We describe a method for differentiating spinal cord human induced pluripotent-derived astrocytes and neurons and their co-culture for electrophysiological recording.
Human pluripotent stem cell-derived astrocytes (hiPSC-A) and neurons (hiPSC-N) provide a powerful tool for modeling Amyotrophic Lateral Sclerosis (ALS) pathophysiology in vitro. Multi-electrode array (MEA) recordings are a means to record electrical field potentials from large populations of neurons and analyze network activity over time. It was previously demonstrated that the presence of hiPSC-A that are differentiated using techniques to promote a spinal cord astrocyte phenotype improved maturation and electrophysiological activity of regionally specific spinal cord hiPSC-motor neurons (MN) when compared to those cultured without hiPSC-A or in the presence of rodent astrocytes. Described here is a method to co-culture spinal cord hiPSC-A with hiPSC-MN and record electrophysiological activity using MEA recordings. While the differentiation protocols described here are particular to astrocytes and neurons that are regionally specific to the spinal cord, the co-culturing platform can be applied to astrocytes and neurons differentiated with techniques specific to other fates, including cortical hiPSC-A and hiPSC-N. These protocols aim to provide an electrophysiological assay to inform about glia-neuron interactions and provide a platform for testing drugs with therapeutic potential in ALS.
Human pluripotent stem cell-derived astrocytes (hiPSC-A) and neurons (hiPSC-N) are powerful tools for modeling Amyotrophic Lateral Sclerosis (ALS) pathophysiology in vitro and provide a translational paradigm for drug discovery strategies1. Researchers have demonstrated that the co-culture of hiPSC-A with hiPSC-N enhances the morphological, molecular, electrophysiological, and pharmacological maturation of both cell types, generating complex neuronal networks and astrocyte-neuron interactions that resemble their in vivo counterparts2,3. Similar co-culture experiments can recapitulate hallmarks of ALS pathobiology such as astrocyte-mediated neurotoxicity4,5 and neuronal hyper-excitability6. Additionally, with advancements in differentiation protocols, human induced pluripotent stem cells (hiPSC) can be differentiated into regionally-specific neural subtypes, including cortical and spinal cord hiPSC-A and hiPSC-N7,8. These strategies provide the potential for modeling cortical and spinal motor neuron pathology in ALS as well as the astrocytic influence on both. However, this requires that there is a reproducible functional assay to determine these effects.
Recently it was shown that multi-electrode array (MEA) recording is particularly suitable for the electrophysiological characterization of neuron-astrocyte co-cultures2. As opposed to single-cell electrophysiological analyses, these high-density electrode arrays passively record extracellular field potentials from large populations of neurons without disrupting culture conditions and preserving cell membranes’ integrity. These platforms are particularly useful for recording the cellular and network activity of cultures over time and in response to pharmacologic manipulation. Finally, when the presence of astrocytes is a culture variable, MEA recordings can provide functional insights into astrocyte-neuron bidirectional interactions2,9.
Presented here is an optimized protocol for the differentiation of hiPSC into spinal cord hiPSC-A and hiPSC-motor neurons (MN) that has been previously validated2. The spinal cord hiPSC-A differentiation protocol consistently results in astrocyte cultures, which are positive for S100 calcium-binding protein B (S100β), glial fibrillary acidic protein (GFAP), and Homeobox B4 (HOXB4) in up to 80%, 50%, and 90% of cells, respectively, indicating a maturing glial and spinal cord specification2,10. The hiPSC-MN differentiation protocol generates neurons that are >90% positive for choline acetyltransferase (ChAT), suggestive of a mature alpha-motor neuron identity2. Additionally, the protocol describes techniques for the generation of hiPSC-A/MN co-cultures previously demonstrated to result in neurons with enhanced morphological complexity by Scholl analysis and immunofluorescent microscopy when compared to neuronal cultures without astrocytes or with rodent astrocytes2. While these descriptions are specific to spinal cord hiPSC-A and hiPSC-N, a unique advantage is that the initial independent culture of astrocytes and neurons followed by steps to co-culture at later time points can be translated to study the effects of neuron-astrocyte interactions from other specific regions as well as disease-specific cells7,8. Finally, the protocol describes how to grow these cultures on MEA plates so that functional activity as a factor of co-culture composition can be studied over time with the ability to manipulate cellular composition as well as culture conditions.
The goal of these protocols is to provide a functional assay to investigate astrocyte-neuron interactions, examine disease-specific changes, and test drugs with therapeutic potential in the field of ALS. Video instructions are provided for the most challenging steps of this protocol.
1. Cell culture media preparation
2. Maintaining and passaging non-confluent human induced Pluripotent Stem Cells (hiPSC)
3. Freezing hiPSC
4. Thawing hiPSC
5. Differentiating hiPSC into spinal cord Neural Progenitor Cells (NPC)
6. Thawing NPC cultures
7. Differentiating spinal cord NPC into motor neurons
8. Differentiating NPC into spinal cord astrocytes
9. Co-culturing MN and spinal cord astrocytes in multi-electrode array plates
10. Multi-electrode array recording
11. Pharmacological assays on multi-electrode array
The spinal cord patterning protocol for the generation of hiPSC-MN and spinal cord hiPSC-A is outlined in Figure 1. In this protocol, hiPSCs are maintained and passaged as non-confluent colonies (Figure 2A). Neurogenesis is initiated (neural induction) through dual SMAD inhibition by the addition of LDN193189 and SB431542, inactivating the bone morphogenetic protein (BMP) and the transforming growth factor-beta (TGF-β) pathways, respectively. A monolayer-based method is used for this step where hiPSC are plated onto an adherent matrix, i.e., basement membrane matrix, to generate a neuroepithelium-like 2D culture. Morphologically, neural induction is marked by the transition from iPSC with large nuclei and round shape to neuroepithelial (stem) cells that are tightly compacted, with a large cytoplasm and cylindrical shape. These cells will divide horizontally as well as vertically, generating a multilayered epithelium (Figure 2B,i).
The preference for a 2D, as opposed to a 3D strategy (i.e., embryoid body-based), relies on literature evidence11,12 suggesting that the former may promote spinal cord patterning by introducing cell-extrinsic environmental cues13. Neuroepithelial cells are then temporally and spatially patterned to generate region-specific, i.e., spinal cord, neural progenitor cells (NPC) (Figure 1). Two key morphogens are used for this purpose: retinoic acid, which determines caudalization, and purmorphamine, a hedgehog signaling agonist, which causes ventral specification. In their absence, the cell-intrinsic regional identity of NPC would be rostral and dorsal14,15,16.
From DIV 15 to DIV 25-DIV 30, the regional identity of these NPC is reinforced by supplementing media with the morphogens. At the same time, the early exposure to neuronal growth factors and a gliogenic cytokine such as ciliary neurotrophic factor (CNTF) generates a mixed population of NPC (Figure 2B,ii). Morphologically these cells are less compacted than neuroepithelial cells, displaying few short processes, arranged in a monolayer after disruption of the columnar epithelium after passaging at DIV 12. At a closer view, some of these cells are elongated, while others display multiple processes, which indicates early commitment toward a glial vs. neuronal fate, respectively.
Neurons will emerge spontaneously from this mixed population unless the gliogenic switch is activated (Figure 1 and Figure 2B,iii). The addition of the Notch pathway inhibitor, Compound E, will enhance lower MN differentiation17. The spinal cord identity of these cells is supported by high levels of ChAT expression (Figure 2C,i). Additionally, it has been previously shown2 that a subpopulation of these motor neurons is also ISL1+.
Astrocyte differentiation is induced by the gliogenic switch through activation of JAK/STAT pathway (Figure 1 and Figure 2B,iv). CNTF and likely other cytokines contained in fetal bovine serum serve this purpose in the proposed protocol. Time is an essential factor as well, since gliogenesis will spontaneously follow after neuronogenesis due to cell-intrinsic clues. After DIV 90, hiPSC-A display a maturing phenotype as indicated by S100β and GFAP expression2,11, while their spinal cord regional identity is supported by >90% expression of HOXB4 (Figure 2C, ii)2,11,14.
The method for co-culturing hiPSC-MN and hiPSC-A is notable for simultaneous plating of these cell subtypes as opposed to techniques where neurons are serially plated on the top of astrocyte cultures. In these simultaneous co-cultures, cells will rearrange spontaneously, with astrocytes creating a feeding layer at the bottom and neurons connecting in networks at the top (Figure 2C,iii). This strategy has previously shown2 to allow for a more uniform distribution of neurons than other co-culture strategies, where these cell types tend to cluster.
Human iPSC-MN are plated alone or in co-culture with hiPSC-A on a 24-well MEA plate (n = 12 per condition), with each well containing 16 electrodes (Figure 3A). Phase-contrast images of either mono- or co-cultures and raster plots of spiking activity from two representative wells are shown (Figure 3B,C). Human iPSC-A enhances the electrophysiological maturation of hiPSC-MN, as shown by significantly higher degrees of spiking and bursting activity in the co-cultures (Figure 3D). This parallels the effects of hiPSC-A on the morphological and molecular maturation of hiPSC-MN that has been previously demonstrated2.
Some of the technical challenges of this protocol are detailed in the video instructions, where the techniques for plating and recording using MEA platforms as well as representative recordings from these cultures, are shown.
Figure 1: Protocol for the generation of spinal cord hiPSC-MN and hiPSC-A and their co-culture. (A) Timeline of the differentiation protocol detailing critical stages. In the insert, focus on the 30-day protocol for generating spinal cord NPC, with day-by-day actions (E exchange medium, P passage, or no action), cell culture base, and supplements (for additional information on the composition, see Table 1 and Table of Materials). (B) Lineages and cell fate commitment are schematically represented. Co-cultures are generated when mature hiPSC-MN (i.e., DIV 60) and hiPSC-A (i.e., DIV 90) are mixed and plated simultaneously. (Illustration created with BioRender). Please click here to view a larger version of this figure.
Figure 2: Morphological changes during the spinal cord differentiation protocol. (A) Phase contrast microscope images of hiPSC. Panels (i) (lower power, 10x) and (ii) (higher power, 20x) show normal hiPSC, panels (iii) (10x) and (iv) (20x) show a differentiated hiPSC colony. Scale bars = 200 µm and 50 µm. (B) Representative phase contrast microscope images of neuroepithelial cells at DIV 5 (i), NPCs at DIV 21 (ii), and mature DIV 60 motor neurons (iii) and DIV 90 astrocytes (iv). Scale bars = 100 µm and 50 µm (C) Representative immunocytochemistry images on 40x oil of hiPSC-MN (i), hiPSC-A (ii) and their co-culture (iii). Maturation and region-specific markers were targeted. Scale bar = 20 µm. Please click here to view a larger version of this figure.
Figure 3: Multi-electrode array recording of hiPSC-MN. (A) Heat map of mean spiking activity from a single MEA plate (n = 24 wells) at DIV 18 after plating. Wells in rows A and B (n = 12 wells / technical replicates) represent cultures of hiPSC-MN alone, while wells in rows C and D (n = 12 wells) are co-cultures of hiPSC-MN/hiPSC-A. (B) Representative phase-contrast images of hiPSC-MN alone (MN) and in co-culture with hiPSC-A (MN+SCA) on MEA plates. Neurons aggregate in large cell clusters when cultured alone, while the co-culture of hiPSC-MN with hiPSC-A results in evenly distributed monolayers. Scale bar = 50 µm. (C) Representative raster plot of spiking activity over 120 s recording time from a single well with hiPSC-MN alone and hiPSC-MN in co-culture with hiPSC-A. Network burst activity across all electrodes is highlighted with a purple box. (D) Quantification and comparison of spiking and bursting activity between neurons alone and neurons in co-culture with astrocytes from the MEA plate shown in panel A. (*** p < 0.001, **** p < 0.0001). Please click here to view a larger version of this figure.
To date, hiPSC- and MEA-based methods for electrophysiological recordings of astrocyte-neuron co-cultures have found limited application in the field of ALS6 and still not in fully human platforms, in contrast to their more widespread use for in vitro modeling of epilepsy9. This platform, however, has the potential to address pathophysiologically relevant questions in ALS research, such as the mechanisms of neuronal hyperexcitability, astrocyte contribution to neurotoxicity, or the role of network activity in the progression of the disease. Additionally, this platform allows for the collection of prospective electrophysiological data for over 9 months in vitro and, therefore, provides an approach for testing compounds with therapeutic potential2.
Protocols for the generation of hiPSC-N and hiPSC-A found in the literature differ extensively in relation to media conditions, the timing of cultures, yield, and maturation profiles, among other factors. A major advantage of the proposed platform is that astrocytes and neurons are differentiated separately and cultured together at later time points by simultaneously plating the two cell types. After optimization for cell densities and timing, this method can be translated to be used with cell types derived from other protocols as well, including other region-specific cells.
Paramount to the proposed platform is the regional specification of motor neurons and astrocytes. While the proposed protocol is spinal cord specific in its ability to generate ChAT+ MN and HOXB4+ astrocytes, well-established differentiation techniques can be used to generate hiPSC-derived neurons and astrocytes displaying cortical identities, such as CTIP2+ layer V cortical motor neurons18 and OTX2+ forebrain astrocytes14. Thus, the proposed platform has the potential to model neural circuits of the cortex and the spinal cord, as well as their connections.
Multiple systems are available for MEA recording. Plastic 24-well MEA plates with 16 electrodes per well with CO2 and temperature control were preferred for the study. This platform is particularly suitable for high-throughput screening as opposed to systems based on single well recordings. The main limitation of plastic plates is that the surface may be more susceptible to degradation over repeated uses. Systems based on glass MEA plates have the advantage of being reusable multiple times after appropriate treatments as outlined above (up 20 to times), without significant loss of data recording quality. However, these treatments and the coating methods are more time-consuming and technically challenging, given the hydrophobic surface of these plates.
One of the main obstacles of iPSC- and MEA-based methods for electrophysiological recording is the variability and reproducibility of experimental findings. Previous studies show that MEA activity of neurons is dependent upon maturation that is influenced by multiple factors, including, but not limited to, appropriate differentiation techniques used in the generation of both astrocytes and neurons, cell density of neurons, the ratio of astrocytes to neurons throughout differentiation and maturation, sequence of astrocyte and neuron co-cultures2. Standardizing these variables as proposed in this protocol is one way to ensure reproducibility. The choice of multiwell MEA systems generating high throughput data will account for experimental variability. If the electrophysiological activity of a culture is less than what is expected at a given time point, it is important to determine that successful differentiation has occurred and sister cultures that can be immunocytochemically or biochemically analyzed are helpful. Given that a very small volume of cells is seeded initially, small pipetting errors can result in relatively large changes in cell numbers and, therefore, density. Using MEA plates that allow for direct visualization and assessment of cell density is important as well.
The authors have nothing to disclose.
This manuscript was supported by the following: 2019 MSCRFF 5119 (AT). K08NS102526 NIH/NINDS (CWH), 2020 Doris Duke Charitable Foundation Clinical Scientist Career Development Award (CWH). 1R01NS117604-01NIH/NINDS (NJM), DOD ALSRP W81XWH2010161 (NJM), 2019 MSCRFD 5122 (NJM). We thank Dr. Raha Dastgheyb and Dr. Norman Haughey for providing the MEA platform and data analysis software we have utilized to validate the described electrophysiological platform. We would like to thank Khalil Rust for their assistance with protocol demonstration and filming.
10 cm sterile culture plates | Falcon | 353003 | |
25 cm2 sterile culture flasks | Falcon | 353136 | |
2-Mercaptoethanol (β-ME) | Thermofisher | 21985023 | Working concentration 110 µM |
500 mL 0.2 µm CA Filter System | Corning | 430769 | |
5 mL pipette | Falcon | 357543 | |
6 well sterile culture plates | Falcon | 3046 | |
Amphotericine B | Gibco | 15290018 | Working concentration 2.0 μg/mL |
Anionic detergent with protease enzyme – Terg-A-Zyme | Sigma-Aldrich | Z273287 | Working concentration 1% m/v |
Ascorbic acid (ASAC) | Sigma | A4403 | Dissolve 100 mg into 250 mL of dH2O to get 0.4mg/ml stock. Sterile filter through a 0.22 µm filter, aliquot and freeze at -80 ºC. Dilute at 1:1000 for use. (working concentration 0.4 µg/mL). |
Axion CytoView MEA 24 plates (M384-tMEA-24W) | Axion | OPT-24 | |
Axion Edge MEA platform | Axion | Maestro Edge | |
Basement Membrane matrix – Matrigel | Corning | 354277 | Details in the protocol |
Benchtop microscope (sterile under cell culture hood) | Zeiss | 415510-1100-000 | Primo Vert |
Bicuculline | Sigma Aldrich | 14340 | Working concentration10 μM |
Ciliary neurotrophic factor (CNTF) | Peprotech | 450-13 | Dissolve 100 µg in 1mL of sterile PBS, and then add 9 mL of sterile 0.1% BSA-PBS to 10 µg/mL. Aliquot and freeze at -80 ºC. Dilute at 1: 1000 for use. (working concentration 10 ng/mL). |
CO2 tanks and regulator for Axion Edge | AirGas/Harris | 9296NC | |
Compound E | Abcam | ab142164 | Dissolve 250 µg in 2.0387 mL of DMSO to get a 250 µM stock. Aliquot and freeze at -80 ºC. Dilute 1:2000 for use. (working concentration 125 nM). |
Cyanquixaline (CNQX) | Sigma Aldrich | C239 | Working concentration 50 μM |
Dihydrokainic acid (DHK) | Tocris | 111 | Working concentration 50 μM and 300 μM |
DMEM/F12 | Thermofisher | 113300 | Working concentration 1x |
Essential 8 Medium + Essential 8 Supplement | Thermofisher | A1517001 | Combine 10 mL of Essential 8 (10x) supplement with 500 mL of Essential 8 Growth Medium (1x) |
Fetal Bovine Serum (FBS) | Thermofisher | 16140071 | Working concentration 1x |
Glial cell line-derived neurotrophic (GDNF) | Peprotech | 450-10 | Dissolve 100 µg in 1mL of PBS to 100 µg/mL, and then add 9 mL of sterile 0.1% BSA-PBS to 10 µg/mL. Aliquot and freeze at -80 ºC. Dilute at 1: 1000 for use. (working concentration 10 ng/mL). |
Hemocytometer | Election Microscopy Sciences | 63510-20 | |
Heparin | Millipore-sigma | H3149-100KU | Dissolve 200 mg in 100 mL of PBS to get 2 mg/mL stock solution. Aliquot the stock in 15 mL and 500 µL tubes. Dilute 1:1000 for use. (working concentration 2 µg/mL). |
Humidity controlled Cell culture incubator | ThermoFisher | 370 | set to 37 ºC, 5 % CO2 |
Insulin-like growth factor 1 (IGF-1) | R&D systems | 291-G1-200 | Dissolve 200 µg in 2 mL of PBS to 100 µg/mL, then add 18 mL of 0.1%BSA-PBS to make 10 µg/mL stock and store at -80 ºC. Aliquot and freeze at -80 ºC. Dilute 10 µg/mL stock at 1: 1000 for use. (working concentration 10 ng/mL). |
Kainc acid | Abcam | ab144490 | Working concentration 5 μM |
Knockout Serum Replacement (KSR) | Thermofisher | 10828 | Working concentration 1x |
Laminin | Thermofisher | 23017-015 | Stock solution 1 mg/mL, working concentration 10 µg/mL (coating), 1 µg/mL (cell media) |
LDN193189 | Stemgent | 04-0074 | Dissolve 2 mg of LDN into 500 µL of Chloroform to get 10 mM stock. Aliquot this and freeze at -80 ºC. For using, dilute the stock 1 to 10 into DMSO [to 1 mM] first, then dilute 1:5000 of 1 mM into the desired media to get 0.2 µM working solution |
L-Glutamine | Thermofisher | 25030 | Working concentration 100x |
MEA glass plates | MultiChannel Systems | 60MEA200/30iR-Ti-gr | |
Multichannel Pipet P200 | Gilson | PJ22224 | |
Neurobasal | Thermofisher | 21103049 | Working concentration 1x |
Non-Essential Amino Acids (NEAA) | Thermofisher | 11140050 | Working concentration 100x |
Pencillin/Streptomycin | Thermofisher | 15140122 | Working concentration 100x |
Polyornithine (PLO) | Sigma-Aldrich | P3655 | Dissolve 100 mg in 1 mL of ddW to get 100 mg/mL stock solution. Aliquot the stock in 100 µL tubes. (working concentration 100 µg/mL) |
Potassium chloride (KCl) | NA | NA | Working concentration100 mM |
Purmorphamine (PMN) | Millipore-Sigma | 540223 | Dissolve 5 mg in 9.6 mL of DMSO to get 1 mM solution. Aliquot and freeze at -80 ºC. Dilute 1:1000 for use. (working concentration 1 µM). |
Recombinant human-brain-derived neurotrophic factor (BDNF) | Peprotech | 450-02 | Centrifuge briefly before reconstitution. Dissolve 100 µg in 1 mL of PBS to 100 µg/mL, and then add 9 mL of sterile 0.1% BSA-PBS to 10 µg/mL. Aliquot and freeze at -80 ºC. Dilute at 1: 1000 for use. (working concentration 10 ng/mL). |
Retinoic acid (RA) | Sigma | R2625 | Dissolve 100 mg into 3.3 mL of DMSO to get 100 mM stock solution. Aliquot the stock 100 µL/tube and freeze at -80 ºC. Take 200 µL of 100 mM stock and dilute 10x (add 1.8 mL of DMSO) to make 10 mM stock. Aliquot 50 µL/tube and store at -80 ºC. Dilute at 1:10000 for use. (working concentration 2 µM). |
ROCK-I nhibitor | Peprotech | 1293823 | Dissolve 5 mg in 1480 µL of dH2O to get 10 mM stock, aliquot and freeze at -80 ºC. Dilute at 1: 500 for use. (working concentration 20 µM). |
SB431542 | Sigma | S4317 | Dissolve 5mg into 1.3 mL of DMSO to get 10 mM stock solution. Aliquot and freeze at -80 ºC. Dilute at 1:1000 for use. (working con 10 µM) |
Sterile cell culture hoods | Baker Company | SG-600 | |
Supplement B – B27 Supplement | Thermofisher | 21985023 | Working concentration 50x |
Supplement N – N2 Supplement | Thermofisher | 17502048 | Working concentration 100x |
Table top cell culture centrifuge | ThermoFisher | 75004261 | Sorvall Legend X1R |
Thermoplastic film – Parafilm | PARAFILM | P7793 | |
Tissue dissociation protease – Dispase | StemCell Technologies | 7923 | Working concentration 1x |
Trypsin Inhibitor | Sigma | T6522-1G | Dissolve 1g in 100mL ddH2O to get 10 mg/mL stock. Aliquot and store at 4 ºC. Dilute 1:10 of trypsin volume for use. (working concentration 1 mg/mL). |
Trypsin-EDTA (0.05%) | Thermofisher | 2530054 | Working concentration 1x |
Waterbath | ThermoFisher | 2332 | Isotemp |