A protocol to generate a co-culture system consisting of neurons derived from induced pluripotent stem cells (iPSCs), primary cortical neurons and astrocytes is described. This co-culture system allows detection of the formation of synaptic contacts and circuits between new, iPSC-derived neurons and pre-existing cortical neurons expressing channelrhodopsin-2.
Here we describe a protocol to generate a co-culture consisting of 2 different neuronal populations. Induced pluripotent stem cells (iPSCs) are reprogrammed from human fibroblasts using episomal vectors. Colonies of iPSCs can be observed 30 days after initiation of fibroblast reprogramming. Pluripotent colonies are manually picked and grown in neural induction medium to permit differentiation into neural progenitor cells (NPCs). iPSCs rapidly convert into neuroepithelial cells within 1 week and retain the capability to self-renew when maintained at a high culture density. Primary mouse NPCs are differentiated into astrocytes by exposure to a serum-containing medium for 7 days and form a monolayer upon which embryonic day 18 (E18) rat cortical neurons (transfected with channelrhodopsin-2 (ChR2)) are added. Human NPCs tagged with the fluorescent protein, tandem dimer Tomato (tdTomato), are then seeded onto the astrocyte/cortical neuron culture the following day and allowed to differentiate for 28 to 35 days. We demonstrate that this system forms synaptic connections between iPSC-derived neurons and cortical neurons, evident from an increase in the frequency of synaptic currents upon photostimulation of the cortical neurons. This co-culture system provides a novel platform for evaluating the ability of iPSC-derived neurons to create synaptic connections with other neuronal populations.
In recent years, our understanding of neuron and neuronal circuit function have been revolutionized by several optogenetic tools that control neuronal excitability. One such tool is the light-activated cation channel, channelrhodopsin-2 (ChR2): neurons expressing ChR2 fire action potentials in response to light stimulation1-3. Because neurons are not ordinarily sensitive to light, this remarkable ability opens up new avenues for the precise temporal and spatial control of the activity of genetically-defined populations of neurons4 and has greatly accelerated studies of brain function. ChR2 has been applied to stem cell research for different purposes, from monitoring neuronal development5 to regulating activity of neural networks6.
Here we used ChR2 to increase the utility of a neuronal co-culture system. Co-culture systems enable the assessment of interactions between different cell types. Co-cultures of neurons and glia, the main cell types of the nervous system, are commonly used to investigate signaling between these different cell types, as well as to evaluate the significance of this signaling for physiological responses, propagation of damage and to examine the molecular mechanisms that are potentially involved in this signaling7. A previous study revealed that human iPSCs differentiate very quickly into neurons in the presence of rat cortical primary culture containing neurons, astrocytes, oligodendrocytes, and microglia8.
In this protocol, we demonstrate a method that utilizes ChR2 in a co-culture system to interrogate synaptic connections between rat cortical neurons and neurons derived from human induced pluripotent stem cells (iPSCs). We describe steps to dissect and harvest brain tissues9 as well as to maintain and differentiate mouse neural progenitor cells (NPCs) into astrocytes and human NPCs into neurons to create the co-culture system. To establish this co-culture system, we used the integration-free reprogramming method described by Okita et al.10 to generate human iPSCs. These iPSCs were then efficiently differentiated into NPCs in chemically-defined conditions using small-molecule inhibitors as described by Li et al.11
In the co-cultures, the different populations of neurons can be identified via detection of ChR2 expression in cortical neurons and tagging of iPSC-derived neurons via the florescent protein, tandem dimer Tomato (tdTomato). Combining electrophysiological recordings from iPSC-derived neurons with optogenetic photostimulation of the cortical neurons allows assessment of the ability of these two types of neurons to form circuits with each other. Specifically, photostimulation of ChR2-expressing cortical neurons evoke synaptic responses in iPSC-derived neurons that have formed synaptic circuits with the photostimulated cortical neuron network. We demonstrate that increased postsynaptic currents are indeed detected in iPSC-derived neurons under such conditions. This protocol allows the assessment of synaptic circuitry under a variety of experimental conditions, including recapitulation of different neurological disease models by using iPSCs derived from fibroblasts from disease patients.
NOTE: All experiments are performed following protocols approved by the Institutional Animal Care and Use Committee at SingHealth.
1. Harvesting Early Postnatal Mouse Brains
2. Hippocampal Tissue Dissociation
3. Maintenance of Adherent Mouse Hippocampal NPCs
4. Differentiation of Mouse NPCs into Astrocytes
5. Harvesting Embryonic Rat Brains and Cortical Tissue Dissociation
6. Electroporation and Plating of Primary Cortical Neurons
NOTE: Gene transfer can be achieved using several methods, including electroporation, calcium phosphate transfection and viral transduction. In this protocol, we describe electroporation to deliver ChR2 construct into primary cortical neurons.
7. Generation of Induced Pluripotent Stem Cells
NOTE: This method has been adapted from Okita et al.10
8. Neural Induction and Maintenance of Human NPCs
NOTE: This method has been described by Li et al.11
9. Differentiation of Human NPCs into Neurons in Co-cultures
10. Electrophysiological Recordings of iPSC-derived Neurons
Here, a protocol describing the multiple steps involved in generating a co-culture consisting astrocytes, cortical neurons and human neurons is illustrated. Human iPSCs were obtained by reprogramming fibroblasts using episomal vectors10 and specified into a neural lineage with a cocktail of small-molecule inhibitors, making NPCs11 which were maintained at high density (Figure 1A). Several steps are required to generate the co-culture: differentiation of mouse NPCs into astrocytes, plating of rat cortical neurons expressing ChR2, and seeding of human NPCs tagged with tdTomato (Figure 1B). At day 2 of differentiation, Glial Fibrillary Acidic Protein (GFAP)+ astrocytes can readily be observed in our cultures (Figure 1C). Mouse NPCs were allowed to differentiate for at least a week before plating cortical neurons expressing ChR2 onto the astrocyte monolayer (Figure 1D). After 3 to 4 weeks of human NPC differentiation, tdTomato+ neurons were detected in the cultures (Figure 1E).
This co-culture system allows consistent detection of postsynaptic currents from individual iPSC-derived neurons in response to light stimulation (Figure 2A). When stimulated by light, action potentials were generated in cortical neurons expressing ChR2 (Figure 2B). iPSC-derived neurons are also excitable (Figure 2C), showing increased action potential firing as the amplitude of depolarizing current pulses was increased (Figure 2D). Thus, these cells exhibit mature neuronal properties. In the absence of light, iPSC-derived neurons received spontaneous synaptic inputs (Figure 2E). These inward postsynaptic currents were predominantly mediated by AMPA receptors, because currents through GABA receptors would be outward and NMDA receptors do not carry current at the holding potential of -70 mV. At least some of these inputs were from presynaptic cortical neurons expressing ChR2, because light stimulation greatly increased the frequency of postsynaptic currents (Figure 2E). The time course of this increase in synaptic input is shown in Figure 2F: photostimulation of the cortical neurons was sustained throughout the 30 sec long light flash. While the frequency of PSCs was elevated when ChR2-expressing cortical neurons were photostimulated (Figure 2G), their amplitude was unaffected (Figure 2H). In summary, these results indicate that the iPSC-derived neurons exhibited mature neuronal properties and could form synaptic connections with presynaptic cortical neurons.
Figure 1: Schematic diagrams for the generation of co-culture system. (A) Timeline for reprogramming human fibroblasts into iPSCs and neural induction to NPCs. (B) Timeline for terminal differentiation of human NPCs into mature neurons on cortical neuron and astrocyte cultures. (C) Mouse NPCs rapidly differentiate into GFAP+ astrocytes after 2 days in vitro. (D) After one week, cortical neurons expressing ChR2 were plated onto GFAP+ astrocytes. (E) At 4 to 5 weeks after differentiation of human NPCs, electrophysiological recordings were performed on tdTomato+ neurons. Scale bars represent 100 µm.
Figure 2: Optogenetic analysis of functional synaptic connectivity. (A) iPSC-derived cells tagged with tdTomato (red) were co-cultured with cortical neurons expressing the light-activated channel, ChR2 (green). Recordings from iPSC-derived neurons revealed postsynaptic current (PSC) responses, which could come from either the cortical neurons or other iPSC-derived neurons. (B) Illumination (blue bar) evoked a series of action potentials in cortical neurons expressing ChR2. (C) iPSC-derived cells are evoked by current injection. (D) Firing vs current curve of iPSC-derived cells. (E) Photostimulation of the ChR2-expressing cortical neurons (blue bar) increased the frequency of PSCs in iPSC-derived neurons. (F) Time course of the increase in PSC frequency produced by photostimulation. Blue bar indicates time of photostimulation. (G, H) While PSC frequency was increased by photostimulation (G), PSC amplitude was unaffected (H). * indicates p <0.05 compared to control with student’s t-test.
Optogenetics provides temporal and spatial precision for activation of defined populations of neurons13. In our experiments, the entire field of the microscope objective was illuminated for 30 sec to photo-stimulate only cortical neurons expressing ChR2. This allowed us to determine whether synaptic connections were formed between different populations of neurons within the co-culture. Our results revealed that photostimulation increases PSC frequency in iPSC-derived neurons, which demonstrates that these neurons receive synaptic input from presynaptic cortical neurons expressing ChR2. This, in turn, established that the iPSC-derived neurons successfully incorporated into circuits with the cortical neurons.
There are several critical steps for the generation of this co-culture system. It is important to ensure that the mouse NPC-derived astrocyte cultures are healthy, with minimal cell death, before proceeding with plating of other cell types. Cortical neurons and human NPCs attach well onto healthy astrocytes and electrophysiological recordings depend heavily on culture conditions. The other key steps in this protocol are the electroporation and plating of cortical neurons onto astrocyte cultures. As a large number of cells die after electroporation, it is important to ensure that at least 6 million cortical neurons are electroporated for plating onto astrocyte cultures in 24-well plates. We have observed that when fewer than 6 million cells were electroporated, the number of neurons that survive did not form a dense neuronal network, which affected electrophysiological recordings. The number of electroporated cortical neurons should also be consistent for each co-culture experiment to allow comparison between different studies. For all steps involving enzymatic digestion, it is essential not to leave cells in digestive solution for too long as it may lead to cell death.
This approach can be used to screen the capability of neurons derived from patient-specific fibroblasts to form synaptic contacts with other neurons. Neurons derived from fibroblasts from patients suffering from the neurodevelopmental disorder Rett Syndrome (RTT) exhibit synaptic transmission defects: RTT neurons exhibit a significant decrease in PSC frequency and amplitude compared to healthy neurons, presumably due to less synaptic connectivity14. Our co-culture system allows examination of drug effects on neurons displaying developmental defects. This can be carried out via addition of compounds of interest during seeding of diseased human NPCs as well as during continuous exposure to compounds throughout the time of neuronal differentiation.
While this co-culture system can also be used as a model for drug testing, a limitation with this approach is that the process of investigating the effects of each candidate compound can be long and tedious as it is not a high-throughput screening method. Following treatment with candidate compounds, iPSC-derived neurons have to be individually patched to determine the effects of each compound on PSCs. Moreover, there are currently not many available fibroblasts and iPSCs from patients. Therefore, it will be of interest in the near future to have more patient cells and also to adapt this protocol for high throughput screening. For example, by labelling iPSC-derived neurons with an activity indicator, such as genetically-encoded sensors of ions or membrane potential, it would be possible to increase the throughput rate substantially by side-stepping the time-consuming electrophysiological procedure. As the whole process of generating this co-culture system, from reprogramming fibroblasts to performing electrophysiological recordings, requires more than 90 days, it is recommended that either the human iPSCs or NPCs be expanded, after the reprogramming and neural induction steps respectively, and cryopreserved to reduce the time needed for future experiments.
In our experiments, we expressed ChR2 in cortical neurons under the synapsin1 promoter. Because this promoter is pan-neuronal, we presumably were photostimulating both inhibitory and excitatory cortical neurons. If the goal of future experiments is to specifically interrogate either inhibitory or excitatory circuits, more specific promoters could be used. Alternatively, cortical neurons could be obtained from transgenic (or virus-injected mice) where ChR2 expression is specifically targeted to genetically-defined subpopulations of neurons. For example, harvesting cortical tissues from a mouse that has Cre recombinase expressed in parvalbumin-containing interneurons (PV interneurons) that is crossed with a mouse with floxed ChR2 will yield a situation where the only cortical neurons expressing ChR2 will be PV interneurons15. The use of such ChR2-expressing interneurons in our co-culture system will enable the determination of whether a patched iPSC-derived neuron is able to incorporate into a circuit with PV interneurons.
Based on a previous study9, cortical neurons are mature with overlapping dendrites after 14 days in vitro. Thus, if photostimulation does not elicit a response from a patched iPSC-derived neuron, it should indicate that the neuron does not have the ability to make synaptic connections with cortical neurons. An iPSC-derived neuron is able to receive synaptic input either from other iPSC-derived neurons or cortical neurons. If traditional patch clamp recordings were used, it would not be possible to distinguish where the synaptic input is coming from. The advantage of our system is that postsynaptic responses from cortical presynaptic input can be selectively evoked and identified. The procedures outlined here provide a simple approach to evaluate the capability of iPSC-derived neurons to integrate into a defined neuronal network.
The authors have nothing to disclose.
We thank K. Deisseroth and S. Je for the ChR2 and Synapsin1-tdTomato lentiviral constructs respectively, C. Chai for sharing expertise and protocol on iPSC generation, W.Y. Leong for technical support, members of the Goh lab for sharing reagents and expertise. This work was supported by Abbott Nutrition and the Academic Centre of Excellence (ACE) research award from GlaxoSmithKline (GSK) to E.L.G., by the National Research Foundation Singapore under its Competitive Research Program (NRF 2008 NRF-CRP 002-082) to E.L.G. and G.J.A., and by the World Class Institute (WCI) Program of the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology of Korea (MEST) (NRF Grant Number: WCI 2009-003) to G.J.A.
Name of Material/ Equipment | Company | Catalog Number | Comments/Description |
Neurocult Proliferation Kit (Mouse) | STEMCELL Technologies | 5702 | Contains NSC Basal Medium and NSC Proliferation Supplement |
Epi5 Episomal iPSC Reprogramming Kit | Invitrogen | A15960 | |
DMEM/F12 | Lonza | 12719F | |
Matrigel | BD Biosciences | 354277 | |
Neurobasal medium | Invitrogen | 21103049 | |
MEM without glutamine | Invitrogen | 11090081 | |
N2 | Invitrogen | 17502048 | |
B27 | Invitrogen | 17504044 | |
L-glutamine | Invitrogen | 25030081 | |
GlutaMAX supplement | Invitrogen | 35050061 | |
PenStrep | Invitrogen | 15140122 | |
BSA | Sigma | A7906 | |
Human LIF | Invitrogen | PHC9484 | |
CHIR99021 | Miltenyi Biotech | 130103926 | |
SB431542 | Sigma | S4317 | |
Compound E | Merck Millipore | 565790 | |
EGF | Merck Millipore | 324856 | |
FGF2 | R&D Systems | 233FB025 | |
Research Grade FBS | Thermo Scientific | SV30160.03 | For astrocyte differentiation medium |
Defined FBS | Thermo Scientific | SH30070.03 | For primary neuron medium |
PBS | Thermo Scientific | SH30028.02 | |
Glucose | Sigma | G8270 | For primary neuron medium |
Sucrose | Sigma | S0389 | |
Heparin | EMD Millipore | 375095 | |
Papain | Worthington | 3126 | |
HBSS | Invitrogen | 14175095 | |
DNase I | Roche | 10104159001 | |
Dispase II | STEMCELL Technologies | 7923 | |
HEPES | Gibco | 15630080 | For harvesting brains |
EBSS | Invitrogen | 14155063 | |
L-Cysteine | Sigma | C7352 | |
Trypsin-EDTA | Invitrogen | 25200056 | |
70 µm cell strainer | BD Biosciences | 352350 | |
20 µm filter unit | Sartorius Stedim | 16534K | |
NaCl | Sigma | S7653 | |
KCl | Sigma | P5405 | |
NaHCO3 | Sigma | S5761 | |
NaH2PO4 | Fluka | 71505 | |
MgCl2 | Sigma | M8266 | |
CaCl2 | Sigma | C1016 | |
D(+)-Glucose | Sigma | G7520 | For electrophysiological studies |
K-gluconate | Sigma | P1847 | |
KOH | KANTO Chemical | 3234400 | |
HEPES | Sigma | H3375 | For electrophysiological studies |
EGTA | Sigma | E3889 | |
Na2ATP | Sigma | A7699 | |
Na3GTP | Sigma | G8877 | |
Borosilicate glass capillaries | World precision instruments, Inc | 1604323 | |
15 ml centrifuge tube | BD Biosciences | 352096 | |
50 ml centrifuge tube | BD Biosciences | 352070 | |
24-well plates | Corning | 9761146 | |
6-well plates | Corning | 720083 | |
T25 flasks | Corning | 430641 | |
12 mm cover glasses | Marienfeld-Superior | 111520 | |
AMAXA Rat Neuron Nucleofector Kit | Lonza | VPG1003 | For electroporation of primary cortical neurons |
Neon Transfection System | Invitrogen | MPK5000 | For electroporation of human fibroblasts |
Mini Analysis | Synaptosoft | Mini Analysis v.6 | For measurement of PSCs |
Normal Dermal Human Fibroblasts | PromoCell | C12300 | |
pLenti-Synapsin-hChR2(H134R)-EYFP-WPRE | Addgene | 20945 | |
Mercury short-arc lamp | OSRAM | HBO 103 W/2 | |