Functional Comparison of Neuronal Cells Differentiated from Human Induced Pluripotent Stem Cell–Derived Neural Stem Cells under Different Oxygen and Medium Conditions

Abstract

Because neurons are difficult to obtain from humans, generating functional neurons from human induced pluripotent stem cells (hiPSCs) is important for establishing physiological or disease-relevant screening systems for drug discovery. To examine the culture conditions leading to efficient differentiation of functional neural cells, we investigated the effects of oxygen stress (2% or 20% O₂) and differentiation medium (DMEM/F12:Neurobasal-based [DN] or commercial [PhoenixSongs Biologicals; PS]) on the expression of genes related to neural differentiation, glutamate receptor function, and the formation of networks of neurons differentiated from hiPSCs (201B7) via long-term self-renewing neuroepithelial-like stem (lt-NES) cells. Expression of genes related to neural differentiation occurred more quickly in PS and/or 2% O₂ than in DN and/or 20% O₂, resulting in high responsiveness of neural cells to glutamate, N-methyl-d-aspartate (NMDA), α-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA), and (S)-3,5-dihydroxyphenylglycine (an agonist for mGluR_1/5), as revealed by calcium imaging assays. NMDA receptors, AMPA receptors, mGluR₁, and mGluR₅ were functionally validated by using the specific antagonists MK-801, NBQX, JNJ16259685, and 2-methyl-6-(phenylethynyl)-pyridine, respectively. Multielectrode array analysis showed that spontaneous firing occurred earlier in cells cultured in 2% O₂ than in 20% O₂. Optimization of O₂ tension and culture medium for neural differentiation of hiPSCs can efficiently generate physiologically relevant cells for screening systems.

Keywords

human induced pluripotent stem cell–derived neural stem cells neural differentiation oxygen tension medium functional neurons

Introduction

The advent of induced pluripotent stem cell (iPSC) technology has changed the study of human diseases. Drug development using cells differentiated from human iPSCs (hiPSCs), including patient-derived and genome-edited hiPSCs,^1,2 has recently become feasible. Neurons are among the most sought-after cell types for target differentiated cells because they are not easily accessible from healthy donors or patients with neurological disorders. Neurons are electrically excitable cells that process and transmit electrical and chemical signals, forming networks connected via synapses. These characteristics of neurons are mainly mediated by voltage-gated ion channels and synaptic receptors, such as glutamate receptors. When creating neurons from hiPSCs for functional assays, it is crucial to ensure that the differentiated cells are both morphologically and functionally representative of neurons. This requires both optimizing the differentiation conditions to produce functional neurons and examining to what degree and how early these channels and receptors are functionally expressed.

Oxygen (O₂) is an important physiological factor for cell function. An O₂ level of 20% has generally been used in conventional in vitro cell culture systems. In mammalian brains, however, interstitial tissue O₂ levels range from ~1% to 5%.³ Recent studies have revealed that O₂ plays crucial roles in the proliferation and differentiation of stem cells. For example, hypoxic conditions (2% O₂) enhance the synthesis of cartilage proteins (collagen I and II and glycosaminoglycans) and mechanical functionality (tensile and compressive properties) in chondrogenic differentiation from human embryonic stem cells (ESCs)⁴; 3% O₂ inhibits maturation of differentiating cardiomyocytes derived from mouse iPSCs⁵; and 1% O₂ hampers adipogenic, osteogenic, and chondrogenic differentiation from human bone marrow mesenchymal stem cells.⁶

Hypoxic conditions also affect the differentiation of neuron subtypes from neural stem cells (NSCs) and neural progenitor cells (NPCs). Several papers have shown that hypoxic conditions promote differentiation of dopaminergic neurons from NSCs/NPCs—either collected from rat embryonic mesencephalon³ or human fetal mesencephalon and striatum,⁷ or derived from mouse ESCs.⁸ Horie et al.⁹ reported that anoxia/hypoxia (0%–3% O₂) in mouse NSCs collected from the ganglionic eminence shifted the neuron subtype of differentiation from γ-aminobutyric acid (GABA)–positive neurons to glutamate-positive neurons. Xie et al.¹⁰ published findings that hypoxia (2% O₂) shifted differentiation to glial fibrillary acidic protein (GFAP)–positive glia from βIII-tubulin-positive neurons during neural differentiation from human ESC- and hiPSC-derived NPCs.

In addition to O₂, compounds and factors used to supplement culture media can affect neural differentiation and maturation. For example, dibutyl cyclic adenosine monophosphate (db-cAMP) increased the βIII-tubulin-positive cell population in NSCs/NPCs collected from the subventricular zone of adult rats.¹¹ Lepski et al.¹² showed that forskolin or 3-isobutyl-1-methylxanthine treatment increased the number of microtubule-associated protein 2 (MAP2)–positive cells and facilitated electrophysiological neuronal maturation of NSCs obtained from the subventricular zone of rat embryos. Ascorbic acid increases the percentage of βIII-tubulin-, tyrosine hydroxylase–, and serotonin-positive cells during neuronal differentiation from mouse ESCs.¹³ Indeed, cAMP derivatives and ascorbic acid are used as supplements in some neural differentiation media to improve differentiation.¹⁴

These results of the effects of O₂ and other factors on neuronal differentiation were mainly determined with immunocytochemical methods. However, there have been few studies on how hypoxia affects the function of the differentiated neuronal cells. To establish physiological or disease-relevant assay systems using human neuronal cells in vitro, it is important that they not only exhibit neuronal morphology and express immunological markers, but also function normally. Therefore, in this study we compared the expression of genes involved in neural differentiation, the functions of ionotropic and metabotropic glutamate receptors, and the formation of networks of neurons differentiated from hiPSC-derived NSCs. We generated long-term self-renewing neuroepithelial-like stem (lt-NES) cells as an hiPSC-derived source of NSCs, because lt-NES cells have unique characteristics, including continuous expandability, ease of culturing (because they are adherent cells), and stable neuronal and glial differentiation potency.¹⁵ These characteristics are suitable for the construction of systems for screening compounds during drug discovery. In this study, we differentiated lt-NES cells in different media (one prepared in house and one commercial) and O₂ conditions (20% and 2% O₂) to investigate the effects of these conditions on functional differentiation and maturation of hiPSC-derived neural cells in order to establish more efficient and robust screening systems.

Materials and Methods

Reagents

N-Methyl-d-aspartate (NMDA), (+)-MK-801 hydrogen maleate, (S)-3,5-dihydroxyphenylglycine hydrate (DHPG), 2-methyl-6-(phenylethynyl)-pyridine hydrochloride (MPEP), NBQX hydrate, and GABA were purchased from Sigma-Aldrich (St. Louis, MO). α-Amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA), JNJ16259685 [(3,4-dihydro-2H-pyrano[2,3-b]quinolin-7-yl)-(cis-4-methoxycyclohexyl)-methanone], and VU0285683 (3-fluoro-5-[3-(2-pyridinyl)-1,2,4-oxadiazol-5-yl]benzonitrile) were obtained from Tocris Bioscience (Bristol, UK). Sodium hydrogen l(+)-glutamate monohydrate and tetrodotoxin (TTX) were obtained from Wako (Osaka, Japan). ADX47273 [(S)-(4-fluorophenyl)-{3-[3-(4-fluorophenyl)-[1,2,4]-oxadiazol-5-yl]-piperidin-1-yl}-methanone] was synthesized by Eisai Europe Ltd. (Hatfield, UK). Dimethyl sulfoxide (Wako) was used as the solvent for DHPG, MK-801, NBQX, MPEP, JNJ16259685, VU0285683, and ADX47273; distilled water (Thermo Fisher Scientific, Waltham, MA) was used for NMDA, AMPA, GABA, glutamate, and TTX.

Cell Line

We used an hiPSC line, 201B7,¹⁶ which was obtained at passage 25 (P25) from Kyoto University, Kyoto, Japan. In addition, we utilized an hiPSC line, P11025 at P13 introduced from Cellectis AB (now Takara Bio Europe, Gotebörg, Sweden). P11025 iPSCs were generated from primary fibroblasts of a healthy Caucasian male, reprogrammed by using episomal vector combined with SOX2, KLF4, OCT4, LIN28, L-MYC, and GLIS-1. These hiPSCs were maintained in Essential 8 (Thermo Fisher Scientific) at 37 °C in a humidified atmosphere of 20% O₂. This study was approved by the Eisai Research Ethics Committee (approval numbers 2013-0013, 2013-193, 2014-193, 2014-0211, 2015-193, 2015-0211, 2016-193, and 2016-0211).

Generation of lt-NES Cells

We generated human lt-NES cells from 201B7 and P11025 cells according to the methods of Falk et al.¹⁵ The generation procedures were as follows: 1.5 × 10⁶ cells were plated in Aggrewell 800 plates (STEMCELL Technologies, Vancouver, Canada) in STEMdiff Neural Induction medium (STEMCELL Technologies). Neural aggregates were harvested on day 5 and cultured in wells coated with poly-l-ornithine (Sigma-Aldrich) and laminin (Sigma-Aldrich). On day 12, neural rosettes were isolated by using STEMdiff Neural Rosette selection reagent (STEMCELL Technologies) and cultured in suspension for 3 days. Subsequently, the neural aggregates were transferred to poly-l-ornithine/laminin-coated dishes and maintained in lt-NES medium (DMEM/F12 [Wako] supplemented with glucose [26.5 mM; Sigma-Aldrich], 1× N2 supplement [Thermo Fisher Scientific], 0.05× B27 supplement [Thermo Fisher Scientific], 1× penicillin/streptomycin [Thermo Fisher Scientific], 10 ng/mL basic fibroblast growth factor [PeproTech, Rocky Hill, NJ], and 10 ng/mL epidermal growth factor [PeproTech]) at 37 °C in a humidified atmosphere of 20% O₂. The medium was changed every 2 days.

Neural Differentiation of lt-NES Cells

The following two neural differentiation media were investigated: DMEM/F12:Neurobasal-based medium^17-20 (hereafter referred to as DN medium or simply DN) and a commercial neural differentiation medium (Neural Differentiation Medium for pan-neural differentiation, PhoenixSongs Biologicals, Branford, CT; hereafter referred to as PS medium or simply PS). The DN medium contained a 1:1 mixture of DMEM/F12 and Neurobasal (Thermo Fisher Scientific) supplemented with 0.5× N2 supplement, 0.5× B27 supplement, brain-derived neurotropic factor (BDNF, 20 ng/mL; Wako), glial cell–derived neurotropic factor (GDNF, 20 ng/mL; Wako), neurotrophin-3 (10 ng/mL; Wako), and laminin (1 μg/mL). PS was supplemented with supplement cocktail (base supplement, 20 ng/mL BDNF, 20 ng/mL GDNF, and 1 μg/mL laminin). Lt-NES cells were differentiated in DN or PS under 20% O₂ or 2% O₂. After neural differentiation, cells were used for immunocytochemical examination, microarray, RT-qPCR, calcium flux assays, and multielectrode array (MEA; see Supplementary Materials and Methods).

Results

Immunocytochemistry

Lt-NES cells were positive for the neural progenitor markers Nestin and SOX2 (Suppl. Fig. S1A). In addition, we observed immunoreactivity for the rosette markers ZO-1 and PLZF¹⁵ (Suppl. Fig. S1B). After 4 weeks of neural differentiation, MAP2⁺ cells (neurons) were produced in all four conditions (DN/20% O₂, DN/2% O₂, PS/20% O₂, and PS/2% O₂) ( Fig. 1A ). Few GFAP⁺ (glial) cells were found in DN regardless of the O₂ condition. In contrast, GFAP⁺ cells were observed in PS under both 20% and 2% O₂, although MAP2⁺ cells dominated in both conditions (~70%–80% of all cells, as determined by the imaging analysis).

Figure 1.

(A) Neural cells differentiated from hiPSC-derived lt-NES cells under different medium (DN or PS) and O₂ (2% or 20%) conditions for 4 weeks. Expression of MAP2 indicates neurons, GFAP indicates astrocytes, and Hoechst labels cell bodies. Bar, 100 μm. (B) Heat map and dendrograms of gene expression related to neural development and function in lt-NES cells and lt-NES cell–derived neural cells differentiated in the four conditions shown in A. Differentiation time– and condition-dependent gene expression changes were observed. The genes used in this heat map are described in the Supplementary Materials and Methods. Max, Average, and Min indicate maximal, average and minimal expression.

Microarray Results

The expression of genes related to neural differentiation differed by medium, O₂ tension, and time in culture (1, 2, 3, or 4 weeks) ( Fig. 1B ). The dendrogram shows five main clusters, suggesting that neural differentiation progressed more rapidly in 2% O₂ than in 20% O₂, and in PS than in DN. But there were no increases of gene expression related to oligodendrocytes and oligodendrocyte progenitor cells.

RT-qPCR Analysis

To confirm the microarray results, we quantified gene expression with RT-qRCR. Changes in representative genes are illustrated in Figure 2 . Detailed values and statistical results from two-way factorial analysis of variance (ANOVA) and Tukey’s test are shown in Supplementary Table S2.

Figure 2.

Expression levels of representative genes involved in neural differentiation and maturation in hiPSC-derived lt-NES cells and lt-NES cell–derived neural cells differentiated in PS or DN under 2% or 20% O₂ for 4 weeks. (A) MAP2. (B) GFAP. (C) GRIA1 (glutamate receptor, ionotropic, AMPA 1). (D) GRIN1 (glutamate receptor, ionotropic, NMDA 1). (E) GRM1 (glutamate receptor, metabotropic 1). (F) GRM5 (glutamate receptor, metabotropic 5). (G) GABRA1 (GABA A receptor, subunit α1). (H) SYP (synaptophysin). (I) CACNA1A (calcium channel, voltage dependent, P/Q type, α_1A subunit). (J) CACNA1B (calcium channel, voltage dependent, N type, α_1B subunit). (K) SLC12A2 (solute carrier family 12 member 2; sodium-potassium-chloride transporter). (L) SLC12A5 (solute carrier family 12 member 5; potassium-chloride transporter). Values are means ± SEM relative to those of DM/20% O₂ conditions (set to 1.00). n = 3. Results were evaluated by two-way factorial ANOVA, followed by Tukey’s post hoc test. *, #, and $ indicate significant difference vs DN/20% O₂, DN/2% O₂, and PS/20% O₂, respectively. One, two, and three symbols indicate p < 0.05, p < 0.01, and p < 0.001, respectively. Two-way ANOVA results are indicated in Supplementary Table S2.

The expression of the neuronal marker gene, MAP2, increased in 2% O₂ ( Fig. 2A ). The medium had a clear effect on the astrocyte marker, GFAP, although O₂ effects were also observed; the highest gene expression was found in PS/20% O₂ ( Fig. 2B ). Several genes related to AMPA receptors (GRIA1) ( Fig. 2C ) and kainate receptors (GRIK3 and GRIK5) showed increased expression in 2% O₂, some of which were also affected positively by PS, resulting in additive increases of gene expression. The combination of 2% O₂ and PS produced synergistic increases for NMDA receptor-related genes (GRIN1 [ Fig. 2D ], GRIN2A, and GRIN3A). For metabotropic glutamate receptors, GRM1 expression was elevated by hypoxia ( Fig. 2E ), and GRM5 expression was amplified by the combination of 2% O₂ and PS ( Fig. 2F ).

Additive increases of gene expression by the combination of 2% O₂ and PS were seen for GABA receptor genes (GABRA1 [ Fig. 2G ], GABRA2, GABRB2, GABRB3, and GABRG2) and synapse-related proteins (SYP [ Fig. 2H ], SYN1, and DLG4). Gene expression of the P/Q- and N-type calcium channels (CACNA1A [ Fig. 2I ] and CACNA1B [ Fig. 2J ], respectively) was increased synergistically by the combination of 2% O₂ and PS. Expression of SLC12A2 and SLC12A5, which encode a sodium-potassium-chloride transporter and a potassium-chloride transporter, respectively, showed contrasting changes: SLC12A2 expression decreased in 2% O₂ ( Fig. 2K ), whereas SLC12A5 expression synergistically increased under the combined influence of 2% O₂ and PS ( Fig. 2L ).

Calcium Flux Imaging for Receptor Function Assays

Glutamate application increased calcium signals in a concentration-dependent manner ( Fig. 3A ). Large responses were observed in PS/2% O₂ and PS/20% O₂. In DN/2% O₂, glutamate-induced calcium flux was also observed, but the maximum response was only ~30% of that in PS/2% O₂. The calcium rise was smallest in DN/20% O₂. The maximum responses and EC₅₀ values are summarized in Supplementary Table S3. The same effects of the media were observed in neural cells differentiated from P11025-derived lt-NES cells (Suppl. Fig. S2).

Figure 3.

Calcium rise assays of ionotropic and metabotropic glutamate receptor function in human neural cells differentiated from lt-NES cells. Lt-NES cells were differentiated in DN or commercial neural differentiation medium (PS) under 2% or 20% O₂ for 4 weeks. Concentration–response curves were drawn based on the means of three independent experiments (n = 3 in each experiment). Values are expressed as means ± SEM. Cells were stimulated with (A) glutamate, (B) NMDA, (C) AMPA, or (D) (S)-3,5-dihydroxyphenylglycine (DHPG, an mGluR₁ and mGluR₅ agonist). Plots show values relative to the maximum (max.) values for PS/2% O₂, which were set to 100%. Maximal activities and EC₅₀ values are shown in Supplementary Table S3.

To further characterize the responses of lt-NES cell–derived neuronal cells to glutamate, we treated the cells with specific agonists. Response patterns to NMDA ( Fig. 3B ) and AMPA ( Fig. 3C ) in these culture conditions were similar to those with glutamate treatment. In general, the induced calcium flux was greatest in PS/2% O₂ and PS/20% O₂, followed by DN/2% O₂, and then DN/20% O₂. The NMDA-induced signal was particularly small in DN/20% O₂, corresponding to only ~10% of that in PS/2% O₂. The calcium flux induced by NMDA and AMPA (30 μM in both cases) was completely inhibited by specific antagonists for NMDA receptors (MK-801, Fig. 4A ) and AMPA receptors (NBQX, Fig. 4B ). IC₅₀ values for MK-801 and NBQX were 0.53 μM and 1.20 μM, respectively, when using neural cells differentiated in PS/2% O₂. In the case of PS/20% O₂, IC₅₀ values for MK-801 and NBQX were 0.50 μM and 0.80 μM, respectively.

Figure 4.

Validation of ionotropic and metabotropic glutamate receptor function in neural cells differentiated from lt-NES cells by using antagonists against these receptors. (A) Concentration-dependent inhibition by MK-801, an NMDA receptor antagonist, of the calcium rise evoked by 30 μM NMDA. The IC₅₀ value was 0.53 μM. (B) Concentration-dependent inhibition by NBQX, an AMPA receptor antagonist, of the calcium rise evoked by 30 μM AMPA. The IC₅₀ value was calculated as 1.2 μM. (C) Concentration-dependent inhibition by MPEP, an mGluR₅ antagonist, of the calcium rise evoked by 50 μM DHPG. About 50% of the DHPG-evoked calcium rise remained at the maximum inhibition. (D) Concentration-dependent inhibition by JNJ16259685, an mGluR₁ antagonist, of the calcium rise evoked by 50 μM DHPG. About 50% of the DHPG-evoked calcium rise remained at the maximum inhibition. (E,F) Effects of cotreatment with JNJ16259685 and MPEP on the calcium rise evoked by 50 μM DHPG. (E) Addition of 0.1 μM JNJ16259685 to samples treated with 50 μM DHPG and MPEP. (F) Addition of 10 μM MPEP to samples treated with 50 μM DHPG and JNJ16259685. The calcium rise evoked by 50 μM DHPG was almost completely inhibited by cotreatment with MPEP and JNJ16259685. (G) Concentration-dependent calcium rise evoked by ADX47273, a positive allosteric modulator of mGluR₅, in the presence of 10 μM DHPG. (H) Concentration-dependent inhibition by VU0285683, a negative allosteric modulator of mGluR₅, of the calcium rise evoked by 50 μM DHPG. Neural cells differentiated for 4 weeks in 2% O₂ and PS neural differentiation medium were used. Values are expressed as means ± SEM. n = 3–4.

We also examined the responsiveness of metabotropic glutamate receptors by using DHPG, an agonist for mGluR_1/5. Neural cells cultured in the four conditions responded to DHPG in a concentration-dependent fashion ( Fig. 3D ), but EC₅₀ values of DHPG in DN/2% O₂ and DN/20% O₂ were ~3-fold larger than those in PS/2% O₂ and PS/20% O₂ (Suppl. Table S3). The maximum activity of the response curve of PS/20% O₂ was ~200% of that of PS/2% O₂. However, the responsiveness under PS/20% O₂ varied greatly. Next, we investigated the receptors responsible for the DHPG-induced calcium flux: mGluR₁, mGluR₅, or both. To test this, we used specific antagonists, MPEP for mGluR₅ and JNJ16259685 for mGluR₁. MPEP alone or JNJ16259685 alone reduced DHPG-induced signals to ~50% of baseline in a concentration-dependent fashion ( Fig. 4C and D for MPEP and JNJ16259685, respectively). The DHPG-induced activity remaining after treatment with 0.1 μM JNJ16259685 or 10 μM MPEP was inhibited almost completely by further treatment with MPEP ( Fig. 4E ) and JNJ16259685 ( Fig. 4F ), respectively, in a concentration-dependent manner. We further investigated mGluR₅ function by using ADX47273 (an mGluR₅ positive allosteric modulator [PAM]) and VU0285683 (an mGluR₅ negative allosteric modulator [NAM]). ADX47273 did not act as an agonist in the absence of DHPG (data not shown) but produced a concentration-dependent calcium flux in the presence of 10 μM DHPG. ADX47273 at 1 and 3 μM reached ~90% of the activity of 300 μM DHPG at maximum ( Fig. 4G ). VU0285683 reduced the calcium flux evoked by 50 μM DHPG in a concentration-dependent manner to ~50% of the activity of 50 μM DHPG ( Fig. 4H ).

MEA Recordings

We recorded the neural activity of lt-NES cell–derived neurons differentiated in PS under 20% or 2% O₂ for 7 weeks. More spikes and bursts were visible in the raster plot at week 6 than at week 4, and they occurred more frequently under 2% O₂ than under 20% O₂ conditions ( Fig. 5A ). Synchrony of spikes was observed in lt-NES cell–derived neurons differentiated under 2% O₂ at week 6, indicating formation of a neuronal network. Spontaneous spikes occurred in the third week of differentiation under both conditions ( Fig. 5B ). The number of electrodes detecting spikes and bursts depended on the O₂ condition (main effects: F_1,4 = 7.912, p = 0.0482 for number of electrodes detecting spikes; F_1,4 = 8.696, p = 0.0420 for those detecting bursts) and differentiation time (main effects: F_3,12 = 10.00, p = 0.0014 for numbers of electrodes detecting spikes; F_3,12 = 5.962, p = 0.0099 for those detecting bursts; Suppl. Table S4). In addition, synergistic effects were observed between the O₂ condition and differentiation time (interaction: F_3,12 = 4.452, p = 0.0254 for numbers of electrodes detecting spikes; F_3,12 = 3.871, p = 0.0379 for those detecting bursts; Suppl. Table S4). Significantly higher numbers of electrodes detecting spikes ( Fig. 5B ) and bursts ( Fig. 5C ) were observed in differentiation weeks 6 (p < 0.01 in both cases) and 7 (p < 0.05 in both cases). The number of spike-detecting electrodes (active electrodes) increased as a function of differentiation time (main effect: F_3,12 = 3.725, p = 0.0421; Fig. 5D and Suppl. Table S4). The O₂ condition had a marginally significant effect on burst duration (main effect: F_1,4 = 6.003, p = 0.0704; Fig. 5E and Suppl. Table S4).

Figure 5.

MEA analysis of spontaneous firing in neurons differentiated from lt-NES cells. (A) Representative raster plots of spikes detected from each of the 64 electrodes (channels [Ch.] 1–64) in neurons cultured in commercial neural differentiation medium (PS) under 20% and 2% O₂ for 4 and 6 weeks. Time-dependent changes in the (B) number of electrodes detecting spikes (active electrodes), (C) number of electrodes detecting bursts, (D) number of spikes per active electrode, and (E) burst duration. Values are expressed as means ± SEM. n = 3. Data were analyzed by two-way repeated-measures ANOVA (Suppl. Table S4), followed by Bonferroni’s test. *, p < 0.05; **, p < 0.01.

To investigate whether inhibitory effects of GABA were reproduced in lt-NES neurons differentiated in PS/2% O₂, we treated neurons with GABA and measured firing numbers with the MEA. After treatment with vehicle, we observed bursts in addition to many spikes ( Fig. 6A ). GABA treatment (0.2, 1, 5, and 20 μM) reduced spike counts. Although bursts were still seen with 0.2 and 1 μM GABA, burst duration and spike counts per burst decreased with increasing GABA concentration, with bursts disappearing at 5 μM GABA. The remaining spikes were almost completely abolished by the addition of 100 nM TTX. GABA treatment significantly reduced spike counts in a concentration-dependent manner from 0.2 μM (p < 0.001 vs the vehicle treatment [0.1% distilled water]; Fig. 6B ).

Figure 6.

Effects of GABA on spontaneous firing in neurons differentiated from lt-NES cells. Lt-NES cells were differentiated in commercial neural differentiation medium (PS) under 2% O₂ for 7 weeks. After baseline levels of spontaneous firing were measured, the neurons were treated with GABA at increasing concentrations (0.2, 1, 5, and 20 μM), and then with TTX at 100 nM. (A) Representative firing patterns of lt-NES neurons treated with GABA or TTX. The traces were taken from the same electrode of a MEA chip. Bursts (black triangles) were also observed in the cells treated with vehicle or with 0.2 μM GABA. (B) GABA significantly reduced spike counts (shown as percent of vehicle treatment). Data were analyzed by one-way repeated-measures ANOVA, followed by Dunnett’s multiple comparison test. n = 3. ***, p < 0.001 vs the vehicle treatment (0.1% distilled water).

Discussion

Immunocytochemical and gene expression analyses revealed that both the medium and the O₂ condition profoundly affected the direction of differentiation of neural cells from hiPSC-derived lt-NES cells and the expression levels of genes involved in neural differentiation and maturation. Lt-NES cells differentiated into neurons in both DN and PS media, but more GFAP⁺ cells were generated in PS medium than in DN medium. PS medium also supported higher expression levels of many genes related to receptors (e.g., AMPA, NMDA, kainate, and GABA receptors) and synapse formation (e.g., synapsin and PSD-95) than did DN medium, and expression of these genes under 2% O₂ was higher than under 20% O₂. Clustering analysis of gene expression in lt-NES cells and lt-NES cell–derived neural cells differentiated under these different conditions showed that PS medium and hypoxia each promoted neural differentiation separately and in an additive or synergistic manner.

SLC12A5 and SLC12A2 encode a potassium-chloride transporter (KCC2) and a sodium-potassium-chloride transporter (NKCC1), respectively. NKCC1 is expressed in neurons and nonneuronal cells, whereas KCC2 is neuron specific and is particularly expressed in mature neurons, where it contributes to mature dendritic spine formation and functional excitatory synapses.²¹ A developmental switch from NKCC1 to KCC2 occurs during neuronal maturation,²² and we observed that 2% O₂ reduced SLC12A2 expression and increased SLC12A5 expression, further supporting the contribution of hypoxia to neuronal maturation and network formation (e.g., more frequent spontaneous synchrony of firings). Moreover, the neuronal maturation process was evidenced by the inhibition of spontaneous firing in lt-NES neurons by GABA, indicating that the developmental shift from depolarization to hyperpolarization in response to GABA had occurred.²²

Of the four conditions tested, PS/2% O₂ generated the most mature neurons. Recently, Sun et al.²³ demonstrated that culturing mouse NPCs in 5% O₂ enhanced their neuronal differentiation and survival relative to when mouse NPCs were cultured in 20% O₂. In 5% O₂, mitochondrial oxidative metabolism was high, decreasing glycolysis and reducing the concentration of reactive oxygen species (ROS). It was reported that ROS levels were increased in neural differentiation in 20% O₂.^24,25 Therefore, inhibition of differentiation-induced ROS by hypoxia may be beneficial for neural differentiation of NSCs. The enhanced neuronal differentiation and function under 2% O₂ observed in our study are consistent with these findings.

In calcium imaging assays, glutamate, NMDA, AMPA, and DHPG all produced concentration-dependent calcium rises in neural cells differentiated from human lt-NES cells in all conditions tested. The activities of NMDA, AMPA, and DHPG were inhibited by treatment with the specific antagonists MK-801, NBQX, and JNJ16259685 and MPEP in a concentration-dependent fashion. Thus, NMDA and AMPA receptors, as well as mGluR₁ and mGluR₅, were pharmacologically functional in these cells. The maximum calcium signal evoked by glutamate, NMDA, or AMPA stimulation was highest in the order of PS/2% O₂ ≥ PS/20% O₂ > DN/2% O₂ > DN/20% O₂. For DHPG treatment, the maximum activity in PS/20% O₂ was about twofold that in PS/2% O₂, but this difference was not statistically significant because of the large variation in the PS/20% O₂ samples. The order of EC₅₀ values was similar to that of the maximum activities. The maximum activities are presumably affected by the functional maturation of neural cells and the number of living cells on which assays are performed. Fewer cells were present in DN than in PS medium, which may have resulted in lower maximum activities. However, the degree of responsiveness to glutamate receptor agonists corresponded to the expression level of genes related to the receptors, suggesting that the responsiveness was not merely an artifact of cell number.

mGluR₅ was functional in human neural cells derived from 201B7 hiPSC-derived lt-NES cells. This receptor has been a therapeutic target molecule for several disorders of the central nervous system, such as schizophrenia and cognitive deficits associated with drug abuse. PAMs of mGluR₅ have been developed because of increased receptor subtype selectivity and reduced potential for receptor sensitization.^26,27 The NMDA receptor and mGluR₅ are coupled physiologically and biochemically.²⁸ Activation of mGluR₅ leads to functional potentiation of the NMDA receptor, which is expected to alleviate dysfunction of glutamatergic receptors.^28-30 To examine whether candidate PAMs of mGluR₅ actually enhance neuronal activity by potentiating NMDA receptors in human neurons, assays that use cell lines that overexpress human mGluR₅ or human NMDA receptor, or both, and human astrocytes are inappropriate, because these cells cannot provide readouts reflecting neuronal function, such as electric signals. Human neurons expressing both functional endogenous NMDA receptors and mGluR₅ will be indispensable as a robust and relevant evaluation model, but there have been few reports regarding such neurons derived from human pluripotent stem cells. In the present study, we observed that both mGluR₅ and NMDA receptors were functional in 201B7-derived human neural cells, particularly those cultured in PS/2% O₂ or PS/20% O₂. These cells may therefore be useful for basic studies and drug development that target mGluR₅. Moreover, we detected mGluR₅ NAM activity using differentiated lt-NES neural cells. mGluR₅ NAMs may be useful for treating a variety of conditions, including pain, anxiety, fragile X syndrome, autism, and depression.³¹ Therefore, lt-NES neural cells will also be useful for screening compounds that possess mGluR₅ NAM activity.

We found that the medium affected astrocyte differentiation: PS medium induced more GFAP⁺ cells and higher GFAP gene expression than did DN medium. Together with the neuronal differentiation and maturation described above, these results suggest that PS medium contains factors that induce the differentiation of both neurons and astrocytes. Although the detailed ingredients and compositions of PS medium have not been published, PhoenixSongs Biologicals has provided information that it contains db-cAMP and ascorbic acid. Because DN medium does not contain either compound, both db-cAMP and ascorbic acid could play important roles in the acceleration of neural differentiation of lt-NES cells. Novel neural differentiation media have been developed to improve the quality of neurons derived from human pluripotent stem cells. Under these circumstances, Bardy et al.¹⁴ recently reported that classic basal media (DMEM/F12 and Neurobasal) and serum impair electrophysiological activity and synaptic function of neurons derived from hiPSCs. This prompted them to design a novel medium (BrainPhys containing a serum-free supplement), in which concentrations of inorganic salts, neuroactive amino acids, and energy substances were adjusted to overcome these impairments.

Our results show that PS medium has several advantages over DN for screening compounds: PS medium requires fewer lt-NES cells and reduces the differentiation period. The effects of medium (or its components) and O₂ condition on gene expression, functional maturation of glutamate receptors, and neural activity can directly affect signal-to-noise ratios and Z factors in compound screening. Therefore, the conditions under which we cultivate cells may be as important as the type of cells we use to optimize screening procedures. Optimizing the combination of medium and O₂ condition will enable us to establish a novel compound screening system. Such a system could provide functional hiPSC-derived neurons or neurons derived from patient iPSCs or genome-edited iPSCs, which are physiologically relevant and potentially related to specific diseases.

Footnotes

Acknowledgements

We thank Marsha Roach of PhoenixSongs Biologicals for providing us information on the contents of the PS medium.

Supplementary material for this article is available on the Journal of Biomolecular Screening Web site at .

Declaration of Conflicting Interests

The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: H.O. is a scientific consultant for San Bio Co., Ltd. All other authors declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was partially supported by grants from the Program for Intractable Disease Research Utilizing Disease-Specific iPS Cells funded by the Japan Science and Technology Agency (JST)/Japan Agency for Medical Research and Development (A-MED) to H.O., and the Ministry of Health, Labor and Welfare (MHLW) of Japan to H.O. J.K. and H.O. received research funding from Eisai Co., Ltd.

References

Imaizumi

Okano

Modeling Human Neurological Disorders with Induced Pluripotent Stem Cells. J. Neurochem. 2014, 129, 388–399.

Okano

Yamanaka

iPS Cell Technologies: Significance and Applications to CNS Regeneration and Disease Research. Mol. Brain 2014, 7, 22.

Studer

Csete

Lee

S.-H.

. Enhanced Proliferation, Survival, and Dopaminergic Differentiation of CNS Precursors in Lowered Oxygen. J. Neurosci. 2000, 20, 7377–7383.

Koay

E. J.

Athanasiou

K. A.

Hypoxic Chondrogenic Differentiation of Human Embryonic Stem Cells Enhances Cartilage Protein Synthesis and Biomechanical Functionality. Osteoarthr. Cartil. 2008, 16, 1450–1456.

Medley

T. L.

Furtado

Lam

N. T.

. Effect of Oxygen on Cardiac Differentiation in Mouse iPS Cells: Role of Hypoxia Inducible Factor-1 and Wnt/Beta-Catenin Signaling. PLoS One 2013, 8, e80280.

Cicione

Muiños-López

Hermida-Gómez

. Effects of Severe Hypoxia on Bone Marrow Mesenchymal Stem Cells Differentiation Potential. Stem Cells Int. 2013, 2013, 232896.

Liu

Tian

Yin

. Generation of Dopaminergic Neurons from Human Fetal Mesencephalic Progenitors after Co-Culture with Striatal-Conditioned Media and Exposure to Lowered Oxygen. Brain Res. Bull. 2009, 80, 62–68.

Kim

T.-S.

Misumi

Jung

C.-G.

. Increase in Dopaminergic Neurons from Mouse Embryonic Stem Cell-Derived Neural Progenitor/Stem Cells Is Mediated by Hypoxia Inducible Factor-1α. J. Neurosci. Res. 2008, 86, 2353–2362.

Horie

Moriya

. Effects of Oxygen Concentration on the Proliferation and Differentiation of Mouse Neural Stem Cells In Vitro. Cell Mol. Neurobiol. 2008, 28, 833–845.

10.

Xie

Zhang

Lin

. Defining the Role of Oxygen Tension in Human Neural Progenitor Fate. Stem Cell Reports 2014, 3, 743–757.

11.

Kim

Zahir

Tator

C. H.

. Effects of Dibutyryl Cyclic-AMP on Survival and Neuronal Differentiation of Neural Stem/Progenitor Cells Transplanted into Spinal Cord Injured Rats. PLoS One 2011, 6, e21744.

12.

Lepski

Jannes

C. E.

Nikkhah

. cAMP Promotes the Differentiation of Neural Progenitor Cells In Vitro via Modulation of Voltage-Gated Calcium Channels. Front. Cell. Neurosci. 2013, 7, 155.

13.

Shin

D.-M.

Ahn

J.-I.

Lee

K.-H.

. Ascorbic Acid Responsive Genes during Neuronal Differentiation of Embryonic Stem Cells. Neuroreport 2004, 15, 1959–1963.

14.

Bardy

van den Hurk

Eames

. Neuronal Medium That Supports Basic Synaptic Functions and Activity of Human Neurons In Vitro. Proc. Natl. Acad. Sci. U.S.A. 2015, 112, E2725–E2734.

15.

Falk

Koch

Kesavan

. Capture of Neuroepithelial-Like Stem Cells from Pluripotent Stem Cells Provides a Versatile System for In Vitro Production of Human Neurons. PLoS One 2012, 7, e29597.

16.

Takahashi

Tanabe

Ohnuki

. Induction of Pluripotent Stem Cells from Adult Human Fibroblasts by Defined Factors. Cell 2007, 131, 861–872.

17.

Koch

Opitz

Steinbeck

J. A.

. A Rosette-Type, Self-Renewing Human ES Cell-Derived Neural Stem Cell with Potential for In Vitro Instruction and Synaptic Integration. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 3225–3230.

18.

Kim

J.-E.

O’Sullivan

M. L.

Sanchez

C. A.

. Investigating Synapse Formation and Function Using Human Pluripotent Stem Cell-Derived Neurons. Proc. Natl. Acad. Sci. U.S.A. 2011, 108, 3005–3010.

19.

Pang

Z. P.

Yang

Vierbuchen

. Induction of Human Neuronal Cells by Defined Transcription Factors. Nature 2011, 476, 220–223.

20.

Ladewig

Mertens

Kesavan

. Small Molecules Enable Highly Efficient Neuronal Conversion of Human Fibroblasts. Nat. Methods 2012, 9, 575–578.

21.

Llano

Smirnov

Soni

. KCC2 Regulates Actin Dynamics in Dendritic Spines via Interaction with β-PIX. J. Cell Biol. 2015, 209, 671–686.

22.

Blaesse

Airaksinen

M. S.

Rivera

. Cation-Chloride Cotransporters and Neuronal Function. Neuron 2009, 61, 820–838.

23.

Sun

Voloboueva

L. A.

Stary

C. M.

. Physiologically Normal 5% O₂ Supports Neuronal Differentiation and Resistance to Inflammatory Injury in Neural Stem Cell Culture. J. Neurosci. Res. 2015, 93, 1703–1712.

24.

Le Belle

J. E.

Orozco

N. M.

Paucar

A. A.

. Proliferative Neural Stem Cells Have High Endogenous ROS Levels That Regulate Self-Renewal and Neurogenesis in a PI3K/Akt-Dependant Manner. Cell Stem Cell 2011, 8, 59–71.

25.

Xavier

J. M.

Morgado

A. L.

Solá

. Mitochondrial Translocation of p53 Modulates Neuronal Fate by Preventing Differentiation-Induced Mitochondrial Stress. Antioxid. Redox Signal. 2014, 21, 1009–1024.

26.

Cleva

R. M.

Olive

M. F.

Positive Allosteric Modulators of Type 5 Metabotropic Glutamate Receptors (mGluR5) and Their Therapeutic Potential for the Treatment of CNS Disorders. Molecules 2011, 16, 2097–2106.

27.

Stauffer

S. R.

Progress toward Positive Allosteric Modulators of the Metabotropic Glutamate Receptor Subtype 5 (mGlu₅). ACS Chem. Neurosci. 2011, 2, 450–470.

28.

Chen

H.-H.

Liao

P.-F.

Chan

M.-H.

mGluR5 Positive Modulators Both Potentiate Activation and Restore Inhibition in NMDA Receptors by PKC Dependent Pathway. J. Biomed. Sci. 2011, 18, 19.

29.

Awad

Hubert

G. W.

Smith

. Activation of Metabotropic Glutamate Receptor 5 Has Direct Excitatory Effects and Potentiates NMDA Receptor Currents in Neurons of the Subthalamic Nucleus. J. Neurosci. 2000, 20, 7871–7879.

30.

Pisani

Gubellini

Bonsi

. Metabotropic Glutamate Receptor 5 Mediates the Potentiation of N-Methyl-D-Aspartate Responses in Medium Spiny Striatal Neurons. Neuroscience 2001, 106, 579–587.

31.

Emmitte

K. A.

mGlu5 Negative Allosteric Modulators: A Patent Review (2010–2012). Expert Opin. Ther. Pat. 2013, 23, 393–408.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.43 MB