Abstract
Aims:
Human embryonic stem cell derived-cardiovascular progenitor cells (hESC-CVPCs) are a promising cell source for cardiac repair, while the underlying mechanisms need to be elucidated. We recently observed cardioprotective effects of human pluripotent stem cell (hPSC)-CVPCs in infarcted nonhuman primates, but their effects on inflammation during early phase of myocardial infarction (MI) and the contribution of such effect to the cardioprotection are unclear.
Results:
Injection of hESC-CVPCs into acutely infarcted myocardium significantly ameliorated the functional worsening and scar formation, concomitantly with reduced inflammatory reactions and cardiomyocyte apoptosis as well as increased vascularization. Moreover, hESC-CVPCs modulated cardiac macrophages toward a reparative phenotype in the infarcted hearts, and such modulation was further confirmed in vitro using human cardiovascular progenitor cell (hCVPC)-conditioned medium (hCVPC-CdM) and highly contained interleukin (IL)-4/IL-13. Furthermore, signal transducer and activator of transcription 6 (STAT6) was activated in hCVPC-CdM- and IL-4/IL-13-treated macrophages in vitro and in hESC-CVPC-implanted MI hearts, resulting in the polarization of macrophages toward a reparative phenotype in the post-MI hearts. However, hESC-CVPC-mediated modulation on macrophages and cardioprotection were abolished in STAT6-deficient MI mice.
Innovation:
This is the first report about the immunoregulatory role played by hESC-CVPCs in the macrophage polarization in the infarcted hearts, its importance for the infarct repair, and the underlying signaling pathway. The findings provide new insight into the mechanism of microenvironmental regulation of stem cell-based therapy during acute MI.
Conclusions:
Implantion of hESC-CVPCs during the early phase of MI promotes infarct repair via the modulation of macrophage polarization through secreted cytokine-mediated STAT6 activation. The findings suggest a therapeutic potential by modulating macrophage polarization during acute phase of MI.
Introduction
Myocardial infarction (MI) and the resulting heart failure (HF) are major causes of morbidity and mortality worldwide. During cardiac remodeling after acute MI (AMI), an ischemia-induced inflammatory cascade occurs via recruitment of a mixture of both protective and cytotoxic cell types as a driver for healing and exacerbating scar formation (19, 48). Cumulated evidence suggests that modulation of the inflammatory response during the early phase of MI improves infarct repair (13, 15, 40); however, clinically useful approaches targeting the modulation of inflammation toward the healing process are lacking (13).
Innovation
Human pluripotent stem cell (hPSC)-derived cardiovascular progenitor cells (hPSC-CVPCs) are a promising cell source for ischemic cardiac repair, but their roles and mechanisms in acute myocardial infarction remain largely unknown. Here, we demonstrate that human embryonic stem cell-derived cardiovascular progenitor cells (hESC-CVPCs) repair infarcted hearts. Such effect is at least achieved by the hESC-CVPC-polarized macrophages to a reparative phenotype through hESC-CVPC-secreted cytokines via activation of signal transducer and activator of transcription 6. These findings extend the knowledge of the benefits and mechanisms of hPSC-CVPCs in the infarct repair, and provide further evidence to support the view that priming macrophages toward a reparative phenotype might be a potential therapeutic approach to facilitate infarct healing.
Macrophages function as primary responder cells in modulating the inflammatory response to AMI (4, 34). After AMI, monocytes/macrophages are activated by chemotactic factors, and polarized into either classically activated inflammatory macrophages (M1-like) or alternatively activated reparative macrophages (M2-like) (4, 34). M1-like macrophages are characterized by secretion of interleukin (IL)-1α, IL-1β, IL-6, and tumor necrosis factor-alpha (TNF-α) that amplify the inflammation cascade, while M2-like macrophages predominantly secrete anti-inflammatory factors IL-10, IL-13, transforming growth factor-β, and proangiogenic factors that can improve cell survival and promote fibroblast activation necessary for scar formation and angiogenesis (18, 42, 53). Recent studies suggest that shifting macrophages away from a M1-like phenotype and/or polarizing macrophages into a M2-like phenotype improve infarct repair (4, 7, 13, 34, 53), while the direct impact of modulating macrophage subpopulations on infarct repair remains to be fully investigated; and effective regulators and approaches targeting macrophage plasticity need to be further developed for priming macrophages toward a reparative phenotype during the early phase of AMI to improve cardiac remodeling and promote the infarct healing.
Human embryonic stem cell-derived cardiovascular progenitor cells (hESC-CVPCs) improve cardiac function of rodent infarcted hearts when implanted during the subacute stage of ischemia/reperfusion (I/R) (17). Besides, nonhuman primate ESC-derived SSEA1+ CVPCs can differentiate into ventricular myocytes and reconstituted part of the scar when transplanted into nonhuman primate hearts at 2-week post-MI (5). However, the precise mechanisms and time windows for the treatment remain largely unknown, and whether hESC-CVPCs have beneficial effects when administered during the early phase of MI remains elusive. We recently demonstrated that transplantation of SSEA1+ hESC-CVPCs into the AMI hearts of nonhuman primates improves recovery of left ventricular (LV) function, but the engraftment rate declines to ∼0.4% on day 3 after delivery (64), suggesting the involvement of paracrine action in their cardioprotection. However, it is unknown how these cells protect damaged hearts through paracrine action when administered during acute phase of MI, whether and how the inflammation, especially macrophage plasticity, is regulated by these cells, and whether such regulation contributes critically to the infarct repair of these cells.
Accordingly, in this study, using a murine MI model involving implantation of SSEA1+ hESC-CVPCs during AMI, combining in vivo and in vitro studies of macrophage polarization and signal transducer and activator of transcription 6 (STAT6) knockout (STAT6KO), we investigated (i) whether hESC-CVPCs delivered during AMI improve the infarct healing, (ii) the effects and regulatory mechanisms of hESC-CVPCs on macrophage plasticity, and (iii) the contribution of such modulation to infarct repair. Our data reveal previously unrecognized important roles and regulatory mechanisms of hESC-CVPCs in priming macrophage polarization toward a reparative phenotype during the early phase of MI and their impact on the infarct healing.
Results
Characteristics of SSEA1+ hESC-CVPCs (hCVPCs) and human cardiac fibroblasts
SSEA1+ hCVPCs were generated from H9 ESCs and displayed CVPC markers mesoderm posterior BHLH transcription factor 1 (MESP1), insulin gene enhancer protein (ISL1), GATA binding protein 4 (GATA4), NK2 homeobox 5 (NKX2.5), and myocyte enhancer factor 2 (MEF2C) (Supplementary Fig. S1A) as previously reported (8, 9, 64). Fluorescence-activated cell sorting (FACS) analysis showed that the purity of human cardiovascular progenitor cells (hCVPCs) was between 96.1% and 97.6% when evaluated for stage-specific embryonic antigen 1 (SSEA1) expression, a surface marker of hCVPCs (5, 36), and ∼98% of hCVPCs were enhanced green fluorescent protein (EGFP) positive (Supplementary Fig. S1B). Human cardiac fibroblasts (hcFbs) expressed fibroblast markers Vimentin and discoidin domain receptor-2 (DDR2) (Supplementary Fig. S1C) as reported (1).
hCVPCs improve contractile function and reduce scar size
To determine whether hESC-CVPCs improve cardiac function after implantation during AMI phase, we compared the functional outcome at days 2, 7, 14, and 28 post-MI as indicated in Figure 1A. The death rate of MI mice did not significantly differ between the PuramatrixTM (PM) control and hcFbs groups (6 deaths of 20 each) during the 4-week follow-up period, but it was significantly reduced by 43% in the hCVPCs group (3 deaths of 18) (Fig. 1B). The MI was confirmed by echocardiography, and all MI groups showed comparably reduced LV ejection fraction (LVEF), LV fractional shortening (LVFS), and LV systolic dimension (LVDs) at day 2 post-MI (Fig. 1C). These parameters worsened in the PM and hcFbs groups up to day 28 post-MI; however, they were significantly improved during day 7 through day 28 post-MI in the hCVPCs group (Fig. 1C). Hemodynamic changes were further analyzed in a terminal study at day 28 post-MI using a pressure–volume loop. The MI-induced deterioration in LV developed pressure (LVDP), LV end-diastolic pressure (LVEDP), LV maximum ascending rates of pressure (+dp/dt max), and stroke volume (SV) was significantly improved by hCVPCs but not hcFbs (Fig. 1D). MI induced a rightward shift of the LV end-systolic pressure–volume relationship (ESPVR) with a shallow slope, while these changes were partially reversed by hCVPCs but not by PM and hcFbs (Fig. 1E). Consistently, all MI animals showed scar formation at day 28 post-MI as analyzed by Masson's trichrome staining, while the scar size in hCVPC-treated hearts was smaller than that in the PM and hcFbs groups (Fig. 1F). These data further indicate that hCVPCs not only improve LV function but also prevent the progressive deterioration of LV remodeling.
hCVPCs improve cardiomyocyte survival and promote angiogenesis
The reduction of scar size may be at least partially due to a reduction of cardiomyocyte death and/or an increase of angiogenesis. We thus examined the number of terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL+) cardiomyocytes, CD31+ endothelial cells, and alpha-smooth muscle actin (α-SMA+) vessels in the infarcted hearts. At day 3 post-MI, TUNEL+ cardiomyocytes within the infarct and border zones in the hCVPCs group were significantly reduced compared with the PM group (Fig. 2A, B). Consistently, the MI-increased caspase-3 cleavage in the infarct zones was reduced by the hCVPCs but not the PM (Fig. 2C and Supplementary Fig. S2). At day 28 post-MI, immunohistochemistry (IHC) showed an increased number of CD31+ endothelial cells (Fig. 2D) and α-SMA+ vessels (Fig. 2E) in the infarct and border zones of hCVPC-treated MI hearts compared with the PM.
Myocardial sections were then stained for mouse cardiac troponin T (cTnT) and GFP antibody to distinguish whether these improvements are related to the engraftment of implanted cells. Transplanted hCVPCs were easily detected in the border zone at day 1, significantly decreased at day 3, and hardly detected at day 7 after cell delivery (Supplementary Fig. S3A). Analysis of human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) DNA levels also showed that the engraftment rate of hCVPCs declined quickly to a level of 0.05% at day 7 after cell delivery (Supplementary Fig. S3B, C). Such a rapid decline in the number of transplanted cells in the host heart hardly allows differentiation; instead, the beneficial effects of hCVPCs on myocardial repair may result from paracrine action in the regulation of inflammatory responses during early phase of MI.
hCVPCs attenuate inflammatory responses and shift macrophages to a reparative phenotype
To test the above hypothesis, we then analyzed the infiltration of inflammatory cells and expression of inflammatory factors in the infarcted LV. Hematoxylin-eosin (HE) staining demonstrated that AMI-induced infiltration of inflammatory cells was reduced by hCVPCs on day 3 but not day 7 post-MI (Supplementary Fig. S4). This was confirmed by IHC staining, showing that hCVPCs significantly reduced infiltration of CD11b+ cells (neutrophils and monocytes/macrophages) and CD68+ cells (monocytes/macrophages) in the infarct and border zones on day 3 but not day 7 post-MI (Fig. 3A, B). Consistently, on day 3 post-MI, AMI-enhanced gene expression of proinflammatory cytokines IL-6, TNF-α, and IL-1β [products of inflammatory macrophages (42)] in the infarct and border zones was markedly reduced by the hCVPCs, whereas on day 7 post-MI these levels markedly declined and did not differ between the PM and hCVPC groups (Fig. 3C). Interestingly, the anti-inflammatory factor IL-10 commonly secreted from reparative macrophages (4, 16, 17, 63) was significantly enhanced by the hCVPCs on day 3 post-MI (Fig. 3C). These results suggest that hCVPCs attenuate inflammatory responses in the infarcted myocardium and might polarize macrophages toward a reparative phenotype.
To determine whether the hCVPCs affect the polarization of macrophages after MI, the proportion of inflammatory macrophages and reparative macrophages in the infarcted hearts were examined by FACS (Supplementary Fig. S5A, B). The treated hearts were digested and stained simultaneously with antibodies of CD11b, CD45, Ly6G, F4/80, and CD206 (Supplementary Fig. S5A). CD45+CD11b+Ly-6G−F4/80+ macrophages were divided based on CD206- (M1-like) and CD206+ (M2-like) (Supplementary Fig. S5B). The proportion of CD45+CD11b+Ly-6G−F4/80+CD206− M1-like macrophages in the infarcted LV was significantly higher than that in the Sham group, but the CD45+CD11b+Ly-6G−F4/80+CD206+ M2-like macrophages were decreased at day 3 post-MI, and both of them returned to the level of the Sham group at day 7 post-MI (Fig. 3D). Interestingly, at day 3 post-MI, the MI-induced opposite changes in the proportions of CD45+CD11b+Ly-6G−F4/80+CD206− M1-like macrophages and CD45+CD11b+Ly-6G−F4/80+CD206+ M2-like macrophages in the hCVPC groups were both returned to a level similar to that in the Sham, while no difference was detected at day 7 post-MI between the PM and hCVPC groups (Fig. 3D). These data indicate that the hCVPCs promote the transition of macrophages toward a reparative phenotype but shift from an inflammatory phenotype during the early phase of MI.
hCVPCs modulate macrophage polarization through the paracrine effect
Next, we examined whether the hCVPC-modulated macrophage polarization is mediated by the paracrine effect. Isolated murine peritoneal macrophages were differentiated into M1-like (interferon gamma [IFNγ] and lipopolysaccharide [LPS]) or M2-like (IL-4 and IL-13) population with or without treatment of hCVPC-conditioned medium (hCVPC-CdM) for 24 h. Morphologically, hCVPC-CdM-treated macrophages were similar to those treated with IL-4/IL-13, while they differed from those treated with IFNγ/LPS or Dulbecco's Modified Eagle Medium (DMEM) (Fig. 4A). Quantitative reverse transcription–polymerase chain reaction (RT-qPCR) analysis showed that the hCVPC-CdM significantly reduced the expression of M1 markers TNFα, inducible nitric oxide synthase (iNOS), and IL-1α but increased the expression of M2 markers Arg1, IL-10, and CD206 (Fig. 4B). Western blot analysis further confirmed that IFNγ/LPS-activated macrophages expressed a higher level of M1 marker iNOS that was largely suppressed by the hCVPC-CdM, while IL-4/IL-13-stimulated macrophages expressed the M2 marker CD206 (Fig. 4C and Supplementary Fig. S6). Interestingly, the latter was mimicked by the hCVPC-CdM and was further elevated when treated with both (Fig. 4C and Supplementary Fig. S6). However, FACS analysis of the macrophages uptaking FITC-beads did not detect the effect of hCVPCs-CdM on macrophage phagocytosis (Supplementary Fig. S7A–C). These data suggest that hCVPCs may secrete certain stimulators to activate macrophages toward a M2-like status. The human antibody array revealed abundant proteins (Supplementary Fig. S8A) and the large amount of IL-4 and IL-13 contained in the hCVPC-CdM (Supplementary Fig. S8B). RT-qPCR analysis further detected the higher expression of M2 stimulators IL-4 and IL-13 in the hCVPCs than in the hESCs (Supplementary Fig. S8C).
hCVPC-polarized M2-like macrophages via the cytokine-activated STAT6
We next sought to identify signaling pathways involved in the regulation of hCVPC-CdM on the macrophage polarization with or without IFNγ/LPS or IL-4/IL-13 in the presence or absence of hCVPC-CdM on cultured macrophages. The signaling pathways protein kinase B (Akt) (47), extracellular signal-regulated kinases 1/2 (Erk1/2) (57), STAT3 (21, 54), and STAT6 (25, 44, 55) related to the macrophage regulation were examined. The treatments did not affect total protein levels of Akt, Erk1/2, STAT3, and STAT6, while IFNγ/LPS upregulated phosphorylation of Erk1/2 and STAT3 but not Akt and STAT6, and these upregulations were not affected by hCVPC-CdM (Supplementary Figs. S9A, B, S10, and S11 and Fig. 5A). In contrast, IL-4/IL-13 and hCVPC-CdM alone did not affect the phosphorylation of Akt, Erk1/2, and STAT3 but increased STAT6 phosphorylation, and the latter was further enhanced when given together for 3 h (Supplementary Figs. S9A, B, S10 and S11 and Fig. 5A). This was further confirmed by Western blot analysis, showing that IL-4/IL-13 and hCVPC-CdM upregulated the level of M2 marker CD206, and STAT6 phosphorylation after 24 h of treatment in the cultured peritoneal macrophages isolated from wild type (WT) mice was abolished in the cells isolated from STAT6KO mice (Fig. 5B and Supplementary Fig. S12). RT-qPCR analysis showed that the STAT6KO suppressed hCVPC-CdM-increased expression of M2 markers CD206, IL-10, and Arg-1 (Fig. 5C), and hCVPC-CdM-reduced the expression of M1 markers TNF-α, IL-1β and iNOS was not observed in STAT6KO macrophages (Fig. 5D). These data indicate that the hCVPCs shift the macrophages toward a M2-like phenotype via the paracrine effect-activated STAT6 signaling pathway.
STAT6 mediates myocardial wound repair of hCVPCs
Next, we sought to determine whether hCVPCs activate STAT6 in the macrophages of infarcted hearts. The STAT6 phosphorylation in the border zone of infarcted hearts was enhanced by hCVPCs when compared with that of the PM group at day 3 post-MI (Fig. 6A, B).
We then verified whether the activation of STAT6 contributes to the hCVPC-mediated benefits in the infarcted hearts in the WT and STAT6KO mice (Fig. 7A). The efficiency of STAT6KO in the heart was confirmed by RT-PCR (Supplementary Fig. S13A) and Western blot (Supplementary Fig. S13B). The LVEF and LVDs were similar between the WT and STAT6KO mice (Fig. 7B) as well as between the infarcted WT and STAT6KO mice at day 2 post-MI (p > 0.05) (Fig. 7C), but the STAT6KO MI mice exhibited worsened LVEF and LVDs (Fig. 7C) and larger scar size (Fig. 7D) than those observed in the WT MI mice at day 28 post-MI. Moreover, the functional improvement (Fig. 7C) and the limitation in scar size (Fig. 7D) by the hCVPCs were abolished in STAT6KO mice at day 28 post-MI, though these parameters were comparable at day 2 post-MI among the groups.
We then investigated whether the STAT6KO affects the benefits of hCVPCs in promoting angiogenesis and antiapoptosis of cardiomyocytes. The apoptotic responses induced by 0.5 and 1 h of oxygen glucose deprivation (OGD) treatment in cardiomyocytes isolated from adult WT and STAT6KO mice were comparable, but the majority of cardiomyocytes lost their myofilament phenotype and died after a 3-h OGD treatment in both WT and KO cardiomyocytes (Supplementary Fig. S14A, B). However, STAT6KO increased MI-induced apoptosis and abolished hCVPC-reduced apoptotic cardiomyocytes in the border zone detected with TUNEL staining at day 3 post-MI (Fig. 8A, B, left panel). Moreover, STAT6KO reduced the number of CD31+ endothelial cells (Fig. 8C, E, middle panel) and α-SMA+ vessels (Fig. 8D, F, right panel) in the infarct and border zones of the PM hearts, and blocked hCVPC-promoted angiogenesis (Fig. 8C–F).
Next, to confirm whether lack of STAT6 interferes with the polarization of macrophages after the MI and the hCVPC treatment, we analyzed CD206+ and CD68+ macrophages at the border zone of infarcted hearts with and without treatment of hCVPCs by IHC staining (Fig. 9A–C). In the infarcted hearts, STAT6KO did not affect the number of CD68+ macrophages at day 3 and 7 post-MI (Fig. 9B), while it decreased the ratio of CD206+/CD68+ at day 3 as well as day 7 post-MI compared with those observed in the WT mice (Fig. 9C). Moreover, on day 3 post-MI, the hCVPC-reduced CD68+ macrophage population and increased CD206+/CD68+ ratio were canceled in STAT6KO mice (Fig. 9C).
Discussion
hESC-CVPCs represent an attractive cell type for cardiac repair, but their roles in AMI and mechanisms of action remain largely unknown. This is the first report showing that the infarct repair of SSEA1+ hESC-CVPCs implanted during the acute phase of MI is concomitantly decreasing the number of inflammatory macrophages and increasing reparative macrophages in the infarcted hearts. These effects are achieved by hESC-CVPC-secreted cytokines via activation of STAT6, resulting in the improvement of existing cardiomyocyte survival and vascularization in the infarcted hearts. The findings extend the benefits of human pluripotent stem cell (hPSC)-CVPCs to ischemic myocardial damage, and provide new knowledge into the mechanisms mediating infarct repair via modulating inflammatory responses and the involved signaling. The findings also provide further evidence to support the view that the priming macrophages toward to a reparative phenotype might be a potential therapeutic approach to facilitate infarct healing (4, 7, 13, 34, 53).
Transplantation of hESC-CVPCs during the subacute stage of I/R or chronic stage of MI has been shown to improve recovery of cardiac function from I/R in rodent (17) and nonhuman primate models (5). Here, we demonstrate that SSEA1+ hESC-CVPC delivered at the acute phase of MI confers cardioprotection characterized by improving cardiac performance, survival of existing cardiomyocytes, promoting angiogenesis, and limiting scar size. These observations are consistent with the findings in nonhuman primates, in which the transplantation of SSEA1+ hESC-CVPCs significantly reduced apoptotic cardiomyocytes in the infarcted hearts at day 3 post-MI and improved LV function at day 28 post-MI (64). These findings indicate that the acute stage of MI is an important window for hESC-CVPCs therapy targeting at saving cardiomyocyte and promoting vascularization in addition to their benefits at the subacute and chronic stages of MI (5, 17). Supportively, implantation of ESC-derived ISL1+-CVPCs (3) and cardiomyocytes (56), mesenchymal stromal cells (MSCs) (4), or cardiosphere-derived cells (CDCs) during early reperfusion (13, 30) or AMI significantly improves healing process of infarct hearts. Thus, the acute phase of MI is a critical window for cell therapy targeting outcome of cardiac repair.
The critical role of inflammatory modulation in the hESC-CVPC-mediated myocardial repair after ischemic injury is revealed in this study. Adequate evidence suggests that the activation of reparative macrophages during acute phase of MI is critical for the healing process. It seems that modulation of macrophage polarization to a reparative phenotype is an important mechanism for stem cell/cytokine therapy. This is supported by the reports showing that transforming macrophages away from inflammatory macrophages to reparative macrophages by delivery of CDCs (6, 13), MSCs (4), IL-4 (51, 53), IL-10 (29), or inactivation of cyclic GMP–AMP synthase (7) improves the infarct repair after MI. Interestingly, hESC-CVPCs not only increase the population of reparative macrophages but also reduce the population of inflammatory macrophages, while MSCs seem to mainly activate reparative macrophages in the infarcted hearts (4). These findings extend previous observations (4, 13, 20, 24, 34, 35, 43, 58), indicating that the hESC-CVPCs may serve as a modulator of the innate immune responses to favor the infarct repair, and the transition of inflammatory macrophages to a reparative phenotype might be an effective target for infarct repair.
The polarization of macrophages to a reparative phenotype during the acute phase of MI may decrease the MI scar via the following mechanisms: (i) improving the survival of cardiomyocytes via the downregulation of proinflammatory cytokines inducing apoptosis of cardiomyocytes, such as IL-1β and TNF-α (49, 61), when suppressing M1 macrophage polarization (27) and the enhancement of anti-inflammatory cytokines attenuating apoptosis of cardiomyocytes, such as IL-10 (14). Similar changes were observed here in the hCVPC-implanted MI hearts; (ii) promoting angiogenesis through macrophage-secreted proangiogenic factors such as fibroblast growth factor-2, vascular endothelial growth factor alpha, insulin-like growth factor, and platelet-derived growth factor (28, 32, 33, 37). We observed that STAT6KO-abolished hCVPC-reduced number of CD68+ macrophages and increased the ratio of CD206+/CD68+ in the infarcted hearts (Fig. 9C), which are accompanied by the blockage of hCVPC-promoted angiogenesis by STAT6KO (Fig. 8C–F), suggesting that the proangiogenic effect induced by SSEA1+ hCVPC-modulated macrophages may contribute to the healing process of the infarcted heart. The precise mechanisms behind the hCVPC-educated macrophage-stimulated angiogenesis and antiapoptosis of cardiomyocytes need to be further elucidated in future. Besides, the hCVPCs might stimulate proangiogenic effect and exert antiapoptotic effect via a macrophage-independent way, and this possibility needs to be explored.
Our results also indicate that benefits associated with the hESC-CVPCs in cardiac repair are mediated by paracrine effects due to the change of microenvironment. A rapid decline in the engraftment rate of the transplanted hESC-CVPCs (Supplementary Fig. S3) as seen in our recent study using the infarcted primates (64) suggesting the paracrine-related anti-inflammatory effects and modulation of macrophage phenotypes may play as a major mechanism for the cardioprotection of hPSC-CVPCs when administered during early phase of MI. This is supported by the experimental data, showing that (i) hESC-CVPCs secrete abundant proteins, including cytokines IL-4 and IL-13 (Supplementary Fig. S8), to promote polarization of macrophages into a reparative phenotype via activating STAT6 (Fig. 5A, B); (ii) the hCVPC-CdM modulates the macrophage phenotype and activates STAT6 in the macrophages (Figs. 4 and 5); (iii) at day 3 post-MI, the macrophages are polarized into a reparative phenotype via activation of STAT6 in the infarcted hearts by implanted hCVPCs (Figs. 3C and 6), while at that time majority implanted cells are gone (Supplementary Fig. S3); and (iv) these processes are associated with the hCVPC-promoted healing process. Thus, hESC-CVPCs provide an approach to modulate macrophage phenotypes and functions by changing the microenvironment in the infarcted hearts. In addition, SSEA1+ hPSC-CVPC-secreted vesicles recapitulate the beneficial effects of their parent cells in the treatment of chronic postinfarct HF murine model (31), and paracrine effects play major roles in the salvage of infarcted myocardium by hPSC-derived cardiac lineage cells implanted during acute phase of MI (56). Therefore, further studies are required to dissect their secretome to explore the potential use of a cell-derived but cell-free therapy for cardiac repair. Besides, upon the development of approaches to improve the survival of implanted hPSC-CVPCs, the contribution of differentiated cardiovascular cells to the infarct repair needs to be evaluated.
Another interesting observation is STAT6, serving as a key signal mediating the therapeutic effects of hESC-CVPCs in the infarcted heart. Although STAT6KO mice exhibit normal cardiac function as previously reported (26), the LV function and fibrosis has shown to be worse, and the benefits of hESC-CVPCs in cell survival, angiogenesis, functional recovery, and scar formation are cancelled by STAT6KO. This indicates the critical role of STAT6 in the myocardial repair, especially with respect to the hESC-CVPC-promoted healing process. This protection can be interpreted by the following hypothetical mechanisms: STAT6 may act as a key signal priming macrophages to a reparative phenotype in response to IL-4/IL-13 and hESC-CVPCs in the MI hearts, resulting in cardiac repair (Figs. 5 –9 and Supplementary Fig. S8). Supportively, the STAT6 in macrophages is activated by IL-4/IL-13 as reported (25) but not Akt, Erk1/2, and STAT3 (Supplementary Fig. S9), by hCVPC-CdM in vitro (Fig. 5), and by hESC-CVPCs in the infarcted hearts (Fig. 6), while IL-4/IL-13- or hCVPC-CdM-activated reparative macrophages are suppressed by STAT6KO (Fig. 5) and hESC-CVPC-promoted reparative macrophage polarization in MI mice is cancelled by STAT6KO (Fig. 9). Moreover, isolated STAT6KO cardiomyocytes show similar OGD injury to that in the WT cells, supporting the indirect protection of STAT6 activation on cardiomyocyte survival in the infarcted hearts. These observations provide further evidence, showing the essential role of STAT6 in the activation of reparative macrophages (11, 44) and promoting the infarct healing. Thus, activation of STAT6 might be a potential target for the treatment of AMI via specific polarization of monocytes/macrophages into a reparative phenotype, though the regulatory effects of STAT6 on other cell types in cardiac repair need to be elucidated. Besides STAT6, other proteins could be involved in the modulation of macrophage phenotypes, resulting in the promotion of infarct heart healing. It is thus worthy to explore whether other proteins secreted by hESC-CVPCs would contribute to the cardioprotection via polarization of macrophages into a reparative phenotype in the MI hearts.
There are limitations in this study. Transplantation of the hESC-derived CVPCs in a mouse model could lead to confusing rejection activity. It would be helpful to control these studies with experiments where mouse ESC-derived CVPCs are used, and verify that the cardioprotection of these cells is also dependent on their paracrine effect and related to modulation of macrophage phenotypes. In addition, as the xenograft model and consequent rejection of hESC-CVPCs in this study prevent us from assessing engraftment and subsequent cellular contribution to cardiac recovery, our experiments do not at all exclude other mechanisms by which hESC-CVPCs could help the repair process of the injured heart, for example by engrafting and becoming functional cardiac cells as seen in hESC-derived cardiomyocytes (10, 41) and MSCs (2, 45).
In conclusion, we demonstrate that modulation of macrophage polarization by delivery of hESC-CVPCs into acutely infarcted mice hearts plays a significant role in blunting the worsening of heart function and fibrosis via their paracrine effect-mediated activation of STAT6. The benefits are associated with a reduced inflammatory response, improved survival of existing cardiomyocytes, and promoted angiogenesis (Fig. 10). These findings not only indicate that hESC-CVPCs may be a therapeutic option for infarct repair, but also suggest that modulation of macrophage polarization to a reparative phenotype and activation of STAT6 in the infarcted heart might be a promising therapeutic approach for ischemic heart disease.
Materials and Methods
Cell culture, differentiation, and characterization
H9 hESCs (WiCell) were routinely maintained in commercially available mTeSR1 medium (Stem Cell Technologies) on Matrigel-coated plates (hESC qualified; BD Biosciences) according to manufacturer's instruction. The induction of hESC-CVPCs followed the protocol reported previously (8, 9, 64). In brief, undifferentiated hESCs were dissociated with Accutase (Stem Cell Technologies) and then plated onto Matrigel-coated culture dishes at a density of 5 × 104 cells/cm2 in hESC-CVPC induction medium (DMEM/F12, 1 × B27 supplement without vitamin A, 1%
Immunostaining assays for characterization of hESC-CVPCs and hcFbs were performed as described previously (1, 8, 22). In brief, cells were fixed with 4% paraformaldehyde (PFA), permeabilized in 0.3% Triton X-100, blocked in 10% normal goat serum (Vector Laboratories), and then incubated with primary antibodies against MESP1, MEF2C, ISL1, GATA4, NKX2.5, Vimentin, and DDR2 at 4°C overnight and detected by DyLight 488- or DyLight 549-conjugated secondary antibodies. Nuclei were stained with Hoechest33342 or 4′,6-diamidino-2-phenylindole (DAPI; Sigma-Aldrich). The SSEA1-positive and GFP-positive proportion of hESC-CVPCs were analyzed and quantified by flow cytometry (FACStar Plus Flow Cytometer; BD Bioscience) as reported previously. The antibodies used are listed in Supplementary Table S1.
Plasmid and cell line construction
EGFP-positive hESCs were generated by using the pCDH cDNA cloning and expression lentivector system (System Biosciences) as described previously (64). In brief, the EGFP cDNA was subcloned into a pCDH-EF1-MCS-T2A-Puro plasmid (System Biosciences) between EcoRI and NotI restriction endonuclease sites. Then, the EGFP-containing lentivirus was packaged into 293FT cells (Invitrogen), and lentivirus particles were collected via centrifugation at 130,000 g for 2 h. hESCs (H9-WA09 hESCs, provided by WiCell Research Institute, Wisconsin) were transfected with the particles and cultured with puromycine (1 mg/mL; Life Technologies) for 7 days. Then, the puromycine-resistant (i.e., EGFP-expressing) cells were collected for use in subsequent experiments.
MI surgical procedure and study design
All procedures involving animals were performed in accordance with the Guidelines for Care and Use of Laboratory Animals published by the United States National Institutes of Health (43a), and were approved by the Institutional Animal Care and Use Committee of Shanghai Institutes for Biological Sciences. Adult male C57BL/6 and BALB/c (WT) mice (The Shanghai Slac Laboratory Animal Co. Ltd., Shanghai, China) and BALB/c background STAT6 knockout (STAT6KO) mice (The Jackson Laboratory, Bar Harbor, ME) were used in this study.
Sex- and age-matched animals were used for all experiments. MI was produced as previously described (12, 40). In brief, male mice aged 10–12 weeks were anesthetized via intraperitoneal injection of 50 mg/kg sodium pentobarbital and ventilated with a volume-regulated respirator (SAR830; Cwe Incorporated) at a frequency of 90 bpm, and the tidal volume ranged 200–300 cc/min. The body temperature was maintained at 37°C throughout the surgical procedure. The mice were randomly selected for the Sham group and the induction of MI. The left chest of mice was opened between the second and the third ribs, and the hearts were exposed. Then, Sham-operated mice had a loose suture placed around the left anterior descending (LAD) coronary artery, and the MI was produced through ligation of the LAD coronary artery with 8-0 Prolene suture at the 3–4 mm on the lower edge of the left auricle. Of 10 μL 10% sucrose solution containing 5 × 105 hCVPCs or hcFbs cells were mixed with 10 μL 0.4% self-assembling nanopeptide (PM; 3-D Matrix, Ltd.) to generate 20 μL of 0.2% PM cell mixture, and then they were directly injected into two sites in the border zone of infarcted myocardium (10 μL per site) just after the ligation of LAD coronary artery as previously described (40). The muscle and skin of the chest were then sutured with 6-0 and 5-0 Prolene suture, respectively. The disposition of animals is listed in Supplementary Table S2.
Heart function measurements
Transthoracic echocardiography (Vevo 2100; Visual Sonics) with a 25-MHz imaging transducer was performed on anesthetized animals as previously described (12) with measurement of LVEF, LVFS, and LVDs. The average values were collected from three consecutive cardiac cycles.
Hemodynamic analysis was performed using a pressure–volume loop as previously described (52, 59). In brief, at 28 days post-MI after echocardiography, a polyethylene pressure transducer catheter (Millar Spr-839 1.4F; Millar Instruments, Houston, TX) was inserted from the right carotid artery into LV for hemodynamic study as previously described (52, 59). LVDP, LVEDP, LV maximum ascending rates of LV pressure (+dp/dt max), and SV were recorded simultaneously. Inferior vena cava occlusion was performed with external compression to produce variably loaded beats to determine the LV ESPVR. Chart software (AD Instrument Ltd., Australia) was used for data processing.
Immunohistochemical staining
Hearts were quickly removed from anesthetized mice and washed with 5 mL cold cardiac arresting buffer (10% KCl), phosphate buffered saline (PBS), and 4% PFA solution. Ventricular tissues were embedded in optimal cutting temperature (SAKURA) compound for histological analysis. Transversal sections (5 μm) were prepared at 400-μm intervals.
For fluorescent IHC, fresh frozen sections were fixed with 4% PFA, permeabilized in 0.4% Triton X-100 (Sigma), and stained with anti-CD31, α-SMA, CD11b, CD68, CD206, cardiac troponin I, cTnT, and GFP antibody. Antibodies were detected by fluorescent-conjugated secondary antibodies. Nuclei were stained with DAPI (Sigma-Aldrich). Histological images were blindly measured using a Zeiss inverted microscope and processed using ZEN software. Three microscopic fields were quantified for each slice by two independent observers. The image-processing software (ImageJ) was used as the image quantification software. The antibodies used are listed in Supplementary Table S1.
Apoptosis was evaluated with an In situ Cell Death Detection Kit (Roche Applied Science, Germany) for TUNEL staining as directed by the manufacturer's instructions. The percentage of TUNEL+ cardiomyocytes was quantified as the ratio of TUNEL+ cardiomyocytes to the DAPI-positive cells. The ImageJ is used to quantificate the TUNEL+ nucleus and total nucleus.
Masson's trichrome staining and HE staining
From Masson's trichrome-stained images, morphometric parameters in each heart including total LV area and scar area were analyzed on five slices (from the point of ligation to the apex of the heart) with ImageJ, as described previously (13, 23, 46). Infarcted scar size was calculated as total scar area divided by LV area. The HE staining analysis was performed as described previously (60).
Isolation of cardiac immune cells and FACS analysis
The procedure and analysis were performed as previously described (62). In brief, the hearts were rapidly removed after being anesthetized, and then minced, digested for 80 min with prewarmed buffer containing 2% collagenase type 2 (Worthington), 0.25% elastase (Worthington), and 0.05% DNase I. After the enzymatic reaction was stopped, the cells were filtered through a 70-μm cell strainer. The filtered cells were subjected to density gradient centrifugation with Percoll to obtain inflammatory cells as described previously (62). Isolated cells were stained at 4°C for 40 min simultaneously with the following antibodies: FITC-labeled rat antimouse CD45, PE-labeled rat antimouse Ly-6G, PE-cy7-labeled rat antimouse CD11b, Bv421-labeled rat antimouse F4/80, Alexa Fluor 647-labeled rat antimouse CD206 to analyze the polarization of macrophages. Cell fluorescence was measured with EPICS ALTRA flow cytometer using EXPO32 software, version 1.2 (Beckman Coulter, Inc.). All antibodies have been listed in Supplementary Table S1.
Isolation and culture of abdominal macrophage
Peritoneal macrophages from C57BL/6 or BALB/c or STAT6KO mice were harvested under aseptic conditions at 3 days after intraperitoneal injection with 3% thioglycollate (Sigma-Aldrich) as previously described (39, 50). Red blood cells were lysed in erythrocyte lysis buffer (10 mmol/L Tris-HCl, pH 7.2), containing 150 mmol/L NH4Cl. The resultant peritoneal macrophages were seeded onto six-well tissue culture plates at 1 × 106 cells/well in 1640 medium containing 10% FBS, and 4 h later the nonadherent cells were removed. The macrophages were then cultured for 24 h in DMEM medium containing 1% FBS before treatment with hCVPC-CdM, LPS (100 ng/mL) + IFNγ (50 ng/mL), or IL-4 (10 ng/mL) + IL-13 (10 ng/mL) in serum-free DMEM. After 24-h treatment, the macrophages were collected for further analysis and dissociated with Trizol to isolate total RNA. After the reverse transcription reaction, M1 (TNF-α, iNOS, and IL-1α) and M2 markers (Arg-1, IL-10, and CD206) were measured by RT-qPCR. Proteins were collected after treatment for 3 or 24 h, dissociated with RIPA buffer, and Western blotted with anti-iNOS, anti-CD206, STAT6, p-STAT6, STAT3, p-STAT3, Akt, p-Akt, Erk1/2, p-Erk1/2, or β-actin. Detailed antibody information is listed in Supplementary Table S1.
Preparation of hCVPC-CdM
hESCs were seeded onto 10-cm dishes at a field density of 3 × 104 cells/cm2. After culture for 72 h in hCVPC-induced medium, cells were washed with PBS thoroughly three times, and medium was changed to serum-free DMEM. After incubation for 48 h in serum-free medium, the supernatant was collected as hCVPC-CdM and cells were removed by centrifugation at 2000 g for 10 min.
Protein analysis
The heart was harvested and rinsed in precold PBS. The infarct with border zone was dissected and stored at −80°C until use. Tissues were minced, homogenized, and digested in tissue lysis buffer (with protease and phosphatase inhibitors; Thermo Scientific). The cells were scraped off culture plates on ice and lysed with RIPA (with protease and phosphatase inhibitors; Thermo Scientific). The resulting suspensions were centrifuged at 10,000 g for 30 min at 4°C, and the protein supernatant was collected. Protein concentrations were measured using a BCA assay (Thermo Scientific). Protein samples were prepared for polyacrylamide gel electrophoresis. Proteins were then transferred to a polyvinylidene difluoride membrane (BioRad) for immunoblotting with relevant antibodies. The antibodies and usage dilution are listed in Supplementary Table S1. Bands were visualized after activation with ECL (Thermo Scientific) and exposure on film (Kodak Carestream Biomax; Sigma-Aldrich). Cytokine release was measured in culture supernatant by cytokine antibody array (Raybiotech L-Series Human Antibody Array L-507), according to the manufacturer's instructions (RayBiotech, Inc.).
Reverse transcription–polymerase chain reaction and RT-qPCR
Reverse transcription–polymerase chain reaction (RT-PCR) and RT-qPCR were performed as described previously (8). In brief, total RNA was extracted from the LV tissue using a RNeasy Plus Mini Kit (QIAGEN) following the manufacturer's instructions and treated with DNase I (Promega) for 15 min to eliminate the potential contamination of genomic DNA. cDNA was generated by reverse-transcribed total RNA (1 μg) using oligo (dT) primer and ReverTra Ace reverse transcriptase (Toyobo). PCR was carried out using Taq DNA Polymerase (Takara). qPCR was performed using the ABI PRISM 7900 system (Applied Biosystems) with the SYBR Green Realtime PCR Master Mix plus (TOYOBO) for relative quantification of the indicated genes. The transcript of GAPDH was used for internal normalization. The RT-PCR and RT-qPCR primers are listed in Supplementary Table S3.
Phagocytosis assay
The analysis was performed as previously described (13). To examine the phagocytic capacity of unstimulated and stimulated macrophage with hCVPC-CdM, LPS (100 ng/mL) + IFNγ (50 ng/mL), or IL-4 (10 ng/mL) + IL-13 (10 ng/mL), we performed a fluorescent latex bead assay (Cayman Chemical) according to the manufacturer's protocol. After prestimulation, macrophages were treated overnight with FITC fluorescently labeled microspheres (0.1-μm diameter). Cells were collected and then stained with F4/80-PE (eBioscience) to examine all macrophages. We then performed FACS analysis for quantitation assessment of microsphere uptake in each group.
Isolation and culture of adult mouse cardiomyocytes and OGD treatment
The adult mouse cardiomyocytes (AMCMs) were isolated from adult WT and STAT6KO mice hearts, and cultured as previously described (38). AMCMs were subjected to hypoxia in vitro in an oxygen control cabinet (Ruskinn, England) mounted within an incubator, and equipped with oxygen controller and sensor for continuous oxygen-level monitoring. A mixture of 85% nitrogen, 10% hydrogen, and 5% CO2 was utilized to create hypoxia, and the O2 in the chamber was monitored and maintained at a level <0.1%. The AMCMs were treated with OGD injury by cultured medium without serum and glucose and in hypoxic condition (<0.1% O2) for 0.5, 1, and 3 h. Then, the cardiomyocytes were stained by TUNEL, cTnT, and DAPI to evaluate the level of apoptosis in the AMCMs. The ImageJ is used to quantify the TUNEL+ nucleus and total nucleus.
Statistical analysis
Data are expressed as mean ± standard error of the mean. Statistical significance was analyzed by using unpaired Student's t test or one-way analysis of variance (ANOVA), followed by Bonferroni's multiple as appropriate. Two-way ANOVA was applied with Tukey's multiple comparison for analysis of echocardiographic data. Statistical analyses were performed with Graphpad Prism software (version 6.1). A p-value <0.05 was considered statistically significant.
Footnotes
Acknowledgments
This work was supported by grants from National Key R&D Program of China (2017YFA 0103700 to H.-T.Y. and Z.C., 2016YFC1301204 to H.-T.Y.), the Strategic Priority Research Program of the Chinese Academy of Sciences (No. XDA16010201 to H.-T.Y.), National Natural Science Foundation of China (81520108004, 81470422 to H.-T.Y.), and the National Institute of Health grant HL136025-01A1 (for Y.W.). The authors thank WiCell Research Institute for providing the H9 hESCs, Chunxing Zheng from Key Laboratory of Stem Cell Biology, SIBS for the assistance in mice peritoneal macrophage isolation and cultivation and in the hESC-CVPCs preparation, and Jizhen Lu and Qiao Liu in the Laboratory of Molecular Cardiology.
Author Disclosure Statement
No competing financial interests exist.
Supplementary Material
Supplementary Figure S1
Supplementary Figure S2
Supplementary Figure S3
Supplementary Figure S4
Supplementary Figure S5
Supplementary Figure S6
Supplementary Figure S7
Supplementary Figure S8
Supplementary Figure S9
Supplementary Figure S10
Supplementary Figure S11
Supplementary Figure S12
Supplementary Figure S13
Supplementary Figure S14
Supplementary Table S1
Supplementary Table S2
Supplementary Table S3
Abbreviations Used
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
