Microenvironment Regulation of Pluripotent Stem Cell-Derived Neural Progenitor Aggregates by Human Mesenchymal Stem Cell Secretome

Abstract

Neural progenitor cells (NPCs) derived from pluripotent stem cells (PSCs), including embryonic stem cells (ESCs) and induced PSCs, are promising cell source for the treatment of various neurological diseases. NPC derivation from PSCs is regulated by microenvironment factors that influence cell fate via paracrine and autocrine effects. In this study, ESC-derived NPC aggregates were replated in the secretomes of bone marrow-derived human mesenchymal stem cells (hMSCs) generated under hypoxia or normoxia to investigate the effects of hMSC secretome on NPC cellular behaviors. The results demonstrated that hMSC secretomes stimulated endogenous secretion of extracellular matrices from NPC aggregates and enhanced cell adhesion and proliferation. NPC functional differentiation measured by migration length, neurite extension, and the yield of neural and glial cells were also increased by threefold to fourfold. Inhibition of fibroblast growth factor-2, transforming growth factor-β1, and brain-derived neurotrophic factor signaling differentially reduced the adherent cell number, migration length, and neurite extension, suggesting the regulatory effects of a broad spectrum of hMSC-derived factors. In summary, ESC-derived NPC aggregates in hypoxic hMSC secretomes may represent a suitable combination to promote the engraftment and neurogenesis in vivo.

Introduction

Neural progenitor cells (NPCs) derived from pluripotent stem cells (PSCs), including embryonic stem cells (ESCs) and induced PSCs, are promising cell sources for neurological diseases in central nervous system (CNS).^1–3 PSC-derived NPC transplantations have demonstrated the improved functional recovery of stroke-damaged brain and the injured spinal cord,^4,5 possibly as repopulating cells to recreate functional neuronal circuits during maturation or by stimulating the endogenous neurogenesis through the secretion of neural trophic factors such as nerve growth factor, brain-derived neurotrophic factor (BDNF), glial cell-derived neurotrophic factor, and vascular endothelial growth factor (VEGF).^6,7 Compared to adult NPCs, PSC-derived NPCs provide an unlimited cell source and have the potential for scalable production for clinical applications.⁸ However, the yield of NPC derivation from PSCs is low and such cells are functionally immature, reducing their in vivo efficacy.^9,10 Additionally, in many neurological diseases such as stroke and spinal cord injury, PSC-derived NPCs are exposed to ischemia and oxidative stresses at the wound sites, which reduced cell survival, adhesion, and the integration with surrounding tissues.¹¹ It has been suggested that the lack of endogenous extracellular matrix (ECM) production may be a contributing factor that reduces in vivo cell viability and engraftment in the presence of elevated levels of matrix metalloproteinases (MMPs) and inflammatory cytokines.^12,13

Mesenchymal stem cells (MSCs) have been tested in preclinical and clinical studies for the treatment of neurodegenerative diseases.^14,15 However, transplanted MSCs have short life span in vivo and their capacity to undergo neuronal differentiation for long-term engraftment is being intensely debated.^16,17 Recent studies have shown that the primary mechanism of MSC's therapeutic effects is the secretion of bioactive molecules that induce anti-inflammatory, anti-apoptotic, neural trophic, and immune-modulating effects, which promote tissue healing by modulating the milieu of injury sites, suppressing immune responses,^15,16 and/or stimulating endogenous neurogenesis.^18–20 For examples, co-culturing MSCs with adult NPCs enhanced cell proliferation and promoted neurite extension by upregulating Notch signaling and the expressions of cell adhesion molecules and ECM proteins (i.e., N-cadherin, fibronectin [FN], and laminin [LN]).^21,22 These enhancements are thought to be induced by MSC secretome because addition of conditioned medium (CM) has been found to have similar effects on NPC differentiation.^14,20,23 While these studies demonstrated the benefits of MSC's trophic effects on NPC differentiation and CNS tissue repair, the mechanism underpinning the trophic effects and the impact of hypoxic secretome are not fully understood.¹⁴

NPC derivation from PSCs has been carried out either in monolayer cultures or from so-called neurospheres through the formation of embryoid bodies (EBs).^24,25 NPCs derived from monolayer cultures have a high degree of homogeneity but lack terminal neuronal maturation with limited neurite network.²⁶ In contrast, NPC aggregates with a heterogeneous neural cell population generated a dense neurite network and mature neuronal electrophysiology.^24,26 Additionally, transplantation of NPCs in 3D microenvironment either as neural aggregates or with 3D ECM scaffolds promoted cell retention, adhesion, survival, and differentiated function.^27,28 These studies suggest that the increased ECMs and the formation of proper milieu in the NPC aggregates play important roles in enhancing NPC proliferation, differentiation, and functional neural development.^26,29

The goal of the present study is to investigate the influence of human MSC (hMSC) secretome generated under hypoxia or normoxia on cellular behaviors of ESC-derived NPC aggregates. It is hypothesized that hMSC secretome provides a milieu of macromolecules that promote NPC aggregate attachment, survival, proliferation, and migration and that hypoxic precultures could further enhance the trophic effects of hMSC secretomes. Understanding the role of hMSC secretome in regulating cellular events of PSC-derived NPCs will reveal the paracrine effect of hMSCs on endogenous neurogenesis and provide a novel approach to derive NPCs from PSCs for potential co-transplantation with hMSCs.

Materials and Methods

Undifferentiated ESC culture and generation of NPCs

Murine ES-D3 line (American Type Culture Collection, Manassas, VA) was maintained on 0.1% gelatin-coated six-well plates in a standard 5% CO₂ incubator. The expansion medium is composed of Dulbecco's modified Eagle's medium (DMEM; Invitrogen, Carlsbad, CA) supplemented with 10% ESC-screened fetal bovine serum (FBS; Hyclone, Logan, UT), 1 mM sodium pyruvate, 0.1 mM β-mercaptoethanol, 100 U/mL penicillin, 100 μg/mL streptomycin (all from Invitrogen), and 1000 U/mL leukemia inhibitory factor (Millipore, Billerica, MA). The cells were seeded at 2–4×10⁴ cells/cm² and subcultured every 2–3 days.

To derive NPCs, ESCs were seeded at 1×10⁶ cells into Ultra-Low Attachment six-well plates (Corning Incorporated, Corning, NY) in 3 mL of DMEM-F12 plus 2% B27 serum-free supplement (Invitrogen) using a 4−/4+ protocol.³⁰ At day 4, all-trans retinoic acid (Sigma-Aldrich, St. Louis, MI) was added at 1 μM in the differentiation medium. NPC aggregates were collected at day 8 to assess the effects of hMSC secretomes (Fig. 1).

FIG. 1.

Schematic diagram of the experimental procedure. Embryonic stem cells (ESCs) were seeded in low attachment plates to form embryoid bodies in Dulbecco's modified Eagle's medium (DMEM)-F12 plus B27 for 4 days. Afterwards the medium was supplemented with retinoic acid for another 4 days to obtain neural progenitor cell (NPC) aggregates. NPCs were replated on gelatin-coated dishes in conditioned media under hypoxia (H-CM), conditioned media under normoxia (N-CM) or unconditioned media (U-CM) and cultured for 3 days before the characterizations. ROS, reactive oxygen species. Color images available online at www.liebertpub.com/tea

hMSC culture and collection of CM

Standardized frozen hMSCs from multiple donors were obtained from the Tulane Center for Gene Therapy and cultured as previously described.^31,32 The MSCs were isolated from the bone marrow of healthy donors ranging in age from 19 to 49 years based on plastic adherence, negative for CD34, CD45, and CD117 (all <2%) and positive for CD29, CD44, CD49c, CD90, CD105, and CD147 markers (all >95%), and possess tri-lineage differentiation potential upon induction in vitro.^33,34 Briefly, hMSCs were expanded at a density of 1.7×10³ cells/cm² using αMEM (Invitrogen) supplemented with 10% FBS and 1% penicillin/streptomycin in a standard 5% CO₂ incubator. At ∼80% confluence, adherent cells were harvested with 0.25% trypsin–EDTA (Sigma-Aldrich) and further propagated. Passage 5 cells were used for collecting CM.

CM collection was performed as previously described.³⁵ Briefly, hMSCs were seeded at a density of 1500 cells/cm² and incubated under hypoxia (2% O₂) or normoxia (20% O₂). After 7 days, the cultures were gently washed three times with phosphate-buffered saline (PBS), and fed with serum-free αMEM supplemented with 1% penicillin/streptomycin. After 24 h, conditioned media under normoxia (N-CM) or hypoxia (H-CM) were collected and filtered through a 0.2-μm filtration device. The αMEM without exposing to hMSCs was used as control (unconditioned media: U-CM). For neural differentiation, N-CM, H-CM, or U-CM were mixed at 50% v/v with DMEM-F12 plus B27 media.

NPC aggregate culture in the presence of hMSC secretomes

At day 8, NPC aggregates were replated on 0.1% gelatin-coated 24-well plates and cultured for 3 days in N-CM, H-CM, or U-CM. Adherent cell numbers were determined at 24, 48, and 72 h using a hemocytometer after trypsin/EDTA dissociation. Floating aggregates in the culture supernatants were also collected, dissociated by trypsin/EDTA, and counted for cell numbers. To assess the role of growth factor signaling in hMSC secretome during NPC differentiation, the following inhibitors were added in the media: PD173074 (20 nM; Sigma-Aldrich), a FGFR1/FGFR3 inhibitor,³⁶ to test the effect of fibroblast growth factor (FGF)-2; SB431542 (10 μM; Sigma-Aldrich), a SMAD2 inhibitor,³⁷ to test the effect of transforming growth factor (TGF)-β1; rabbit anti-BDNF (1:100; Millipore) to test the influence of BDNF. The cultures were treated with the inhibitors or anti-BDNF for 72 h and evaluated for cell growth, migration, and neurite length during this period.

Bromodeoxyuridine proliferation assay

The proliferation of the cells surrounding NPC aggregates at day 2 after replating was assessed using bromodeoxyuridine (BrdU) staining. Briefly, the cells were incubated in a DMEM-F12 plus B27 medium containing 10 μM BrdU (Sigma-Aldrich) for 90 min. The cells were then fixed with 70% cold ethanol, followed by a denaturation step using 2 N HCl/0.5% Triton X-100 for 30 min in the dark. The samples were reduced with 1 mg/mL sodium borohydride for 5 min and incubated with mouse anti-BrdU (1:100; Invitrogen) in blocking buffer (0.5% Tween 20/1% bovine serum albumin in PBS), followed by the incubation with Alexa Fluor^® 488 goat anti-Mouse IgG₁ (1:100; Molecular Probes, Grand Island, NY). The cells were mounted with 4′,6-diamidino-2-phenylindole (DAPI) and visualized using a fluorescent microscope (Olympus IX70, Melville, NY). Both the samples without incubation with BrdU and the samples treated with BrdU but stained with secondary antibody only were used as negative controls.

Cell cycle analysis

Two days after replating, the distribution of attached NPCs in various phases of cell cycle was assessed after the trypsin/EDTA dissociation.³⁸ Cells were fixed and permeabilized with 70% ethanol at 4°C for 40 min. The samples were then centrifuged and resuspended in a solution containing 20 μg/mL propidium iodide (PI; Sigma-Aldrich) plus 0.5 mg/mL RNAse A (Sigma-Aldrich). The samples without PI staining were used as negative control. After the incubation and washing, the cells were acquired with BD FACSCanto™ II flow cytometer and the results were analyzed using FlowJo software.

Neurite length and migration analysis

The migration length was assessed using a light microscope (Olympus IX70). The distance between the edge of the attached aggregates and the furthest migrated cells was determined at five distinct locations per aggregates using ImageJ software. For the neurite length, the distance between the edge of attached aggregates and the extremities of at least eight neurites per aggregates were calculated using ImageJ software. At least three spheres per condition were analyzed. The cells from U-CM group were used as negative control.

Live/Dead, reactive oxygen species, and poly-caspase assays

Live/Dead assay

Cell viability was assessed using LIVE/DEAD^® staining kit (Molecular Probes). The attached cells were incubated in DMEM-F12 containing 1 μM calcein AM (green color for live cells) and 2 μM ethidium homodimer I (red color for dead cells) for 30 min. The samples were then washed and imaged under a fluorescent microscope. The live cell controls were the normal ESC stock cultures. The dead cell controls were prepared by treating the cells with 70% methanol for 30 min.

Reactive oxygen species assay

Reactive oxygen species (ROS) was assessed using Image-iT™ Live Green ROS Detection kit (Molecular Probes).³⁹ Briefly, the aggregates were washed in Hank's Balanced Salt Solution and incubated in a solution of 25 μM carboxy-H₂DCFDA for 30 min at 37°C. The samples were imaged under fluorescence microscope to visualize ROS location in the aggregates. For quantification, the fluorescence intensity was analyzed using ImageJ software. As the positive control, the cells were incubated in 100 μM tert-butyl hydroperoxide (TBHP) prior to staining with carboxy-H₂DCFDA. The negative controls were the cultures without staining with carboxy-H₂DCFDA.

Caspase assay

Caspases were detected using Image-iT Live Green Caspase Detection Kits (Molecular Probes). Briefly, the aggregates were incubated for 1 h with fluorescent inhibitor of caspases (FLICA) reagent and imaged under fluorescence microscope. The fluorescence intensity was analyzed by ImageJ software. The cultures without staining with FLICA were used as negative controls.

Immunocytochemistry and enzyme-linked immunosorbent assay

For ECM expression, adherent or suspended NPC aggregates were fixed with 4% paraformaldehyde (PFA) and permeabilized with 0.2–0.5% Triton X-100. The samples were then blocked and incubated with primary ECM antibodies at 1:100 dilutions, including rabbit polyclonal FN, LN, collagen type IV (COL IV), and vitronectin (VN) (Abcam, Cambridge, MA).^35,39 After washing, the samples were incubated with Alexa Fluor 488 goat anti-Rabbit IgG (1:100; Molecular Probes), counterstained with DAPI, and visualized using a fluorescence microscope. To detect differentiation markers, the cells were incubated with mouse or rabbit primary antibody against the following: Nestin (1:200; Sigma-Aldrich) and Musashi-1 (1:200; Abcam) for progenitors, β-tubulin III (1:200; Millipore) for neurons, and glial fibrillary acidic protein (GFAP, 1:200; Millipore) for astrocytes. After washing, the cells were incubated with the corresponding secondary antibody: Alexa Fluor 488 goat anti-Mouse IgG₁ (for GFAP and β-tubulin III) or Alexa Fluor 488 goat anti-Rabbit IgG (for Nestin and Musashi-1) at 1:100 dilutions, and the cells were observed under fluorescence microscope. The samples with secondary antibody only were used as negative controls.

To quantify ECM contents,⁴⁰ the adherent or suspended NPC aggregates in different conditions were incubated with donkey anti-Rabbit IgG conjugated with horseradish peroxidase (HRP; Rockland Immunochemicals, Inc., Gilbertsville, PA) after the incubation with primary antibodies. After washing, 1 mL of 3,3′,5,5′-tetramethylbenzidine (TMB) substrate (Thermo scientific, Hudson, NH) was added to the samples, incubated for 5–25 min, and stopped by a solution of 0.1 M sulfuric acid. The absorbance units (AUs) were measured using a microplate reader (Bio-Rad, Richmond, CA) at a wavelength of 405 nm with background subtraction at 655 nm. The AU values were corrected by subtracting the values of negative control stained with HRP-IgG only. The measurements (i.e., indicating total ECM contents) were also normalized by the cell numbers to obtain specific ECM contents. To evaluate the effect of FGF signaling on extracellular signal-regulated kinase (ERK) pathway, cell lysates were collected and the levels of pERK were quantified by enzyme-linked immunosorbent assay (ELISA) following the manufacturer's instructions (Enzo Life Sciences, Farmingdale, NY). The specific pERK expression was calculated with AU values normalized to cell numbers. To quantify the growth factors in CM, the concentrations of FGF-2 and TGF-β1 in N-CM and H-CM were measured by ELISA according to the manufacturers' instructions (R&D Systems, Minneapolis, MN for FGF-2; Life Technologies for TGF-β1). The presence of BDNF was evidenced in blocking assay using anti-BDNF.

Flow cytometry

For flow cytometry, 1×10⁶ cells/sample were fixed with 4% PFA and permeabilized with 100% cold methanol. After blocking, the samples were incubated with primary antibodies against Nestin (1:200), β-tubulin III (1:200), GFAP (1:200), or Musashi-1 (1:200) followed by the corresponding secondary antibody (1:100 dilutions): Alexa Fluor 488 goat anti-Mouse IgG₁ (for GFAP and β-tubulin III) or Alexa Fluor 488 goat anti-Rabbit IgG (for Nestin and Musashi-1). The cells were acquired with BD FACSCanto II flow cytometer and analyzed against isotype controls using FlowJo software.

Statistical analysis

Each experiment was carried out at least three times. The average values of two or three independent experiments were presented and the results are expressed as (mean±mean absolute deviation). In each experiment, triplicate samples were used. To assess the statistical significance, ANOVA followed by Fisher's LSD post hoc tests. A p-value<0.05 was considered statistically significant.

Results

hMSC secretome promoted ECM secretion of ESC-derived NPC aggregates

After 8 days of differentiation, NPC aggregates were replated in the presence of N-CM, H-CM, or U-CM for an additional 3 days. The expression of endogenous ECMs in the attached cells was evaluated (Fig. 2A–H). FN was able to spread on the gelatin-coated surface to the area without cells, while LN, COL IV, and VN were mainly detected close to the cell membrane. Because the culture medium was serum-free and there was no exogenous FN, LN, COL IV, and VN used for coating, the expression indicated the endogenous ECM secretion from NPCs. The specific expressions of FN, LN, COL IV, and VN were similar for attached NPCs cultivated in N-CM, H-CM, and U-CM, while the expressions of total ECM proteins were enhanced by threefolds in N-CM/H-CM versus U-CM (Fig. 2I, J). The floating NPC aggregates were also assessed for endogenous ECM expression. N-CM and H-CM conditions displayed about twofold increase in specific expression of ECM proteins compared with U-CM (Fig. 2K). The expressions of total ECM proteins were also increased (1.3- to 2.0-folds) in N-CM/H-CM compared with U-CM (Fig. 2L). Hence, hMSC secretomes promoted endogenous ECM secretion of ESC-derived NPC aggregates.

FIG. 2.

Extracellular matrix (ECM) expression of ESC-derived NPC aggregates in N-CM, H-CM, and U-CM. (A–D) ECM expressions around aggregates: (A) Fibronectin (FN); (B) laminin (LN); (C) collagen type IV (COL IV); (D) vitronectin (VN). (E–H) ECM expressions in migrating cells: (E) FN; (F) LN; (G) COL IV; (H) VN. Scale bar: 100 μm. Quantitative analysis of ECM expression for adherent NPCs: (I) specific ECM expression; (J) total ECM expression. Quantitative analysis of ECM expression for NPC aggregates in suspension: (K) specific ECM expression; (L) total ECM expression. *p-value<0.05. Color images available online at www.liebertpub.com/tea

hMSC secretome promoted adhesion and proliferation of ESC-derived NPC aggregates

The impact of hMSC secretomes on adhesion and proliferation of ESC-derived NPC aggregates were evaluated. N-CM and H-CM significantly increased cell attachment and outgrowth compared with U-CM based on morphology observation (Fig. 3A). In all conditions, the cells around the aggregates displayed a heterogeneous morphology composed of flat cells and bipolar cells with neurite projections, suggesting the presence of a mixed population of mature neurons and neural progenitors.⁴¹ The adherent cell numbers (including the migrating single cells and the attached aggregates) were threefold to fivefold higher for N-CM and H-CM than U-CM, while no significant difference was observed between N-CM and H-CM (Fig. 3B). Consistently, the percentage of floating cells in U-CM was also higher (90% vs. 40–60%) compared with N-CM and H-CM (Fig. 3C), suggesting that N-CM and H-CM favored NPC aggregate adhesion.

FIG. 3.

Adhesion and proliferation of ESC-derived NPC aggregates in N-CM, H-CM, and U-CM. (A) Morphology of the replated NPC aggregates; (B) adherent cell number; (C) percentage of floating cells; (D) images of BrdU⁺ cells around NPC aggregates; (E) representative bromodeoxyuridine (BrdU) expression; (F) cell cycle distribution. Scale bar: 100 μm. *p-value<0.05 for the comparison of N-CM or H-CM versus U-CM. Color images available online at www.liebertpub.com/tea

To further analyze the status of adherent cells, BrdU incorporation was evaluated to detect the cells in S phase. More BrdU⁺ cells (about 2.8-folds) were observed for cells around NPC aggregates in N-CM and H-CM conditions compared with U-CM (Fig. 3D, E). To confirm this observation, cell cycle analysis was performed 24 h postreplating. About 28–29% NPCs in N-CM and H-CM were in S-phase, while only 17% for U-CM (Fig. 3F). Conversely, 74% of cells in G0/G1 were observed for U-CM compared with 60–63% in N-CM/H-CM conditions. Hence, N-CM and H-CM promoted NPC proliferation, which also contributed to the higher adherent cell numbers.

hMSC secretome regulated the survival of ESC-derived NPC aggregates

To interrogate whether the higher cell numbers in N-CM/H-CM versus U-CM is related to cell survival, cell viability, ROS, and caspase expression were assessed. NPCs cultivated in N-CM/H-CM displayed higher cell viability compared with U-CM (75% vs. 45%) (Fig. 4A, B). Similarly, caspase expression for NPCs in N-CM/H-CM was about 1.7-fold lower compared with U-CM (Fig. 4C). Because ROS may induce caspase expression, the ROS levels were also measured (Fig. 4D–F). ROS expression was about 1.8-fold lower in H-CM compared with N-CM and U-CM in the absence of TBHP (Fig. 4E). Upon ROS induction using TBHP, about threefold less ROS expression was observed in H-CM (Fig. 4F), indicating that H-CM may provide better antioxidant effect compared with N-CM and U-CM.

FIG. 4.

Cell survival, caspase, and ROS expression of ESC-derived NPC aggregates in N-CM, H-CM, and U-CM. (A) Representative Live/Dead staining images; (B) comparison of cell viability; (C) comparison of caspase expression; (D) representative ROS images; (E) comparison of ROS expression; (F) ROS expression after tert-butyl hydroperoxide (TBHP) treatment. Scale bar: 100 μm. *p-value<0.05. Color images available online at www.liebertpub.com/tea

Effects of hMSC secretome on neural differentiation of ESC-derived NPC aggregates

After growing for 3 days, NPCs were assessed for the expression of neural markers Nestin (progenitor), Musashi-1 (progenitor), GFAP (astrocyte), and β-tubulin III (neuron) (Fig. 5). Similar percentages of marker expression were observed for all three conditions: 44–48% β-tubulin III, 20–24% GFAP, 58–68% Nestin, and 82–85% Musashi-1 (Fig. 5A, B). Before the treatment, day 8 NPCs expressed about 60–70% Nestin, 80–90% Musashi-1, and low levels of GFAP and β-tubulin III (<10%), and thus NPC aggregate replating in hMSC secretomes resulted in neural differentiation. About 2.5- to 3.0-fold increase in the positive cell numbers for each marker was observed for N-CM/H-CM conditions compared with U-CM condition (Fig. 5C), which may be directly linked to the increased total cell numbers in N-CM/H-CM.

FIG. 5.

Neural differentiation of ESC-derived NPC aggregates in N-CM, H-CM, and U-CM. (A) Representative fluorescence images of neural markers. The images were from cells in H-CM culture; (B) representative flow cytometry histograms of neural markers. The histograms were from cells in H-CM culture. Scale bar: 100 μm. (C) Percentages of positive cells expressing neural markers; (D) Total numbers of positive cells expressing neural markers. *p-value <0.05. Color images available online at www.liebertpub.com/tea

Although the percentages of neural markers were similar, quantitative analysis indicated about threefold to fourfold longer neurite length in N-CM/H-CM compared with U-CM (i.e., 600 μm vs. 180 μm at 72 h) (Fig. 6A, C). In addition, NPC migration length from the aggregates was increased by twofold to threefold in N-CM/H-CM compared with U-CM condition (i.e., 340–450 μm vs. 170 μm at 72 h) (Fig. 6B, D). No significant difference in neurite length was observed between N-CM and H-CM treated cells, while migration length was slightly longer for H-CM than N-CM treated cells.

FIG. 6.

Neurite extension and cell migration of ESC-derived NPC aggregates in N-CM, H-CM, and U-CM. (A) Neurite length; (B) migration length. *p-value <0.05 for the comparison of N-CM or H-CM versus U-CM; (C) representative morphology of neurite extensions in different groups; (D) representative morphology of cell migration around NPC aggregates. Scale bar: 100 μm.

Effects of FGF-2, TGF-β1, and BDNF signaling

To further delineate the effects of hMSC secretomes, the presence of FGF-2 and TGF-β1 in N-CM and H-CM were analyzed (Table 1). H-CM contained higher concentration of FGF-2 than N-CM but has similar level of TGF-β1. To illustrate the effects of FGF-2 and TGF-β1, inhibitors for FGF-2 and TGF-β1 were tested in the cultures. As BDNF was another important growth factor in MSC-CM,⁴² the blocking effect of anti-BDNF was also evaluated. After PD173074 (a FGFR inhibitor) treatment, the adherent cell numbers significantly decreased (8–10×10⁵ vs. 2–4×10⁵) in N-CM/H-CM or modestly (3–4×10⁵ vs. 0.5–1×10⁵) in U-CM (Fig. 7A). The decrease in U-CM condition indicated the inhibition of endogenous FGF signaling from NPCs.⁴³ No difference in neurite length was observed with PD173074 treatment (Fig. 7B), while migration length decreased (400–500 μm vs. 200–300 μm) for N-CM and H-CM (Fig. 7C). Because FGF-2 binding triggers the phosphorylation of ERK activity,³⁵ pERK expression was also evaluated. A decrease in pERK activity was observed after PD173074 treatment in all three conditions, further confirming the presence of FGF signaling (Fig. 7D).

FIG. 7.

Effect of fibroblast growth factor-2 signaling on ESC-derived NPC aggregates in N-CM, H-CM, and U-CM. The cells with or without treatment of PD173074 were compared. (A) Adherent cell number; (B) neurite length; (C) migration length; (D) level of pERK expression; *p-value<0.05. In (A–C), a, b, and c correspond to a p-value <0.05 for the paired comparison of N-CM versus N-CM+PD173074, H-CM versus H-CM+PD173074, or U-CM versus U-CM+PD173074 respectively.

Table 1.

Growth Factor Concentrations in N-CM and H-CM

Growth factors	Media	Concentration (pg/mL) ^a
FGF-2^b	N-CM	6.6±1.3
	H-CM	10.0±2.0
TGF-β1	N-CM	16.1±13.7
	H-CM	6.7±2.3
BDNF	N-CM	Presence was evidenced by anti-BDNF treatment
	H-CM

Indicates average concentrations from three collections of conditioned media.

Indicates significant difference (p<0.05).

BDNF, brain-derived neurotrophic factor; FGF, fibroblast growth factor; H-CM, conditioned media under hypoxia; N-CM, conditioned media under normoxia; TGF, transforming growth factor.

TGF-β1 inhibitor SB431542 significantly reduced the adherent cell numbers (3.5–4.2×10⁵ vs. 0.5–1.8×10⁵) in N-CM and H-CM (Fig. 8A). The cell number in U-CM was also reduced with SB431542 treatment, indicating the presence of endogenous TGF-β1 signaling from NPCs. In addition, reduced neurite length (630–690 μm vs. 160–190 μm) and migration length (570–600 μm vs. 160–200 μm) were observed in the presence of SB431542, indicating the impact of TGF-β1 signaling on neural cell functions (Fig. 8B, C).

FIG. 8.

Effect of transforming growth factor-β1 signaling on ESC-derived NPC aggregates in N-CM, H-CM, and U-CM. The cells with or without treatment of SB431542 were compared. (A) Adherent cell number; (B) neurite length; (C) migration length. In (A–C), a, b, and c correspond to a p-value <0.05 for the paired comparison of N-CM versus N-CM+SB431542, H-CM versus H-CM+SB431542, or U-CM versus U-CM+SB431542 respectively.

Inhibition of BDNF signaling with anti-BDNF significantly reduced the adherent cell numbers (3.2–3.6×10⁵ vs. 1.1–1.5×10⁵) in N-CM/H-CM and in U-CM (1.7×10⁵ vs. 0.8×10⁵) (Fig. 9A). The decrease of adherent cell number in U-CM indicated endogenous BDNF secretion by NPCs. The neurite extension was reduced after treating with BDNF antibody (480–650 μm vs. 300–450 μm), but no significant difference was observed for migration length (Fig. 9B, C). The inhibition effects of these three molecules were summarized in Table 2. Taken together, N-CM/H-CM conditions enhanced FGF-2, TGF-β1, and BDNF signaling and differentially promoted NPC proliferation, migration, and neurite extension.

FIG. 9.

Effect of brain-derived neurotrophic factor (BDNF) signaling on ESC-derived NPC aggregates in N-CM, H-CM, and U-CM. The cells with or without treatment of anti-BDNF antibody were compared. (A) Adherent cell number; (B) neurite length; (C) migration length. In (A–C), a, b, and c correspond to a p-value <0.05 for the paired comparison of N-CM versus N-CM+anti-BDNF, H-CM versus H-CM+anti-BDNF, or U-CM versus U-CM+anti-BDNF respectively.

Table 2.

Percentage of Inhibition After Treatment with PD 173074, SB 431542, or BDNF Antibody

Inhibitor type	N-CM (%)	H-CM (%)	U-CM (%)
PD173074
Adherent cell number	62.5±10.6	58.2±27.1	67.5±17.0
Neurite length	0	0	20.0±13.7
Migration length	37.8±12.5	31.7±14.0	22.4±4.5
SB431542
Adhesion cell number	51.7±23.3	72.5±17.3	49.6±4.8
Neurite length^a	67.0±10.0	73.3±3.7	21.0±10.3
Migration length^a	66.6±0.2	68.1±4.0	0
BDNF antibody
Adhesion cell number^b	42.2±12.3	63.1±6.6	56.7±2.8
Neurite length^a	21.3±12.0	42.1±12.2	0
Migration length	0	0	0

Percentage of inhibition was calculated using the average values of 48 and 72 h after replating (from Figs. 7 –9) as follows:

\documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} $$\% \ inhibibation = \frac { without \ inhibition - with \ inhibition } { without \ inhibtion } \times 100 \%$$ \end{document} .

Indicates significant difference (p<0.05) between N-CM/H-CM and U-CM groups.

Indicates significant difference (p<0.05) between N-CM and H-CM groups.

U-CM, unconditioned media.

Discussion

NPC derivation from PSCs is a promising approach to obtain renewable neural cells but the derivation process is less than robust and the NPC properties remain to be improved. Methods such as niche-dependent and ECM-supported differentiations have been tested to improve derivation efficiency and promote NPC maturation.^26,44 hMSC secretome contains a diverse range of bioactive molecules that regulate the cellular events underpinning NPC differentiation and maturation. The results of the current study revealed that hMSC secretome enhanced ESC-derived NPC proliferation, migration, and neurite extension primarily through regulating the molecular milieu in NPC aggregates including ECM proteins and growth factors.

The presence of hMSC secretome influenced the ratio of adherent to nonadherent NPC aggregates derived from ESCs. The heterogeneity of ESC-derived NPC aggregates is a result of differential cell adhesion properties, while treatment with hMSC secretome appeared to increase NPC adhesion due to the upregulation of ECM proteins such as LN and FN especially in floating aggregates. In vitro, ECM proteins have been shown to regulate the survival, proliferation, and neurite outgrowth of human ESC-derived NPCs in a dose-dependent manner through integrin-ECM signaling.^41,45 In vivo, ECM proteins are involved in neural development and have shown to support neurogenesis from the transplanted NPCs or endogenous progenitors by promoting cell survival and providing a neuro-permissive environment.^28,46 The mechanism for the enhanced ECM secretion in ESC-derived NPC aggregates may involve multiple factors and regulatory pathways emanating from hMSC secretome. FGF signaling is known to induce the transcription of LN-1 and COL IV in EBs, affecting the ECM microenvironment.⁴⁷ The presence of tissue inhibitor of metalloproteinase-1 in hMSC secretome may also play a role because it could inhibit ECM degradation by regulating MMP-9 activity.⁴⁸ Although the exact balance of the increased ECM secretion versus inhibition of ECM degradation requires further study, multiple factors in hMSC secretome and their concerted effects are likely to be involved.^47–49

Increased ECM secretion in the 3D aggregates may serve as reservoirs and in turn enhance local enrichment of growth factors such as FGF-2 and TGF-β1.^35,50 FGF-2 is known to bind heparin sulfate proteoglycans and influence cell fate through the activation of mitogen-activated protein kinase with consequent phosphorylation of ERK1/2.³⁵ FGF-2 is a mitogenic factor and has been commonly used to maintain NPCs at progenitor stage.⁴ Indeed, blocking FGF-FGFR signaling using PD173074 abolished the positive effects of hMSC secretomes on cell adhesion and migration, but not neurite extension as shown in our study. TGF-β1 is also a matrix-associating protein and its sequestration and release from ECMs is an important mechanism that regulates its bioactivity.^51,52 MSCs are significant source of TGF-β1,^37,43,53 which has been shown to promote glial cell differentiation of adult NPCs.³⁷ TGF-β1 also exerted anti-apoptotic and growth/migration-enhancing activities on NPCs through SMAD2 (for growth and migration) and ERK1/2 (for anti-apoptotic behavior) signaling.⁵⁴ In our study, SB431542 treatment reduced NPC adhesion/proliferation, migration, and neurite extension, suggesting the involvement of TGF-β1 signaling in neural development. BDNF, secreted by hMSCs at a higher level than NPCs, is a primary trophic factor that supports the growth and differentiation of neuronal lineage.^42,43 BNDF binding to collagen enhanced the neurotrophic effect and promoted axonal regeneration compared with BDNF without collagen-binding domain.⁵⁵ In the current study, inhibition of BDNF signaling resulted in the reduced cell number and neurite length but not migration, indicating its contribution in neuronal cell maturation. Together, the results suggested that multiple secretome-derived factors are involved in regulating NPC behaviors from initial adhesion to later neurite extension and that ECM may be an important facilitator for the growth factor-mediated enhancement.

Apart from the factors in the hMSC secretome, endogenous factors secreted from NPCs may also have autocrine effects that regulate NPC fate. NPC aggregates treated with inhibitors in U-CM also showed the reduced adherent cell number (50–68%) as in N-CM (42–63%) and H-CM (58–73%) groups (Table 2). However, the more significant decrease in adherent cell number, neurite length, and migration length following growth factor inhibition in the N-CM/H-CM groups compared with the U-CM group (Figs. 7 –9) suggested more pronounced trophic effects originating from the hMSC secretomes.

hMSC secretomes significantly reduced caspase expression and promoted NPC aggregate survival from our results. Besides FGF-2, TGF-β1, and BDNF, MSC secretomes are reported to contain anti-apoptotic and antioxidant factors such as stanniocalcin-1,¹⁴ VEGF, and stromal cell-derived factor-1.⁵⁶ These factors may together contribute to the protection of NPCs from ROS damage and caspase activation,^14,54,57 thereby increasing cell viability.

In this study, hMSC secretomes did not affect the percentage of differentiating cells but rather increased the yield of neural cells, presumably due to the increased cell adhesion, survival, and proliferation. Whether the restricted effects are due to the inherent secretome properties or variations in secretome collection and treatment time remains to be investigated.^58,59 The day 8 NPC aggregates used in this study may be at primitive stage, and a longer treatment with hMSC secretome may be required to influence mature neural populations. The collection time of hMSC secretome at 24 h may also limit its effect on neural differentiation.⁵⁸ Nevertheless, the results demonstrated the potential of hMSC secretome to enhance the functions of differentiated NPCs.

Hypoxic secretome H-CM did not enhance NPC adhesion and proliferation compared to N-CM. Although hypoxia could increase the secretion of FGF-2, BDNF, and pro-angiogenic factors such as VEGF compared with normoxia,^35,60 the dose level (pg-ng/mL) may not be sufficient to elicit a difference in cell adhesion/proliferation compared to N-CM. However, the more significant decline in NPC numbers by BDNF inhibition in H-CM (63% vs. 42%, Table 1) indicates a higher contribution of hMSC-derived BDNF in H-CM. The reduced ROS expression by NPCs in H-CM compared with N-CM under oxidative conditions (i.e., using TBHP) also suggests an increased antioxidant effect in H-CM compared with N-CM.

Compared to the restricted effects by H-CM in NPC derivation, hypoxia directly influences multiple cellular events during neurogenic differentiation of ESCs.⁶¹ Hypoxia (2% O₂) was shown to accelerate early commitment of ground-state ESCs toward primitive neuroectoderm but appear to negatively impact cell viability after 6 days of neural commitment.⁶² The effects of hypoxia on the neurosphere derivation are mediated by activation of hypoxia-inducible factors (HIF) and the target genes of erythropoietin, VEGF, and Bcl-2 family members, although HIF-independent pathways may also play a role.^63,64 Hypoxia also affects the stage of lineage specification from NPCs. Mild hypoxia (2.5–5% O₂) was shown to enhance human neural stem cell differentiation into neuronal and oligodendroglial cells, while under normoxia (20% O₂) the cells had the preferential commitment to the astrocyte lineage, possibly due to anti-apoptotic effects indicated by the reduced caspase expression.⁶⁵ Compared to H-CM, oxygen tension directly influences neurogenic differentiation of ESCs via multiple mechanisms at molecular levels.

Further study is required to address the temporal effects of the secreted factors, their interactions, and inhibitory effects from the known or unidentified factors in the secretome. To this end, modulating hMSC culture condition is an important means to control the secretome properties. For example, negative factors such as MMP-2 endogenously secreted by hMSCs is known to mediate ECM degradation and negatively influence NPC derivation.⁴⁸ Pretreatment by tumor necrosis factor alpha or hypoxia, however, could inhibit the activity of hMSC-secreted MMP-2, highlighting the effects of culture conditions on hMSC secretome.^48,66 Recent studies suggest that hMSC culture as 3D aggregates dramatically influences their secretory profiles and produces secretome with enhanced anti-inflammatory properties.^67,68 Moreover, studies are also required to address the interplay between multiple factors that coexist in the secretome. It is known that the growth factors in hMSC secretome are able to exert their effects at concentration range well below those of exogenous growth factors typically used in in vitro studies.^14,44 Whether the regulatory molecules in hMSC secretome have additive effects on cellular behaviors is yet to be investigated. Understanding these interactions may reveal novel regulatory pathways and lead to novel approach to enhance NPC derivation. To this end, system biology approach is desirable because of its potential to provide a global view of regulatory network during lineage-specific differentiation.⁶⁹

Conclusions

hMSC secretomes enhanced endogenous ECM secretion from NPCs and modulated adhesion, proliferation, survival, migration, and neurite extension of ESC-derived NPC aggregates. These regulatory functions were mediated in part by hMSC-derived FGF-2, TGF-β1, and BDNF and possibly their interactions with ECM proteins. H-CM treatment displayed similar cellular responses compared to N-CM condition but had better antioxidant effects. The results demonstrated the potential of hMSC secretome to regulate neural development of PSC-derived NPCs.

Footnotes

Acknowledgments

The authors would like to thank Ms. Ruth Didier of FSU Department of Biomedical Sciences for her help in flow cytometry analysis. This work is supported by FSU startup fund, FSU GAP award, and FSU student technology committee. Funding support in part from the James King Biomedical Research Program (4KB09) is also acknowledged.

Disclosure Statement

No competing financial interests exist.

References

, Gautam

, Yang

, Qiu

, Melkoumian

, Weber

, Telukuntla

, Srivastava

, Whiteley

E.M.

, and Brandenberger

Differentiation of oligodendrocyte progenitor cells from human embryonic stem cells on vitronectin-derived synthetic peptide acrylate surface. Stem Cells Dev, 22, 1497, 2013.

Chambers

S.M.

, Fasano

C.A.

, Papapetrou

E.P.

, Tomishima

, Sadelain

, and Studer

Highly efficient neural conversion of human ES and iPS cells by dual inhibition of SMAD signaling. Nat Biotechnol, 27, 275, 2009.

Kim

D.S.

, Lee

J.S.

, Leem

J.W.

, Huh

Y.J.

, Kim

J.Y.

, Kim

H.S.

, Park

I.H.

, Daley

G.Q.

, Hwang

D.Y.

, and Kim

D.W.

Robust enhancement of neural differentiation from human ES and iPS cells regardless of their innate difference in differentiation propensity. Stem Cell Rev, 6, 270, 2010.

Oki

, Tatarishvili

, Woods

, Koch

, Wattananit

, Mine

, Monni

, Prietro

D.T.

, Ahlenius

, Ladewig

, Brustle

, Lindvall

, and Kokaia

Human induced pluripotent stem cells form functional neurons and improve recovery after grafting in stroke-damaged brain. Stem Cells, 30, 1120, 2012.

Rossi

S.L.

, Nistor

, Wyatt

, Yin

H.Z.

, Poole

A.J.

, Weiss

J.H.

, Gardener

M.J.

, Dijkstra

, Fischer

D.F.

, and Keirstead

H.S.

Histological and functional benefit following transplantation of motor neuron progenitors to the injured rat spinal cord. PLoS One, 5, e11852, 2010.

De Feo

, Merlini

, Laterza

, and Martino

Neural stem cell transplantation in central nervous system disorders: from cell replacement to neuroprotection. Curr Opin Neurol, 25, 322, 2012.

Martino

, and Pluchino

The therapeutic potential of neural stem cells. Nat Rev Neurosci, 7, 395, 2006.

Daadi

M.M.

, and Steinberg

G.K.

Manufacturing neurons from human embryonic stem cells: biological and regulatory aspects to develop a safe cellular product for stroke cell therapy. Regen Med, 4, 251, 2009.

Hibaoui

, and Feki

Human pluripotent stem cells: applications and challenges in neurological diseases. Front Physiol, 3, 267, 2012.

10.

S.M.

, and Hochedlinger

Harnessing the potential of induced pluripotent stem cells for regenerative medicine. Nat Cell Biol, 13, 497, 2011.

11.

Love

Oxidative stress in brain ischemia. Brain Pathol, 9, 119, 1999.

12.

Yong

V.W.

, Power

, Forsyth

, and Edwards

D.R.

Metalloproteinases in biology and pathology of the nervous system. Nat Rev Neurosci, 2, 502, 2001.

13.

Vroemen

, Aigner

, Winkler

, and Weidner

Adult neural progenitor cell grafts survive after acute spinal cord injury and integrate along axonal pathways. Eur J Neurosci, 18, 743, 2003.

14.

Teixeira

F.G.

, Carvalho

M.M.

, Sousa

, and Salgado

A.J.

Mesenchymal stem cells secretome: a new paradigm for central nervous system regeneration? Cell Mol Life Sci, 70, 3871, 2013.

15.

Giordano

, Galderisi

, and Marino

I.R.

From the laboratory bench to the patient's bedside: an update on clinical trials with mesenchymal stem cells. J Cell Physiol, 211, 27, 2007.

16.

Joyce

, Annett

, Wirthlin

, Olson

, Bauer

, and Nolta

J.A.

Mesenchymal stem cells for the treatment of neurodegenerative disease. Regen Med, 5, 933, 2010.

17.

Maltman

D.J.

, Hardy

S.A.

, and Przyborski

S.A.

Role of mesenchymal stem cells in neurogenesis and nervous system repair. Neurochem Int, 59, 347, 2011.

18.

Hardy

S.A.

, Maltman

D.J.

, and Przyborski

S.A.

Mesenchymal stem cells as mediators of neural differentiation. Curr Stem Cell Res Ther, 3, 43, 2008.

19.

Rivera

F.J.

, Couillard-Despres

, Pedre

, Ploetz

, Caioni

, Lois

, Bogdahn

, and Aigner

Mesenchymal stem cells instruct oligodendrogenic fate decision on adult neural stem cells. Stem Cells, 24, 2209, 2006.

20.

Rivera

F.J.

, Kandasamy

, Couillard-Despres

, Caioni

, Sanchez

, Huber

, Weidner

, Bogdahn

, and Aigner

Oligodendrogenesis of adult neural progenitors: differential effects of ciliary neurotrophic factor and mesenchymal stem cell derived factors. J Neurochem, 107, 832, 2008.

21.

Q.M.

, Fu

Y.M.

, Shan

Z.Y.

, Shen

J.L.

, Zhang

X.M.

, Lei

, and Jin

L.H.

MSCs guide neurite directional extension and promote oligodendrogenesis in NSCs. Biochem Biophys Res Commun, 384, 372, 2009.

22.

Wang

, Tu

, Lou

, Xie

, Lai

, Guo

, and Deng

Mesenchymal stem cells regulate the proliferation and differentiation of neural stem cells through Notch signaling. Cell Biol Int, 33, 1173, 2009.

23.

Bai

, Caplan

, Lennon

, and Miller

R.H.

Human mesenchymal stem cells signals regulate neural stem cell fate. Neurochem Res, 32, 353, 2007.

24.

Nori

, Okada

, Yasuda

, Tsuji

, Takahashi

, Kobayashi

, Fujiyoshi

, Koike

, Uchiyama

, Ikeda

, Toyama

, Yamanaka

, Nakamura

, and Okano

Grafted human-induced pluripotent stem-cell-derived neurospheres promote motor functional recovery after spinal cord injury in mice. Proc Natl Acad Sci U S A, 108, 16825, 2011.

25.

Ying

Q.L.

, Stavridis

, Griffiths

, Li

, and Smith

Conversion of embryonic stem cells into neuroectodermal precursors in adherent monoculture. Nat Biotechnol, 21, 183, 2003.

26.

Illes

, Theiss

, Hartung

H.P.

, Siebler

, and Dihne

Niche-dependent development of functional neuronal networks from embryonic stem cell-derived neural populations. BMC Neurosci, 10, 93, 2009.

27.

Mothe

A.J.

, Kulbatski

, Parr

, Mohareb

, and Tator

C.H.

Adult spinal cord stem/progenitor cells transplanted as neurospheres preferentially differentiate into oligodendrocytes in the adult rat spinal cord. Cell Transplant, 17, 735, 2008.

28.

Tate

C.C.

, Shear

D.A.

, Tate

M.C.

, Archer

D.R.

, Stein

D.G.

, and LaPlaca

M.C.

Laminin and fibronectin scaffolds enhance neural stem cell transplantation into the injured brain. J Tissue Eng Regen Med, 3, 208, 2009.

29.

Kearns

S.M.

, Laywell

E.D.

, Kukekov

V.K.

, and Steindler

D.A.

Extracellular matrix effects on neurosphere cell motility. Exp Neurol, 182, 240, 2003.

30.

Okada

, Shimazaki

, Sobue

, and Okano

Retinoic-acid-concentration-dependent acquisition of neural cell identity during in vitro differentiation of mouse embryonic stem cells. Dev Biol, 275, 124, 2004.

31.

Zhao

, and Ma

Perfusion bioreactor system for human mesenchymal stem cell tissue engineering: dynamic cell seeding and construct development. Biotechnol Bioeng, 91, 482, 2005.

32.

Grayson

W.L.

, Ma

, and Bunnell

Human mesenchymal stem cells tissue development in 3D PET matrices. Biotechnol Prog, 20, 905, 2004.

33.

Munoz

, Kim

, Liu

, Logan

T.M.

, and Ma

Gas chromatography-mass spectrometry analysis of human mesenchymal stem cell metabolism during proliferation and osteogenic differentiation under different oxygen tensions. J Biotechnol, 169, 95, 2014.

34.

Grayson

W.L.

, Zhao

, Izadpanah

, Bunnell

, and Ma

Effects of hypoxia on human mesenchymal stem cell expansion and plasticity in 3D constructs. J Cell Physiol, 207, 331, 2006.

35.

Kim

, and Ma

Autocrine fibroblast growth factor 2-mediated interactions between human mesenchymal stem cells and the extracellular matrix under varying oxygen tension. J Cell Biochem, 114, 716, 2013.

36.

Kim

, and Ma

Regulation of autocrine fibroblast growth factor-2 signaling by perfusion flow in 3D human mesenchymal stem cell constructs. Biotechnol Prog, 28, 1384, 2012.

37.

Robinson

A.P.

, Foraker

J.E.

, Ylostalo

, and Prockop

D.J.

Human stem/progenitor cells from bone marrow enhance glial differentiation of rat neural stem cells: a role for transforming growth factor beta and Notch signaling. Stem Cells Dev, 20, 289, 2011.

38.

Sart

, Schneider

Y.J.

, and Agathos

S.N.

Influence of culture parameters on ear mesenchymal stem cells expanded on microcarriers. J Biotechnol, 150, 149, 2010.

39.

Sart

, Ma

, and Li

Cryopreservation of pluripotent stem cell aggregates in defined protein-free formulation. Biotechnol Prog, 29, 143, 2013.

40.

Sart

, Ma

, and Li

Extracellular matrices decellularized from embryonic stem cells maintained their structure and signaling specificity. Tissue Eng Part A, 20, 54, 2014.

41.

, Tavakoli

, Derby

, Serebryakova

, Rao

M.S.

, and Mattson

M.P.

Cell-extracellular matrix interactions regulate neural differentiation of human embryonic stem cells. BMC Dev Biol, 8, 90, 2008.

42.

, Zhang

, Xue

, Ma

, Lu

, and Yu

Bone mesenchymal stromal cells stimulate neurite outgrowth of spinal neurons by secreting neurotrophic factors. Neurol Res, 34, 172, 2012.

43.

Hawryluk

G.W.

, Mothe

A.J.

, Chamankhah

, Wang

, Tator

, and Fehlings

M.G.

In vitro characterization of trophic factor expression in neural precursor cells. Stem Cells Dev, 21, 432, 2012.

44.

Axell

M.Z.

, Zlateva

, and Curtis

A method for rapid derivation and propagation of neural progenitors from human embryonic stem cells. J Neurosci Methods, 184, 275, 2009.

45.

, Deng

, Wang

, and Xu

X.M.

Effects of extracellular matrix molecules on the growth properties of oligodendrocyte progenitor cells in vitro. J Neurosci Res, 87, 2854, 2009.

46.

Tate

C.C.

, Tate

M.C.

, and LaPlaca

M.C.

Fibronectin and laminin increase in the mouse brain after controlled cortical impact injury. J Neurotrauma, 24, 226, 2007.

47.

, Chen

, Scheele

, Arman

, Haffner-Krausz

, Ekblom

, and Lonai

Fibroblast growth factor signaling and basement membrane assembly are connected during epithelial morphogenesis of the embryoid body. J Cell Biol, 153, 811, 2001.

48.

Ranganath

S.H.

, Levy

, Inamdar

M.S.

, and Karp

J.M.

Harnessing the mesenchymal stem cell secretome for the treatment of cardiovascular disease. Cell Stem Cell, 10, 244, 2012.

49.

Flanders

K.C.

, Ren

R.F.

, and Lippa

C.F.

Transforming growth factor-betas in neurodegenerative disease. Prog Neurobiol, 54, 71, 1998.

50.

Discher

D.E.

, Mooney

D.J.

, and Zandstra

P.W.

Growth factors, matrices, and forces combine and control stem cells. Science, 324, 1673, 2009.

51.

Wells

R.G.

, and Discher

D.E.

Matrix elasticity, cytoskeletal tension, and TGF-beta: the insoluble and soluble meet. Sci Signal, 1, pe13, 2008.

52.

Chen

, Sivakumar

, Barley

, Peters

D.M.

, Gomes

R.R.

, Farach-Carson

M.C.

, and Dallas

S.L.

Potential role for heparan sulfate proteoglycans in regulation of transforming growth factor-beta (TGF-beta) by modulating assembly of latent TGF-beta-binding protein-1. J Biol Chem, 282, 26418, 2007.

53.

Montes

, Ligero

, Sanchez

, Catalina

, de la Cueva

, Nieto

, Melen

G.J.

, Rubio

, Garcia-Castro

, Bueno

, and Menendez

Feeder-free maintenance of hESCs in mesenchymal stem cell-conditioned media: distinct requirements for TGF-beta and IGF-II. Cell Res, 19, 698, 2009.

54.

Park

S.M.

, Jung

J.S.

, Jang

M.S.

, Kang

K.S.

, and Kang

S.K.

Transforming growth factor-beta1 regulates the fate of cultured spinal cord-derived neural progenitor cells. Cell Prolif, 41, 248, 2008.

55.

Liang

, Han

, Jin

, Xiao

, Huang

, Ni

, Chen

, Kong

, Wu

, and Dai

The promotion of neurological recovery in the rat spinal cord crushed injury model by collagen-binding BDNF. Biomaterials, 31, 8634, 2010.

56.

Wang

, Yasuhara

, Shingo

, Kameda

, Tajiri

, Yuan

W.J.

, Kondo

, Kadota

, Baba

, Tayra

J.T.

, Kikuchi

, Miyoshi

, and Date

Intravenous administration of mesenchymal stem cells exerts therapeutic effects on parkinsonian model of rats: focusing on neuroprotective effects of stromal cell-derived factor-1alpha. BMC Neurosci, 11, 52, 2010.

57.

Ishii

, Okada

, Kadoya

, Matsuzaki

, Shimazaki

, and Okano

Stromal cell-secreted factors promote the survival of embryonic stem cell-derived early neural stem/progenitor cells via the activation of MAPK and PI3K-Akt pathways. J Neurosci Res, 88, 722, 2010.

58.

Ribeiro

C.A.

, Salgado

A.J.

, Fraga

J.S.

, Silva

N.A.

, Reis

R.L.

, and Sousa

The secretome of bone marrow mesenchymal stem cells-conditioned media varies with time and drives a distinct effect on mature neurons and glial cells (primary cultures). J Tissue Eng Regen Med, 5, 668, 2011.

59.

Salgado

A.J.

, Fraga

J.S.

, Mesquita

A.R.

, Neves

N.M.

, Reis

R.L.

, and Sousa

Role of human umbilical cord mesenchymal progenitors conditioned media in neuronal/glial cell densities, viability, and proliferation. Stem Cells Dev, 19, 1067, 2010.

60.

Wei

, Fraser

J.L.

, Lu

Z.Y.

, Hu

, and Yu

S.P.

Transplantation of hypoxia preconditioned bone marrow mesenchymal stem cells enhances angiogenesis and neurogenesis after cerebral ischemia in rats. Neurobiol Dis, 46, 635, 2012.

61.

Stacpoole

S.R.

, Bilican

, Webber

D.J.

, Luzhynskaya

, He

X.L.

, Compston

, Karadottir

, Franklin

R.J.

, and Chandran

Efficient derivation of NPCs, spinal motor neurons and midbrain dopaminergic neurons from hESCs at 3% oxygen. Nat Protoc, 6, 1229, 2011.

62.

Fernandes

T.G.

, Diogo

M.M.

, Fernandes-Platzgummer

, da Silva

C.L.

, and Cabral

J.M.

Different stages of pluripotency determine distinct patterns of proliferation, metabolism, and lineage commitment of embryonic stem cells under hypoxia. Stem Cell Res, 5, 76, 2010.

63.

Francis

K.R.

, and Wei

Human embryonic stem cell neural differentiation and enhanced cell survival promoted by hypoxic preconditioning. Cell Death Dis, 1, e22, 2010.

64.

Siddiq

, Aminova

L.R.

, Troy

C.M.

, Suh

, Messer

, Semenza

G.L.

, and Ratan

R.R.

Selective inhibition of hypoxia-inducible factor (HIF) prolyl-hydroxylase 1 mediates neuroprotection against normoxic oxidative death via HIF- and CREB-independent pathways. J Neurosci, 29, 8828, 2009.

65.

Santilli

, Lamorte

, Carlessi

, Ferrari

, Rota Nodari

, Binda

, Delia

, Vescovi

A.L.

, and De Filippis

Mild hypoxia enhances proliferation and multipotency of human neural stem cells. PLoS One, 5, e8575, 2010.

66.

Lozito

T.P.

, and Tuan

R.S.

Mesenchymal stem cells inhibit both endogenous and exogenous MMPs via secreted TIMPs. J Cell Physiol, 226, 385, 2011.

67.

Ylostalo

J.H.

, Bartosh

T.J.

, Coble

, and Prockop

D.J.

Human mesenchymal stem/stromal cells cultured as spheroids are self-activated to produce prostaglandin E2 that directs stimulated macrophages into an anti-inflammatory phenotype. Stem Cells, 30, 2283, 2012.

68.

Sart

, Tsai

A.-C.

, Li

, and Ma

Three-dimensional aggregates of mesenchymal stem cells: cellular mechanisms, biological properties, and applications. Tissue Eng Part B Rev, 2013. DOI: 10.1089/ten.teb.2013.0537.

69.

Mallanna

S.K.

, and Rizzino

Systems biology provides new insights into the molecular mechanisms that control the fate of embryonic stem cells. J Cell Physiol, 227, 27, 2012.