Long-Term Hypoxia Maintains a State of Dedifferentiation and Enhanced Stemness in Fetal Cardiovascular Progenitor Cells

Early-stage mammalian embryos survive within a low oxygen tension environment and develop into fully functional, healthy organisms despite this hypoxic stress. This suggests that hypoxia plays a regulative role in fetal development that influences cell mobilization, differentiation, proliferation, and survival. The long-term hypoxic environment is sustained throughout gestation. Elucidation of the mechanisms by which cardiovascular stem cells survive and thrive under hypoxic conditions would benefit cell-based therapies where stem cell survival is limited in the hypoxic environment of the infarcted heart. The current study addressed the impact of long-term hypoxia on fetal Islet-1+ cardiovascular progenitor cell clones, which were isolated from sheep housed at high altitude. The cells were then cultured in vitro in 1% oxygen and compared with control Islet-1+ cardiovascular progenitor cells maintained at 21% oxygen. RT-PCR, western blotting, flow cytometry, and migration assays evaluated adaptation to long term hypoxia in terms of survival, proliferation, and signaling. Non-canonical Wnt, Notch, AKT, HIF-2α and Yap1 transcripts were induced by hypoxia. The hypoxic niche environment regulates these signaling pathways to sustain the dedifferentiation and survival of fetal cardiovascular progenitor cells.


Introduction
Cardiovascular progenitor cells (CPCs) have been evaluated for numerous applications to treat a variety of cardiovascular diseases from congenital heart defects to myocardial infarctions [1,2]. This population of cells is capable of differentiating into the three cardiac lineages: cardiomyocytes, endothelial, and vascular smooth muscle cells. In addition to their multipotent capacity, these cells can be harvested on a patient-specific basis and rapidly expanded to obtain adequate cell numbers for in vitro experimentation and therapeutic applications [3,4]. To complement these intrinsic qualities in therapeutic applications, it is essential that we understand how to program these cells during and throughout both culture and clinical application. The field of stem cell programming continues to develop to meet these needs.
The microenvironment of a cell includes not only chemical differentiation signals but also physical differentiation factors such as sheer force, osmolarity, pressure, stretch, and oxygen tension. These physical factors remain a popular topic for experimentation, as each factor plays an individual role in controlling differentiation. An optimal method of controlled differentiation is yet to be established; however, recent findings suggest that oxygen tension regulates differentiation and migratory capacity during fetal development [5].

Cardiovascular Progenitor Cell Characterization
Fetal CPCs from both normoxic and hypoxic sheep were harvested for this study. Monoclonal cell populations were expanded and grouped based on their expression of Islet-1, c-KIT, CD105, and SSEA-4 via flow cytometry ( Figure 1). Islet-1 is a known marker of early lineage, multipotent CPCs. The expression of SSEA-4 is representative of the invasiveness of the cells in this population. The cardiovascular progenitor clones chosen for this study were selected based on the triple positive expression of Islet-1, c-kit, and SSEA-4, as stemness and invasiveness enhance therapeutic potential.

Wnt, Notch, and FAK Function Synergistically to Promote Cell Survival
Pathways that have been previously connected to hypoxic regulation in other models were evaluated in Islet-1+ early CPC clones following long term hypoxia exposure. These pathways include NRG/ERBB, HIF-1α, non-canonical Wnt, Notch, and FAK. Using RT-PCR, we found that NRG and ERBB expression remained unchanged, suggesting that this pathway does not play a role in long-term hypoxic regulation (Figure 2A). Neither HIF-1α, which has been demonstrated to increase in short-term hypoxia [3,11], or its downstream regulator PDK3 were affected by long-term hypoxia ( Figure 2B). The non-canonical Wnt pathway has an active role in promoting cell survival and preventing apoptosis. In our hypoxia model, non-canonical Wnt5a expression increased 98-fold, while concurrently, the expression of Wnt11, a member of the canonical Wnt pathway, was inhibited 0.36-fold ( Figure 2C). Canonical Wnt signaling downregulation attenuates cell fate specification. Increased Notch signaling fosters cellular growth and survival [20,21]. Here, despite low oxygen tensions, Notch signaling was induced, suggesting enhanced survival in cardiovascular progenitor cells cultured in hypoxic conditions ( Figure 2D). The expression of FAK, MAPK1, and PKC transcripts was upregulated following hypoxic conditioning ( Figure 2E). Interestingly, these genes, as well as NOTCH1, promote AKT signaling activation. The upregulation of the PI3K/AKT pathway could also explain the lack of HIF-1α induction, as AKT is known to counter these effects in cardiomyoblast populations [29].

Wnt, Notch, and FAK Function Synergistically to Promote Cell Survival
Pathways that have been previously connected to hypoxic regulation in other models were evaluated in Islet-1+ early CPC clones following long term hypoxia exposure. These pathways include NRG/ERBB, HIF-1α, non-canonical Wnt, Notch, and FAK. Using RT-PCR, we found that NRG and ERBB expression remained unchanged, suggesting that this pathway does not play a role in long-term hypoxic regulation (Figure 2A). Neither HIF-1α, which has been demonstrated to increase in short-term hypoxia [3,11], or its downstream regulator PDK3 were affected by long-term hypoxia ( Figure 2B). The non-canonical Wnt pathway has an active role in promoting cell survival and preventing apoptosis. In our hypoxia model, non-canonical Wnt5a expression increased 98-fold, while concurrently, the expression of Wnt11, a member of the canonical Wnt pathway, was inhibited 0.36-fold ( Figure 2C). Canonical Wnt signaling downregulation attenuates cell fate specification. Increased Notch signaling fosters cellular growth and survival [20,21]. Here, despite low oxygen tensions, Notch signaling was induced, suggesting enhanced survival in cardiovascular progenitor cells cultured in hypoxic conditions ( Figure 2D). The expression of FAK, MAPK1, and PKC transcripts was upregulated following hypoxic conditioning ( Figure 2E). Interestingly, these genes, as well as NOTCH1, promote AKT signaling activation. The upregulation of the PI3K/AKT pathway could also explain the lack of HIF-1α induction, as AKT is known to counter these effects in cardiomyoblast populations [29].

FAK and Notch Directly Upregulate PI3K/AKT in the Hypoxic Niche
AKT signaling activation is associated with enhanced proliferation and survival. Notch regulation of AKT signaling stems from the induction of Hes1, which directly activates AKT and its downstream effectors [21,23]. Similarly, the increased expression of FAK, MAPK, and PKC observed in the hypoxic niche promote AKT-dependent survival and resistance to oxidative stress [19]. To further elucidate the mechanism by which the AKT signaling pathway is induced, we confirmed the upregulation of several additional transcripts in the AKT signaling pathway via RT-PCR. For example, CCND1 and SOD2 were upregulated in CPCs cultured in hypoxia ( Figure 3A,B). These genes are associated with increased cell cycle activity and the protection against reactive oxygen species released during oxidative stress, respectively [18,19]. Next, we assessed AKT phosphorylation via Western blot ( Figure 3C,D). The hypoxic niche increased the ratio of phosphorylated AKT relative to its non-phosphorylated counterpart, confirming an activation of AKT signaling.

FAK and Notch Directly Upregulate PI3K/AKT in the Hypoxic Niche
AKT signaling activation is associated with enhanced proliferation and survival. Notch regulation of AKT signaling stems from the induction of Hes1, which directly activates AKT and its downstream effectors [21,23]. Similarly, the increased expression of FAK, MAPK, and PKC observed in the hypoxic niche promote AKT-dependent survival and resistance to oxidative stress [19]. To further elucidate the mechanism by which the AKT signaling pathway is induced, we confirmed the upregulation of several additional transcripts in the AKT signaling pathway via RT-PCR. For example, CCND1 and SOD2 were upregulated in CPCs cultured in hypoxia ( Figure 3A,B). These genes are associated with increased cell cycle activity and the protection against reactive oxygen species released during oxidative stress, respectively [18,19]. Next, we assessed AKT phosphorylation via Western blot ( Figure 3C,D). The hypoxic niche increased the ratio of phosphorylated AKT relative to its non-phosphorylated counterpart, confirming an activation of AKT signaling.

FAK and Notch Directly Upregulate PI3K/AKT in the Hypoxic Niche
AKT signaling activation is associated with enhanced proliferation and survival. Notch regulation of AKT signaling stems from the induction of Hes1, which directly activates AKT and its downstream effectors [21,23]. Similarly, the increased expression of FAK, MAPK, and PKC observed in the hypoxic niche promote AKT-dependent survival and resistance to oxidative stress [19]. To further elucidate the mechanism by which the AKT signaling pathway is induced, we confirmed the upregulation of several additional transcripts in the AKT signaling pathway via RT-PCR. For example, CCND1 and SOD2 were upregulated in CPCs cultured in hypoxia ( Figure 3A,B). These genes are associated with increased cell cycle activity and the protection against reactive oxygen species released during oxidative stress, respectively [18,19]. Next, we assessed AKT phosphorylation via Western blot ( Figure 3C,D). The hypoxic niche increased the ratio of phosphorylated AKT relative to its non-phosphorylated counterpart, confirming an activation of AKT signaling. The AKT pathway promotes survival through many downstream effectors including Yap1, a member of the Hippo pathway. The phosphorylation of Yap1 results in cytoplasmic retention and eventually apoptosis or degradation. The nuclear translocation of non-phosphorylated Yap1 allows downstream targets to fulfill their role of driving cell proliferation and enhancing cell survival [30]. In response to hypoxia, CPCs presented with an increase in Yap1 gene expression ( Figure 4A,B). While an increase of Yap1 expression was observed via RT-PCR, phosphorylated Yap1 was statistically unchanged, according to Western blotting, contrary to what we hypothesized ( Figure 4C,D). In long-term hypoxia, the levels of phosphorylated/inactive Yap1 are not significantly elevated, suggesting that active Yap1 continues to enter the nucleus and promote cell survival. eration and enhancing cell survival [30]. In response to hypoxia, CPCs presented wit increase in Yap1 gene expression ( Figure 4A,B). While an increase of Yap1 expression observed via RT-PCR, phosphorylated Yap1 was statistically unchanged, accordin Western blotting, contrary to what we hypothesized ( Figure 4C,D). In long-term hypo the levels of phosphorylated/inactive Yap1 are not significantly elevated, suggesting active Yap1 continues to enter the nucleus and promote cell survival.

Islet-1+ Cardiovascular Progenitor Cells Remain Dedifferentiated under Long-Term Hyp Conditions
HIF-2α is a known regulator of OCT4, SOX2, and Nanog expression in embry stem cells subjected to hypoxia through the nuclear translocation and activation of poxic response elements [6,31,32]. The long-term cell culture of CPCs corroborated data, as verified by PCR analysis and predicted by Ingenuity Pathway Analysis ( ( Figure 5A,B and Supplemental Figure S1). HIF-2α, OCT4, SOX2, and Nanog were up ulated in hypoxic Islet-1+ cell clones and indicate that long-term hypoxia induces the differentiation of fetal CPCs. Nanog transcript levels were found to decline following sequent exposure to normoxic conditions (Supplemental Figure S2). Nestin is a kn stemness marker in progenitor cells [33] and was elevated under hypoxic conditions. ilarly, CXCR4 was upregulated in hypoxic conditioning, suggesting increased stem [34]. Long-term hypoxia induced c-Kit expression and maintained Islet-1 expression

Islet-1+ Cardiovascular Progenitor Cells Remain Dedifferentiated under Long-Term Hypoxic Conditions
HIF-2α is a known regulator of OCT4, SOX2, and Nanog expression in embryonic stem cells subjected to hypoxia through the nuclear translocation and activation of hypoxic response elements [6,31,32]. The long-term cell culture of CPCs corroborated this data, as verified by PCR analysis and predicted by Ingenuity Pathway Analysis (IPA) ( Figure 5A,B and Supplemental Figure S1). HIF-2α, OCT4, SOX2, and Nanog were upregulated in hypoxic Islet-1+ cell clones and indicate that long-term hypoxia induces the dedifferentiation of fetal CPCs. Nanog transcript levels were found to decline following subsequent exposure to normoxic conditions (Supplemental Figure S2). Nestin is a known stemness marker in progenitor cells [33] and was elevated under hypoxic conditions. Similarly, CXCR4 was upregulated in hypoxic conditioning, suggesting increased stemness [34]. Long-term hypoxia induced c-Kit expression and maintained Islet-1 expression, as shown by flow cytometry (Figure 5C-E). SSEA-4, typically known as a marker of invasiveness [35], was not induced, suggesting that fetal cells in a hypoxic niche stay harbored in said niche.

Cell Cycle Progression and Migration Are Unaffected by Long Term Hypoxia
To assess whether cell cycle progression is impacted by long-term hypoxic conditions, we conducted cell cycle analysis by flow cytometry. A representative tracing of both normoxic and hypoxic populations can be seen in Figure 6A,B. The data were quantified using a Dean-Jett-Fox model. As shown in Figure 6C-E, no significant difference was identified when comparing cell cycle progression in normoxic and hypoxic CPCs. Changes in oxygen tension similarly had no effect on invasion, suggesting that the upregulated signaling pathways did not affect the invasive capacity of fetal CPCs.

Cell Cycle Progression and Migration Are Unaffected by Long Term Hypoxia
To assess whether cell cycle progression is impacted by long-term hypoxic conditions, we conducted cell cycle analysis by flow cytometry. A representative tracing of both normoxic and hypoxic populations can be seen in Figure 6A,B. The data were quantified using a Dean-Jett-Fox model. As shown in Figure 6C-E, no significant difference was identified when comparing cell cycle progression in normoxic and hypoxic CPCs. Changes in oxygen tension similarly had no effect on invasion, suggesting that the upregulated signaling pathways did not affect the invasive capacity of fetal CPCs.

Discussion
Short-term hypoxic culture beneficially impacts CPCs such that survival and invasiveness are enhanced via AKT activation and SDF-1α sensitization [3]. Hypoxia inducible factors appear to be at the forefront of short-term hypoxic survival, as they interact with hypoxia response elements on the DNA and upregulate transcripts associated with survival [11,32]. We report here that the long-term maintenance of CPCs under hypoxic conditions promotes stemness and a state of dedifferentiation. These adaptations are the direct result of the upregulation of several interconnected signaling pathways. The longterm hypoxic response and the mechanisms associated with cell survival differ from the response to short-term hypoxia in early-stage CPCs. According to our research, there exists a common theme of AKT upregulation that connects both short-and long-term hypoxia, but it seems that differences arise when evaluating hypoxia inducible factors. Short term hypoxia results in the upregulation of HIF-1 and downstream survival molecules such as BCL2 and HMOX [3]. This current research suggests that long term hypoxic culture deals with survival through differing pathways including non-canonical Wnt, FAK,

Discussion
Short-term hypoxic culture beneficially impacts CPCs such that survival and invasiveness are enhanced via AKT activation and SDF-1α sensitization [3]. Hypoxia inducible factors appear to be at the forefront of short-term hypoxic survival, as they interact with hypoxia response elements on the DNA and upregulate transcripts associated with survival [11,32]. We report here that the long-term maintenance of CPCs under hypoxic conditions promotes stemness and a state of dedifferentiation. These adaptations are the direct result of the upregulation of several interconnected signaling pathways. The long-term hypoxic response and the mechanisms associated with cell survival differ from the response to short-term hypoxia in early-stage CPCs. According to our research, there exists a common theme of AKT upregulation that connects both short-and long-term hypoxia, but it seems that differences arise when evaluating hypoxia inducible factors. Short term hypoxia results in the upregulation of HIF-1 and downstream survival molecules such as BCL2 and HMOX [3]. This current research suggests that long term hypoxic culture deals with survival through differing pathways including non-canonical Wnt, FAK, Notch, and Yap. HIF-2α and its downstream dedifferentiation factors are unique to long-term hypoxia and may need additional time and stress to activate dedifferentiation. OCT4, SOX2, and Nanog transcripts are not elevated in short-term hypoxic CPCs [3,36].
The hypoxic fetal microenvironment, as well as long-term culture under hypoxic conditions in vitro, regulates survival and dedifferentiation pathways. Non-canonical Wnt, Notch, FAK, AKT, and Yap signaling contributes to cell survival despite incredibly low oxygen tensions. The pathways discussed here interact to assert an anti-apoptotic state, reduce oxidative stress, and promote viability despite the altered nutrient intake of the cell in the hypoxic niche. Often, these pathways are uncontrollably upregulated in the cancer microenvironment, leading to unfettered proliferation, anti-apoptosis, and survival [10,37]. In contrast, the long-term hypoxic developmental niche maintains a level of control despite the low oxygen tension stressor of the microenvironment.
HIF-2α remains upregulated, leading to the increased expression of hypoxic response elements and the transcription of genes responsible for SOX2, OCT4, and Nanog. The reexposure of hypoxic cells to a normoxic niche downregulated Nanog expression, confirming the level of control hypoxia has over dedifferentiation. These adaptations suggest that hypoxia promotes dedifferentiation in fetal ovine CPCs [6]. Dedifferentation could potentially broaden the multipotency of fetal CPCs, thereby widening their potential for therapeutic applications. The exact lineage staging must be further evaluated in order to fully understand the possibilities of long-term hypoxia in stem cell reprogramming.
We have elucidated five different pathways that contribute to sustaining a dedifferentiated survival state in fetal cardiovascular progenitor cells when subjected to a long-term hypoxic niche. In Figure 7, we propose and support a mechanism by which these pathways interconnect with a network identified using IPA . These interactions influence the endpoints of hypoxic regulation and are connected either directly or indirectly acting through intermediates such as Hes1, FAK, and NFk-B [50,61,69,73]. Notch and AKT both interact with Hes1 to activate genes that regulate antiapoptosis, oxidative resistance, and dedifferentiation [73][74][75]. Yap1 and FAK influence AKT, achieving the same goals [40]. NFk-B interacts with non-canonical Wnt, Notch and AKT [61,69,73] Understanding the role of individual signaling pathways that function together in maintaining a dedifferentiated survival state provides insight into cellular adaptation to hypoxia. This information is relevant both in organismal development and as a potential pretreatment for therapeutic application.
Cardiovascular progenitors are currently being evaluated for their potential use in novel therapeutic applications following physical differentiation protocols involving sheer force, transmural pressure, and oxygen tension [3,4]. These physical differentiation factors can potentially introduce unwanted differentiation in multipotent stem cells, hindering their efficacy for regenerative applications. Long-term hypoxia could potentially augment therapies at the pre-treatment stage and broaden the therapeutic potential of progenitor cells by maintaining multipotency. The new information provided in this study addresses the impact of the microenvironment on differentiation while contributing to our understanding of stem cell potential for therapeutic applications.
resistance, and dedifferentiation [73][74][75]. Yap1 and FAK influence AKT, achieving the same goals [40]. NFk-B interacts with non-canonical Wnt, Notch and AKT [61,69,73] Understanding the role of individual signaling pathways that function together in maintaining a dedifferentiated survival state provides insight into cellular adaptation to hypoxia. This information is relevant both in organismal development and as a potential pretreatment for therapeutic application.

Animal Selection
The high altitude long-term hypoxic sheep were housed at Barcroft Laboratory, White Mountain Research Station, Bishop, CA, USA (altitude 3820 m; PaO 2 60 ± 2 mmHg) beginning at 30 days gestation. The pregnant females were transported from altitude prior to delivery and subsequently sacrificed at sea level. Fetal atrial tissue was harvested for the study. For the control group, pregnant females were reared at sea level for the entirety of gestation and sacrificed near term, comparable to the long-term hypoxic group. The body and organ weights of the long-term hypoxic fetuses did not differ significantly from the normoxic controls. These studies were approved by the IRB and the Animal Care and Use Committee of Loma Linda University, Loma Linda, California under protocol #8110004 on 4 December 2013.

Cell Isolation
Islet-1+ cardiovascular progenitor cell clones were isolated from 5 fetal Suffolk sheep, as previously described by our laboratory [76]. Briefly, atrial cardiac tissue from either normoxic or hypoxic fetal sheep was broken down into 1 mm 3 sections which were then digested by collagenase (Roche Applied Science, Indianapolis, IN, USA) for 2 h at 37 • C. The cardiac tissue suspension was next filtered through a 40µm cell strainer to isolate CPCs. The resultant cardiovascular progenitor populations were clonally expanded after diluting cells to 0.8 cells per well in a 96-well plate. These monoclonal populations were evaluated for their expression of cardiovascular progenitor cell markers, including Islet-1, c-kit, and SSEA-4, to verify the desired CPC lineage. Cell clones expressing these three markers were used for all of the experiments once monoclonally expanded.

Hypoxic Cell Culture
The monoclonal populations that were isolated from hypoxic sheep were continuously cultured exclusively under hypoxic conditions when expanded in vitro. Hypoxic CPC clones were maintained in a separate 37 • C incubator under 5% CO 2 and 1% oxygen. The normoxic/control CPCs arising from normoxic sheep were only cultured under normoxic conditions with 5% CO 2 and 22% atmospheric oxygen. The culture media and cell expansion procedures remained the same between the normoxic and hypoxic cell clones, and all of the clones were analyzed at low passage.

Quantitative RT-PCR
The normoxic and hypoxic cell lines were trypsinized and stored in RNAProtect (Qiagen, Valencia, CA, USA) until RNA isolation could proceed with the RNeasy Mini Kit (Qiagen, Valencia, CA, USA). RNA quality and quantity were assessed using a Nanodrop 2000 spectrophotometer (Thermo Fischer Scientific, Rochester, NY, USA) and gel electrophoresis. cDNA was prepared using 2µg of RNA and Superscript III, following the manufacturer's instructions (Life Technologies, Carlsbad, CA, USA). Quantitative realtime polymerase chain reaction (qRT-PCR) was performed using Go-Taq qPCR Mastermix (Promega, Madison, WI, USA) and the iCycler iQ TM 5 PCR Thermal Cycler (Bio-Rad, Hercules, CA, USA), following a protocol of 94 • C for 10 min and 45 cycles of 94 • C for 15 s, 58 • C for 60 s, and 72 • C for 30 s. RT-PCR products were run on 2% agarose gels with a low mass ladder (Invitrogen, Carlsbad, CA, USA) to confirm the amplification of the appropriate gene. Primers were designed using the National Center for Biotechnology Information Primer-BLAST program and purchased from Integrated DNA Technologies (Coralville, IA, USA). Primer sequences can be found in Table S1.

Western Blot
The cells were detached via a cold trypsin protocol, as previously described by our laboratory [3], and stored in a solution of RIPA buffer, protease inhibitor cocktail, sodium fluoride, sodium orthovanadate, and 0.5M EDTA. This cell solution was agitated for 2 h at 4 • C and subsequently centrifuged at 14,000× g and aliquoted for quantification using the Micro BCA Protein Assay Kit (Thermo Fischer, Waltham, MA, USA). These aliquots were run on an automated, gel-free Western blot system (ProteinSimple Wes, San Jose, CA, USA) to quantify specific protein expression across our cell lines. The antibodies used can be found in Table S2.

Ingenuity Pathway Analysis
We input our genetic regulation data into Ingenuity Pathway Analysis (Qiagen, Valencia, CA, USA) to evaluate and assess potential and known pathway interactions between the targeted survival and dedifferentiation pathways. IPA assembled a network of direct and indirect pathway interactions that we used to validate our proposed mechanism of hypoxic regulation. The networks were generated through the use of IPA (QIAGEN Inc., https://www.qiagenbioinformatics.com/products/ingenuity-pathway-analysis).

Statistical Analysis
We used a two-tailed, paired t-test to compare the mean of all normally distributed data and a Wilcoxon matched-pairs signed rank test to compare the mean of all nonnormally distributed data. For cell cycle and migration, samples in each group were pooled and a two-tailed, unpaired t-test was used for normally distributed data, whereas a Mann-Whitney U test was used to compare the mean of all non-normally distributed data. Prism version 7 was used for all statistical analysis, and all data are represented as the mean ± the standard error of the mean. p-Values < 0.05 were considered to be statistically significant.

Conclusions
The generation of the mitotically stable cell lines of ovine fetal cardiovascular progenitors that are characterized by long-term hypoxia-induced enhancements in stemness has the potential to be of great importance for creating sheep models for human cell/tissue engineering and regenerative medicine. The latter are aimed at cardiovascular therapies and cardiosurgical treatments of a variety of cardioangiopathies [3,4]. In addition, the establishment of permanent cardiovascular stem cell clones that exhibit augmented stemness properties as a result of expressing the genotypic and phenotypic traits related to their tissue-specific genomic, epigenomic and proteomic profiles could also be valuable for the generation of nuclear donor cells for modern assisted reproductive technologies (ARTs) such as cloning sheep and other mammalian species by somatic cell nuclear transfer (SCNT) [77][78][79]. Future work in these areas is needed to address these possibilities.

Data Availability Statement:
The data presented in this study are available upon reasonable request.

Conflicts of Interest:
The authors declare no conflict of interest.