Generation of CD31+ endothelial cells derived from hiPSCs and their in vitro characterization

Several previous studies reported successful generation of ECs from hiPSCs (hiPSC-ECs) using a combination of small molecules, including a GSK3 inhibitor22. Based on previous reports, we generated hiPSC-ECs from hiPSCs using the GSK3 inhibitor CHIR99021 (Supplementary materials and methods, Supplementary Fig. 1a). To produce hiPSC-ECs expressing green fluorescence protein (GFP) to facilitate cell tracking in the heart tissues in further experiments, we produced hiPSCs expressing GFP signals by transfecting GFP lentiviral particles, enriched them by FACS based on GFP expression, and used them to differentiate into ECs (Supplementary Fig. 1b). qRT-PCR results verified that the expression level of OCT4, a pluripotency marker, was significantly reduced, and the expression level of CD31, a specific marker for EC, was significantly increased in the hiPSCs differentiating into the EC lineage (Supplementary Fig. 1c, Supplementary Table 1). On differentiation Day 7, we observed that approximately 25.08% of the differentiating hiPSC-ECs were positive for human CD31 antibody, and subsequently, we enriched these CD31+ cells by FACS. Following FACS, the enriched CD31+ hiPSC-ECs were maintained in human endothelial serum-free medium with cytokines, including VEGF, to maintain their characteristics as EC lineage cells (Supplementary Fig. 1c). The CD31+ hiPSC-ECs displayed a typical cobblestone-like EC morphology and expressed similar mRNA levels of EC-specific markers, such as cluster of differentiation 31 (CD31), vascular endothelial cadherin (VE-Cadherin), Von Willebrand factor (vWF) and vascular endothelial growth factor receptor 2 (VEGFR2), compared with human umbilical cord endothelial cells (HUVECs) (Supplementary Fig. 1c, d). The results from flow cytometry analysis further demonstrated that the CD31+ hiPSC-ECs were 97.19% and 85.53% positive for CD31 and CD144, respectively (Supplementary Fig. 1e). In addition, the immunofluorescence results confirmed that the CD31+ hiPSC-ECs expressed abundant levels of the CD31 and vWF proteins (Supplementary Fig. 1f). At the functional level, the CD31+ hiPSC-ECs displayed the capacity for uptake of Ac-LDL (Supplementary Fig. 1g) and the formation of a capillary-like network on top of Matrigel (Supplementary Fig. 1h).

To determine if the CD31+ hiPSC-ECs (hiPSC-ECs afterward) could form de novo vessels via a vasculogenesis-dependent mechanism in MI-induced hearts, we intramyocardially injected hiPSC-ECs at two different sites in the border zone of the MI-induced rat hearts. MI was generated by ligation of the left anterior descending (LAD) artery in the heart. hiPSC-ECs continuously expressing the green fluorescence (GFP) signal were used for tracking purposes. To visualize the functional vessels in the MI-induced hearts, we performed perfusion staining with isolection-B4 (IB4) conjugated with a red fluorescent dye, rhodamine, into the heart prior to tissue harvest 8 weeks after injection with hiPSC-ECs. Fluorescent image analyses showed that the number of IB4+ capillaries in the hiPSC-EC-injected hearts was significantly higher than that in the MI control hearts (Fig. 1a).

a Representative images of blood vessels stained with IB4-rhodamine (red) in the infarct zone, border zone, and remote zone and at 8 weeks after injection of hiPSC-ECs and their quantification summary. For quantification, the number of capillaries in five randomly selected fields (mm2) in each heart was counted. n=5. *p<0.05. Scale bars: 100m. b Representative image of blood vessels newly formed by iPSC-ECs-GFP (green), IB4-rhodamine (red) and DAPI (blue). Scale bars: 20m. cj Rats undergoing MI were intramyocardially injected with hiPSC-ECs or control cells, followed by echocardiography analysis. c The schematic timeline from MI modeling and transplantation of iPSC-EC to measurement of cardiac function. d Left ventricular ejection fraction (EF), (e) left fractional shortening (FS), (f) left ventricular internal diastolic dimension (LVIDd), (g) left ventricular internal systolic dimension (LVIDs), (h) septal wall thickness (SWT), (i) posterior wall thickness (PWT), and (j) relative wall thickness (RWT). n=6. *p<0.05. k Representative images showing cardiac fibrosis after staining with Massons trichrome in the hearts harvested 8 weeks after cell treatment. Quantification results of cardiac fibrosis (l) and viable myocardium (m). n=5. *p<0.05. Scale bars: 2000m.

Next, to evaluate the potential and magnitude of the contribution of hiPSC-ECs to vasculogenesis in the MI hearts, we traced the GFP and RFP signals from hiPSC-ECs within cardiac tissues. Confocal microscopy images demonstrated a considerable number of vessels, double-positive for both IB4 and GFP signals from hiPSC-ECs, in the infarct region of the heart tissues receiving hiPSC-ECs at 8 weeks post-injection. Interestingly, a substantial number of hiPSC-ECs were incorporated into the host capillary network, and many of them were located in the perivascular area (Fig. 1b). The results clearly suggest that hiPSC-ECs could reconstruct de novo vessels in ischemic hearts.

Given that vascular regeneration improved through vasculogenesis leads to functional recovery from MI, we hypothesized that intramyocardial injection of hiPSC-ECs into MI hearts may promote cardiac function. Subsequently, we performed serial echocardiography to evaluate left ventricular (LV) function and cardiac remodeling from PRE (1-week post-MI and prior to cell treatment), and 2, 4, and 8 weeks after cell treatment. In this study, we employed a MI model that cells were transplanted one week after induction of MI to mimic the clinical situation of MI patients as close as possible. The results of echocardiography demonstrated that both ejection fraction (EF) and fractional shortening (FS) in all experimental groups were significantly lower compared with the sham group that did not receive any intervention. (Supplementary Fig. 2ag). Of importance, the hearts receiving hiPSC-ECs displayed significantly higher EF and FS than those in the MI control group until 8 weeks after the cell treatment (Fig. 1cd). Among several parameters for cardiac remodeling, such as left ventricular internal diastolic dimension (LVIDd), left ventricular internal systolic dimension (LVISd), septal wall thickness (SWT), posterior wall thickness (PWT), and relative wall thickness (RWT), the LVIDd and LVIDs in the hiPSC-EC-treated hearts were significantly lower than those in MI control hearts, indicating that hiPSC-ECs protected the hearts from adverse cardiac remodeling. (Fig. 1ei and Supplementary Fig. 2h). Similarly, the results of Massons trichrome staining obtained using cardiac tissue harvested at 8 weeks post-cell treatment showed that the area of fibrosis (%) in the hiPSC-EC-injected group was considerably smaller and the viable myocardium (%) was larger than that in the MI control group (Fig. 1jm). Based on these results, we confirmed that hiPSC-ECs can directly contribute to de novo vessel formation in vivo in MI-exposed hearts, resulting in enhanced cardiac function.

Subsequently, we investigated our central hypothesis of whether simultaneous induction of both vasculogenesis and angiogenesis could lead to comprehensive vascular regeneration and functional improvement in the MIhearts. Since we already verified that hiPSC-ECs successfully achieved vasculogenesis in the MI hearts, we sought to identify an additional cellular source that can induce complementary angiogenesis from the blood vessels in the host heart and finally decided to test genetically modified human mesenchymal stem cells engineered to continuously release human SDF-1 protein (SDF-eMSCs)23. The SDF-eMSCs were indistinguishable from normal BM-MSCs. The SDF-eMSCs exhibited a homogeneous spindle-shaped cell morphology, representing hMSCs (Supplementary materials and methods, Supplementary Fig. 3a). The SDF-eMSCs had a high proliferative potential based on the gradual increase in population doubling levels (PDL) during the culture times compared to normal BM-MSCs24 (Supplementary Fig. 3b). The SDF-eMSCs expressed several markers specific for human MSCs, such as CD90, CD44, CD105 and CD73, without the expression of CD34, CD11b, CD19, CD45 and HLA-DR (Supplementary Fig. 3c). The SDF-eMSCs stably secreted human SDF-1 protein, as determined by human SDF-1 enzyme-linked immunosorbent assay (ELISA) analysis (Supplementary Fig. 3d). The results from SDF-eMSC karyotyping revealed a normal human karyotype of the SDF-eMSCs without chromosomal abnormalities, suggesting the genetic stability of the SDF-eMSCs (Supplementary Fig. 3e).

To investigate whether SDF-eMSCs could augment the angiogenic potential of ECs, we performed various types of in vitro experimental analyses with SDF-eMSCs. Among the first, to determine whether SDF-eMSCs influenced the gene expression associated with ECs and angiogenic properties, we treated 30% conditioned media (CM) harvested from cultured SDF-eMSCs or BM-MSCs to the cultured hiPSC-ECs for 3 days and performed qRT-PCR analyses. The expression levels of stromal-derived factor-1 alpha (SDF-1), tyrosine kinase with Ig and epidermal growth factor homology domain 2 (Tie-2), vWF, E-selectin (CD62), and intercellular adhesion molecule-1 (ICAM-1) were significantly higher in the hiPSC-ECs treated with SDF-eMSC-CM than in the hiPSC-ECs exposed to BM-MSC-CM (Fig. 2a). In particular, the increased expression of E-selectin and ICAM-1 is known to be involved in angiogenesis in the presence of activated ECs25,26,27,28,29. Next, in EC migration assays, as shown in Fig. 2, the addition of conditioned media from the SDF-eMSCs (SDF-eMSC-CM) significantly enhanced the migration of hiPSC-ECs or HUVECs compared with the migration of the ECs treated with CM from human bone marrow-derived MSCs (BM-MSC-CM), suggesting that cytokines released from the SDF-eMSCs bolster the mobility of ECs (Fig. 2b and Supplementary Fig. 4a). In addition, to test whether SDF-eMSCs directly promote the angiogenic potential of ECs, we performed Matrigel tube formation assays, a representative experiment to evaluate the vessel formation potential of cells. The results from Matrigel tube formation assays demonstrated that the number of branches formed in both the hiPSC-ECs and the HUVECs treated with 30% CM harvested from the cultured SDF-eMSCs was significantly greater than that in the BM-MSC-CM-treated ECs (Fig. 2c and Supplementary Fig. 4b). Interestingly, treatment with SDF-eMSC-CM not only promoted tube formation by the hiPSC-ECs but also contributed to the maintenance of vessels formed from the hiPSC-ECs. Unlike the hiPSC-EC-generated vessels exposed to BM-MSC-CM that began to disrupt the vessel structure within 24h of vessel formation, treatment with SDF-eMSC-CM supported the integrity of vessels for up to 48h.

a qRT-PCR analysis of relative mRNA expression associated with ECs and angiogenesis in the hiPSC-ECs treated with the conditioned media (CM) from cultured bone marrow mesenchymal stem cells (BM-MSC-CM) or SDF engineered MSCs (SDF-eMSC-CM) for 3 days. The y-axis represents the relative mRNA expression of target genes to glyceraldehyde-3-phosphate dehydrogenase (GAPDH). n=3. *p<0.05. b EC migration assay. Representative images of migrated hiPSC-ECs and quantification of the migrated area (%). The hiPSC-ECs were placed in transwells (top), and regular media (EGM, EBM) or the conditioned media (CM) collected from different cell sources (BM-MSC-CM and SDF-eMSC-CM) were placed in transwells (bottom) for 7h. n=3. *p<0.05. c Tube formation assay. The hiPSC-ECs were cultured in 24-well plates coated with Geltrex with regular media (EGM, EBM) or conditioned media (CM) (BM-MSC-CM and SDF-eMSC-CM) for 9, 24, or 48h. Representative images of tube formation and quantification summary for the number of junctions. n=3. *p<0.05.

To provide a cellular reservoir where SDF-eMSC-PAs can constantly release SDF-1 to MI hearts, we produced a patch encapsulating SDF-eMSC (SDF-eMSC-PA) by mixing SDF-eMSCs with a 2% heart-derived decellularized extracellular matrix (hdECM)-based bioink and loaded it onto the polycaprolactone (PCL) mesh (Fig. 3 and Supplementary Fig. 5). Subsequently, to confirm whether SDF-eMSC-PAs are functional and can efficiently release SDF-1, we cultured SDF-eMSC-PAs in vitro for 28 days (Supplementary Fig. 5a) and collected supernatants at various time points for three days to generate the release kinetics of the SDF1-eMSC-PAs using the SDF-1 ELISA kit. The cumulative release curve showed that although the initial concentration of SDF-1 was higher in the SDF-cytokine-PAs (300ng/ml) than in the SDF-eMSC-PAs on Day 0, no SDF-1 was detectable in the SDF-cytokine-PAs from Day 7. However, the expression of SDF-1 released from the SDF-eMSC-PAs increased consistently until Day 21 (Supplementary Fig. 5b), suggesting that SDF-eMSCs continuously secreted SDF-1 within the patch.

a Procedures for manufacturing a cardiac patch encapsulating SDF-engineered MSCs (SDF-eMSC-PA) with a polycaprolactone (PCL) platform produced by a 3D printing system. b Optical image within the hdECM patch. SDF-eMSC-PAs were prelabeled with the red florescence dye DiI for tracing. Scale bars: 1mm. c Image of epicardially transplanted SDF-eMSC-PAs in the MI-induced heart. d Macroscopic view of hearts at 8 weeks after PA transplantation.

To finally determine whether simultaneous induction of both vasculogenesis and angiogenesis by using hiPSC-ECs and SDF-eMSC-PAs could lead to comprehensive vascular regeneration and functional improvement in MI-induced hearts, we induced MI by LAD ligation after the formation of five experimental groups as follows: (1) MI control, (2) SDF-eMSC-PA implanted epicardium of MI hearts (PA only, 1106), (3) hiPSC-ECs, intramyocardial injection (EC only, 1106), and (4) combined platform of hiPSC-ECs and SDF-eMSC-PA (EC+PA, 1106 in each) (Fig. 4). We first performed serial echocardiography for all experimental groups at pre, 2, 4 and 8 weeks after cell treatment. All experimental groups were significantly reduced compared with that in the sham group (Supplementary Fig. 6ag). Of interest, cardiac function in the EC+PA group was significantly preserved until 8 weeks compared with its cardiac function at pre, but cardiac function in other groups, such as the control, hiPSC-EC alone and SDF-eMSC-PA alone groups, continuously decreased until 8 weeks. (Fig. 4ad). Adverse cardiac remodeling determined by the LVIDd, LVIDs, SWT, PWT, and RWT was notably reduced in the EC+PA group compared with the other groups (Fig. 4ei and Supplementary Fig. 6h). To further evaluate cardiac function more precisely, we performed LV hemodynamic measurements using an invasive pressure-volume (PV) catheter, which can measure the hemodynamic pressure and volume of the LV. The results of the PV loop at 8 weeks post-cell treatment showed that the EC+PA group had significantly improved cardiac function and prevented adverse cardiac remodeling compared with the other groups (Fig. 5). The two parameters of general cardiac function, stroke volume (SV) and cardiac output (CO), were significantly higher (Fig. 5ac), and the maximum volume (V max), which is the cardiac remodeling index measured at the maximum diastole, was significantly lower in the EC+PA group than in the other groups (Fig. 5d). Although the pressure max (P max) measured at the maximum systole did not differ significantly between groups, the maximum rate of pressure change (dP/dtmax) and the minimum rate of pressure change (dP/dtmin), which indicate the pressure change in LV per second, were increased in the EC+PA group. (Fig. 5ef and Supplementary Fig. 7a). Temporal variation in the occluded inferior vena cava (IVC) was used to evaluate load-independent intrinsic cardiac contractibility. The end-diastolic pressure-volume relationship (EDPVR), which indicates the absence of diastolic dysfunction, did not differ between the groups, whereas the slope of the end-systolic pressure volume relationship (ESPVR), which indicates cardiac contractibility, was significantly improved in the EC+PA group compared with the other groups (Fig. 5gh and Supplementary Fig. 7b). Collectively, these results from LV hemodynamic measurements consistently demonstrate that treatment with the combined platform with hiPSC-ECs and SDF-eMSC-PAs improves cardiac repair in MI hearts.

a Left ventricular ejection fraction (EF). b EF delta change at 8 weeks after cell treatment. c Left fractional shortening (FS). d FS delta change at 8 weeks after cell treatment. e Left ventricular internal diastolic dimension (LVIDd). f Left ventricular internal systolic dimension (LVIDs). g Septal wall thickness (SWT). h Posterior wall thickness (PWT). i Relative wall thickness (RWT). n=611. *p<0.05.

a Representative images of the hemodynamic pressure and volume (PV) curve at steady state at 8 weeks post-cell treatment. b Stroke volume (SV). c Cardiac output (CO). d Volume max (V max) defining the amount of blood volume in the LV at end-diastole. e dP/dtmax refers to the maximal rate of pressure changes during systole. f The minimal rate of pressure changes during diastole (dP/dtmin). g Slope of end-systolic pressure volume relationship (ESPVR) indicating the intrinsic cardiac contractibility as measured by transient inferior vena cava (IVC) occlusion. h Slope of end-diastolic pressure volume relationship (EDPVR). n=4. *p<0.05.

Next, we investigated the in vivo behavior of intramyocardially implanted hiPSC-ECs in the presence or absence of SDF-eMSC-PAs. Since hiPSC-ECs constantly express GFP, we could track their fate in heart tissue sections. Confocal microscopic examination of heart tissues harvested at 8 weeks after cell treatment demonstrated that implantation of SDF-eMSC-PAs significantly improved the retention and engraftment of intramyocardially injected GFP-positive hiPSC-ECs. Quantitatively, the proportion of GFP-positive hiPSC-ECs in the EC+PA group was substantially higher than that in the EC group (Fig. 6a). Of interest, while the hiPSC-ECs in the hiPSC-ECs alone group were localized near the injection sites, the hiPSC-ECs in the EC+PA group were distributed throughout the regions of the left ventricle. Given the ability of SDF-eMSC-PAs to improve the survival and retention of injected hiPSC-ECs, we sought to examine whether SDF-eMSC-PAs exerted direct cytoprotective effects in hiPSC-ECs in vitro. Ischemic injury was simulated by exposing hiPSC-ECs to H2O2 (500M). Administration of CM from SDF-eMSCs (SDF-eMSC-CM) significantly improved the viability of both hiPSC-ECs and HUVECs as determined by the LIVE/DEAD assay and CCK-8 assay (Supplementary Fig. 8af). Treatment with SDF-eMSC-CM substantially increased the number of viable cells, suggesting that SDF-eMSC-CM exerts direct cytoprotective effects on ECs against ischemic insults. Subsequently, we performed thorough histological analyses using heart tissues harvested 8 weeks post-cell treatment to examine whether SDF-eMSC-PAs could concurrently promote hPSC-EC-dependent vasculogenesis as well as angiogenesis of host blood vessels. IB4 conjugated with rhodamine was systemically injected to identify the functional endothelium in these experiments. Initially, confocal images demonstrated that the number of total IB4-positive (IB4+) capillaries in both the border zone and infarct zone of the hearts in the EC+PA group was substantially higher than that in the other groups, including the EC group (Fig. 6b). The number of vessels that were GFP negative but positive for IB4 (GFP-/IB4+) was also significantly higher than that in other groups, including the EC-only group (Fig. 6c). These results suggest that the combined approach significantly promoted the angiogenesis of host vessels in MI hearts. More importantly, the number of de novo vessels formed by hiPSC-ECs-GFP+ was substantially higher in the EC+PA group than in the EC-only group, indicating that SDF-eMSC-PAs facilitates hiPSC-EC-dependent vasculogenesis (Fig. 6de and Supplementary Fig. 9a). Notably, the number of larger blood vessels (diameter range: >5 m), one of the indicators of functional blood vessels in the EC+PA group, was significantly higher than that in the EC group. Of interest, many of those larger vessels in the EC+PA group displayed abundant expression of -SMA, a marker for smooth muscle cells, suggesting that these larger vessels (CD31+/-SMA+) may be arteriole-like vessels, indicating that SDF-eMSC-PA played certain roles in vascular ingrowth and maturation (Fig. 6eg and Supplementary Fig. 9b).

a Representative image of hiPSC-ECs-GFP within the infarct area at 8 weeks post-cell treatment and their quantification summary. n=3. *p<0.05. Scale bars: 1000m. b Representative images of blood vessels stained with IB4-rhodamine (red) in the infarct zone (IZ), border zone (BZ), and remote zone at 8 weeks after cell treatment and a summary of their quantification. n=57. *p<0.05. Scale bars: 100m. c Representative images of blood vessels negative for GFP but positive for IB4 (GFP-/IB4+) in the infarcted area and their quantification summary. hiPSC-ECs-GFP (green), IB4-rhodamine (red) and DAPI (blue). n=5. *p<0.05. Scale bars: 20m. d, e Representative images of GFP and IB4 (GFP+/IB4+)-positive blood vessels in the infarcted area and their quantification. hiPSC-ECs-GFP (green), IB4-rhodamine (red) and DAPI (blue). n=5. *p<0.05. Scale bars: 20m. f, g Diameter of hiPSC-EC-derived GFP-positive blood vessels in the infarcted area and border zone. n=5. *p<0.05.

To further investigate whether the vascular regeneration achieved by the combined platform (EC+PA) was sufficient to salvage the myocardium from ischemic insult, we quantified the viable myocardium by immunostaining for cardiac troponin T (cTnT) antibody using the heart tissues harvested from all experimental groups at 8 weeks post-cell treatment. The number of viable cTnT+ cardiomyocytes in the EC+PA group was significantly higher than that in the other groups (Fig. 7a). These results from histological analyses using heart tissues motivated us to test whether SDF-eMSCs (Supplementary Fig. 10a) could confer direct cytoprotective effects on cardiomyocytes against ischemic insults in vitro. Ischemic injury was simulated by exposing cardiomyocytes to H2O2 (500M). The results from both the LIVE/DEAD assay and the cholecystokinin-8 (CCK-8) assay demonstrated that the administration of SDF-eMSC-conditioned media (CM) significantly improved the viability of cultured cardiomyocytes isolated from neonatal rats (NRCM) against H2O2 treatment compared with other treatments. These results also suggest that SDF-eMSCs have direct cytoprotective effects against ischemic insults (Supplementary Fig. 11ac).

a Representative immunostaining images of myocardium stained with cTnT (green) and DAPI (blue) at 8 weeks after cell treatment and quantification of the number of cTnT-positive cardiomyocytes. SDF-eMSCs labeled with DiI within the cardiac patch (red) n=5. *p<0.05. Scale bar: 300m. b Representative images of Massons trichrome staining using heart tissues harvested 8 weeks after cell treatment. c, d Quantification summary of a percentage of fibrosis and viable myocardium. n=5. *p<0.05. Scale bars: 2000m.

Consequently, the combined treatment group showed a significant decrease in cardiac fibrosis. The results of Massons trichrome staining using cardiac tissue harvested at 8 weeks exhibited an area of fibrosis (%), which was significantly lower in the combined treatment groups than in the other groups (Fig. 7bd). Taken together, our results clearly suggested that the combined treatment resulted in comprehensive cardiac repair through enhanced vascular regeneration and that the SDF-eMSCs contributed at least to some extent indirect protection of myocardium from ischemic injury via consistent secretion of cytoprotective SDF cytokines.

Visit link:
Enhancement strategy for effective vascular regeneration following myocardial infarction through a dual stem cell approach | Experimental &...

Comments are closed.