We modify and implement a previously published protocol describing the rapid, reproducible, and efficient differentiation of human induced Pluripotent Stem Cells (hiPSCs) into excitatory cortical neurons12. Specifically, our modification allows for control of neuronal cell density and use on micro-electrode arrays to measure electrophysiological properties at the network level.
Neurons derived from human induced Pluripotent Stem Cells (hiPSCs) provide a promising new tool for studying neurological disorders. In the past decade, many protocols for differentiating hiPSCs into neurons have been developed. However, these protocols are often slow with high variability, low reproducibility, and low efficiency. In addition, the neurons obtained with these protocols are often immature and lack adequate functional activity both at the single-cell and network levels unless the neurons are cultured for several months. Partially due to these limitations, the functional properties of hiPSC-derived neuronal networks are still not well characterized. Here, we adapt a recently published protocol that describes production of human neurons from hiPSCs by forced expression of the transcription factor neurogenin-212. This protocol is rapid (yielding mature neurons within 3 weeks) and efficient, with nearly 100% conversion efficiency of transduced cells (>95% of DAPI-positive cells are MAP2 positive). Furthermore, the protocol yields a homogeneous population of excitatory neurons that would allow the investigation of cell-type specific contributions to neurological disorders. We modified the original protocol by generating stably transduced hiPSC cells, giving us explicit control over the total number of neurons. These cells are then used to generate hiPSC-derived neuronal networks on micro-electrode arrays. In this way, the spontaneous electrophysiological activity of hiPSC-derived neuronal networks can be measured and characterized, while retaining interexperimental consistency in terms of cell density. The presented protocol is broadly applicable, especially for mechanistic and pharmacological studies on human neuronal networks.
The development of human induced Pluripotent Stem Cells (hiPSCs) differentiation protocols to generate human neurons in vitro has provided a powerful new tool for studying neurological disorders. Until recently, the study of these disorders was severely hampered by the lack of model systems using human neurons. Although rodents can be used to study neurological disorders, the results of such studies cannot be translated easily to humans1. Given these limitations, hiPSC-derived neurons are a promising alternative model that can be used to elucidate molecular mechanisms underlying neurological disorders and for in vitro drug screening.
In the past decade, several protocols to convert hiPSCs into neurons have been developed2-8. However, these protocols are still limited in many ways. First, the protocols are often time-consuming: generating neurons with adequate maturation (i.e. synapse formation) and functional activity requires months of culturing procedures, which renders large-scale studies difficult9. In addition, hiPSC-to-neuron conversion efficiency is low. Protocols often yield a heterogeneous population of neurons, and thus do not allow studies of specific subsets of neuronal cells. Moreover, the protocols are not reproducible, yielding different results for different iPSC lines10,11. Lastly, the maturation stage and functional properties of the resulting neurons are also variable 10.
To address these problems, Zhang et al. (2013)12 developed a protocol that rapidly and reproducibly generates human neurons from hiPSCs by overexpressing the transcription factor neurogenin-2. As reported by the authors, differentiation occurs relatively quickly (only 2 – 3 weeks after inducing expression of neurogenin-2), the protocol is reproducible (neuronal properties are independent of the starting hiPSC line), and the hiPSC-to-neuron conversion is highly efficient (nearly 100%). The population of neurons generated with their protocol is homogeneous (resembling upper-layer cortical neurons), allowing the investigation of cell-type specific contributions to neuronal disorders. Furthermore, their hiPSC-derived neurons exhibited mature properties (e.g., the capability to form synapses and robust functional activity) after only 20 d.
Characterizing the electrophysiological properties of hiPSC-derived neurons at the network level is an important prerequisite before hiPSC technology can be exploited for the study of human diseases. For this reason, many research groups have recently begun to investigate stem-cell-derived neurons at the network level using micro-electrode array (MEA) devices (Multichannel Systems, Reutlingen, Germany)13-16. The electrodes of a MEA are embedded in a substrate on which neuronal cells can be cultured. MEAs can be used to explore the electrophysiological properties of neuronal networks and the in vitro development of their activity. Currently, MEAs are used only in combination with differentiation protocols that take several months to yield mature networks. Hence, combining MEAs with a rapid differentiation protocol should facilitate the use of this technology in large-scale studies of neurological disorders.
Here, we present a modification of the Zhang et al. (2013)12 protocol and adapt it for use on MEAs. In particular, rather than relying on an acute lentiviral transduction, we instead created hiPSC lines stably expressing rtTA/Ngn2 before inducing differentiation. We did this primarily to have reproducible control over the neuronal cell density, since the neuronal cell density is critical for neuronal network formation, and for good contact between the neurons and the electrodes of the MEA17,18. Although the Zhang et al. protocol is very efficient with respect to conversion of transduced hiPSCs, it is inherently variable with respect to the final yield of neurons from the number of hiPSCs plated initially (see Figure 2E in Zhang et al.)12. With a stable line, we eliminate many issues causing variability, such as lentiviral toxicity and infection efficiency. We then optimized the parameters that reliably produce hiPSC-derived neuronal networks on MEAs, obtaining network maturation (e.g., synchronous events involving a majority of the channels) by the third week. This rapid and reliable protocol should enable direct comparisons between neurons derived from different (i.e. patient-specific) hiPSC lines as well as provide robust consistency for pharmacological studies.
All experiments on animals were carried out in accordance with the approved animal care and use guidelines of the Animal Care Committee, Radboud University Medical Centre, the Netherlands, (RU-DEC-2011-021, protocol number: 77073).
1. Glia Cell Isolation and Culture
NOTE: The protocol presented here is based on the work of McCarthy and de Vellis19, and a very similar detailed protocol for mouse astrocytes is available20. To generate primary cultures of cortical astrocytes from embryonic (E18) rat brains, a pregnant rat needs to be sacrificed, the embryos need to be harvested from the uterus, and the brains need to be isolated from the embryos. To fill a T75 flask, the cortices from 2 embryonic brains need to be combined. As an alternative, commercially available purified and frozen astrocytes can be purchased.
2. Generation of rtTA/Ngn2-positive hiPSCs
NOTE: The hiPSCs used for our experiments were generated in-house by lentiviral transduction of human fibroblasts with the reprogramming factors cMYC, SOX2, OCT4 and KLF4.
NOTE: For the generation of rtTA/Ngn2-positive hiPSCs, lentiviral vectors are used to stably integrate the transgenes into the genome of the hiPSCs. The protocol for the production of the lentivirus has been published previously22. The details of the lentiviral packaging vectors that are used to produce the rtTA and Ngn2 lentivirus particles are provided in the Table of Materials/Equipment. The transfer vector used for the rtTA lentivirus is pLVX-EF1α-(Tet-On-Advanced)-IRES-G418(R); i.e. this vector encodes a Tet-On Advanced transactivator under control of a constitutive EF1α promoter and confers resistance to the antibiotic G418. The transfer vector used for the Ngn2 lentivirus is pLVX-(TRE-thight)-(MOUSE)Ngn2-PGK-Puromycin(R); i.e. this vector encodes the gene for murine neurogenin-2 under control of a Tet-controlled promoter and the puromycin resistance gene under control of a constitutive PGK promoter. Hence, by using these two transfer vectors, an hiPSC line can be created for which the expression of murine neurogenin-2 can be induced by supplementing the medium with doxycycline. For the transduction of the hiPSCs, the supernatant with the lentivirus particles is used (referred to as 'lentivirus suspension' in the remainder of the text), i.e. without concentrating the particles using ultracentrifugation.
Final concentration of G418 | Final concentration of puromycin | |
Day 4 | 250 µg/mL | 2 µg/mL |
Day 5 | 250 µg/mL | 2 µg/mL |
Day 6 | 250 µg/mL | 1 µg/mL |
Day 7 | 250 µg/mL | 1 µg/mL |
Day 8 | 250 µg/mL | 1 µg/mL |
Table 1: Concentrations of Antibiotics during the Selection Period. Concentrations of the puromycin and G418 during the 5 d of the selection period.
3. Differentiation of rtTA/Ngn2-positive hiPSCs to Neurons on 6-well MEAs and Glass Coverslips
NOTE: In this protocol, the details are provided for differentiating rtTA/Ngn2-positive hiPSCs on two different substrates, i.e. 6-well MEAs (devices composed of 6 independent wells with 9 recording and 1 reference embedded microelectrodes per well) and glass coverslips in the wells of a 24-well plate. The protocols, however, can easily be adapted for larger substrates (e.g., for the wells of 12- or 6-well plates), by scaling up the mentioned values according to the surface area.
4. Establish the Neurophysiological Profile of hiPSC-derived Neurons
NOTE: Two to three weeks after the induction of differentiation, the hiPSC-derived neurons can be used for different downstream analyses. In this section, examples of some downstream analyses are given that can be performed to establish the neurophysiological profile of the hiPSC-derived neurons.
Here we have successfully modified a protocol in which hiPSCs are differentiated directly into cortical neurons by over-expressing the transcription factor neurogenin-212 and we have adapted it for the use of MEAs. This approach is fast and efficient allowing us to obtain functional neurons and network activity already during the third week after the induction of differentiation.
During the course of the differentiation protocol, the cells morphologically started to resemble neurons: small processes were formed and neurons started connecting to each other (Figure 1A). We established a neurophysiological profile of the neurons derived from a healthy-control hiPSC line, by measuring their neuronal morphology and synaptic properties during development. hiPSC-derived neurons were stained for MAP2 and synapsin-1/2 at different days after the start of differentiation (Figure 2A). The derived neurons show mature neuronal morphology already 3 weeks after the induction of differentiation. The number of synapsin-1/2 puncta (a measure for the number of synapses) was quantified based on synapsin-1/2 immunocytochemistry stainings. The number of synapsin-1/2 puncta increased over time, suggesting that the level of neuronal connectivity is also increasing (Figure 2B). The number of synapsin-1/2 puncta 23 days after the induction of differentiation was similar in two independent IPS lines (Figure 2C). At 23 DIV most synapsin1/2 puncta were juxtaposed to PSD-95 puncta, which is indicative for functional synapses (Figure 2D).
Consistent with the results described by Zhang et al., we generated a population of excitatory upper layer cortical neurons, confirmed by pan-neuronal and subtype-specific cortical markers such as BRN2 and SATB2(layer II/III). We did not observe neurons that were positive for deep layer neurons CTIP2 (layer V) or Foxp2 (layer VI) (Figure 2E and F)
To characterize the electrophysiological activity of the hiPSC-derived neurons, we used whole-cell current and voltage clamp recordings, i.e. intrinsic properties and excitatory input onto these neurons were measured during development. The neurons were able to generate action potentials already one week after the of differentiation and the percentage of spiking cells was increasing over time (Figure 2G and H). Furthermore, the neurons received excitatory synaptic input already a week after the induction of differentiation: both frequency and amplitude of the excitatory synaptic input increased during development (Figure 2I – K).
To better understand how single-cell activity combines to form network-level functions, it is essential to study how neurons work in concert. In vitro neuronal networks cultured on MEAs constitute a valuable experimental model for studying the neuronal dynamics. We recorded 20 min of electrophysiological network activity of neurons derived from a healthy-control hiPSC line cultured on 6-well MEAs (Figure 2M). Few weeks after the induction of differentiation, the neurons derived from healthy-control hiPSCs formed functionally active neuronal networks, showing spontaneous events (0.62 ± 0.05 spike/s; Figure 2N). At this stage of development (i.e. 16 d after the start of the differentiation) no synchronous events involving all the channels of the MEAs are detected (Figure 2O). The level of network activity increased during the development: during the fourth week after the induction of differentiation, the neuronal networks showed high level of spontaneous activity (2.5 ± 0.1 spike/s; Figure 2N) in all the wells of the device. The networks also exhibited synchronous network bursts (4.1 ± 0.1 burst/min, Figure 2O) with long duration (2,100 ± 500 ms).
Figure 1: hiPSC Differentiation into Neurons. A. Three time points of hiPSCs differentiation into neurons on coverslips. B. Plating of hiPSCs on MEAs. C. Astrocytes at 100% confluency in T75 flask (note that the cells form a tessellated monolayer). Scale bars: 150 µm. Please click here to view a larger version of this figure.
Figure 2. hiPSC-derived Neurons Characterization. A. hiPSC-derived neurons were stained for MAP2 (green) and synapsin-1/2 (red) at different days after the start of differentiation. Scale bar: 10 µm. B. Quantification of synapsin puncta in two independent experiments. In each experiments at least 10 cells were analyzed. C. Quantification of synapsin puncta at DIV23 in neurons derived from two independent IPS lines. D. hiPSC-derived neurons were stained for PSD-95 (green) and synapsin-1/2 (red) 23 days after the start of differentiation. Synapsin puncta are juxtaposed to PSD-95 puncta. E. hiPSC-derived neurons were stained for MAP2 (green) and BRN2 (red) or SATB2 (red) 23 d after the start of differentiation. F. Percentage of MAP2 positive cells that were positive for indicated markers. G. Representative current clamp recordings showing that action potentials can be generated as early as 7 d after the start of differentiation. H. Percentage of cells at different days after the induction of differentiation that show one or more action potentials. I. Representative traces of excitatory postsynaptic currents (EPSCs) received by hiPSC-derived neurons at different days after differentiation. J. Frequency of excitatory postsynaptic currents during development. K. Amplitude of excitatory postsynaptic currents during development. L. Representative traces of EPSC recordings without (control) and with CNQX (CNQX). M. Neurons derived from one hiPSC line were cultured on a 6-well MEA and network activity is shown for hiPSC-derived neuronal networks 16 and 23 d after the induction of differentiation. The activity recorded from each well (sampling rate of 10 kHz) is indicated with a different color (5 min of the 20 min of recording are shown). N. Firing rate 16 and 23 d after the induction of differentiation. O. Bursting rate 16 and 23 d after the induction of differentiation. Please click here to view a larger version of this figure.
Given the results, the quality of the resulting hiPSC-derived neurons can be assessed by making a neurophysiological profile of the cells. That is, three to four weeks after the start of the differentiation, the morphology, synapsin-1/2 expression and electrophysiology of the neurons can be assessed. At that time point, the hiPSC-derived neurons are expected to show a neuronal-like morphology, to be MAP2, synapsin/PSD-95 positive when performing immunocytochemistry, and to exhibit spontaneous electrophysiological activity (both at the single-cell and network level).
Here we have implemented an efficient hiPSC-differentiation protocol published by Zhang et al. (2013)12 for measuring the network activity of hiPSC-derived neuronal networks on MEAs. We adapted the original protocol by creating an rtTA/Ngn2-positive hiPSC line before inducing neuronal differentiation. This additional step allows us to control the neuronal cell density on the MEA. Control over the neuronal density was an important pre-requisite for adapting the protocol to MEAs and for ensuring consistency. To measure the activity of neuronal networks using MEAs, the neurons need to form dense networks directly on top of the MEA electrodes17,18. This necessarily requires tight control over the plating density of the neurons. The rtTA/Ngn2-positive hiPSC line allows for control of neuron density because this tactic does not rely on acute lentiviral transductions of hiPSCs prior to differentiation; the rtTA/Ngn2-positive hiPSC line therefore nearly eliminates any variation in the final yield due to, for example, lentiviral toxicity and variable infection efficiency.
Another critical step of the experimental procedure is the number of the rat astrocytes that are cocultured with the differentiating hiPSCs. Astrocytes actively contribute to the refinement of developing neural circuits by controlling synapse formation, maintenance, and elimination, all of which are important processes for neuronal functioning. The protocol presented in this paper is highly astrocyte-dependent: to fully mature and form functional synapses, the neurons require support from the astrocytes. We experienced that the number of astrocytes should be roughly equal to the number of hiPSC-derived neurons to support the maturation of the neurons and the formation of neuronal networks exhibiting spontaneous activity. Since our astrocyte protocol yields primary cell cultures with a limited life span, the isolation of rat astrocytes has to be performed regularly.
Our adaptation of the protocol published by Zhang et al. (2013)12 for use with MEA technology will likely significantly improve our ability to study the network activity of hiPSC-derived networks. Previously, protocols used for studying hiPSC-derived neuronal networks with MEAs relied on time-consuming differentiation procedures13-16. The protocol from Zhang et al. (2013) provides a rapid alternative, and our modification removes a source of variability, which makes it now more feasible to use hiPSC-derived neurons in combination with MEA technology, especially in high-throughput or pharmacological studies. In addition, because the method published by Zhang et al. (2013)12 yields a homogeneous population of upper-layer cortical neurons, our adapted protocol makes possible focused studies into the network activity of this particular neuronal subset.
Nonetheless, this approach has also several limitations. First, the homogeneity of the cultures can also be considered a disadvantage, because the cultures are less likely to resemble in vivo networks, where different classes of neurons (i.e. inhibitory and excitatory neurons) constitute a heterogeneous network. To further enhance the use of the hiPSC-derived neurons with MEA technology, it will be important to develop rapid (transgene-based) differentiation protocols for other neuronal cell populations. If protocols become available, the in vitro networks would mimic in vivo networks more closely. Second, at present rat astrocytes must be added to the hiPSC-derived neurons for growth support, and therefore the resulting neuronal network is not a human neuronal network sensu stricto. Reliable protocols for differentiating hiPSCs into astrocytes may in the future solve this problem26. Third, two-dimensional neuronal networks, as described here, are a limited model for studying complex three-dimensional in vivo neuronal networks. Fortunately, protocols describing three-dimensional cultures of rat primary neurons in combination with MEA technology are already available27,28. Prospectively, the combination of rapid differentiation protocols for obtaining hiPSC-derived neurons and astrocytes with three-dimensional culture techniques and MEA technology should provide novel insight into the biological mechanisms underlying neurological disorders.
The authors have nothing to disclose.
The authors thank Jessica Classen for performing the whole-cell patch-clamp experiments. The hiPSCs used in our experiments were kindly provided by Huiqing Zhou and Willem van den Akker from the Radboud University Nijmegen. The transfer vectors used in this protocol were kindly provided by Oliver Brüstle, Philipp Koch and Julia Ladewig from the University of Bonn Medical Centre.
Lebovitz's L-15 medium | Gibco | 11415-064 | |
B-27 supplement | Gibco | 0080085SA | |
Poly-D-lysine | Sigma-Aldrich | P6407 | |
Ca2+/Mg2+-free HBSS | Gibco | 14175-095 | |
0.05 % Trypsin-EDTA | Gibco | 25300-054 | |
2.5 % Trypsin | Gibco | 15090-046 | |
High-glucose DMEM | Gibco | 11965-092 | |
FBS | Sigma-Aldrich | F2442-500ML | |
Penicillin/Streptomycin | Sigma-Aldrich | P4333 | |
70 µm cell strainer | BD Falcon | 352350 | |
DPBS | Gibco | 14190-094 | |
psPAX2 lentiviral packaging vector | Addgene | Plasmid #12260 | |
pMD2.G lentiviral packaging vector | Addgene | Plasmid #12259 | |
Basement membrane matrix | Gibco | A1413201 | |
DMEM/F12 | Gibco | 11320-074 | |
Cell detachment solution | Sigma-Aldrich | A6964 | |
E8 medium | Gibco | A1517001 | |
ROCK inhibitor | Gibco | A2644501 | Alternatively, ROCK inhibitors like thiazovivin can be used. |
Polybrene | Sigma-Aldrich | H9268-5G | |
G418 | Sigma-Aldrich | G8168-10ML | |
Puromycin | Sigma-Aldrich | P9620-10ML | |
Vitronectin | Gibco | A14700 | |
6-well MEAs | Multi Channel Systems | 60-6wellMEA200/30iR-Ti-tcr | |
Glass coverslips | VWR | 631-0899 | |
Poly-L-ornithine | Sigma-Aldrich | P3655-10MG | |
Laminin | Sigma-Aldrich | L2020-1MG | |
Doxycyclin | Sigma-Aldrich | D9891-5G | |
N-2 supplement | Gibco | 17502-048 | |
Non-essential amino acids | Sigma-Aldrich | M7145 | |
NT-3, human recombinant | Promokine | C66425 | |
BDNF, human recombinant | Promokine | C66212 | |
Trypsin-EDTA | Gibco | 25300-054 | |
L-alanyl-L-glutamine | Gibco | 35050-038 | |
Neurobasal medium | Gibco | 21103-049 | |
Cytosine β-D-arabinofuranoside | Sigma-Aldrich | C1768-100MG | |
Straight fine-tipped forceps | Fine Science Tools | 11251 | |
Fine-tipped spring scissors | Fine Science Tools | 91500-09 |