen M.J.
- Xiang G.,
- Schuster M.D.,
- Seki T.,
- et al.