Platelet-derived growth factor receptor-alpha positive cardiac progenitor cells derived from multipotent germline stem cells are capable of cardiomyogenesis in vitro and in vivo

Cardiac cell therapy has the potential to revolutionize treatment of heart diseases, but its success hinders on the development of a stem cell therapy capable of efficiently producing functionally differentiated cardiomyocytes. A key to unlocking the therapeutic application of stem cells lies in understanding the molecular mechanisms that govern the differentiation process. Here we report that a population of platelet-derived growth factor receptor alpha (PDGFRA) cells derived from mouse multipotent germline stem cells (mGSCs) were capable of undergoing cardiomyogenesis in vitro. Cells derived in vitro from PDGFRA positive mGSCs express significantly higher levels of cardiac marker proteins compared to PDGFRA negative mGSCs. Using Pdgfra shRNAs to investigate the dependence of Pdgfra on cardiomyocyte differentiation, we observed that Pdgfra silencing inhibited cardiac differentiation. In a rat myocardial infarction (MI) model, transplantation of a PDGFRAenriched cell population into the rat heart readily underwent functional differentiation into cardiomyocytes and reduced areas of fibrosis associated with MI injury. Together, these results suggest that mGSCs may provide a unique source of cardiac stem/progenitor cells for future regenerative therapy of damaged heart tissue.


INTRODUCTION
Current therapeutic approaches for end-stage heart failure are limited to pharmacological therapies, mechanical ventricular assist devices (VAD), and cardiac transplantation. Unfortunately, even the availability of a donor hearts is limited by complications associated with a lifetime of immunosuppression. Stem cell-based therapies offer a novel approach to overcome these limitations by replacing damaged or lost myocardial tissues and restore cardiac functions. Several candidate cell types used in preclinical animal models and humans studies include, embryonic stem cells (ESCs), induced pluripotent stem cells (iPSCs), neonatal cardiomyocytes, skeletal myoblasts (SKMs), endothelial progenitor cells (EPCs), and mesenchymal stem cells (MSCs) [1][2][3][4]. A consensus, however, on the ideal cell type for the treatment of heart disease has yet to be reached.

Research Paper: Cardiology
Oncotarget 29644 www.impactjournals.com/oncotarget The success of stem cell therapy in cardiac regeneration relies, in part, upon identifying cell surface makers that enable reliable enrichment of a cardiac stem/ progenitor cell population. Among potential candidate markers for cardiac progenitors is platelet-derived growth factor receptor-alpha (Pdgfra); a cell-surface protein expressed on cardiac progenitor cells [5,6]. Differentiation of mouse ESCs into cardiomyocyte reveals that this subpopulation co-expresses PDGFRA and another marker FLK1 [7].
Initially thought to be unipotent, germline stem cells (GSCs) cultured under defined conditions are capable of acquiring pluripotency [8][9][10][11][12][13]. Recently, we demonstrated that multipotent germline stem cells (mGSCs) express markers of pluripotency, can differentiate into derivatives of all three germ layers in vitro, and are capable of forming teratomas in immune deficient mice [14]. Additionally, mGSCs are capable of differentiatiate into cardiomyocytes and endothelial cells in vitro, and in vivo studies suggest that these cells have ability to restore functions in damaged hearts of animal models [5,15].
In this study, we used defined culture conditions to derive cardiac stem/progenitor cells from mouse mGSCs. Particularly, we found that isolation of PDGFRA expressing cardiac stem/progenitor cells were capable of effective differentiation into cardiomyocytes in vitro, and displayed in vivo functional properties when transplanted in the hearts of a rat model of myocardial infarction. Together these findings suggest that mGSCs are a potential stem cell source from which to derive cardiac stem/ progenitor cells capable of repairing damaged myocardial tissue.

Effects of differentiation medium on mGSCs cardiac induction
Our first steps were to determine the optimal culture conditions that promote cardiac differentiation of mGSCs. As such, embryoid bodies (EBs) derived from mGSCs were cultured for 3 days in either IMDM/FBS, KO-DMEM/KSR, KO-DMEM/FBS, or N2/B27 medium. To evaluate the temporal changes in gene expression associated with early cardiogenesis, we assessed the expression of Brachyury, a T-box domain-containing transcription factor expressed in embryonic mesoderm that is down-regulated following initiation of tissuespecific patterning [16]. Exposure of mGSC-derived EBs to N2/B27 medium was associated with a marked increase in Brachyury gene expression (Supplementary Figure 1). This up regulation is consistent with previous findings showing that EBs display a characteristic spike in Brachyury expression at the onset of cardiac differentiation [16].

Analysis of FLK1 and PDGFRA expression during differentiation
We next evaluated cardiac differentiation of mGSCderived EBs following exposure to N2/B27 culture medium (without growth factors) by using flow cytometry to assess PDGFRA and FLK1 expressing populations. Following exposure to N2/B27 culture medium (without growth factors), we observed the fraction of PDGFRA + cells increase by 0.1%, 9.6%, and 13.3% after 3, 4, and 5 days, respectively. In contrast, FLK1 + expressing cells accounted for only 0.2%, 0.5%, and 1.0% of this same population (Supplementary Figure 2A). Culture of mGSC-derived EBs in MEMα containing 10% FBS promoted a 1.3%, 7.9%, and 13.8% increase in FLK1+ expressing cells after 3, 4, and 5 days, but was conversely associated with only a small fraction of PDGFRA+ cells (Supplementary Figure 2B).

Analysis of cardiac lineage differentiation potential of PDGFRA + population
After 5 days of culturing mGSCs in N2/B27 culture medium, the cells were FACS sorted by gating for PDGFRA + or PDGFRA − cell populations ( Figure 1A). These respective cell populations were then collected and plated on 0.1% gelatin-coated 24-well culture dishes in N2/B27 medium containing 30 ng/mL bFGF and 10 ng/mL VEGF. Two days after plating, the expression of Pou5f1, a marker of pluripotency was assessed. Specifically, the mGSCs used in these experiments were derived from transgenic mice expressing Enhanced Green Fluorescent Protein (EGFP) under the control of the Pou5f1 promoter and distal enhancer elements. Whereas POU5f1 mediated EGFP expression was not observed in PDGFRA + cells, PDGFRA − derivatives showed robust EGFP expression. This suggests that undifferentiated mGSCs are contained within the PDGFRA − population ( Figure 1B-1E). Further analysis of Pou5f1 gene expression corroborated this finding, as Pou5f1 transcript levels were significantly lower (P < 0.05) in PDGFRA + cells compared to PDGFRAcells ( Figure 1F).
Suspecting that an undifferentiated mGSC population was contained within the PDGFRA − population, we subcutaneously transplanted sorted PDGFRA + and PDGFRAcells into mice. Within 4 months, Ki67 + teratomas were observed in all mice transplanted with PDGFRA − cells (Figure 2A-2F). This suggests that pluripotent characteristics retained by PDGFRAcells derived from mGSCs are not immediately amendable for use in cardiac cell therapy. In contrast, mice transplanted with PDGFRA + cells did not form teratomas, even as far out as 8 months post-implantation ( Figure 2A).
We next investigated whether PDGFRA + cells were capable of efficient cardiomyogenesis. In vitro cultures of Oncotarget 29645 www.impactjournals.com/oncotarget  Oncotarget 29647 www.impactjournals.com/oncotarget mGSCs along with control cultures of mESCs and iPSCs, were maintained in N2/B27 differentiating medium for 5 days, following which PDGFRA + and PDGFRA − populations were sorted, collected, and re-plated in the presence of N2/B27 differentiating medium for an additional 5 days. The collected cells were analyzed for the expression of cTnT, a transcription factor associated with cardiomyocytes differentiation. Unlike PDGFRAcell populations, PDGFRA + cells derived from mGSCs, mESC, or iPSCs were associated with a significant upregulation of cTnT gene expression ( Figure 3). This suggests that compared to PDGFRAcells, PDGFRA + cells have greater cardiac stem/progenitor-like potential, and more capable of differentiating into cardiomyocytes.

Effect of BMP4 and FBS on derivation PDGFRA + population
The BMP signaling pathways play a pivotal role in cardiogenesis [17,18], and our previous findings show that the culture of mGSCs with BMP4 (under serum-free conditions) promotes cardiomyocyte differentiation [14].

PdgfR-a signaling is critical for cardiac differentiation of PSCs
To further investigate the in vitro dependency of mGSCs on Pdgfra signaling for cardiac differentiation, we used shRNA to knockdown Pdgfra expression (shPdgfra). Parallel control experiments used mESCs and iPSCs. Stable shRNA-mediated Pdgfra knockdown was confirmed by qRT-PCR ( Figure 5). After 5 days in the EB-inducing culture medium, either empty vector control or shPdgfra cell populations were evaluated for the expression levels of cardiac lineage-specific markers (i.e. Mesp1, Isl1, Mef2c, Nkx2-5, and Tbx5; Figure 5). Results show that following Pdgfra knockdown, there is a substantial reduction of the cardiac lineage-specific gene expression levels in EBs derived from mESC, iPSC, and mGSC cell types. These findings suggest that activated Pdgfra signaling is necessary for the expression of cardiacrelated gene expression during cardiac differentiation.

Effect of growth factors in PdgfR-a expression during cardiac differentiation of PSCs
We next sought to determine whether the proportion of Pdgfra-expressing cardiac stem/progenitor cells derived from mGSCs could be influenced by additional growth factors. mGSCs, along with pluripotent control cultures of ESCs and iPSCs, were maintained in N2/ B27 medium and were subsequently treated with various concentrations of factors known to promote cardiac differentiation. These included: gamma-secretase inhibitor (GSI), activin, BMP4, and Noggin [19][20][21]. Compared to untreated cells cultured in N2/B27 medium alone, the exposure of mGSCs, ESCs, or iPSCs, to BMP4 (6.25 ng/ mL) showed significant increases in the proportion of PDGFRA expressing cells (mGSC = 11.0 ± 2.7% vs. 37.9 ± 2.5%; ESC = 9.6 ± 0.5% vs. 43.2 ± 2.0%; iPSCs = 8.2 ± 0.2% vs. 39.2 ± 2.5%; P < 0.05). Higher concentrations of BMP4 resulted in further increases to the proportion of PDGFRA expressing cells ( Figure 6). In contrast, no significant increases to the number of PDGFRA + cells was observed among mGSCs cells treated with GSI, activin, or Noggin ( Figure 6). Similar results observed with control mESCs and iPSCs further substantiated these finding, and suggested that the expansion of a PDGFRA expressing population is strongly influenced by BMP4 signaling.
To evaluate the cardiogenic potential of mGSCderived EBs, we expanded the number of PDGFRA + cells by exposing cells to BMP4 (50 ng/mL) ( Figure 7A). The expanded PDGFRA + cell population was subsequently enriched by cell sorting, and the cells seeded on 96-well ultra-low attachment plates at a density of 5 × 10 3 cells/ well in N2/B27 medium containing 10 ng/mL VEGF and 30ng/mL bFGF. In order to establish monolayers of differentiating cardiomyocytes, the 2-day culture of re-aggregated EBs ( Figure 7B) were re-plated on 0.1% gelatin-coated tissue culture plates. After another 4 days in culture, the BMP4-induced PDGFRA + population were found to robustly express the cardiomyocytes marker, cTnT ( Figure 7C-7E). There was no statistically significant difference in the expression of cTnT across either PDGFRα + or FLK1 + populations derived from mGSCs (Supplementary Figure S4). Also, spontaneous beating was observed as day 5 post plating (S1 Video). These findings suggest that PDGFRA + and FLK1 + cell population may have similar potential to undergo cardiomyocyte differentiation.

Therapeutic regeneration for damaged heart by transplanting PDGFRA + cardiac progenitors
The therapeutic efficacy of the PDGFRA + cells generated from mGSC-derived EBs was evaluated using an in vivo rat MI model. Following a 5-day induction with BMP4 (50 ng/mL), the PDGFRA + population was enriched by cell sorting and subsequently injected into peri-infarct zones of the infarcted rat hearts. After 4 weeks, the fibrotic area of infarcted hearts was evaluated by Mason Trichrome staining ( Figure 8A-8D). Compared to the hearts of control rats administered with PBS, quantification of fibrotic regions in the hearts of rats that were transplanted with PDGFRA + cells showed significantly smaller areas of fibrosis, ( Figure 8E).
To further evaluate the ability for PDGFRA + cells to engraft within recipient heart tissue, donor cells were labeled with PKH26 dye for in vivo tracking. Under fluorescence microscopy, we observed numerous PKH26-labeled cells that co-expressed cTnT. These findings provide strong evidence that the PDGFRA + cells

DISCUSSION
To advance cardiac stem cell therapy, and provide effective repair of damaged heart tissue, the ideal candidate cell should share characteristics with cardiac stem/progenitor cells and be capable of functionally differentiating into all cardiac-specific lineages. Using the ligand BMP4, we demonstrate that cardiomyocytes can be efficiently generated by expanding a PDGFRA expressing cell population derived from mouse mGSCs. Moreover, we show that this PDGFRA expressing cell population has the potential to promote cardiac regeneration in vivo. Notably, we observed that PDGFRA + cells have a greater potential to differentiate into cardiomyocytes in vitro than their PDGFRA − cells. Rather, PDGFRA − cells derived from mGSCs readily formed teratomas in mice, indicating that this cell population harbors unwanted pluripotency that would have dire implications if introduced into the clinic [22,23].
Several reports suggest that the expression of Pdgfra is associated with the cardiomyocyte lineage, as its expression is observed not only in the developing mesoderm at embryonic d7.5, but is also expressed by pluripotent stem cells undergoing cardiac differentiation in vitro [6,7,24]. We found that Pdgfra is necessary for cardiac-specific differentiation, as abrogation of Pdgfra expression using shRNA dramatically reduced the expression of genes associated with cardiac-specific differentiation. Particularly, shPdgfra knockdown resulted in a significant reduction in Isl1 expression in mGSCs cultures. As a LIM homeodomain transcription factor, Isl1 expression is required within a subset of undifferentiated cardiac progenitors for the development of cardiomyocytes [25]. These findings provide further support for Pdgfra signaling as having a critical role for in cardiac cell differentiation.
Flk1 is an early mesodermal surface marker that plays a central role not only in hematopoietic and endothelial differentiation but also in cardiomyocyte generation [26,27]. In the current study, we observed Oncotarget 29652 www.impactjournals.com/oncotarget that cardiac progenitors derived from either FLK1 + or PDGFRA + cell populations were each capable of undergoing cardiomyocyte differentiation in vitro; however, the expansion of PDGFRA + and FLK1 + cell populations each required different culture medium and appeared to respond differently to serum. Particularly, under a serum-free differentiation system, there was very little expansion of FLK1 + cells, but a marked increase in the PDGFRA expressing population. Conversely, in serum-containing MEMα culture medium, there was an increase in the FLK1 + population but not in the PDGFRA + cell population. This discrepancy may be explained by the inherent complexity of a serum, which limits our ability to completely understand the key factors and mechanisms that regulate the differentiation processes. Subsequent experiments aim to further elucidate the difference between PDGFRA + and FLK1 + cells, and characterize their respective potential to undergo cardiomyocyte differentiation, both in vitro and in vivo.
Particularly, inactivation of Notch signaling by treatment with GSI, promotes ESC differentiation into cardiac mesodermal cells [19]. GSI effectively blocks signaling pathways by preventing the cleavage of the intracellular fragment of the Notch receptor. In our study, we observed that use of GSI in the culture system had a limited impact on promoting the expansion of the PDGFRA + cell population. Similarly, the sequential treatment of hESCs with activin A and BMP4 for 5 days is associated with improving the overall efficiency of cells undergoing cardiac differentiation [36]. That said, the generation of cardiac progenitor cells from hESC and hiPSC cultures can occur in the absence of activin. In this study we observed that activin had little to no impact on promoting increased expansion of the PDGFRA expressing population.
Here we report on the derivation of cardiac progenitor-like cells from mGSCs by specifically enriching for a PDGFRA expressing cell population. Importantly, we show that in a rat MI model, transplanted PDGFRA + cells derived from mGSCs were capable of differentiate into cardiomyocytes and significantly reduced areas of fibrotic and damaged heart tissue. Our findings open the door for Oncotarget 29653 www.impactjournals.com/oncotarget cardiac progenitors derived from mGSCs, and potentially circumvent ethical quandaries surrounding ESCs, as well as the challenges associated with non-autologous stem cell transplantation. Additionally, there are many genderspecific differences in cardiac diseases that have yet to be clarified [37]. Subsequent use of mGSC-based cell therapy for cardiac regeneration may also provide an alternative approach to addressing gender-specific cardiac pathology.

MATERIALS AND METHODS
All procedures were performed according to guidelines for the ethical treatment of animals and approved by Institutional Animal Care and Use Committee in Chung-Ang University, Seoul, Korea.

PdgfR-a knockdown
PSCs were seeded at a density of 5 ×10 4 cells/well in 24-well plates and were incubated for 24 hours. The cells were then transfected with a single short hairpin RNA [shRNA; Empty Vector Control (TRC pLKO; GE Healthcare Dharmacon, Lafayette, CO) or ShPdgfra (TRCN0000001423; E Healthcare Dharmacon, Lafayette, CO)] vector construct targeting Pdgfra expression. According to the manufacturer's instructions, transfected cells were cultured in puromycin (0.2 µg/mL; Sigma, St. Louis, MO) for approximately two weeks in order to allow stable selection.

Real-time quantitative reverse transcriptasepolymerase chain reaction (qRT-PCR)
Total RNA was isolated and prepared from cells using the PureLink RNA Mini Kit (Invitrogen). RNA was reverse transcribed using the Superscript III Reverse Transcriptase (Invitrogen) according to the manufacturer's instructions. qRT-PCR was performed using a 7500 Real-Time PCR System (Applied Biosystems, Carlsbad, CA) and the synthesized cDNA was amplified using TaqMan Gene Expression Master Mix (Applied Biosystems). All gene expression levels were normalized to levels of GAPDH. All TaqMan primers and probes used were commercially obtained from Applied Biosystems.

Flow cytometry and cell sorting
Dissociated cells were stained with the following antibodies purchased from eBioscience: conjugated anti-mouse PDGFRA (CD140a), phycoerythrin (PE), conjugated anti-mouse FLK1-Allophycocyanin (APC), rat igG2a isotype control-APC, and rat igG2a isotype control-PE. The dissociated cells were suspended in phosphatebuffered saline (PBS) supplemented with 1% FBS, 10 mM [4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid] HEPES, 1 mM pyruvate, antibiotics (50 U/mL penicillin and 50 µg/mL streptomycin), 1 mg/mL glucose (PBS-S). The cells were then incubated with the appropriate antibodies for 20 minutes on an ice bath and washed twice with excess PBS-S for FACS analysis. After the final wash, the cells were resuspended (1 ×10 6 cells/mL) in PBS-S containing 1 µg/mL propidium iodide (PI; Sigma) and kept in dark on an ice bath until further analysis. Flow cytometric analyses and cell sorting were performed using the Dual-Laser FACS Aria II (BD Biosciences, Center for Research Facilities, Chung-Ang University). The sorted cells were centrifuged and plated in 0.1% gelatin-coated coverslip or V-shaped ultra-low attachment 96-well plates (Corning) at a density of 5 × 10 3 cells/well in N2/ B27 supplemented medium (Invitrogen) containing 10