Hypoxia Induces Early Neurogenesis in Human Fetal Neural Stem Cells by Activating the WNT Pathway

Fetal neural stem cells (FNSCs) present in the human fetal brain differentiate into cells of neuronal and glial lineages. The developing fetus is exposed to lower oxygen concentrations than adults, and this physiological hypoxia may influence the growth and differentiation of the FNSCs. This study aimed to evaluate the effect of hypoxia on the differentiation potential of human FNSCs isolated from the subventricular zone of aborted fetal brains (n = 5). FNSCs were isolated, expanded, and characterized by Nestin and Sox2 expression using immunocytochemistry and flow cytometry, respectively. These FNSCs were exposed to 20% oxygen (normoxia) and 0.2% oxygen (hypoxia) concentrations for 48 h, and hypoxia exposure (n = 5) was validated. Whole transcriptome analyses (Genespring GX13) of FNSCs exposed to hypoxia (Agilent 4 × 44 K human array slides) highlighted that genes associated with neurogenesis were enriched upon exposure to hypoxia. The pathway analysis of these enriched genes (using Metacore) showed the involvement of the WNT signaling pathway. Microarray analyses were validated using neuronal and glial lineage commitment markers, namely, NEUROG1, NEUROG2, ASCL1, DCX, GFAP, OLIG2, and NKX2.2, using qPCR (n = 9). DCX, ASCL1, NGN1, and GFAP protein expression was analyzed by Western blotting (n = 3). This demonstrated upregulation of the neuronal commitment markers upon hypoxia exposure, while no change was observed in astrocytic and oligodendrocyte lineage commitment markers. Increased expression of downstream targets of the WNT signaling pathway, TCF4 and ID2, by qPCR (n = 9) and increased protein expression of CTNNB1 (β-catenin) and ID2 by Western blot (n = 3) indicated its involvement in mediating neuronal differentiation upon exposure to hypoxia.


Introduction
Neural stem cells (NSCs) are multipotent cells that can differentiate into neurons, astrocytes, and oligodendrocytes. Fetal neural stem cells (FNSCs) are neural crest-derived cells located in the subventricular zone and dentate gyrus of the fetal brain. NSCs are also found in the cortex, striatum, and subependymal zone of the adult brain [1][2][3]. In addition to cues provided by the NSC niche, oxygen concentration can also influence the growth and differentiation potential of NSCs and plays a critical role during embryonic development [4,5]. As the fetus develops inside the uterus, the difference in oxygen concentration between maternal and fetal circulation shows that the developing fetus is normally exposed to lower oxygen concentrations, and thus, despite fetal hemoglobin having a greater affinity for oxygen and the existence of other compensatory mechanisms, there is a possibility that the fetal brain may be exposed to a hypoxic environment in utero [6][7][8]. The partial pressure of oxygen in arterial blood in adults is 85-95 mmHg, while cerebral tissue oxygen concentrations in the adult correspond to 30-50 mmHg [7]. Partial pressure of oxygen in fetal blood is still lower , and in the fetal brain it can be less than 7.6 mmHg [8]. Studies have shown that hypoxia may influence NSC development and plasticity [5,9]. It has been reported that mild hypoxia (5% O 2 ) activates molecular pathways such as Wnt/beta-catenin and Notch, which regulate the self-renewal and proliferation of stem cells, including NSCs [5,10,11].
This study aimed to understand the role of hypoxia on the differentiation potential of human FNSCs. It also elucidates the possible mechanism by which hypoxia may influence lineage commitment in human FNSCs.

Sample Collection
Aborted fetal samples were collected from the Department of Obstetrics and Gynecology, AIIMS, New Delhi, India. Informed consent was obtained from mothers undergoing Medical Termination of Pregnancy (MTP) in their second trimester of pregnancy (12-20 weeks) for maternal indications. Mothers undergoing MTP for fetal indications (such as chromosomal anomalies) were excluded from the study. Approval was taken from the Institutional Ethics Committee and Institutional Committee for Stem Cell Research before starting the study. The study was carried out in accordance with the Helsinki Declaration.

Isolation of Human Fetal Neural Stem Cells (FNSCs)
Isolation of human FNSCs from the brains of aborted fetuses was performed as per a published protocol [12]. Briefly, tissue from the subventricular zone of the brain was isolated and plated onto poly-D-lysine-coated culture flasks in neural stem cell media containing neurobasal media (GIBCO, NY, USA) with 1% N2 supplement (GIBCO, NY, USA), 2% neural survival factor-1 (Lonza, IA, USA), 1% glutamax (GIBCO, NY, USA), 5 mg/mL bovine serum albumin (Sigma, MO, USA), penicillin (50 IU/ml), streptomycin (50 µg/ml), and gentamicin (2 µg/ml). Tissue demonstrating cells radiating from the core was gently dissociated and subcultured onto poly-Dlysine-coated flasks to generate monolayers of FNSCs. For the neurosphere assay, human FNSCs at passages 2-3 were plated onto non-adherent culture dishes at a density of 1000 cells/ cm 2 in neural stem cell media (prepared as described above).

Flow Cytometry
Human FNSCs were fixed with 2% paraformaldehyde and permeabilized with 1% BSA containing 0.1% Triton X-100. Cells were blocked with 2% BSA for half an hour and subsequently stained with (intracytoplasmic) mouse anti-human SOX2 antibody conjugated with V450 (BD Biosciences, cat. no. 561610) using appropriate controls. Cells were washed and resuspended in 2% paraformaldehyde, and data were acquired using BD LSR Fortessa (BD Biosciences, San Jose, CA, USA) and analyzed using FlowJo v10 software.

Immunocytochemistry
Human FNSCs (at passages 3-4) were plated onto coverslips coated with poly-D-lysine. They were washed with PBS and fixed with 2% PFA. Cells were incubated for 1 h in blocking solution (1% BSA with 0.1% Triton X-100) and then washed with PBS. The cells were incubated overnight at 4 °C with primary antibody (mouse anti-Nestin 1:1000, Cat no. 33475, CST, MA, USA; rabbit anti-SOX2 1:1000, Cat no. 23064, CST, MA, USA). The cells were washed three times with PBS and then incubated with secondary antibody (mouse anti-rabbit FITC, 1:1000: Cat no. A11008, Invitrogen; goat anti-mouse Alexa Fluor 594, Cat no. A-11005, Invitrogen) for 1 h at room temperature. The cells were then washed three times with PBS and mounted onto glass slides using Vectashield mountant containing DAPI. The slide was allowed to dry overnight. Images were taken on a Nikon Eclipse Ti-S fluorescence microscope (Tokyo, Japan) and analyzed with NIS-Elements BR software.

Exposure of Human FNSCs to Different Oxygen Concentrations
Human FNSCs were exposed to oxygen concentrations mimicking normoxia (20% oxygen) and hypoxia (0.2% oxygen) for 48 h at 37 °C and 5% CO 2 , which was created using an Anoxomat hypoxia induction system (Advanced Instruments, Norwood, MA, USA). Hypoxia exposure was validated by evaluating the expression of the hypoxia-responsive genes CA9, VEGF, and PGK-1.

RNA Isolation, cDNA Synthesis, and qPCR
Total RNA was extracted from the cells after exposure to different oxygen concentrations using Tri-Reagent (Sigma, MO, USA) and quantified by a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, MA, USA). cDNA was synthesized with 1 μg total RNA using M-MuLV-RT (Thermo Fisher Scientific, MA, USA) and random hexamer primers (IDT, IL, USA). The expression of various genes was evaluated in the cells (in triplicate) using gene-specific primers (IDT, IL, USA) ( Table 1) and a DyNAmo Flash SYBR Green qPCR kit (Thermo Fisher Scientific, MA, USA) using a CFX96 Touch™ Real-Time PCR Detection System (Bio-Rad, CA, USA). 18S rRNA was used as an internal reference gene for normalization. The relative fold change in gene expression was calculated using the 2 −ΔΔCT method. Human FNSCs exposed to normoxia (20% oxygen) were used as controls for hypoxia experiments.

Western Blots
Fetal neural stem cells exposed to different concentrations of oxygen were lysed in RIPA lysis buffer with proteasephosphatase inhibitors, and the protein concentration in lysates was determined using the bicinchoninic acid assay (BCA). Fifty micrograms of lysate was loaded into each well for SDS-PAGE (4% stacking gel, 12-15% resolving gel). Normoxic (20%) and hypoxic (0.2%) samples were run on the same gel. Proteins were blotted onto a 0.22-µm nitrocellulose membrane, followed by blocking in 5% non-fat milk (NFM) in TTBS. Blots were incubated overnight at  CGT TGA ACC CCA TT  CCA TCC AAT CGG TAG TAG CG  151  DCX  GGG GGT GTG GGC ATA AAG AA  CCT GCT CTT TAC CAG CCT CC  149  ASCL1  TCC CCC AAC TAC TCC AAC GA  GTT GTG CGA TCA CCC TGC TT  193  Ngn1  TCT TGG TCT GTT TCT CCG GC  GGG TCA GTT CTG AGC CAG TC  120  Ngn2  TGA CTG ACA GAC AGA CAC GC  TGA CGA ACA TCT TAG TTG GCTC  257  Olig2  TCG CAT CCA GAT TTT CGG GT  AAA AGG TCA TCG GGC TCT (Table 2) diluted in 1% NFM, followed by three washes with TTBS. Blots were then incubated for 1 h with the corresponding secondary antibody diluted in 1% NFM. This was followed by five washes in TTBS. The blot was incubated with Luminol and peroxidase (Abbkine SuperLumia ECL Plus Kit, Hubei, China), and chemiluminescence detection was performed using an Azure Biosystems C280 gel documentation system (Dublin, CA, USA), followed by analysis with ImageJ software. Normalization was performed using β-actin protein levels.

Gene Expression Microarrays
RNA concentration and integrity were analyzed using Nanodrop (Thermo Fisher Scientific, MA, USA) and Bioanalyzer (Agilent, Santa Clara, CA, USA), respectively. Expression microarrays were performed on two biological replicates each of human FNSCs exposed to normoxia and hypoxia on 4 × 44 K human expression array slides (G2519F) (Agilent, Santa Clara, CA, USA) following the manufacturer's protocol. Briefly, a total of 200 ng of total RNA per sample (n = 2) was subjected to cDNA conversion and linear amplification with fluorescent labeling to prepare Cy3-labeled cRNA. Complementary hybridization (17 h at 65 °C) was followed by washing. Slides were scanned at 3-µm resolution, followed by feature extraction using feature extraction software version 10.7.1.1 (Agilent Technologies, Santa Clara, USA). Analysis was performed using Genespring Software v14.9.1 (Agilent Technologies, Santa Clara, USA). Principal component analysis and hierarchical clustering analysis were performed to identify the gene target distribution, which showed a consistent difference in expression between normoxia-and hypoxia-exposed cells. This was followed by differential gene expression analysis, which was performed using an unpaired t-test with Genespring software (GX v14.9.1). The list of differentially expressed genes was imported into the Gene Ontology consortium and Metacore for gene ontology and pathway analysis.

Statistical Analysis
Statistical analysis was performed using GraphPad Prism v6. Statistical differences between the normoxia-and hypoxiaexposed groups were estimated using the Mann-Whitney test. A p value < 0.05 was considered statistically significant.

Isolation and Characterization of Human Fetal Neural Stem Cells
Human fetal neural stem cells (FNSCs) were observed to radiate out from the core of tissue isolated from the subventricular zone of the brain (Fig. 1a). Upon dissociating and subculturing these cells, a small, unipolar monolayer of human FNSCs was obtained at passages 2-3 (Fig. 1b). The neurosphere assay displayed their ability to form neurospheres after 3-4 days of plating onto a nonadherent surface (Fig. 1c). Immunocytochemical staining at passages 3-4 helped characterize the human FNSCs and demonstrated the expression of NESTIN and SOX2 (Fig. 1d). Flow cytometry indicated that more than 90% of FNSCs expressed SOX2 (Fig. 1e  and 1f).

Exposure of FNSCs to Different Oxygen Concentrations for 48 Hours
Human FNSCs were exposed to different oxygen concentrations mimicking normoxia (20%) and hypoxia (0.2%) for 48 h. Hypoxia exposure was validated by evaluating CA9, PGK1, and VEGF expression by qPCR. The mean fold change ± SEM in CA-9, PGK1, and VEGF expression in the FNSCs exposed to hypoxia were 661.6 ± 211 (p < 0.0001), 27.61 ± 11.48 Gene ontology and pathway analysis of human FNSCs exposed to hypoxia (0.2% oxygen). a Gene Ontology analysis of human FNSCs exposed to hypoxia analyzed in the Gene Ontology consortium. b Path-way map of genes enriched in neurogenesis after exposure to hypoxia. c Network map of genes enriched in neurogenesis (p < 0.001), and 42.25 ± 0.44 (p < 0.01), respectively ( Fig. 2a, b, c). Exposure to hypoxia was also validated by measuring HIF1α protein expression by western blot (n = 3) (Fig. 2d). The normalized protein expression of HIF1α increased from 0.09 ± 0.04 in normoxic controls to 1.75 ± 0.62 (p = 0.10) in FNSCs exposed to 0.2% hypoxia (Fig. 2e).

Gene Ontology and Pathway Analysis
The analyses of differentially expressed genes (DEGs) in FNSCs exposed to hypoxia compared to normoxia showed that 974 genes were upregulated, while 368 genes were downregulated (Supplementary Tables 1 and  2). Gene ontology analysis of differentially expressed genes in FNSCs exposed to hypoxia (Supplementary Tables 3-6) showed that pathways pertaining to cell development and cell differentiation were enriched. The genes involved in these GO terms were further analyzed for biological processes and showed that regulation of neuron projection development, positive regulation of neurogenesis, neuron projection guidance, cell morphogenesis involved in neuron differentiation, regulation of neuron differentiation, neuron projection morphogenesis, neuron projection development, neuron development, and generation of neurons were enriched and had a role in cell differentiation of human FNSCs exposed to hypoxia (Fig. 3a). The genes involved in neuron development, neuron differentiation, and generation of neurons were then evaluated for enriched pathways (Table 3) and networks using Metacore software. This analysis showed that the Wnt-beta catenin canonical network was involved, with candidate genes such as Wnt and Frizzled being upregulated (Fig. 3b). It was also seen that a subnode of a VEGF pathway was linked with the Wnt-beta catenin canonical network (Fig. 3c).

Discussion
This study used fetal neural stem cells as a model system to partially mimic the physiological hypoxic conditions in utero and their possible influence on the development of the human fetal brain. In this study, human fetal neural stem cells (FNSCs) were isolated from the subventricular zone of aborted fetal brains. These multipotent stem cells, derived from the neuroectoderm, have the potential to differentiate into neurons, astrocytes, and oligodendrocytes. The isolated FNSCs displayed the ability to form neurospheres and expressed the characteristic neural stem cell markers Nestin and Sox2, as reported earlier [12,13]. Despite the existence of compensatory mechanisms, there is a possibility of physiological hypoxia in the developing fetal brain in utero [6,8]. To partially mimic physiological hypoxia in utero, human FNSCs were exposed to normoxia (20% oxygen) and hypoxia (0.2% oxygen) for 48 h. Hypoxia mediates its action through HIF1α; therefore, its elevated protein levels and increased gene expression of its downstream targets, carbonic anhydrase (CA9), phosphoglycerate kinase (PGK-1), and vascular endothelial growth factor (VEGF), indicate that the FNSCs were exposed to a hypoxic environment. This finding also agrees with previously published reports related to downstream targets of HIF1α [14].
This study investigated the expression of early markers of neurogenesis, such as NEUROG1, NEUROG2, ASCL1, and DCX, in human FNSCs exposed to hypoxia. These are also considered to be lineage commitment markers of neurogenesis. The expression of NEUROG1, ASCL1 and DCX was found to be increased in human FNSCs exposed to hypoxia at both transcriptional and translational levels. However, the expression of NEUROG2 was found to be slightly decreased. Interestingly, the increase in NEUROG1 expression was much greater than the decrease in NEUROG2 expression. This composite picture showing increased expression of NEUROG1, ASCL1, and DCX, supported by an increase in protein expression of DCX, signifies that hypoxia exposure in human FNSCs may be responsible for initiating neurogenesis and thus promoting FNSCs to commit to neuronal lineage. Our findings are supported by reports indicating the essential role of NEUROG1, ASCL1, and DCX in neurogenesis [15][16][17]. Interestingly, there are reports of different hypoxic conditions enhancing the expression of NEU-ROG1 and DCX [15,18,19]. However, there are also a few studies reporting the downregulation of neurogenesis and ASCL1 by hypoxia, which are contrary to our findings [4,20]. Again, studies also corroborate our findings by reporting that hypoxia or HIF-1α stimulates neurogenesis [21,22].
Our findings also indicate that the protein expression of GFAP, an astrocyte lineage commitment marker, was decreased, and the gene expression of OLIG2 and NKX2.2, oligodendrocyte lineage commitment markers, was not influenced by exposing FNSCs to hypoxia. This indicates that hypoxia does not induce FNSCs to commit to glial cell lineages. Our findings are further supported by the fact that even though there are studies reporting reactive astrocytosis and gliosis in hypoxic injury, there are no reports suggesting the promotion of neurogenesis in neural stem cells [23,24].
As FNSCs were exposed to hypoxia, whole genome transcriptomic changes and Gene Ontology analysis revealed that genes related to cell development and differentiation were being modulated. Upon further analysis of these genes for biological processes, it was found that genes involved in neurogenesis were upregulated when FNSCs were exposed to hypoxia. Considering that the human brain is more evolved and complex than the rodent brain, our study is the first to report the induction of neurogenesis in human FNSCs upon exposure to hypoxia. Few studies have reported that hypoxia stimulates neural stem cell proliferation but have not identified its influence on neurogenesis [24][25][26].
Pathway analysis of our data indicated that the Wnt-betacatenin signaling pathway may be implicated during this differentiation. This finding corroborates previous reports implicating this pathway in hypoxia-mediated proliferation of neural stem cells [5,26,27].
Pathway analysis of microarray data displaying genes enriched in cell development and neurogenesis in our study showed that the canonical Wnt-beta catenin signaling pathway was involved in promoting the commitment of FNSCs to the neuronal lineage after exposure to hypoxia. In this study, the expression of some critical regulators of the Wnt-beta catenin signaling pathway, viz., beta catenin, TCF4, and ID2, were elucidated. The protein expression of beta-catenin was found to be increased, even though its gene expression was slightly downregulated. The expression of downstream effector targets of the Wnt signaling pathway, TCF4 and ID2, was Fig. 4 Expression of lineage commitment markers in FNSCs exposed to hypoxia. qPCR analysis of gene expression of neural cell fate markers a Neurogenin 1 (NEUROG1), b Neurogenin 2 (NEUROG2), c Doublecortin (DCX), and d Achaete-scute homolog 1 (ASCL1); astrocytic cell fate marker e glial fibrillary acidic protein (GFAP); oligodendrocyte lineage markers f Oligodendrocyte transcription factor 2 (Olig2) and g NK2 homeobox 2 (Nkx2.2) in FNSCs exposed to hypoxia as compared to controls (n = 6-9). 18S rRNA was used as reference gene. h Representative Western blots and densitometric quantification (n = 2-3) of i doublecortin (DCX) and j glial fibrillary acidic protein (GFAP) k achaete-scute homolog 1 (ASCL1) and (l) Neurogenin 1 (Ngn1) protein expression in FNSCs exposed to hypoxia as compared to controls. Data are represented as mean ± SEM. **p < 0.01 ◂ also increased in FNSCs exposed to hypoxia, indicating its involvement in mediating the effects of hypoxia. Interestingly, a few studies have reported that Wnt signaling is responsible for stem cell maintenance as well as neuronal differentiation, lineage commitment, axon guidance, and neurite outgrowth [28,29]. It has also been reported that Wnt signaling facilitates neurological recovery in experimental stroke, thereby establishing its role in neurogenesis [28,30].
Beta catenin, one of the regulators of the Wnt signaling pathway, translocates to the nucleus and interacts with TCF to activate the transcriptionally active complex. Interestingly, target genes of this complex are NEUROG1 and NEUROG2 [31]. NEUROG1/2 then binds to p300/CBP coactivator proteins to promote neuronal differentiation [32]. The expression of TCF4 was found to be increased in this study, which might also explain the increase seen in the composite picture of increased expression of lineage commitment markers of neurogenesis, such as NEUROG1, DCX, and ASCL1. This is the likely mechanism involved in influencing FNSCs to differentiate into neurons after they are exposed to hypoxia. Recent studies have also confirmed the involvement of the Wnt signaling pathway in mediating neurological recovery Fig. 5 Expression of critical regulators of the Wnt signaling pathway in FNSCs exposed to hypoxia. qPCR analysis of the gene expression of Wnt signaling pathway members a β-catenin (CTNNB1), b Transcription factor 4 (TCF4), and c inhibitor of DNA binding 2 (ID2) in FNSCs exposed to hypoxia compared to controls. 18S rRNA was used as a reference gene. d Representative Western blots and densitometric quantification of e β-catenin and f ID2 protein expression in FNSCs exposed to hypoxia compared to controls. Data are represented as the mean ± SEM. *p < 0.05, **p < 0.01 in stroke and epilepsy, lending credence to our findings of hypoxia promoting neurogenesis in FNSCs [10,28,30].
To the best of our knowledge, this is the first study to show that there was an increase in lineage commitment markers of neurogenesis, viz., NEUROG1, DCX and ASCL1, on exposing FNSCs to hypoxic conditions, while observing no change in the astrocytic and oligodendrocytic lineage markers. The mechanism attributed to the increased neurogenesis may be due to an increase in downstream effectors of the Wnt signaling pathway, viz., TCF4 and ID2 indicate the involvement of the Wnt signaling pathway in mediating this action.