Author + information
- Received June 22, 2012
- Revision received July 31, 2012
- Accepted August 6, 2012
- Published online November 20, 2012.
- Kai Kang, MD⁎,†,
- Lu Sun, MD⁎,†,
- Yun Xiao, BSc‡,
- Shu-Hong Li, MD, MSc†,
- Jun Wu, MD, MSc†,
- Jian Guo, MD, PhD†,
- Shu-Ling Jiang, MD⁎,
- Lei Yang, MD⁎,
- Terrence M. Yau, MD, MSc†,
- Richard D. Weisel, MD†,
- Milica Radisic, PhD, PEng‡ and
- Ren-Ke Li, MD, PhD⁎,†,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Ren-Ke Li, MaRS Centre, Toronto Medical Discovery Tower, 101 College St., Room 3-702, Toronto, Ontario M5G 1L7, Canada
Objectives This study investigated whether cytokine enhancement of a biodegradable patch could restore cardiac function after surgical ventricular restoration (SVR) even when seeded with cells from old donors.
Background SVR can partially restore heart size and improve function late after an extensive anterior myocardial infarction. However, 2 limitations include the stiff synthetic patch used and the limited healing of the infarct scar in aged patients.
Methods We covalently immobilized 2 proangiogenic cytokines (vascular endothelial growth factor and basic fibroblast growth factor) onto porous collagen scaffolds. We seeded human mesenchymal stromal cells from young (50.0 ± 8.0 years, N = 4) or old (74.5 ± 7.4 years, N = 4) donors into the scaffolds, with or without growth factors. The patches were characterized and used for SVR in a rat model of myocardial infarction. Cardiac function was assessed.
Results In vitro results showed that cells from old donors grew slower in the scaffolds. However, the presence of cytokines modulated the aging-related p16 gene and enhanced cell proliferation, converting the old cell phenotype to a young phenotype. In vivo studies showed that 28 days after SVR, patches seeded with cells from old donors did not induce functional recovery as well as patches seeded with young cells. However, cytokine-enhanced patches seeded with old cells exhibited preserved patch area, prolonged cell survival, and augmented angiogenesis, and rats implanted with these patches had better cardiac function. The patch became an elastic tissue, and the old cells were rejuvenated.
Conclusions This sustained-release, cytokine-conjugated system provides a promising platform for engineering myocardial tissue for aged patients with heart failure.
- cardiac patch
- mesenchymal stromal cells
- myocardial infarction
- surgical ventricular restoration
Heart failure continues to be an important cause of death and disability from cardiovascular disease (1). Patients who have had an extensive myocardial infarction (MI) are at risk of heart failure because of adverse ventricular remodeling. Surgical ventricular restoration (SVR) can normalize the size and shape of the ventricle and improve heart function (2). However, the STICH (Surgical Treatment for Ischemic Heart Failure) trial demonstrated that SVR and coronary artery bypass grafting (CABG) restored ventricular volumes but did not improve symptoms, exercise ability, or global ventricular function compared with CABG alone (3). Two potential reasons for the limited improvement observed with SVR include the use of a stiff, synthetic patch, which renders the infarct scar and adjacent border zone fibrotic and nonelastic, and the advanced age of the patients undergoing the procedure, which renders them unable to adequately heal the repaired region and maintain the restored ventricular function because of dysfunctional stem cells. Therefore, we investigated a new intervention intended to restore elasticity to the ventricular repair and rejuvenate the aged phenotype.
Pre-clinical studies demonstrated that SVR was more effective than medical therapy after an extensive MI (4–6); however, the use of stiff, synthetic patches was associated with recurrent ventricular dilation (4–7). Initial studies with biodegradable scaffolds demonstrated that they induced cellular recruitment and engraftment, resulting in viable tissue after the patch dissolved (8). Seeding the biodegradable patches with cells increased the thickness of the construct compared with nonseeded patches by increasing cellular infiltration and engraftment (9,10). However, thicker patches have limited perfusion and the cells in the center die, limiting the benefit of these grafts. Immobilization of cytokines within the biodegradable biomaterial has been reported to provide sustained cytokine release to induce angiogenesis, improve cell survival, and enhance cell recruitment (11,12). We previously developed a method to covalently immobilize angiogenic cytokines to a collagen scaffold (11,12), resulting in their sustained release and producing an angiogenic effect in both the biomaterial and the surrounding tissue. This platform, consisting of conjugated cytokines but no cells, was successfully used to repair defects of the right ventricular outflow tract in a rat model (12).
Preliminary studies in children demonstrated that cell-seeded biodegradable vascular patches resulted in the formation of viable tissue after the biomaterial dissolved (13). However, this approach may be difficult in older patients with extensive coronary artery disease. We and others demonstrated that bone marrow cells from older individuals do not restore heart function as well as cells from younger individuals and that older recipients of cell therapy do not respond as well as younger recipients (14,15). In the current study, we created a cytokine-enhanced cardiac patch by immobilizing vascular endothelial growth factor (VEGF) and basic fibroblast growth factor (bFGF) onto a collagen scaffold and seeding it with human mesenchymal stromal cells (hMSCs) isolated from younger or older patients with cardiac disease. We evaluated the potential of this cytokine-enhanced, cell-seeded patch to improve recovery of ventricular function after SVR.
Please refer to the Online Methods section for further details.
Bone marrow collection
The collection of human samples was approved by the Research Ethics Board of University Health Network. Each patient provided informed consent. Bone marrow aspirates were obtained from the sternum of patients undergoing CABG at Toronto General Hospital. “Young” hMSCs were isolated from patients age ≤57 years (50.0 ± 8.0 years, N = 4); “old” hMSCs were isolated from patients age ≥66 years (74.5 ± 7.4 years, N = 4).
Preparation of the collagen scaffolds
The collagen scaffolds were prepared and immobilized with cytokines as previously described (11,12). Four groups of scaffolds were tested with and without growth factors: fresh (scaffolds right after preparation), blank (scaffolds incubated in media for 3 days), young (scaffolds seeded with young hMSCs and incubated in media for 3 days), and old (scaffolds seeded with old hMSCs and incubated in media for 3 days). The tensile strength of the scaffolds was tested with the ElectroForce 5200 BioDynamic Test Instrument (Bose, Eden Prairie, Minnesota) using the 22 N load sensor, and their structure was examined by scanning electron microscopy. Immobilized and free VEGF and bFGF levels in the phosphate-buffered saline supernatant and scaffolds were quantified with enzyme-linked immunosorbent assay (ELISA).
Bone marrow cell–engineered scaffolds
Isolated hMSC seeding was performed as previously described (11,12). Cell numbers in the scaffolds were evaluated with the MTT assay after 2 and 4 days of culture. Proliferating cells were labeled with bromodeoxyuridine (BrdU), and after culturing for 48 h, the scaffolds were stained with BrdU antibody. The BrdU-positive cells in the scaffolds were counted in 5 randomly selected high-power (20×) fields and expressed as a percentage. Staining with antibodies against alpha-smooth muscle actin (α-SMA) and connexin 43 was performed after 4 days of culture. The scaffolds were evaluated in 5 randomly selected high-power (20×) fields, and the number of positively stained cells was counted and expressed as a percentage. The mRNA profiles of collagen type I and III, cyclin-dependent kinase inhibitor 2A (CDKN2A or p16), and regucalcin (RGN) were assessed with real-time polymerase chain reaction (PCR) after 4 days of culture.
All animal procedures were performed in accordance with the Guide for the Care and Use of Laboratory Animals (National Institutes of Health, revised 1996) and approved by the Animal Care Committee of University Health Network. To avoid immune rejection, all rats were given 5 mg/kg cyclosporine A (Novartis, Basel, Switzerland) intraperitoneally each day from 3 days before to 28 days after patch implantation. A transmural MI was generated in adult rats, and 2 weeks later, echocardiography was used to screen the rats on the basis of infarct size (Online Fig. S1). Only those animals with an akinetic infarct area greater than 25% but less than 35% of the left ventricular (LV) free wall were randomly assigned to one of the following groups and underwent SVR: Control = collagen patch; GF = collagen patch covalently immobilized with growth factors; Old = collagen patch seeded with old hMSCs; Young = collagen patch seeded with young hMSCs; Old + GF = collagen patch covalently immobilized with growth factors and seeded with old hMSCs; Young + GF = collagen patch covalently immobilized with growth factors and seeded with young hMSCs.
Patch implantation was performed 14 days after coronary ligation, as previously described (9,16). LV function was evaluated by echocardiography before MI (pre-ligation baseline), before patch implantation (day 0), and 7, 14, and 28 days afterward. Assessment of patch morphology and cell survival and vascular density was performed 28 days after patch implantation. Identification of the cells in the patch was performed by immunostaining. Details are provided in the Online Methods section.
Statistical analysis was performed with GraphPad Prism v4. All data are expressed as mean ± SD. Comparisons among groups were made with 1-way or 2-way analysis of variance or analysis of covariance. If the F test was significant, pairwise tests of individual group means were carried out using the Newman-Keuls post-test, Bonferroni post-test, or Duncan's multiple range test. A p value <0.05 was considered statistically significant.
We first characterized the scaffolds to ensure that cytokine immobilization did not adversely affect the stiffness, strength, or porosity of the scaffolds and, therefore, their usefulness for SVR. Stress–strain curves demonstrated that cytokine-free and -enhanced scaffolds exhibited similar behaviors during the breaking process (Fig. 1A). Young's modulus, a measure of stiffness, did not differ with the addition of growth factors to the scaffold (Fig. 1B). Ultimate tensile strength, the maximum stress a material can withstand while being stretched, was lower in cytokine-enhanced scaffolds than in cytokine-free scaffolds, with the exception of the Old group (Fig. 1C). However, the ultimate tensile strength was calculated at the final breaking point, at approximately 300% strain. Because this strain is not applicable in vivo, the Young's modulus is more suitable for characterizing the mechanical property of the scaffolds for our application. Scanning electron microscopy showed no significant differences in the porosity or pore structures of the scaffolds with or without cytokines (Fig. 1D).
To ensure covalent immobilization of the growth factors to the collagen scaffolds (and not just physical attachment), we used ELISA to determine the immobilization efficiencies for VEGF and bFGF, which were 42% and 24%, respectively, and the physical bonding efficiencies, which were 2% and 5% (Fig. 1E), demonstrating that both cytokines were chemically immobilized to the scaffolds. The scaffolds were prepared to provide a slow release of cytokines over time after SVR. The release study, based on ELISA, demonstrated that the amount of cytokine released from the scaffolds over 4 weeks was small, but the release rates were similar for VEGF and bFGF (Fig. 1F). A significant amount of each cytokine remained in the scaffolds after 28 days (Fig. 1G).
hMSC number, proliferation, and differentiation in scaffolds in vitro
Because cell engraftment within the scaffold increases the thickness of the patch and results in a viable tissue after the patch dissolves, we evaluated hMSC proliferation within the scaffold. The MTT assay demonstrated that the number of hMSCs increased within both the cytokine-free and -enhanced scaffolds, whether from young or old donors (Fig. 2A). As expected, the old hMSCs seeded in cytokine-free scaffolds had the least number of cells at both days 2 and 4 after seeding. Culturing the old hMSCs in cytokine-enhanced scaffolds significantly increased cell numbers to the levels seen with young cells. BrdU staining demonstrated fewer proliferating cells in cytokine-free patches seeded with old hMSCS than with young cells (Figs. 2B and 2C). Culturing old hMSCs in cytokine-enhanced scaffolds significantly increased the number of proliferating cells to a level similar to that of young hMSCs.
We also assessed the effect of the cytokines on differentiation of the hMSCs seeded in the scaffolds. As demonstrated in Figure 3, cytokine stimulation produced different responses in hMSCs from young and old donors. Real-time PCR showed that the mRNA expression of collagen I and III, characteristically secreted by fibroblasts, was significantly greater in older cells, and cytokine enhancement dramatically down-regulated this expression. In contrast, mRNA expression of collagen I and III in younger cells increased with cytokine stimulation (Figs. 3A and 3B). A similar trend was observed with immunostaining for α-SMA and connexin 43 (Figs. 3C and 3D).
Cytokine-enhanced scaffolds modulated aging-related genes in old hMSCs in vitro
The advanced age of most patients who require SVR after an extensive MI limits their ability to adequately heal the repaired infarct region of the heart because of dysfunctional stem cells, and the value of cell implantation is restricted because of stem cell exhaustion in older patients. Therefore, we were interested in the specific effects of cytokine stimulation on aging-related genes in old hMSCs. Both p16 and RGN are oncogenes associated with human rejuvenation (17,18) and are conventionally used to assess the degree of aging. Higher expression of p16 suggests less capacity for cell rejuvenation, whereas higher RGN expression suggests greater cell regeneration potential. We evaluated the mRNA expression level of these 2 genes by real-time PCR after culturing hMSCs within the scaffolds for 4 days. The old cells seeded in the cytokine-free scaffolds demonstrated the highest p16 mRNA expression among the groups. Old hMSCs seeded in the cytokine-enhanced scaffolds showed p16 mRNA expression levels similar to those of the young cells, with or without cytokines (Fig. 4A). The mRNA expression of the aging-induced p16 gene was inhibited by the conjugated cytokines. RGN showed a similar, but opposite, pattern; however, the effects of the cytokines did not reach statistical significance (Fig. 4B). We further investigated p16 by Western blot analysis, which confirmed the real-time PCR result (Fig. 4C). Staining the MSCs for p16 showed nuclear and perinuclear localization of the protein (Fig. 4D).
Old hMSCs in cytokine-enhanced patches improved cardiac function and patch morphology after SVR
Our in vivo study in rats assessed repair of the LV defect after infarct excision and SVR using patches with or without cytokines and with or without cells. Coronary artery ligation of the rat hearts resulted in significant LV dilation and progressive ventricular dysfunction, as assessed by echocardiography (Fig. 5A). All echocardiographic parameters were virtually indistinguishable among the animals at the time of patch implantation (day 0) (Figs. 5B to 5F) because of the pre-selection of animals with similar infarct sizes. Two weeks after SVR, fractional shortening, fractional area change, and ejection fraction in the Old group had decreased more than in the Young and Young + GF groups (Figs. 5B to 5D). At 28 days, the protective effect of the sustained cytokine release was reflected in the significant preservation of fractional shortening and ejection fraction in the Old + GF group, which was similar to that in the Young and Young + GF groups (Figs. 5B and 5D). The left ventricular internal diastolic dimension (LVIDd) and left ventricular internal systolic dimension (LVIDs) in the Old group increased by day 14 and continued this trend to day 28 (Figs. 5E and 5F). The sustained cytokine release resulted in smaller LVIDd and LVIDs in the Old + GF group. The Old group was not different than the Control group (blank patch) or GF group (patch with immobilized cytokines) for any of the parameters.
Cardiac function also was assessed with a pressure-volume catheter (Fig. 6A). In contrast to the Old group, the Young group was associated with better pre-load recruitable stroke work and end-systolic pressure-volume relationship, load-independent indices of ventricular function (Figs. 6B and 6C). Likewise, the immobilized cytokines exerted beneficial effects on the patches seeded with cells from old donors. Ejection fraction and dP/dt max, load-dependent indices of systolic function, also showed significant improvement with cytokine enhancement of old donor cells, rendering the Old + GF group similar to both the Young and Young + GF groups (Figs. 6D and 6E). This trend also was observed for the indices of diastolic function (Figs. 6F and 6G), although it was not significant.
Because ventricular dilation has been demonstrated to recur after SVR, we were interested in the patch morphology after repair. A thin, dilated patch indicates initiation of the redilation of the repaired ventricle. Patch area increased in all groups on day 28 compared with the area before implantation (Figs. 7A and 7B). However, the dilatation of the patch was most dramatic in the Old group. Cytokine enhancement of the patches seeded with old cells prevented patch dilation similar to the patches seeded with young cells. Patch thickness was identical in all groups early after implantation. Similar to patch area, at 28 days after implantation, the patches were least thick in the Old group (Figs. 7C and 7D). Cytokine enhancement of the patches seeded with old cells produced a trend to increased patch thickness, but the difference was not significant.
Old hMSCs in cytokine-enhanced patches increased cell survival and vascular density after SVR
We demonstrated that cell proliferation occurred in the scaffolds in vitro; however, cell survival after patch implantation is necessary for cell engraftment and the formation of viable tissue after the biomaterial dissolves. Staining of heart sections for human mitochondria showed that cell survival 28 days after patch implantation was significantly lower in the patches seeded with hMSCs from old donors. However, cytokine enhancement significantly increased old cell survival, but not to the level achieved with young cells (Figs. 8A and 8B).
To identify which types of cells were present in the patch area after SVR, we performed staining with antibodies against CD45, von Willebrand factor–factor VIII (vWF-FVIII), discoidin domain receptor 2, SMA, and sarcomeric actinin. Co-staining for human mitochondria identified transplanted cells (Online Fig. S2). Human mitochondrial negative cells within the patch area were mostly positive for CD45 and vWF-FVIII. Therefore, many host cells infiltrated the patch area and may be contributing to angiogenesis. Transplanted cells (positive for human mitochondria) were mostly positive for SMA and discoidin domain receptor 2, indicating that some of the hMSCs may have been in the process of transdifferentiation to myofibroblasts.
To evaluate the angiogenic effect of the immobilized cytokines in both the patch and surrounding tissue, we stained heart sections for vWF-FVIII and α-SMA to identify capillary structures and mature vessels, respectively. Vascular density was lowest in the Old group, but it increased significantly with the addition of cytokines so that the Old + GF group had vascular densities similar to those of the Young and Young + GF groups (Figs. 8C and 8D).
This study demonstrated what may have been intuitively anticipated—that cardiac patches seeded with cells from old donors did not improve ventricular function after SVR as well as patches seeded with cells from young donors. Previous studies have established the benefits of cytokine enhancement of cell-seeded cardiac constructs to induce angiogenesis and improve cell survival. This is the first study to suggest that cytokine enhancement restored the proliferative capacity of bone marrow cells from old donors and permitted the recovery of ventricular function as well as cells from young donors. The initial clinical trial of cell-seeded cardiac patches (13) is encouraging, and the addition of immobilized cytokines may permit the application of this technique to the patients most in need—elderly survivors of an extensive MI.
A major limitation of cell-seeded cardiac patches has been low cell survival and proliferation. To overcome this, investigators have attempted to create an interior vasculature within the biomaterial to facilitate metabolic exchange between the implanted cells and the microenvironment (19,20) or to pre-condition the cell-seeded patch in vitro to improve oxygen delivery to the bioengineered tissue (21). The most promising approach has been the delivery of growth factors (22) or genes (23) to shorten the period between implantation and vascularization of the matrices. The development of functionally mature vasculature is a dynamic process that requires multiple cytokines acting sequentially (23), and investigators have attempted to exploit a cell-seeded platform to facilitate the orchestration of cytokine delivery (12,16). VEGF has been documented to be an important growth factor that produces irregular, hyperpermeable capillaries (23), and bFGF enhances maturation of the developing blood vessel network by recruiting smooth muscle cells to the primitive tubular vessels to produce functional, mature vessels (24–26). Combining VEGF and bFGF can synergistically accelerate the functional angiogenesis induced by cell therapy (27).
The biodegradable collagen patch used for SVR in this series is similar to other synthetic scaffolds previously reported for this purpose. The material used in the initial clinical trial (13) induced more tissue formation than the nonbiodegradable materials currently used for LV repair (8). Alternative approaches include decellularized biomaterials (28–30), which retain regenerative cytokines and have been used for cardiac repair. Extensive investigations will be required to determine the best approach to repair the heart by tissue engineering. We have previously demonstrated that cytokine-enhanced biodegradable scaffolds induced greater tissue in-growth and improved recovery of ventricular function when used for SVR than control biodegradable biomaterials not enhanced with cytokines (12–16).
The major finding of the current study was the limited benefit achieved with SVR using patches seeded with cells from old donors and the restoration of that benefit with the immobilization of cytokines within the biomaterial. The primary in vitro difference between the young and old hMSCs was cellular proliferation, and cytokine enhancement increased proliferation of the old cells. To further explore the possible mechanisms of cytokine enhancement of old hMSCs, we assessed expression of the aging-related p16 and RGN genes in the hMSCs and demonstrated a relation between cytokine enhancement and gene expression. This may be the first report to demonstrate that stimulation of aged hMSCs with exogenous VEGF and bFGF reversed the aging phenotype by modulating aging-related gene expression. These preliminary findings require further investigation to fully elucidate the effects of cytokine stimulation on aged hMSCs.
The improvement in functional recovery after SVR with cytokine-enhanced patches seeded with old cells may have been the result of the increased proliferation of the implanted cells and possibly stimulation of the cardiac-resident progenitor cells in the infarct border zone. Young cells and cytokine-stimulated old cells induced angiogenesis not only in the patch but also in the peri-infarct region. The increased angiogenesis enhanced survival of both the implanted cells and the host cells. Cell engraftment was associated with the preservation of patch area, prevention of adverse matrix remodeling, and avoidance of ventricular thinning and dilation. We identified smaller patches seeded with young cells and cytokine-enhanced old cells 4 weeks after implantation. Of note, the preservation of patch area did not compromise new vessel formation or cell survival. As such, the improvement in cardiac function was likely due to the combination of preserved ventricular geometry and improved perfusion, as demonstrated in our previous investigation of SVR (16). Indeed, in the current study, ejection fraction correlated significantly with cell survival, patch thickness, and vascular density (Online Fig. S3).
Restoration of cardiac function was possible with a cytokine-enhanced, tissue-engineered patch that rejuvenated aged cells. Covalent immobilization of 2 proangiogenic cytokines, VEGF and bFGF, onto a collagen scaffold enhanced cell proliferation in vitro and prolonged cell survival and improved angiogenesis to restore ventricular morphology and function in vivo. Of note, the improvement was most obvious with patches seeded with cells from old donors. This novel cytokine-conjugated, sustained-release system provides a practical and promising platform for cardiac repair in elderly survivors of an extensive MI, an important advance in an increasingly aging society.
The authors thank Eileen Mercier and ING Canada Inc., for generous support.
For an expanded Methods section, and supplemental figures and table, please see the online version of this article.
This research was funded by a grant from the Canadian Institutes of Health Research (MOP102535 to Dr. Li) and the Natural Sciences and Engineering Research Council of Canada (CPG-104296 to Dr. Radisic and Dr. Li). Dr. Li holds a Canada Research Chair in Cardiac Regeneration. Dr. Yau holds the Angelo & Lorenza DeGasperis Chair in Cardiovascular Surgery Research. All authors have reported they have no relationships relevant to the contents of this paper to disclose.
Drs. Kang and Sun contributed equally to this work.
- Abbreviations and Acronyms
- alpha smooth muscle actin
- basic fibroblast growth factor
- coronary artery bypass grafting
- enzyme-linked immunosorbent assay
- human mesenchymal stromal cells
- left ventricular
- left ventricular internal diastolic dimension
- left ventricular internal systolic dimension
- myocardial infarction
- polymerase chain reaction
- CDKN2A, cyclin-dependent kinase inhibitor 2A
- regucalcin (senescence marker protein-30)
- surgical ventricular restoration
- vascular endothelial growth factor
- von Willebrand factor–factor VIII
- Received June 22, 2012.
- Revision received July 31, 2012.
- Accepted August 6, 2012.
- American College of Cardiology Foundation
- Lloyd-Jones D.,
- Adams R.J.,
- Brown T.M.,
- et al.
- Jones R.H.,
- Velazquez E.J.,
- Michler R.E.,
- et al.
- Tsuneyoshi H.,
- Nishina T.,
- Nomoto T.,
- et al.
- Heeschen C.,
- Lehmann R.,
- Honold J.,
- et al.
- Zhang H.,
- Fazel S.,
- Tian H.,
- et al.
- Vale P.R.,
- Losordo D.W.,
- Milliken C.E.,
- et al.
- Kenagy R.D.,
- Hart C.E.,
- Stetler-Stevenson W.G.,
- Clowes A.W.
- Losordo D.W.,
- Dimmeler S.
- Spanholtz T.A.,
- Theodorou P.,
- Holzbach T.,
- Wutzler S.,
- Giunta R.E.,
- Machens H.G.
- Zhao Z.Q.,
- Puskas J.D.,
- Xu D.,
- et al.