Increased Angiogenesis and Improved Left Ventricular Function after Transplantation of Myoblasts Lacking the MyoD Gene into Infarcted Myocardium

Skeletal myoblast transplantation has therapeutic potential for repairing damaged heart. However, the optimal conditions for this transplantation are still unclear. Recently, we demonstrated that satellite cell-derived myoblasts lacking the MyoD gene (MyoD−/−), a master transcription factor for skeletal muscle myogenesis, display increased survival and engraftment compared to wild-type controls following transplantation into murine skeletal muscle. In this study, we compare cell survival between wild-type and MyoD−/− myoblasts after transplantation into infarcted heart. We demonstrate that MyoD−/− myoblasts display greater resistance to hypoxia, engraft with higher efficacy, and show a larger improvement in ejection fraction than wild-type controls. Following transplantation, the majority of MyoD−/− and wild-type myoblasts form skeletal muscle fibers while cardiomyocytes do not. Importantly, the transplantation of MyoD−/− myoblasts induces a high degree of angiogenesis in the area of injury. DNA microarray data demonstrate that paracrine angiogenic factors, such as stromal cell-derived factor-1 (SDF-1) and placental growth factor (PlGF), are up-regulated in MyoD−/− myoblasts. In addition, over-expression and gene knockdown experiments demonstrate that MyoD negatively regulates gene expression of these angiogenic factors. These results indicate that MyoD−/− myoblasts impart beneficial effects after transplantation into an infarcted heart, potentially due to the secretion of paracrine angiogenic factors and enhanced angiogenesis in the area of injury. Therefore, our data provide evidence that a genetically engineered myoblast cell type with suppressed MyoD function is useful for therapeutic stem cell transplantation.

The myogenic regulatory factors are a group of skeletal muscle-specific basic helix-loop-helix (bHLH) transcription factors, including MyoD, Myf5, myogenin, and MRF4, that play an essential role in satellite cell activation, proliferation and differentiation [25], [26]. Satellite cell-derived myoblasts lacking the MyoD gene (MyoD 2/2 ) display accelerated growth rates as well as delayed terminal differentiation and muscle regeneration [27]. In addition, MyoD 2/2 myoblasts display more primitive characteristics than wild-type cells and represent an intermediate stage between stem cells and myogenic precursors [27], [28]. Recently, we demonstrated that MyoD 2/2 myoblasts engraft with significantly higher efficiency compared to wild-type myoblasts after injection into injured skeletal muscle [29]. Importantly, many anti-apoptotic genes are up-regulated in the MyoD 2/2 myoblast population, while genes known to promote apoptosis are down-regulated. Consistent with this gene expression profile, MyoD 2/2 myoblasts display remarkable resistance to apoptosis and increased cell survival [30], [31]. Therefore, MyoD 2/2 myoblasts may be useful for the treatment of damaged heart tissue.

Isolation of Wild-type and MyoD 2/2 Myoblasts for Cardiac Repair
Recently, we reported that MyoD 2/2 myoblasts display remarkable resistance to apoptosis and increased cell survival compared to wild-type myoblasts after injection into injured skeletal muscle [29], [30]. To compare cell engraftment and cardiac function after the direct injection of myoblasts into infarcted mouse heart, wild-type and MyoD 2/2 myoblasts were purified from the skeletal muscle of adult mice and passaged 6-8 times before transplantation into infarcted hearts. MyoD 2/2 myoblasts displayed an enlarged cytoplasm and nuclear processes. In contrast, wild-type cells displayed a rounded morphology with a small, compact nucleus (Fig. 1A). Under growth conditions, MyoD 2/2 myoblasts clearly expressed the myogenic marker Pax7, but not the myogenic marker MyoD, while wild-type myoblasts expressed both markers (Fig. 1A). As previously reported, MyoD 2/2 myoblasts, but not wild-type myoblasts, displayed a severe defect in their ability to differentiate into sarcomeric myosin heavy chain (MHC) + elongated myocytes or multinucleated myotubes under low serum conditions (Fig. 1B) [28]. However, MyoD 2/2 myoblasts were able to undergo terminal differentiation when incubation times in low serum conditions were increased [28] or cells were injected directly into injured muscle [29]. Therefore, we concluded that both wild-type and MyoD 2/2 myoblasts had been successfully isolated from adult mice. Following isolation, wild-type and MyoD 2/2 myoblasts were co-transfected with PGK-nlacZ-MAR and PGK-puro plasmids for cell transplantation, and stable transformants were pooled after puromycin selection. Fig. 1C shows that more than 90% of stable wild-type and MyoD 2/2 transformants expressed nuclear lacZ after puromycin selection.

Efficient Engraftment of MyoD 2/2 Myoblasts after Transplantation into Infarcted Heart
Next, isolated myoblasts were used for transplantation after MI. One week after transplantation into the peri-infarct region, X-gal staining showed that MyoD 2/2 myoblasts engrafted at a higher rate in the recipient heart than wild-type myoblasts ( Fig. 2A). Cross sections clearly showed that the progeny of wild-type and MyoD 2/2 myoblasts were visible at the site of injection, near the peri-infarct region of left ventricular (LV), and also in uninjured areas of the heart ( Fig. 2A). Fig. 2B shows the myoblast engraftment rate by 1, 3, 7, 14 and 28 days after cell transplantation. As previously reported [32], both MyoD 2/2 and wild-type myoblast groups displayed a significant reduction in cell number 1 day after injection (MyoD 2/2 myoblasts 13.8062.52%, n = 3, wild-type myoblasts 3.7560.83%, n = 3, p,0.05). This reduction was more pronounced for wild-type myoblasts than for MyoD 2/2 myoblasts. Previous work demonstrated that MyoD 2/2 myoblasts possess remarkable resistance to apoptosis and increased survival compared to wild-type myoblasts after injection into regenerating skeletal muscle [29], [30]. Similar to myoblast transplantation into injured skeletal muscle, antibody staining for caspase-3 clearly showed massive apoptotic cell death of the engrafted myoblasts 3 days after transplantation (Fig. 2C, D). This cell death was more prominent for wild-type myoblasts than MyoD 2/2 myoblasts. Therefore, it is likely that the immediate reduction in cell number is due to post-transplantation myoblast cell death. In addition, engrafted MyoD 2/2 myoblasts displayed higher rates of proliferation by day 3 (Fig. 2C, D). After 1 week, significantly more MyoD 2/2 myoblasts engrafted than wild-type myoblasts (MyoD 2/2 myoblasts 10.0861.41%, n = 5, wild-type myoblasts 3.1160.43%, n = 3, p,0.05) (Fig. 2B). By weeks 2 and 4, the engraftment of MyoD 2/2 myoblasts declined but remained significantly higher than the engraftment of wild-type myoblasts (MyoD 2/2 myoblasts 3.5160.59%, n = 5, wild-type myoblasts 0.8760.29%, n = 3, p,0.05, and MyoD 2/2 myoblasts 1.6860.23%, n = 5, wild-type myoblasts 0.5960.12%, n = 5, p,0.01, respectively). These results clearly demonstrate that larger numbers of MyoD 2/2 myoblasts survive in the infarcted heart at least by 4 weeks post-transplantation, and that a decrease in cell death and increase in cell proliferation may contribute to their higher engraftment rates.

Improved Cardiac Function after MyoD 2/2 Myoblast Transplantation
The higher engraftment rates of MyoD 2/2 myoblasts in recipient heart suggested that they may improve cardiac function more efficiently than wild-type myoblasts. Table 2 and Fig. 3A show the results of echocardiographic assessment 2 and 4 weeks after MI with or without cell transplantation. LV end-diastolic dimensions were significantly lower for hearts transplanted with MyoD 2/2 and wild-type myoblasts compared to hearts that received medium alone. In addition, LV end-systolic dimensions were also significantly lower at 2 and 4 weeks for mice receiving MyoD 2/2 and wild-type myoblasts compared to mice receiving medium alone. This result suggests that LV dilation seen in infarcted hearts receiving medium alone may have been improved as a result of myocardial myoblast transplantation. Myocardial infarction frequently produces LV dilation associated with myocyte hypertrophy and interstitial fibrosis outside the injured myocardium. These changes in LV geometry, referred to as remodeling, contribute to the development of depressed cardiac performance [33]. Importantly, the MyoD 2/2 group showed increased ejection fraction compared to the wild-type group by 2 and 4 weeks (MyoD 2/2 myoblasts 44.661.1%, n = 10, wild-type myoblasts 39.861.5%, n = 9, p,0.05, and MyoD 2/2 myoblasts 43.661.6%, n = 8, wild-type myoblasts 38.160.8%, n = 7, p,0.05, respectively) without any significant differences in the MI sizes among the 3 groups (Table 2 and Fig. 3B). These results suggest that MyoD 2/2 myoblasts improve cardiac systolic function more efficiently than wild-type myoblasts at least by 4 weeks after transplantation. Next, we examined the fate of engrafted myoblasts following myocardial injection (Fig. 4A, B). First, an anti-nestin antibody was used to detect skeletal muscle-derived cells after lacZ staining. Nestin is a component of intermediate filaments expressed in myoblasts and regenerating skeletal muscle fibers but not in the cardiomyocytes of adult mice [34], [35], [36]. The majority of lacZ-positive MyoD 2/2 and wild-type myoblast-derived progeny formed nestin-positive multinucleated skeletal myotubes in the infarcted heart by 2 weeks (wild-type myoblasts 94.761.6%, MyoD 2/2 myoblasts 88.761.5%, n = 3) (Fig. 4A, B). These nestin-positive skeletal myotubes also expressed laminin, a marker of the basal lamina in both skeletal muscle fibers and cardiomyocytes (Fig. 4A). To examine whether engrafted myoblasts can give rise to cardiomyocytes in recipient hearts, we performed immunostaining against several cardiomyocyte markers 2 weeks after cell transplantation. Sections were examined to identify nuclear lacZ/DAPI positive cells co-expressing cardiomyocyte markers. Very few cells derived from MyoD 2/2 myoblasts coexpressed the cardiac sarcomeric protein cardiac-troponin T (cTnT) and the cardiac transcription factor myocyte enhancer factor-2 (MEF2) after transplantation (Fig. 4C). These results suggest that the majority of engrafted wild-type and MyoD 2/2 myoblasts form skeletal muscle fibers upon transplantation into damaged heart tissue. Therefore, improved cardiac function in infarcted heart following myoblast transplantation is not due to the contribution of cardiomyocyte-like cells.

Increased Angiogenesis after Myoblast Transplantation in Infarcted Heart
Recent work suggests that a variety of stem cell types can repair heart function through paracrine effects that induce angiogenesis in the injured heart [22], [3]. Therefore, we examined whether the injection of myoblasts could increase angiogenesis 1 week after transplantation (Fig. 5A, B). The number of CD31 + blood vessels in the injured area was greater after transplantation of MyoD 2/2 myoblasts (656.8627.0%, n = 8) than after transplantation of wildtype myoblasts (562.8623.2%, n = 8) or medium alone (437.3625.1%, n = 7). FACS analysis was performed to measure the number of CD45 2 CD31 + endothelial cells 1 week after MI (Fig. 5C, D). Clearly, MI injected with medium alone induced a marked decrease in the number of CD45 2 CD31 + endothelial cells compared to sham heart (no ligation) (Medium: 22.761.7%, No ligation, 53.162.7%, n = 6). Myoblast injection increased the number of CD45 2 CD31 + endothelial cells in the infarct and border zones. In addition, the number of CD45 2 CD31 + endothelial cells in the infarct and border zones was greater for the MyoD 2/2 myoblast group than for the wild-type myoblast group (MyoD 2/2 myoblasts 32.961.0%, wild-type myoblasts 27.761.0%, n = 6, p,0.05). The infiltration of blood cells and cardiac progenitor cells is known to improve heart histology and function after MI [3]. Therefore, we used FACS analysis to measure the infiltration of CD45 + blood cells and CD45 2 CD31 2 Sca-1 + cardiac progenitor cells (Fig. 5C, D). Following MI, an increase in the infiltration of CD45 + and CD45 2 CD31 2 Sca-1 + cells was detected. However, the number of CD45 + and CD45 2 CD31 2 Sca-1 + cells was not significantly different between wildtype and MyoD 2/2 myoblast groups. Figure 2. Engraftment of wild-type and MyoD 2/2 myoblasts after myocardial injection into infarcted heart. (A) These panels show MI induced by left coronary artery ligation. Wild-type and MyoD 2/2 myoblasts were directly injected into the peri-infarct regions of LV. After 1 week, Xgal staining of whole heart indicated that more MyoD 2/2 myoblasts engrafted than wild-type myoblasts (arrows). Arrowheads indicate left coronary artery ligation points. X-gal staining of cross sections indicated that more MyoD 2/2 myoblasts than wild-type myoblasts engrafted in both injured and uninjured areas of the heart. Arrows indicate engrafted lacZ + wild-type and MyoD 2/2 myoblasts. Scale bars = 1 mm. (B) Engraftment of injected cells in the infarcted heart. More MyoD 2/2 myoblasts than wild-type myoblasts engraft in the infarcted heart 3 day to 4 weeks post-transplantation. Yaxis indicates the survival rates of engrafted cells after an injection of 1610 6 wild-type or MyoD 2/2 myoblasts. (C) By day 3, anti-activated caspase-3 (Casp-3) and anti-phospho-histone H3 (p-His-H3) antibody staining demonstrated that lacZ + wild-type and MyoD 2/2 myoblasts proliferate or undergo apoptotic cell death (n = 4). Nuclei were counter-stained with DAPI (blue). Scale bar = 50 mm. (D) Comparison of the relative numbers of lacZ + /activated caspase-3 + or lacZ + /phospho-histone H3 + cells for wild-type and MyoD 2/2 myoblasts 3 days after transplantation. doi:10.1371/journal.pone.0041736.g002 Co-culture with MyoD 2/2 Myoblasts Stimulates Cell Proliferation of Endothelial Cells MyoD 2/2 myoblast transplantation induced greater angiogenesis in infarcted heart compared to wild-type myoblasts, suggesting that the MyoD 2/2 muscle niche may promote greater endothelial cell proliferation and/or survival compared to wild-type muscle. To examine this possibility, we compared vascular densities between wild-type and MyoD 2/2 adult tibialis anterior skeletal muscles. The MyoD 2/2 muscle niche clearly displayed a larger number of CD31 + blood vessels compared to the wild-type muscle niche (MyoD 2/2 631.6621.1%, wild-type 572.6614.2%, n = 5, p,0.05) (Fig. 6A, B). To test for possible interactions between this skeletal muscle niche and endothelial cells, brain-derived endothelial (bEnd) cells were either cultured alone or co-cultured with mouse 10T1/2 fibroblasts, wild-type myoblasts or MyoD 2/2 myoblasts (Fig. 6C, D). Cells were grown in myoblast growth medium for 3 days and stained for the endothelial cell marker VEcadherin. bEnd cells co-cultured with MyoD 2/2 myoblasts (322.3611.2%, n = 3) displayed more VE-cadherin positive endothelial cells than bEnd cells co-cultured with wild-type myoblasts (263.369.5%, n = 3) or 10T1/2 cells (213.0615.5%, n = 3), indicating that co-culture with MyoD 2/2 myoblasts    cells from cell death in an ischemic environment, wild-type and MyoD 2/2 myoblasts were co-cultured with bEnd cells incubated under hypoxic conditions for 3 days (Fig. 6E, F). While cell numbers for both wild-type and MyoD 2/2 myoblasts decreased over time under hypoxic conditions, survival of MyoD 2/2 myoblasts was substantially higher (32.863.7%, n = 3 by day 3) compared to wild-type myoblasts (18.764.8%, n = 3, p,0.05 by day 3) (Fig. 6E). Therefore, MyoD 2/2 myoblasts are more resistant to hypoxia than wild-type myoblasts, which may account for the higher survival rate of MyoD 2/2 myoblasts after injection into ischemic heart. In contrast, bEnd cells are more sensitive to hypoxic culture conditions by day 3 (10.061.9%, n = 3) (Fig. 6E). Interestingly, bEnd cells had a significantly higher survival rate when co-cultured with MyoD 2/2 myoblasts (31.961.6%, n = 4, p,0.01) and wild-type myoblasts (22.961.8%, n = 4, p,0.05) compared with bEnd cells alone (12.562.8%, n = 4). Co-culture with MyoD 2/2 myoblasts led to increased rates of bEnd survival compared to wild-type myoblast co-culture (p,0.05, Fig. 6F). These results suggest that transplantation of MyoD 2/2 myoblasts can efficiently improve cardiac systolic function by increasing angiogenesis, as well as endothelial cell proliferation and survival.

Paracrine Angiogenic Factors Expressed in MyoD 2/2 Myoblasts
To identify which molecular cascades are altered in MyoD 2/2 myoblasts, we re-visited data generated from Affymetrix Gene-Chip DNA microarrays [29]. Previously, we demonstrated that anti-apoptotic genes are up-regulated in MyoD 2/2 myoblasts, while genes known to execute apoptosis are down-regulated [29]. Table 3 and Fig. 7A show that MyoD 2/2 myoblasts up-regulate a number of growth factors including the angiogenic factors chemokine CXC3 motif ligand 1 (CX3CL1), fibroblast growth factor 7 (FGF7), brain-derived neurotrophic factor (BDNF), vascular endothelial growth factor-D (VEGF-D), stem cell factor (SCF), stromal cell-derived factor-1 (SDF-1) and placental growth factor (PlGF) but down-regulate other growth factors such as insulin like growth factor binding protein 2 (IGFBP2) and 5 (IGFBP5). CX3CL1 is a chemokine involved in inflammation [37]. FGF7 is a mitogen for epithelial cells involved in wound healing [38]. BDNF is a trophic factor for neuronal development and synaptic plasticity [39]. SCF is a cyctokine for hematopoiesis [40]. VEGF-D and PlGF are a VEGF family involved in vasculogenesis and angiogenesis [41]. SDF-1 is involved in the chemo-attraction of many types of stem cells [42]. IGFBP2 and 5 are binding partners of insulin like growth factor (IGF), which inhibits IGF activity [43]. We confirmed the GeneChip microarray data for these growth factor-related genes using semiquantitative RT-PCR (Fig. 7A). These results suggested that MyoD 2/2 myoblasts have a greater survival rate after transplantation due to alterations in apoptotic cascades, the expression of angiogenic factors, and the induction of angiogenesis. We also examined HIF1a expression, a hypoxia inducible transcription factor that regulates VEGF gene expression. However, there was no difference in HIF1a gene expression between wild-type and MyoD 2/2 myoblasts.
Next, we examined whether the acute loss of MyoD in wild-type myoblasts by gene knockdown (KD) is sufficient to induce expression of these angiogenic factors and whether ectopic expression of MyoD in MyoD 2/2 myoblasts can suppress expression of these genes. RT-PCR showed that wild-type myoblasts infected with a lentiviral vector expressing shRNA for MyoD exhibited an extensive reduction in MyoD expression compared to myoblasts infected with a control shRNA vector (Fig. 7B). Importantly, for wild-type myoblasts expression of angiogenic genes (CX3CL1, FGF7, BDNF, VEGF-D, SDF-1 and PlGF) was efficiently induced by MyoD-KD. By contrast, RT-PCR showed that infection with a lentiviral vector expressing MyoD effectively rescued MyoD expression in MyoD 2/2 myoblasts (Fig. 7 B). Ectopic expression of MyoD in MyoD 2/2 myoblasts resulted in a marked reduction in the expression of these angiogenic factors. These results suggest that MyoD is a negative regulator of these angiogenic genes.

SDF-1 and PlGF can Stimulate Endothelial Cell Proliferation and Survival
Microarray data and RT-PCR demonstrated that CX3CL1, FGF7, BDNF, VEGF-D, SDF-1 and PlGF genes are up-regulated in MyoD 2/2 myoblasts. To examine whether these angiogenic factors are able to stimulate angiogenesis, bEnd cells were cultured with these angiogenic factors for 3 days (Fig. 8A). Incubation with SDF-1 and PlGF had a stronger angiogenic effect (131.568.0% n = 3, p,0.05 and 137.365.4% n = 3, p,0.05, respectively) compared to control (10063.6% n = 3). Next, to test whether SDF-1 and PlGF have an influence on the cell survival of cultured endothelial cells, bEnd cells were cultured in medium containing the angiogenic factors under hypoxic conditions for 1-3 days (Fig. 8B). Under hypoxic conditions, SDF-1 and PlGF induced greater survival of bEnd cells (by day 3, 23.662.1% n = 3, p,0.05 and 27.361.3% n = 3, p,0.01, respectively) compared to culture medium alone (by day 3, 11.961.0% n = 3). These results suggest that MyoD 2/2 myoblasts are able to improve cardiac function after transplantation into infarcted heart due, in part, to the secretion of the paracrine angiogenic factors SDF-1 and PlGF and the resulting induction of angiogenesis in the area of injury.

Discussion
Patient specific skeletal myoblasts can be isolated using muscle biopsies and expanded ex vivo before autologous cell transplantation. In addition, myoblasts have been shown to engraft and form skeletal muscle fibers in recipient hearts. Although myoblasts are unable to couple electromechanically with the host myocardium, they can partially improve heart function by increasing angiogenesis, decreasing fibrosis, and directing the differentiation of other muscle fibers [44], [45], [12].
MyoD, a skeletal muscle-specific master transcription factor, plays an essential role in myogenic determination, and myoblasts lacking MyoD may represent an intermediate stage between quiescent satellite cells and myogenic precursors. Recently, our laboratory demonstrated that MyoD 2/2 myoblasts possess remarkable resistance to apoptosis and increased cell survival compared to wild-type myoblasts after injection into injured skeletal muscle [28], [29], [30].
In this study, we demonstrate that MyoD 2/2 myoblasts are a useful cell source for myocardial repair, and that they display higher rates of survival than wild-type myoblasts in vivo following transplantation into an infarcted heart, as well as in vitro under hypoxic conditions. Recently, Di Carlo et al. showed that hypoxia-an increase in the number of CD45 + blood cells (arrows) and CD45 -CD31 2 Sca-1 + cardiac progenitor cells (black boxes) and a decrease in the number of CD45 2 CD31 + endothelial cells (red boxes) in the hearts of mice receiving wild-type myoblasts, MyoD 2/2 myoblasts, or medium alone. (D) The total number of CD45 + blood cells, CD45 -CD31 2 Sca-1 + cardiac progenitors, and CD45 2 CD31 + endothelial cells was estimated using FACS data (n = 3). doi:10.1371/journal.pone.0041736.g005 dependent inhibition of skeletal muscle differentiation was associated with MyoD degradation by the ubiquitin-proteasome pathway [46]. In addition, hypoxic conditions inhibit permanent withdrawal of myoblasts from the cell cycle by producing a marked decrease in MyoD, Myf5 and myogenin levels. Recently, we demonstrated that MyoD 2/2 myoblasts up-regulate many antiapoptotic genes including Pax3, Bcl-2 and Bcl-xL. It is possible that the delayed differentiation exhibited by MyoD 2/2 myoblasts protects against the high levels of apoptotic cell death displayed by wild-type myoblasts undergoing terminal differentiation [29], [30], [47]. In addition to an increase in cell survival, we have shown that transplantation of MyoD 2/2 myoblasts provides functional benefits to damaged heart tissue. Mice receiving MyoD 2/2 myoblasts displayed a greater improvement in ejection fraction by 4 weeks after transplantation, compared to the wild-type myoblasts. In the future, it will be important to determine how long this functional benefit lasts.
Transplantation of skeletal myoblasts into infarcted heart has consistently been shown to improve cardiac function [12]. Transplanted myoblasts survive, proliferate, and eventually differentiate into skeletal muscle fibers but fail to produce cardiomyocytes. In accordance with previous studies, we report here that engrafted myoblast progeny lack the expression of cardiac specific markers, and the majority of MyoD 2/2 myoblastderived cells express nestin and laminin, markers of skeletal muscle myotube differentiation. Although recent work demonstrated that Wnt11 gene transfer enhances cardiomyogenic differentiation of skeletal muscle-derived stem cells [48], it is likely that engrafted myoblasts express cardiomyocyte-markers due to fusion with host cardiomyocytes in the recipient heart [49].  Table 3. Different expression in secreted factor genes in between MyoD -/and wild-type myoblasts.   Although mechanisms leading to functional heart improvement remain unclear, transplanted myoblasts may initiate remodeling of the host tissue by exerting important paracrine effects on the surrounding myocardium [50], [51], [45], [23], [44]. It is possible that transplanted myoblasts, which eventually differentiate into multinucleated skeletal muscle fibers, also exert active contractile effects [52]. However, since skeletal muscle differentiation initiated after myocardial injection may induce cardiac arrhythmia [53], it needs to be cautious to use of skeletal myoblasts for myocardial injection. Importantly, the paracrine effects of angiogenic factors, rather than cardiomyocyte differentiation, must play an important role in cell-based therapy for infarcted heart [54], [55]. Cells shown to produce this result upon transplantation include skeletal myoblasts [44], [45], [56], [23], bone marrow-derived cells [3], [57], [58], mesenchymal stem cells [59], [60], human CD133 + cells [61], human cord blood progenitors [62] and human CD34 + cells [63]. The release of paracrine factors by transplanted cells may exert a protective effect by stimulating angiogenesis within the infarcted and non-infarcted regions. In turn, this increased angiogenesis may prevent cardiomyocytes from undergoing apoptotic cell death, protecting the infarcted heart from further cardiac remodeling and scar tissue formation.
Paracrine factors provide protection to injured heart by eliciting VEGF production, increasing vascular density, increasing blood flow, and decreasing endothelial cell apoptosis [64], [65]. PlGF, a growth factor belonging to the VEGF family, plays a critical role in pathologic angiogenesis. Recent work demonstrated that PlGF expression is up-regulated after myocardial damage, and that PlGF administration can facilitate cardiac repair after myocardial ischemia [66]. In addition, SDF-1 has been shown to induce stem cell homing to injured heart post-MI and induce angiogenesis [67], [68]. An interaction between satellite cells and endothelial cells within skeletal muscle was recently reported [69], and it is possible that reciprocal signals from satellite and endothelial cells may be required for appropriate in vivo cellular function.
Importantly, our work demonstrates that hearts receiving MyoD 2/2 myoblasts display an increase in vascular density compared to hearts receiving wild-type myoblasts or medium alone. These results suggest that transplantation of MyoD 2/2 myoblasts can efficiently improve cardiac systolic function through increased angiogenesis. Supporting this idea, Affymetrix DNA microarrays and RT-PCR demonstrated that MyoD 2/2 myoblasts up-regulate a number of growth factors including the angiogenic factors CX3CL1, FGF7, VEGF-D, SDF-1 and PlGF. In addition, in vitro cell culture experiments showed that SDF-1 and PlGF increase endothelial cell survival and proliferation. MyoD directly regulates the transcription of microRNA expression, suppressing specific targets during myogenic differentiation by blocking protein translation and/or degrading RNA, and it is possible that angiogenic genes up-regulated in MyoD 2/2 myoblasts contain MyoD-regulated microRNA binding sites [70], [30]. Regardless, the substantial increase in the survival of MyoD 2/2 myoblasts after transplantation provides reason to believe they may improve cardiac function more efficiently than their wild-type counterparts.
Recently, Menasche's group demonstrated that low myoblast doses did not improve the regional or global LV function of patients with MI occurring 4 weeks before transplantation [71]. Notably, high doses of myoblasts significantly reduced LV volumes compared to the placebo group, indicating the importance of protecting LV remodeling after myoblast transplantation. However, we used an acute injury model for myoblast transplantation, which differs significantly from Menasche's clinical trials.
Although the generation of human MyoD 2/2 myoblasts for clinical purposes is not possible, MyoD expression or function may be chemically, transcriptionally or post-translationally suppressed in order to recreate the beneficial effects of MyoD 2/2 myoblast transplantation shown here. Bromodeoxyuridine (BrdU), a thymidine analogue, is known to be a strong inhibitor of muscle differentiation through its suppression of MyoD gene expression [72]. Alternatively, suppression of MyoD may be achieved by infection of patient myoblasts with a lentivirus carrying an RNA interference (RNAi) or Id protein expression vector [30]. Id proteins are essential dimerization partners of the MyoD family of transcription factors, and act to suppress their effector functions [73].
In conclusion, our results suggest that transplanted MyoD 2/2 myoblasts are superior to wild-type myoblasts for myocardial cell therapy. These results represent a novel approach to myoblast transplantation, in which the ablation of MyoD expression increases angiogenesis, cell survival, and the functional recovery of infarcted heart.

Animals
Eight to 12-week-old BALB/c female mice weighing 17-20 g were obtained from Harlan Sprague Dawley. BALB/c wild-type and MyoD 2/2 myoblasts were transplanted into BALB/c recipient mice since immune rejection in this system has been reported to be minimal in short term transplantation experiments [74]. The animals were housed in an SPF environment and were monitored by the Research Animal Resources (RAR) of the University of Minnesota. All protocols were approved by the Animal Care and Use Committee (IACUC, Code Number: 1003A79635) of the University of Minnesota.

Myocardial Infarction (MI) and Cell Transplantation
MI was induced by left coronary artery ligation as previously described [78]. Briefly, the left coronary artery was permanently ligated using a 9-0 nylon surgical suture, resulting in an infarct size equivalent to 30% of the mass of the LV. Sham operations were performed by following the same protocol but omitting the ligation step. Immediately after the left coronary artery ligation, viable mice were randomized into the following 3 groups: (1) MyoD 2/2 myoblast recipients; (2) wild-type myoblast recipients; and (3) medium only recipients. For the MyoD 2/2 and wild-type myoblast groups, 1610 6 cells were re-suspended in DMEM supplemented with 2% FBS and injected directly into the peri-infarct region of the myocardium 3-4 times using a 32-gauge needle. After injection, the mice were allowed a recovery period ranging from 1 day to 4 weeks.

Echocardiographic Studies and Measurement of MI Size
Mice were lightly anesthetized with ketamine HCl (50 mg/kg, IP) and xylazine (16.5 mg/kg, IP). Echocardiography was performed using a SONOS 5500 echocardiographic system equipped with a 15.6-MHz phased-array transducer (Phillips, Netherlands). A two-dimensional short-axis view of the LV was obtained at the level of the papillary muscles. The LV internal dimensions, including end-diastolic dimensions (LVDd) and endsystolic dimensions (LVDs), were measured by the leading-edge method using a minimum of 8 consecutive cardiac cycles on the short-axis view of the LV. Ejection fraction (EF) was calculated using the following formula: EF = (LVDd 2 2LVDs 2 )/ LVDd 2 6100. Following echocardiography, mice were euthanized and their hearts removed. MI size is expressed as a percentage of LV surface area as previously described [78].

Evaluation of Engrafted Cell Number in the Infarcted Heart
Whole heart samples were stained with X-gal (Invitrogen) overnight as previously described [29]. Following X-gal staining, hearts were embedded in Tissue-Tek OCT compound (Fisher Scientific) and frozen in liquid nitrogen-cooled isopentane. A cryostat was used to produce 8 mm transverse tissue sections. Cell nuclei were counter-stained with 49,6-diamidino-2-phenylindole (DAPI, Sigma-Aldrich). The number of engrafted cells was determined by examining 10 serial sections of whole heart for X-gal and DAPI double positive nuclei.

Measurement of Vascular Density
Vascular density was measured by anti-CD31 antibody (BD Pharmingen) staining for 3 to 5 randomly selected non-overlapping fields in each cross section of infarct scar area of heart 1 week after myoblast injection, or tibialis anterior (TA) muscle of wildtype and MyoD 2/2 mice. The vascular density was calculated as the total number of CD31 + capillaries per section as previously described [79].

Fluorescent Activated Cell Sorting (FACS)
Dissociated cells were prepared from the border and infarct areas of damaged heart and from sham mouse heart after digestion with collagenase type B and dispase II (Roche Diagnostics) [30], [76]. FACS analysis was performed using a FACS Calibur (BD Biosciences) equipped with double lasers. The following antibodies were used: allophycocyanin (APC)-labeled anti-CD31, phycoerythrin (PE)-labeled anti-Sca-1, and fluorescein isothiocyanate (FITC)-labeled anti-CD45 (all from BD Pharmingen). APClabeled anti-rat IgG, PE-labeled anti-rat IgG, and FITC-labeled anti-rat IgG were used for control experiments (all from BD Pharmingen). Alexa 488 and PE were excited by a 488 nm argon laser and detected using FL1 (530/30) and an FL2 (576/26) filters, respectively. APC was excited by a 633 nm red diode laser and detected using an FL4 filter (620/20). Gates were strictly defined based on single antibody-stained control cells as well as the forward scatter (FSC) and side scatter (SSC) patterns.

Hypoxic Cell Culture
Hypoxia was established in anaerobic culture jars using a GasPak system (Becton Dickinson) with a palladium catalyst capable of producing a 0.2% oxygen level within 24 hours as previously described [80]. bEnd cells co-cultured with 10T1/2 cells, wild-type or MyoD 2/2 myoblasts were maintained under normal and hypoxic conditions for 1-3 days. Viable cells were counted after Trypan Blue (Invitrogen) staining. Endothelial cell survival was calculated as the number of X-gal positive cells after bEnd cell infection with a pMX-lacZ vector. For hypoxic conditions, twice as many wild-type myoblasts were used because of their relatively low survival rate compared to MyoD 2/2 myoblasts (half the survival rate after 3 days of culture under hypoxic conditions). Myoblasts were detected using an anti-Myf5 antibody (Santa Cruz Biotechnologies) and a Vectastain Elite ABC Kit (Vector Laboratories).

Microarray Data Analysis
The microarray data used in this study was previously produced and described [29]. Briefly, total RNA was isolated from 2 independently prepared low passage (passages 6-8) myoblast cultures (wild-type; n = 2, MyoD 2/2 ; n = 2) by acid-phenol extraction. Mu211K Affymetrix arrays contain approximately 13,000 genes and expressed sequence tags (ESTs). Data analysis was performed using GeneSpring 3.2.2 software (Silicon Genetics, Redwood City, CA) as previously described [29]. Each MyoD 2/2 chip (n = 2) was compared to a wild-type chip (n = 2) to examine disparities in gene expression. Difference calls demonstrating reproducible results in average comparisons were extracted for further analysis.

Semi-quantitative RT-PCR
Total RNA was isolated from cells and tissues using TRIZOL (Invitrogen). Purified RNA was reverse-transcribed (Roche: Transcriptor First Strand cDNA Synthesis Kit) followed by 20-35 PCR cycles (Eppendorf Thermal Cycler) using the gene specific primer pairs described in Table 1. Optimal PCR cycles for each pair were determined by several different amplifications of the PCR products. Quantitative analysis was performed using Image-J software from the NIH. Relative expression was calculated using an internal b-actin control.

Statistics
All data are expressed as mean 6 SEM. For transplantation experiments, differences among groups were compared by a twotailed Student's t-test with a two-sample equal variance. A repeated measures two-way analysis of variance (ANOVA) and Bonferroni post-hoc measures were used to determine differences in cardiac function, angiogenesis and cell number between control and experimental groups with time as the repeated factor. Asterisks or double asterisks indicate experimental pairs where differences between the compared values were statistically significant (p,0.05) or (p,0.01), respectively.