Myocardial Ischemia Alters the Phenotypes and Secretome MSCs
MSCs are multipotent tissue-resident cells, and exhibit a selective ability for tissue repair by engrafting and differentiating into desired cells at the damaged tissues [25-27]. To determine the effect of myocardial Isch on resident cardiac MSCs (cMSCs) and peripheral blood MSCs (pbMSCs), we isolated MSCs from male Lewis rats with Isch 30 days after MI (n=10) or sham operation (n=10) as previously described [10, 28]. Rats which were not ligated in the same part of the hearts and served as sham group (Sham). Both cMSCs and pbMSCs, isolated after either MI or sham operation, were plastic-adherent, showed universe spindle like morphology, and had good clonogenicity (Fig. S2A). We next evaluated cell proliferation rate for 96 hours and found that cMSCs from the ischemic hearts had the highest growth rate compared with other groups (Fig. S2B). Immunophenotypic analyses by flow cytometry indicated that compared with other MSCs, the Isch cMSCs were strongly positive for mesenchymal lineage markers, SH2, SH3, and CD147, while negative for hematopoietic lineage marker CD 34, CD45, and CD117 (Fig. S2C). Notably, Thy-1 (CD90) , which mediates fibroblast adhesion and migration, was expressed at lower frequency in Isch cMSCs compared to other MSCs. Last, we performed direct differentiation toward adipocyte, chondrocyte, osteoblast, and vascular cells by growth factor supplementation and growth on defined matrices. Multi-lineage differentiation into osteoblasts, adipocytes, chondrocytes, and blood endothelial cells confirmed the identity of the MSCs (Fig. S2D–G). These resulsts suggest that myocardial ischemia affects the phenotype of MSCs, particularly cardiac MSCs.
MSCs modulate myocardial repair through their paracrine, angiogenesis, anti-fibroblasis, and antioxidation properties after MI . For that purpose, we used a rat angiogenesis array to measure the expression of 60 angiogenesis-related cytokines from the conditioned medium of cultured MSCs. Significantly, cMSCs isolated from Isch hearts secreted higher levels (1.3-2.2 fold) of proinflammatory cytokines, interleukin (IL)-1α, IL-6, matrix metalloproteinase 2 (MMP2) and MMP9, transforming growth factor-β1 (TGF-β1), and tumor necrosis factor-α (TNFα), compared with their Sham counterparts. Conversely, we found cMSCs from Isch hearts showed a much smaller release of anti-inflammatory factor IL-4 than those from Sham hearts (Fig.1A). Importantly, we found that the levels of the pro-angiogenic factors, angiopoietin1 (Ang1), basic fibroblast growth factor (bFGF), hepatocyte growth factor (HGF), vascular endothelial growth factor (VEGF), tunica interna endothelial cell receptor 2 (Tie2), and vascular endothelial growth factor receptor (VEGFR), were significantly decreased in cMSCs from Isch hearts (1.1-2.2 fold lower) compared with Sham cMSCs (Fig. 1B). It is notable that myocardial Isch induced a different secretory profile in pbMSCs. Although the levels of IL-6, MMP9, and TNFα were slightly increased in pbMSCs from Isch rats (about 1.2-fold higher) compared with pbMSCs from Sham rats, most cytokines secreted from Isch pbMSCs, such as IL-1α, IL-4, TGF-β1, MMP2, Ang1, bFGF, HGF, Tie2, VEGF, and VEGFR2, showed no change in cytokine levels, compared with Sham pbMSCs. Similarly, immunofluorescence results reveled that higher expression of IL-1α was related with lower level of IL-4, Ang1, and bFGF in Isch cMSCs (Fig. 1C-F). These results indicate that myocardial Isch induced cMSCs remodeling by decreasing of proangiogenic and anti-inflammatory factors.
Ischemia Mediated Imbalance in Expression of c-Myc and Oct4 in cMSCs
Since myocardial ischemia switched cMSCs toward an inflammatory phenotype, we searched for putative transcription factors (TFs) involved in inflammatory transformation of cMSCs. Our transcriptome comparison between Isch cMSCs and Sham cMSCs showed several known regulators were upregulated in Isch cMSCs, such as IL-6, interleukin-8 (IL-8), TNFα, TGF-β1, MMP2, and MMP9. These genes are all implicated in inflammation and fibrosis. We also found upregulation of c-Myc in Isch cMSCs (Fig. 2A). As c-Myc is one transcription factor of the four Yamanaka factors involved in proliferation, apoptosis, differentiation, immunity and somatic cell transformation . As for other three Yamanaka factors, Oct4 decreased significantly in Isch cMSCs, while other two factors, Klf4 and Sox2, showed no significant difference between Isch cMSCs and Sham cMSCs (Fig. 2A). To ascertain the expression difference, we performed qRT-PCR and western blotting to detect the mRNA and protein levels of these TFs. Myocardial ischemia significantly increased the mRNA and protein expressions of IL-6, IL-8, TNFα, TGF-β1, MMP2, MMP9, and c-Myc in cMSCs. Nevertheless, the expression of Oct4 was much lower in Isch cMSCs than in Sham MSCs. (Fig.2B, C). These expression difference was further confirmed by immunofluorescence staining. Compared with Sham cMSCs, Isch cMSCs showed a significant enrichment of IL-6 and TGF-β1 (Fig. 2D), higher expression of c-Myc, and lower expression of Oct4 (Fig.2E). Importantly, c-Myc are distributed in both the cytoplasm and nucleus of Isch cMSCs, but mainly located in the nuclei of Sham cMSCs, while Oct4 accumulated mainly in the nuclei of both Isch cMSCs and Sham cMSCs (Fig.2E). Together, our results indicate imbalance in expression of Oct4 and c-Myc under ischemic condition regulating genes associated with remodeling of cMSCs. Therefore, to explore the mechanisms involved in cMSCs remodeling under hypoxic or ischemic conditions, we focused on Oct4 and c-Myc in cMSCs from the ischemic hearts.
c-Myc and Oct4 Shows Divergent Phenotypes of cMSCs Under Hypoxia
Both Oct4 and c-Myc regulate pluripotency and stemness of adult stem cells [32, 33], and may regulate genes associated with angiogenesis and inflammatory post-MI [10, 34]. Therefore, we investigated their role in cMSCs remodeling in vitro, and cultured Isch cMSCs transfected with a lentivirus encoding overexpressed (oe) c-Myc (oec-Myc), Oct4 (oeOct4), or vehicle (served as control, CON) for 48 hours under hypoxic conditions. Normoxic cultivation of cMSCs with or without transfection of c-Myc or Oct4 was also performed to clarify these effects under normoxia. We confirmed that c-Myc and Oct4 transfection was efficiently induced in cMSCs under normoxia, and hypoxia further increased their expression (Fig. S3A, B). Notably, dynamic analysis of the percentage of BrdU-positive cells/total number of cells revealed that oec-Myc stimulated cell proliferation kinetically, and oeOct4 further enhanced this proliferation (Fig. S3C, D). Similar to BrdU expression, immunofluorescence showed that the level of Ki67 expression was higher in cMSCs receiving overexpression of c-Myc or Oct4 compared with their respective controls (Fig. S3E, F, I, and J). In line with these findings, CCK-8 assay in optical density (OD) value, which was a quantitative index for the growth capacity, showed the same trend in transfection efficiency as observed for the generation of cell proliferation (Fig. S3G, H). Importantly, compared with normoxia, the enhancement of this index was remarkably observed in the oeOct4 treating cMSCs after 48-h of hypoxic culture. Together, these results suggest that the transfection of c-Myc favors the growth of cMSCs, and Oct4 overexpression further enhances this regularity that is especially significant in cMSCs under hypoxic conditions. Thus, we next focused on evaluating the effects of c-Myc and Oct4 overexpression on the phenotype change of cMSCs under hypoxia.
First, we investigated global changes in gene expression that accompanied angiogenic/ inflammatory phenotypes of c-Myc and Oct4 overexpression in cMSCs, and conducted bulk RNA-seq on cMSCs overexpressed with or without c-Myc or Oct4. Consistent with the above data that c-Myc and Oct4 overexpression have different effects on the growth of cMSCs, gene set enrichment analysis of bulk RNA-seq data using the MSigDb showed about 75% of the up-regulated hallmark and Kyoto Encyclopedia of Genes and Genomes pathways in cMSCsoec-Myc versus cMSCsCON were also up-regulated in cMSCsoeOct4 versus cMSCsCON, and nearly 25% of all upregulated pathways in cMSCsoec-Myc versus cMSCsCONwere down-regulated in the cMSCsoeOct4 versus cMSCsCON(Fig. 3A, B). Then, we calculated the Log2 FC of cMSCsoeOct4 compared to cMSCsoec-Myc. 405 genes were upregulated more than 1 Log2 FC and 712 genes downregulated less than −1 Log 2 FC; in total, 4.2% of genes (Fig. 3C). According to GO terms, upregulated genes in cMSCsoeOct4 included response to angiogenesis, cell activation, immune system process, cell adhesion, cytokine-mediated signaling pathway, and positive regulation of cell activation, and the genes involved with inflammatory response were down-regulated in cMSCsoeOct4(Fig. 3D). From these signaling systems, of note are the following genes: Ang-1, bFGF, HGF, IGF-1 (Insulin-like growth factor-1), Tie2, and VEGF. These genes are all implicated in angiogenesis and MEndoT . Other notable genes are vWF, VEGFR, and Chd1 (chromatin remodelling factor 1). The majority of these genes play roles in proliferation, pluripotency, and angiogenic properties of stem cells [35, 36], while vWF is a gene encoding endothelial cell surface. By contrast, Oct4 overexpression dramatically reduced the mRNA expression levels of the inflammatory factors IL-1α, IL-6, and TNFα, and the fibrosis-related factors MMP2 and MMP9. However, compared with Oct4 overexpression, c-Myc overexpression caused a lower degree of decrease (Fig. 3E). These trends (oeOct4>oec-Myc) were also confirmed by immunofluorescence which revealed more vWF+ cells and less IL-1+TNFα+ cells and MMP2+MMP9+ cells in oeOct4versus CON, but relatively lower levels of vWF and higher levels of inflammation and fibrosis in cMSCsoec-Myc (Fig. 3F). Thus, c-Myc triggers diverse cellular outcomes: hypoxia-induced cellular angiogenesis, and cellular inflammation and fibrosis, whereas Oct4 overexpression promotes angiogenesis in coexistence with inhibition of inflammation and fibrosis.
Oct4 Cooperates with c-Myc to Accelerate cMSCs Remodeling
Oct4 and c-Myc are both stem cell regulators, and play key roles in sustaining and amplifying pluripotent stem cells , however, the functional interaction has proven obscure. Therefore, we next investigated whether Oct4 interacts with c-Myc in vitro by simultaneous transfection of cMSCs with RNA constructs encoding Oct4 and c-Myc, or c-Myc siRNA knockdown (sic-Myc), or vehicle (-), seeded on Matrigel to facilitate angiogenesis, and subjected to serum starvation under hypoxia. After 72 h of hypoxic cultivation, cMSCs were collected, and their angiogenesis, inflammatory, fbroblast were analyzed by immunofluorescence and western blotting. Immunofluorescence showed the highest expressions of the vascular characteristic markers, factor VIII and CD31 in the cMSCs cotransfected with Oct4 and c-Myc, followed by that in the sic-Myc- transfected cMSCs co-treated with oeOct4, and the lowest was observed in the sic-Myc-transfected MSCs (Fig. 4A, B). Conversely, Oct4 overexpression dramatically decreased the protein levels of inflammatory markers, CD80 and CD11b, and fibroblast markers including collagen I and vimentin, in oec-Myc-treated cMSCs, whereas the levels of these marker proteins were significantly higher in oec-Myc-transfected cMSCs than in sic-Myc-transfected cMSCs (Fig. 4A-D). Western blot showed similar results after Oct4 overexpression. Upregulating Oct4 in oec-Myc-transfected cMSCs promoted increase of factor VIII and α-SMA expressions, and down-regulated the expression of CD80, CD11b, collagen I, and vimentin. Moreover, Oct4 overexpression inhibited decrease of angiogenesis but promoted reduction of inflammatory and fibroblast in sic-Myc-transfected cMSCs (Fig. 4E). Overall, Oct4 accelerates cMSCs to switch towards an angiogenesis phenotype, and reduced their inflammatory and fibrosis induced by c-Myc overexpression.
Oct4 Regulates c-Myc Function and Translocation
We next investigated the underlying molecular mechanism responsible for the promotion of Oct4 on c-Myc-induced angiogenesis. It has been reported that co-activation of Myc can drive tumor cell proliferation by programming inflammation and angiogenesis . The global transcriptional regulations were examined in the cMSCs transfected with vehicle or oeOct4 after 72 h of hypoxic culture. We mined the expression of some known c-Myc targets in these cells and found that oeOct4-transfected cMSCs have higher expression of angiogenesis-related signaling (Fig. 5A), which is critical in promoting c-Myc-induced angiogenesis . This observation was confirmed by qRT-PCR (Fig. 5B). Especially, VEGF signaling including VEGF, VEGFR2, MAPK, and AKT  were specifically and significantly activated in oeOct4-transfected cMSCs. To validate the effect of Oct4 in mediating c-Myc functions, we transfected the oec-Myc-treated cells with either Oct4 or a control vector. Although Oct4 overexpression did not affect c-Myc mRNA levels (Fig. 5C), it promoted tube-forming ability in the c-Myc-treated cMSCs (Fig. 5D, E), confirming the role of Oct4 in mediating c-Myc-induced angiogenesis of cMSCs in vitro. To validate these results, we performed silencing assays. In contrast with Oct4 overexpression in cMSCs, silencing Oct4 markedly downregulated c-Myc targets mainly including VEGF, VEGFR2, MAPK, and AKT, compared with the control oligo, resulting in no significant change in c-Myc expression (Fig. 5F). It has been reported that altering Oct4 protein level affects its translocation from nuclear to cytoplasm . We isolated total RNA from cytoplasm and nuclei and measured the levels of Oct4. For cMSCs without Oct4 overexpression, higher levels of Oct4 were detected in nuclei than in cytoplasm, but c-Myc expressed more significantly in cytoplasm than in nuclei (Fig. 5G). For cMSCs transfected with Oct4, significantly higher levels of the ectopic Oct4 were detected in the cytoplasm relative to the nuclei (Fig. 5H). Western blotting using fractionated samples showed a similar distribution of c-Myc: Silencing Oct4 resulted in higher expression of c-Myc in the nuclei than in the cytoplasm, whereas ovexpressing Oct4 produced an opposite pattern of c-Myc distribution (Fig. 5I). By confocal microscopy, we detected mainly cytoplasmic c-Myc in the Oct4-transfected cells and nuclear distribution of c-Myc in the vector-transfected cells (Fig. 5J). To confirm the binding of the c-Myc promoter by Oct4, a chromatin immunoprecipitation (ChIP) assay in cMSCs was performed. It was suggested that c-Myc promoter fragments could only be amplified from chromatin precipitated with Oct4 antibodies (Fig. 5K). Taken together, our study showed that Oct4 overexpression induced cytoplasmic translocation of c-Myc. It has been reported that nuclear c-myc is readily degraded . In our study, c-Myc was mainly translocated to the cytoplasm in oeOct4-treated cMSCs, but we did not detect decreased level of total c-Myc. This suggested that c-Myc was colocalized with Oct4, preventing degradation (Fig. S4).
Expression of Oct4 Activating c-Myc in cMSCs Improves Their Myocardial Repair
To further assess the role of Oct4 -c-Myc in mediating the response of cMSCs to local signals in the injured heart, we injected cMSCs pre-treated with lentiviral vectors encoding Oct4 (oeOct4), Oct4 siRNA (siOct4), vehicle, or in combination with c-Myc siRNA (sic-Myc) or c-Myc overexpression (oec-Myc) were randomly transplanted into the hearts of recipient rats subjected to acute MI. Rats with MI were also randomly assigned to receive PBS injection and served as a control group. The 160 animals were randomly divided into eight groups and underwent serial echocardiography studies. Thereafter, all animals were followed-up for 30 days, during which 36 rats died. The surviving 124 rats were then sacrificed and subjected to pathology and molecular biology tests. We evaluated cardiac function and remodeling by echocardiography at 1 and 30 days after cell transplantation. The rationale for a 30-day follow-up was to determine mycoardial repair of transplanted cMSCs. After 30 days, Kaplan–Meier survival analysis showed higher survival rate in the rats receiving transplantation of cMSCs transfected with both oec-Myc and oeOct4 than in the rats receiving PBS injection and cMSCs therapy alone (95% in the rats receiving both oec-Myc and oeOct4-treated PBMSCs versus 50% in the rats receiving PBS, p=0.016; versus 55% in the rats receiving cMSCs alone, p=0.020), and this effect was canceled in the rats receiving siOct4-treated cMSCs. However, the inhibition of Oct4 was not rescued by c-Myc overexpression. No significant differences were found between the animals receiving cMSCs alone and those receiving cMSCs pretreated with oec-Myc plus siOct4 or sic-Myc plus oeOct4 (Fig. 6A). Echocardiographic studies showed that on day 1 post-infarct, animals in all eight groups developed typical changes of acute heart failure and LV early remodeling, in comparison with data obtained at the baseline levels. These changes included decreased cardiac function index LVFS, dilated LVEDD and LVEDV, and thinning of LVAWd. Significantly, cMSCs transfected with oec-Myc and oeOct4 were more effective than vector-treated cMSCs in the improvement of cardiac function, and the prevention of adverse LV dilatation after MI (Fig.6B-E). Although transplantation of vector-treated cMSCs revealed slight improvement of LVFS, LVEDD, LVEDV, and LVAWd at 30d post-MI, oec-Myc-treated cMSCs were more effective than PBS in the improvement of these indexes. Moreover, compared with vehicle-treated cMSCs and oec-Myc-treated cMSCs, cMSCs treated with both oec-Myc and oeOct4 futher increased the change in LVFS by 3.6- and 2.2-fold, respectively (Fig. 6B), and LVAWd by 75% and 31%, respectively (Fig. 6E), and decreased the changes in LVEDD by 94% and 44%, respectively (Fig. 6C), and LVEDV by 1.0- and 0.5-fold, respectively (Fig. 6D). These beneficial effects were eliminated in the rats receiving sic-Myc-treated cMSCs. The indices in the rats receiving PBS indicated sustained exacerbation with decrease of LVEF, dilation of LVEDD and LVEDV, and thinning of LVAWd. Similarly, TTC staining demonstrated the smallest amount of infarct size in the rats receiving cMSCs co-treatd with oec-Myc and oeOct4 (Fig.6F). However, deficiency of c-Myc or Oct4 cancelled this decrease. Thus, overexpression of Oct4 in c-Myc-transfected cMSCs enhanced their ability to improve cardiac remodeling after MI.
Postmortem morphometry revealed that compared with PBS injection, cMSCs therapy did not significantly reduce MI-induced inflammation, including neutrophil infiltration, myocyte loss, and bleeding. Transplantation of cMSCs co-transfected with Oct4 and c-Myc caused the greatest reduction of myocardial inflammation, accompanying the greatest increase of viable cardiomyocytes. However, this amelioration was abolished by transfection of Oct4 siRNA (Fig. 7A-C). By microscopic examination, infarcted hearts treated with oec-Myc-transfected cMSCs displayed inflammation, fibrosis, and angiogenesis at the infarct site, and co-transfection of the two siRNAs of c-Myc and Oct4 showed extensive fibrosis with a thicker, fibrotic scar 30 days after MI (Fig. 7A, D, E). These findings probably reflect an anti-fibrotic response of Oct4 in collaboration with the implanted cardiac MSCs. It is interesting to note that a microscopic evaluation revealed cardiomyogenesis in the rats receiving transplantation of cMSCs co-transfected with c-Myc and Oct4. This cardiomyogenesis was mainly located around the blood vessels (Fig. 7A).
In the experiments with cMSCs receiving transfection of vehicle, oec-Myc, sic-Myc, oeOct4, or siOct4, we transplanted cells from male donor to female recipient mice hearts. Thus, cell retention was assessed by staining for the sex-determining region Y chromosome (Fig. 8A–C). Indeed, we found significant amounts of donor cells at the site of transplantation of cMSCs transfected with oec-Myc plus oeOct4. In contrast, we identified only a few donor cells in the hearts receiving sic-Myc/siOct4-treated cMSCs (Fig. 8A, B). Quantitative statistical analysis displayed that the number of cMSCs was greater in hearts treated with cMSCs co-transfected with c-Myc and Oct4 than in the hearts treated with cMSCs-alone, or cMSCs transfected with sic-Myc or/ plus siOct4, 30 days after MI (Fig. 8C). To determine the distribution of these transplanted cells in the infarcted heart, we pre-labeled c-MSCs genetically with EGFP before transplantation. Notably, EGFP+ (plasma stained green) cells were not dispersed throughout the left ventricle but were assembled around blood vessels as revealed by double staining these cells with anti-vWF monoclonal antibodies (Fig. 8D, E). vWF is a marker of blood vascular endothelial cells and communicates angiogenesis. The number of vWF+ (plasma stained red) endothelial cells per square millimeter was counted under fluorescence microscope in the peri-infarct regions of hearts from these different treatment groups. The vessel density was increased in oec-Myc-treated cMSCs compared with PBS -treated cMSCs and was even higher in the oec-Myc+oeOct4 treated cMSCs than in the oec-Myc-transfected cMSCs (Fig. 8D-F). Moreover, some EGFP-labeled cMSCs expressed vWF (Fig. 8D, E, arrowhead). These cells seemed to switch into a vascular endothelial phenotype. Overexpression of Oct4 promoted the expression of vWF on oec-Myc-transfected cMSCs (Fig. 8D). In addition, the density of transplanted cells was positively correlative with the number of blood vessels assessed by immunofluorescence (Fig. 8G, r =0.94, p=0.018), probably communicating proangiogenic “rescue me” signals to save cMSCs. Together our findings indicate that Oct4 overexpression cooperated with c-Myc to promote MEndoT of implanted cMSCs in vivo, improve their survival, and enhance cardiomyogenesis, which subsequently led to decreased scar formation and inhibition of LV remodeling.