Adenosine A_2B Receptors on Cardiac Stem Cell Antigen (Sca)-1–Positive Stromal Cells Play a Protective Role in Myocardial Infarction

Abstract

Transplantation of mesenchymal stem-like cells to the heart is known to improve cardiac recovery in animal models of myocardial infarction (MI). Because stimulation of A_2B adenosine receptors on mouse cardiac stem cell antigen (Sca)-1⁺CD31⁻ mesenchymal stem-like cells significantly up-regulates their secretion of pro-angiogenic factors, we hypothesized that ablation of the A_2B receptor signaling in these cells would reduce their ability to improve vascularization of the infarct area seen after transplantation. Wild-type (WT) C57BL/6 mice underwent permanent left coronary artery ligation and received intramyocardial injections of Sca-1⁺CD31⁻ cells generated from WT or A_2B receptor knockout (A_2BKO) mice or the same volume of cell-free saline. Only 12% to 16% of injected cells remained in the ventricles 1 week later; there was no significant difference between WT and A_2BKO cell survival. Transplantation of WT, but not A_2BKO, cells significantly reduced both post-MI decline in cardiac function and adverse remodeling compared with that seen in control hearts. Morphological analysis conducted 4 weeks after MI revealed significantly increased vascularization of the infarct areas and reduced myocardial scarring in animals treated with WT, but not with A_2BKO, cells compared with control. Thus, our study demonstrated that the A_2B receptor signaling linked to up-regulation of pro-angiogenic factors in cardiac Sca-1⁺CD31⁻ stromal cells is essential for overall improvement of cardiac recovery seen after their transplantation to the injured heart.

Mesenchymal stem-like cells of various tissue origins have been proposed to be used in cell-based transplantation therapy to enhance tissue repair and functional recovery after myocardial infarction (MI). Among them, cardiac mesenchymal stem-like cells have been consistently shown to possess superior paracrine potency and myocardial protection efficacy compared with stem/progenitor cells originated from other tissues.^1,2 In the mouse heart, these cells are represented by a population of stromal cells characterized by the expression of stem cell antigen (Sca)-1 on their surface and the absence of the endothelial cell surface marker, CD31.^3–9 Several groups independently reported that the delivery of cardiac Sca-1⁺CD31⁻ cell populations to the heart resulted in improved revascularization of injured tissue and attenuated decline of cardiac function in animal models of MI.^6,9–11 Although the precise mechanism of these protective effects remains unknown, the early assumption that these cells can replace damaged cardiomyocytes has recently given way to the realization that they also, and perhaps mainly, exert a beneficial effect via the release of paracrine factors.^3,12–17 As such, the beneficial properties of transplanted cells are likely regulated by local factors present in the ischemic tissue, including high levels of extracellular adenosine.

Adenosine is an endogenously produced signaling molecule that binds to the family of extracellular G-protein–coupled seven transmembrane receptors, which include A₁, A_2A, A_2B, and A₃ subtypes. Adenosine concentrations in the myocardial interstitium remain low in normal conditions. Myocardial ischemia is known to considerably increase extracellular adenosine levels.^18,19 MI results in even larger levels of extracellular adenosine because of its massive release from damaged cells and catabolism of coreleased adenine nucleotides. Subsequent inflammation in the injured myocardium may further increase extracellular adenosine levels as the result of cell stress and tissue hypoxia.²⁰ Therefore, extracellular adenosine becomes an important part of the tissue microenvironment generated by myocardial ischemia and infarction.

Recently, we found that stimulation of A_2B adenosine receptors on mouse cardiac Sca-1⁺CD31⁻ stromal cells significantly increased secretion of pro-angiogenic factors.²¹ In this study, we tested the hypothesis that the A_2B receptor signaling in these cells is important for their therapeutic effects seen after transplantation to the heart after MI.

Materials and Methods

Mouse Cardiac Sca-1⁺CD31⁻ Stromal Cells

Wild-type (WT) and A_2B receptor knockout (A_2BKO) conditionally immortalized cardiac Sca-1⁺CD31⁻ stromal cell lines were generated, as described previously,²¹ from H-2K^b-tsA58 transgenic mice carrying a thermolabile T antigen crossed with WT and A_2BKO mice, respectively. Cells were propagated on 0.1% gelatin-coated tissue culture dishes in Dulbecco's modified Eagle's medium (high glucose) (Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum, 1× antibiotic-antimycotic solution, 2 mmol/L glutamine (all from Sigma, St. Louis, MO), and 10 ng/mL of interferon-γ (R&D Systems, Minneapolis, MN) under a humidified atmosphere of air/CO₂ (19:1) at a low temperature (33°C). Six days before experiments, cells were replated and cultured in the absence of interferon-γ at a higher temperature (37°C) to allow them to revert to their primary phenotype, as described previously.²¹ Immediately before surgery, cells were harvested by treatment with Accutase-Enzyme Cell Detachment Medium (eBioscience, San Diego, CA), washed, and resuspended in PBS for transplantation. In some experiments, cells were detached, incubated at a concentration of 10⁶ cells/mL in Dulbecco's modified Eagle's medium in the presence of 1 μL/mL of Vybrant 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindodicarbocyanine iodide (DiD) cell-labeling solution (Life Technologies, Grand Island, NY) for 20 minutes, washed, and replated 72 hours before surgery. In parallel, a portion of harvested cells was resuspended in Dulbecco's modified Eagle's medium to analyze adenosine-dependent vascular endothelial growth factor (VEGF) secretion.

Flow Cytometry

Cells were analyzed either freshly isolated from ventricles or after culture, as described previously,²¹ using an LSRII flow cytometer (Becton Dickinson, Franklin Lakes, NJ).

Analysis of Adenosine-Dependent VEGF Secretion

Harvested Sca-1⁺CD31⁻ cells were cultured in 6-well plates overnight at 37°C to allow cell attachment. Cells were washed and incubated in a fresh medium in the presence or absence of the stable adenosine analogue 5′-N-ethylcarboxamidoadenosine (NECA; 10 μmol/L) and 1 U/mL adenosine deaminase (both from Sigma). After collection of culture media, the cells were detached and counted. VEGF concentrations in culture media were measured using enzyme-linked immunosorbent assay kits (R&D Systems).

MI and Cell Transplantation

All studies were conducted in accordance with the Guide for the Care and Use of Laboratory Animals, as adopted and promulgated by the NIH (Bethesda, MD). Animal studies were reviewed and approved by the Vanderbilt Institutional Animal Care and Use Committee. Surgical procedures to produce MI in mice were performed in the Cardiovascular Pathophysiology and Complications Core of the Vanderbilt University Mouse Metabolic Phenotyping Center, as previously described.²² Male 10- to 12-week-old WT C57BL/6 mice (Harlan World Headquarters, Indianapolis, IN) were used in all experiments. MI was produced by permanent ligation of the left coronary artery. A total of 2.5 × 10⁵ WT or A_2BKO cardiac Sca-1⁺CD31⁻ stromal cells, suspended in 25 μL of PBS, were injected into the peri-infarct area. Equal volumes of cell-free PBS were injected in the same manner to a control group of animals.

Echocardiography

Echocardiography was performed on unsedated mice using a Vevo 2100 Imaging System equipped with an 18- to 38-MHz linear-array transducer (MS400; VisualSonics, Inc., Toronto, ON, Canada). Images were acquired at the rate of 300 frames per second. Left ventricular M-mode tracings were obtained at the level of papillary muscles using the two-dimensional parasternal short axis imaging as a guide. All image acquisitions, left ventricular measurements, and calculations of functional parameters were conducted in the Cardiovascular Pathophysiology and Complications Core of the Vanderbilt University Mouse Metabolic Phenotyping Center by an experienced echocardiographer who was blinded to animal groups.

Determination of Scar Size

To measure infarct size, excised hearts were immersion fixed in 10% buffered formalin for 24 hours and transferred to 70% ethanol, after which ventricles were cut into three equal parts (in length) parallel to the atrioventricular groove. From each paraffin-embedded ventricular part, two transverse sections (5 μm thick) were made at different levels (for a total of six sections). After performing Masson-trichrome staining, infarct sizes were determined by the midline length measurement method.²³ All measurements were conducted using ImageJsoftware version 1.45s (NIH) in a blinded manner.

Determination of Vascular Density in the Infarct Area

Paraffin sections at the midventricular level, adjacent to those used for Masson-trichrome staining, were processed for immunoperoxidase staining of endothelial cells using rabbit anti-mouse von Willebrand factor (Sigma) as a primary antibody to visualize blood vessels, as previously described.²⁴ One medium-power (×10 magnification) field in the central and two ×10 fields in the marginal areas of the scar from each slide were selected, and digitized images of these fields (0.88 × 0.67 mm) were used to count all stained vascular structures in a blinded manner. The tissue area was measured in each field, and vascular density was expressed as number of blood vessels per mm² of tissue.

Statistical Analysis

Data were analyzed using GraphPad Prism software version 4.0 (GraphPad Software Inc., San Diego) and presented as means ± SEM. Comparisons between several treatment groups were performed using one-way analysis of variance, followed by appropriate posttests. Comparisons between two groups were performed using two-tailed unpaired t-tests. P < 0.05 was considered significant.

Results

Sca-1⁺CD31⁻ cells used in this study carry a gene encoding the thermolabile T antigen (tsA58Tag), which can be detected by flow cytometry after staining with the monoclonal antibody Pab108 (Pharmingen, San Diego). Figure 1, A and B, shows that >80% of cells under permissive conditions (33°C with interferon-γ) are stained with Pab108, whereas <8% of cells can be stained with Pab108 after 48 hours of culture at a higher temperature of 37°C. No difference in Pab108 staining was seen between WT and A_2BKO cells (Figure 1B). Therefore, we used the temperature-dependent Pab108 staining to trace the fate of injected cells in further experiments. In addition, cells were labeled with the lipophilic cell tracer, DiD. Figure 1, C and D, demonstrates identical levels of DiD fluorescence in labeled WT and A_2BKO cells before their injection. To ensure that coupling of adenosine A_2B receptors to paracrine functions is preserved in cells prepared for transplantation experiments, we tested their response to the stable adenosine analogue, NECA, using VEGF release as a marker. Indeed, only WT, but not A_2BKO, Sca-1⁺CD31⁻ cells responded to stimulation with 10 μmol/L NECA by a significant increase in VEGF release (Figure 1E).

Figure 1.

Characterization of WT and A_2BKO Sca-1⁺CD31⁻ cells before transplantation. A: Representative cytofluorographic outlier contour plots of tsA58Tag expression determined with monoclonal Pab108 antibody (Pab108) in WT and A_2BKO Sca-1⁺CD31⁻ cells cultured under permissive (left panels) and nonpermissive (right panels) conditions. Isotype-matching antibody (Isotype) was used as a negative control. B: Graphic representation of data from flow cytometry analysis of intracellular tsA58Tag staining of WT and A_2BKO Sca-1⁺CD31⁻ cells cultured under permissive (closed bars) and nonpermissive conditions (open bars). Values are means ± SEM of three experiments. C: Flow cytometry histograms of WT (blue) and A_2BKO (red) Sca-1⁺CD31⁻ cells labeled with DiD. Shaded areas represent the fluorescence of unlabeled cells. D: Graphic representation of data from flow cytometry analysis of DiD labeling of WT and A_2BKO Sca-1⁺CD31⁻ cells. Values are means ± SEM of five experiments. MFI, mean fluorescence intensity. E: Up-regulation of VEGF release after stimulation of adenosine receptors in WT, but not A_2BKO, Sca-1⁺CD31⁻ cells. WT or A_2BKO cells were incubated in the absence (basal) or presence of 10 μmol/L NECA (+NECA) for the indicated times, and VEGF was measured in supernatants. Values are means ± SEM of three experiments. Asterisks indicate statistical difference compared with corresponding basal levels: ^∗∗P < 0.01, ^∗∗∗P < 0.001, unpaired two-tailed t-test.

To assess the role of A_2B receptors in the protective effects of Sca-1⁺CD31⁻ cells after MI, three groups of WT mice underwent permanent ligation of the left coronary artery and WT or A_2BKO cells (2.5 × 10⁵ in 25 μL of PBS), or the same volume of PBS without suspended cells was directly injected into the left ventricular myocardium. In an ancillary study, survival of injected cells was evaluated by flow cytometry 1 week after surgery. In myocyte-free single-cell suspensions isolated from ventricles injected with WT or A_2BKO Sca-1⁺CD31⁻ cells, DiD⁺ cells were found only within the CD31⁻CD45⁻ cell population. DiD⁺ cells also retained their Sca-1 expression (Figure 2A). Approximately 12% of injected DiD-labeled cells still remained in the ventricles; no difference in DiD⁺ cell numbers was found between hearts injected with WT and A_2BKO cells (Figure 2B). Of interest, we found significantly more endothelial (CD31⁺CD45⁻) cells in ventricles of hearts injected with WT, but not A_2BKO, cells compared with control (Figure 2C). In a complementary approach, we evaluated survival of injected cells by their expression of tsA58Tag induced by incubation under permissive conditions for 24 hours (Figure 2D). Only 4 ± 0.4 × 10⁴ cells in individual ventricle single-cell preparations, corresponding to approximately 16% of injected Sca-1⁺CD31⁻ cells, were stained with Pab108 under permissive conditions; no difference in Pab108⁺ cell numbers was found between hearts injected with WT and A_2BKO cells (Figure 2E).

Figure 2.

A_2B receptor signaling in Sca-1⁺CD31⁻ stromal cells has no effect on their survival 1 week after transplantation but results in more endothelial cells in the infarcted hearts. A: Representative cytofluorographic outlier contour plots of CD31 and CD45 expression and Sca-1⁺DiD⁺ cells presented as percentage of CD31⁻CD45⁻ cells in single-cell suspensions obtained from ventricles injected with WT or A_2BKO Sca-1⁺CD31⁻ cells labeled with DiD, or cell-free PBS control. B: Total numbers of DiD-labeled WT and A_2BKO Sca-1⁺CD31⁻ cells calculated from data shown in A and total numbers of CD31⁻CD45⁻ cells in each ventricular single-cell suspension. Values are means ± SEM of nine animals in each group. C: Total numbers of CD31⁺CD45⁻ endothelial cells calculated from data shown in A and total numbers of cells in each ventricular single-cell suspension. Values are means ± SEM of nine animals in each group. Asterisks indicate significant difference compared with cell-free control: ^∗∗P < 0.01, one-way analysis of variance with Dunnett's posttest. D: Representative cytofluorographic outlier contour plots of tsA58Tag expression determined with monoclonal Pab108 antibody (Pab108) in single-cell suspensions obtained from ventricles injected with WT or A_2BKO Sca-1⁺CD31⁻ cells and cultured under nonpermissive (left panels) and permissive (right panels) conditions for 48 hours. Isotype-matching antibody (Isotype) was used as a negative control. E: Total numbers of WT and A_2BKO Sca-1⁺ cells expressing tsA58Tag under nonpermissive (open bars) and permissive (closed bars) conditions calculated from data shown in D and total numbers of cells in each ventricular single-cell suspension. Values are means ± SEM of five animals in each group.

Postinfarction changes in cardiac function in all experimental groups were assessed in a blinded manner using echocardiography (Figure 3, A and B). Changes in the left ventricular fractional shortening values (Figure 3C) were calculated from left ventricular internal dimension (LVID) end-diastolic (Figure 3D) and end-systolic (Figure 3E) measurements taken before and 1 and 4 weeks after MI. In agreement with previous reports on protective properties of cardiac Sca-1⁺CD31⁻ cells,^6,9–11 we found that transplantation of WT cells significantly reduced a decrease in fractional shortening observed 1 and 4 weeks after MI compared with PBS control. In contrast, transplantation of A_2BKO cells had no effect on the postinfarction decline in cardiac function seen in the control group (Figure 3C). Similar effects of cell transplantation on cardiac function were found by analysis of left ventricular ejection fraction (Figure 3F). Transplantation of WT, but not A_2BKO, cells also ameliorated adverse changes in LVID measurements, with significantly reduced end-diastolic (Figure 3D) and end-systolic (Figure 3E) LVID on postinfarction week 4 compared with that seen in PBS-treated control hearts.

Figure 3.

Transplantation of WT, but not A_2BKO, Sca-1⁺CD31⁻ cells significantly attenuated the deterioration of cardiac function after MI. A: Timeline of the study protocol. LCA, left coronary artery. B: Representative echocardiograms recorded 4 weeks after transplantations of WT or A_2BKO Sca-1⁺CD31⁻ cells, or vehicle control (PBS). C: LV fractional shortening was significantly improved 1 and 4 weeks after WT, but not A_2BKO Sca-1⁺CD31⁻, cell transplantation compared with cell vehicle control (PBS). D: End-diastolic left ventricular internal diameter (LVIDd) was significantly smaller 4 weeks after WT, but not A_2BKO, Sca-1⁺CD31⁻ cell transplantation compared with cell vehicle control (PBS). E: End-systolic left ventricular internal diameter (LVIDs) was significantly smaller 4 weeks after WT, but not A_2BKO, Sca-1⁺CD31⁻ cell transplantation compared with cell vehicle control (PBS). F: LV ejection fraction was significantly improved 4 weeks after WT, but not A_2BKO, Sca-1⁺CD31⁻ cell transplantation compared with cell vehicle control (PBS). Values are expressed as means ± SEM (n = 6 to 7 animals per group). Asterisks indicate significant difference compared with controls (PBS): ^∗P < 0.05, one-way analysis of variance with Dunnett's posttest.

In agreement with echocardiographic data, morphological analysis conducted 4 weeks after MI revealed significantly reduced myocardial scarring in the animals treated with WT cells compared with the control animal group. In contrast, there was even a tendency for an increase in scar size in the A_2BKO cell-treated group, albeit not statistically significant, when compared with the control PBS-treated group (Figure 4, A and C).

Figure 4.

Transplantation of WT, but not A_2BKO, Sca-1⁺CD31⁻ cells significantly reduced infarct scar size and increased vascular density in the infarct area. A: Representative Masson-trichrome–stained heart sections 4 weeks after transplantations of WT or A_2BKO Sca-1⁺CD31⁻ cells, or vehicle control (PBS). B: Representative micrographs of infarcted areas stained with Masson-trichrome (top panels) or immunostained for von Willebrand factor (brown staining in the middle and bottom panels) 4 weeks after transplantations of WT or A_2BKO Sca-1⁺CD31⁻ cells, or vehicle control (PBS). Boxed areas indicate regions of magnification. Scale bar = 200 μm (middle panels). C: Infarct scar size calculated as percentage of the total midline circumference averaged from six sections per heart. See Materials and Methods for details. Values are expressed as means ± SEM (n = 6 to 7 animals per group). D: Vascular density calculated as a number of blood vessels per mm² of scar tissue. Entire areas of middle-power fields (middle panels) in B were systematically analyzed. Original magnification, ×10. See Materials and Methods for details. Values are expressed as means ± SEM (n = 6 to 7 animals per group). Asterisks indicate significant difference compared with control (PBS): ^∗P < 0.05, ^∗∗P < 0.01, one-way analysis of variance with Dunnett's posttest.

Finally, we examined if the lack of A_2B receptor–mediated pro-angiogenic signaling in the transplanted cardiac Sca-1⁺CD31⁻ stromal cells would translate into reduced stimulation of revascularization of the infarcted area. Figure 4B shows representative micrographs (×4 magnification) of Masson-trichrome–stained infarcted areas, with boxed areas corresponding to the fields (×10 magnification) where all vascular structures identified by von Willebrand factor immunostaining were systematically counted. Magnified images (×40 magnification) of immunostained blood vessels are also presented in Figure 4B. In agreement with previous reports,^6,9–11 we found that transplantation of WT Sca-1⁺CD31⁻ cells significantly increased vascularization of the infarcted areas compared with that seen in the control PBS-treated hearts. In contrast, the vascular density in the infarcted areas of hearts injected with A_2BKO cells was no different from that seen in the control PBS-treated group (Figure 4D). Taken together, our data suggest that A_2B receptor–dependent regulation of pro-angiogenic properties of cardiac Sca-1⁺CD31⁻ stromal cells may play an important role in neovascularization and cardioprotection after MI.

Discussion

Tissue damage and inflammatory reactions after MI trigger release and generation of adenosine, which accumulates at high levels in the extracellular compartment. There is growing evidence that adenosine can modulate multiple facets of myocardial injury and remodeling via binding to the low-affinity A_2B adenosine receptor subtype. Adenosine A_2B receptors have been shown to promote neovascularization in tumors,²⁵ lung,²⁶ retina,²⁷ and skeletal muscles²⁴ by promoting the release of pro-angiogenic factors. We have recently reported that A_2B receptors are essential for adenosine-dependent up-regulation of paracrine functions of cardiac Sca-1⁺CD31⁻ stromal cells, which can serve as an important source of pro-angiogenic stimuli in the heart.²¹ Herein, we explored the relevance of our findings to previously described therapeutic properties of these cells, which reportedly increased neovascularization of the infarct scar area when injected intramyocardially in a mouse model of MI.^6,10,11 We found that injected A_2BKO Sca-1⁺CD31⁻ cells failed to reproduce the pro-angiogenic effects of WT cells, which, in agreement with previous reports,^6,10,11 induced a significant increase in vascularization of infarcted hearts. Furthermore, we found that transplantation of WT, but not A_2BKO, Sca-1⁺CD31⁻ cells into the infarcted hearts resulted in increased endothelial cell numbers already detectable 1 week later. This increase could not be explained by preferential differentiation of injected WT Sca-1⁺CD31⁻ cells to endothelial cells because the cell tracer, DiD, was not found in the endothelial cell population 1 week after injection of DiD-labeled cells. Results of cell fate tracing experiments using DiD and the expression of tsA58Tag suggest that only 12% to 16% of injected cells remain in the infarcted heart at this time point, but no differences in their survival or phenotype were found between WT and A_2BKO cells.

In parallel experiments, we verified that WT and A_2BKO cells used for transplantation differed only in adenosine-dependent, but not basal, secretion of VEGF, a potent pro-angiogenic factor. However, VEGF secretion is not the only function that can be affected by the loss of A_2B receptors in transplanted cells. Cardiac Sca-1⁺CD31⁻ stromal cells represent a rich source of paracrine factors. We have previously shown that stimulation of A_2B receptors on these cells also up-regulates production of CXCL-1 and IL-6,²¹ factors implicated in the regulation of neovascularization.^28,29 It is possible, therefore, that the cumulative effect of several paracrine factors up-regulated by A_2B receptors can contribute to the observed difference between WT and A_2BKO cells in their ability to stimulate neovascularization. Our study also demonstrated that transplanted Sca-1⁺CD31⁻ stromal cells decreased scar size and improved cardiac function after MI only when A_2B receptor signaling was preserved in these cells. It is likely that the loss of pro-angiogenic factors contributes to the loss of cardioprotection when A_2BKO cells are used, given that the development of neovascularization within the infarcted area is an integral part of cardiac repair, leading to long-term salvage and survival of viable myocardium, reduction in collagen deposition, and improvement in cardiac function.

We cannot exclude, however, that other potential mechanisms of therapeutic cell actions may be affected by the loss of A_2B receptors. In addition to their well-recognized pro-angiogenic effects, paracrine factors released by cardiac Sca-1⁺CD31⁻ stromal cells have been implicated in stimulation of receptor-mediated survival pathways.^3,9 Sca-1⁺CD31⁻ stromal cells have been also suggested to play a role of resident progenitor cells in the heart because they can be induced in vitro to transdifferentiate under appropriate conditions into different cell lineages, including cardiomyocytes.^{4–6,8,10,30,31} However, the number of reported cardiomyocytes derived from Sca-1⁺CD31⁻ cells was reportedly too low to account for the significant enhancement of cardiac function in this in vivo model.⁶ Furthermore, we have previously demonstrated that stimulation of A_2B receptors on cardiac Sca-1⁺CD31⁻ stromal cells has no effect on their cardiomyogenic differentiation in vitro.²¹ Therefore, it is unlikely that potential changes in differentiation of Sca-1⁺CD31⁻ cells into cardiomyocytes can contribute significantly to the observed difference between the effects of WT and A_2BKO cell injections on postinfarction cardiac function.

Systemic pharmacological modulation of A_2B receptor signaling has been recently used to analyze the role of these receptors in MI. Paradoxically, inhibition of A_2B receptors with the selective antagonist, GS-6201, early after coronary artery ligation resulted in improved cardiac function compared with vehicle controls.³² On the other hand, stimulation of A_2B receptors starting a week after coronary artery ligation also improved cardiac function compared with vehicle controls.³³ These seemingly contradictory reports may be attributed to potential adverse effects of A_2B receptor signaling in cells active during the initial phase of MI (eg, inflammatory cells) and beneficial effects of these receptors in other cell types active during the healing phase, which occurs later. In the current study, we took advantage of a previously established model of Sca-1⁺CD31⁻ cell transplantation into the infarcted heart to improve cardiac recovery. In contrast to systemic inhibition of A_2B receptors, we abrogated A_2B receptor signaling only in transplanted cardiac Sca-1⁺CD31⁻ stromal cells to analyze its role specifically in these cells, thus bypassing the potential effects of blocking A_2B receptor signaling in other cell types during the acute inflammation phase of MI. In agreement with the observed beneficial effects of A_2B receptor stimulation during the healing phase of MI, our study suggests that stimulation of A_2B receptors on resident Sca-1⁺CD31⁻ stromal cells may lead to an increase in neovascularization of infarcted hearts and contribute to the overall cardioprotection after MI. In this study, we used a previously described and well-reproducible experimental design with injection of the cells immediately after MI induction.^6,9–11 The use of mesenchymal stem-like cells in the clinic would require a considerably longer period between onset of MI and therapeutic cell injections. Considering the reported benefits of pharmacological inhibition of A_2B receptors early after MI,³² it would be interesting to explore potential added benefits of Sca-1⁺CD31⁻ injections shortly after completion of initial treatment with A_2B receptor antagonists. Also, because MI-induced interstitial adenosine levels may be already on decline by that time, it would be important to determine whether up-regulation of the A_2B receptor signaling before cell transplantation could boost their therapeutic effects. This could be accomplished by exposing the cells to hypoxia ex vivo, as we demonstrated previously in different cell types.³⁴

In conclusion, our study provided new in vivo evidence for a pro-angiogenic role of A_2B receptors on cardiac Sca-1⁺CD31⁻ stromal cells. For the first time, to our knowledge, we demonstrated that the A_2B adenosine receptor signaling linked to up-regulation of pro-angiogenic factors in Sca-1⁺CD31⁻ cells is essential for overall cardioprotective properties of these cells in a mouse model of MI. We have recently demonstrated that the human heart harbors a population of mesenchymal stem-like cells similar to the mouse cardiac Sca-1⁺CD31⁻ stromal cells that increase release of pro-angiogenic factors in response to A_2B receptor stimulation.²¹ Therefore, it is likely that our current findings in mice can be translated to humans.