This procedure yields telencephalic neurons by going through checkpoints which are similar to those observed during human development. The cells are allowed to spontaneously differentiate, are exposed to factors which push them towards the neural lineage, are isolated, and are plated onto coverslips to allow for terminal differentiation and maturation.
Here, a stepwise procedure for efficiently generating telencephalic glutamatergic neurons from human pluripotent stem cells (PSCs) has been described. The differentiation process is initiated by breaking the human PSCs into clumps which round up to form aggregates when the cells are placed in a suspension culture. The aggregates are then grown in hESC medium from days 1-4 to allow for spontaneous differentiation. During this time, the cells have the capacity to become any of the three germ layers. From days 5-8, the cells are placed in a neural induction medium to push them into the neural lineage. Around day 8, the cells are allowed to attach onto 6 well plates and differentiate during which time the neuroepithelial cells form. These neuroepithelial cells can be isolated at day 17. The cells can then be kept as neurospheres until they are ready to be plated onto coverslips. Using a basic medium without any caudalizing factors, neuroepithelial cells are specified into telencephalic precursors, which can then be further differentiated into dorsal telencephalic progenitors and glutamatergic neurons efficiently. Overall, our system provides a tool to generate human glutamatergic neurons for researchers to study the development of these neurons and the diseases which affect them.
Human pluripotent stem cells (PSCs), including both human embryonic stem cells (hESCs) and induced pluripotent stem cells (iPSCs), have the capacity to generate every cell type in the body, including neurons1-3. Directed differentiation of various neuronal subtypes from human PSCs holds the key for the application of these cells in regenerative medicine. The generation of functional neuronal subtypes during development is a complex process involving the induction of neural lineage, the specification of regional progenitors along the rostro-caudal axis, and the differentiation of post-mitotic neuron types from the regional progenitors4,5. Beginning in 2001, several systems have been established to generate neural lineage from hESCs, which have provided a platform for the subsequent generation of neuronal subtypes6,7. Based upon developmental principles, several neuron types such as spinal motor neurons8-12, midbrain dopaminergic neurons13-15, and neural retinal cells16,17 have been efficiently specified from human PSCs. This was done by applying critical morphogens which are important for the specification of these neuron types during in vivo development. Other protocols have also been developed to promote the differentiation of hESCs into neurons using either additional factors18-20 such as small molecules or by co-culturing with other cell types to help promote differentiation21.
The human neocortex is highly developed and contains many cell types, including glutamatergic neurons which play an important role in learning, memory, and cognitive function22,23. The first step in generating glutamatergic neurons in culture is to specify telencephalic progenitor cells. Yoshiki Sasai’s group first reported the directed differentiation of telencephalic precursors from mouse ESCs (mESCs) using a serum-free suspension culture in the presence of DKK1 (which inhibits Wnt signaling) as well as LeftyA (which inhibits nodal signaling)24. Subsequently, several groups including ours have also reported the specification of telencephalic precursors from human PSCs in serum free medium 25-27. The generation of telencephalic precursors from human PSCs does not require the use of exogenous morphogens and the efficiency in generating these precursors is much higher than that from mESCs 26,27. Here, a chemically defined system for neural induction which was well established by Zhang’s group7 has been described. Without the addition of exogenous caudalizing factors, this protocol efficiently generates telencephalic precursors from human PSCs27. These progenitors can then be differentiated into dorsal or ventral progenitors by regulating the signaling of Wnt and sonic hedgehog (SHH).The dorsal progenitors can further differentiate into glutamatergic neurons efficiently27. In addition, this protocol also works well for the generation of glutamatergic neurons from human iPSCs28, which allows for the generation of patient-specific neurons that can be utilized to explore the mechanism of action as well as potential therapies for a large array of diseases. Moreover, our system also provides a platform to explore the development and specification of diverse neuronal types in the telencephalon.
1. Generation of Human Pluripotent Stem Cell Aggregates (D1-D4)
2. Induction of Neuroepithelial Cells (D5-D17)
3. Generation of Telencephalic Progenitors (D17-D24)
4. Further Differentiation into the Telencephalic Neurons (D25+)
Here, a protocol to differentiate human PSCs into telencephalic glutamatergic neurons through several critical steps: the formation of PSC aggregates, the induction of neuroepithelial cells, the generation of telencephalic progenitors, and the terminal differentiation of these progenitors into telencephalic neurons (Figure 1) has been described. This system is robust and efficient in the generation of telencephalic progenitors and glutamatergic neurons. As an example (Figure 2), without the addition of caudalizing factors, the hESCs were differentiated into the neural lineage27. At 24 days post differentiation, the majority of neural precursors were positive for FOXG1 (a transcription factor expressed by telencephalon) but negative for HOXB4 (a marker for hindbrain and spinal cord cells) suggesting the telencephalic progenitors were successfully generated (Figure 2D). These telencephalic progenitors possess a dorsal phenotype, as indicated by the expression of PAX6 (a marker expressed by dorsal progenitors) (Figure 2E), but not NKX2.1 (a marker for ventral progenitors) (Figure 2F). Following further differentiation, these telencephalic dorsal progenitors differentiated into glutamatergic neurons, which were positive for the glutamatergic marker TBR1 (around 80%)27. Neurons which stained positive for TBR1 were also positive for neuronal markers βIII-tubulin (Figure 2G) or microtubule-associated protein 2 (MAP2, Figure 2H). These cells also expressed vesicular glutamate transporter 1 (vGLUT1), a marker for mature glutamatergic neurons, suggesting the efficient generation of telencephalic glutamatergic neurons in the culture (Figure 2I).
Figure 1. A schematic timeline for the neuronal differentiation process.
Figure 2. Images highlighting cells during several critical stages during neuronal differentiation. (A) Human ESCs were cultured for 4 days. Phase images showing the formation of ESC aggregates (B) and neuroepithelial cells (C). At 24 days after differentiation from hESCs, the majority of NE cells were FOXG1+ but HOXB4– (D). The telencephalic progenitors (FOXG1+) were PAX6+ (E) but NKX2.1– (F) after 1 month of differentiation. After two additional weeks (6 weeks total) of differentiating the cells on the coverslips, most of them stained positive for the glutamatergic marker TBR1 (G,H) as well as the neuronal markers βIII-tubulin (G) and MAP2 (H). After eight weeks of differentiation, the cells were positive for the mature glutamatergic marker, vGLUT1 (I). Scale bars: 100 μm (A), 50 μm (B-I). (D-I have been adapted from our previous publication27 with permission).
There are several critical steps during the neural differentiation process. It is important to ensure that the human PSCs are pluripotent because otherwise the cells may already be biased towards becoming a non-neuronal lineage. This can be confirmed by staining the human PSCs with antibodies against pluripotency markers such as Oct4, Sox2, Nanog, and Tra-1-60 1-3. If the human PSCs do not attach very well after passaging them, ROCK inhibitor (Y27632) can be added to help. For those having difficulty with keeping their cells pluripotent, some potential issues are the quality/density of the MEF cells, as well as the lot of KOSR being used as there can be variation from batch to batch. Although MEF feeder cells have been used to maintain PSCs in the above protocol, this method also works for PSCs that are cultured using feeder-free systems.
When the cells are being broken up to form the PSC aggregates, it is important to limit their dispase exposure to only a few minutes (the edges will round up). The length of time that it takes before the edges of the iPSCs (up to 10 min) is usually longer than that of the ESCs (3-5 min). It is optimal to add the dispase to only one plate of human PSCs at a time to ensure that the cells are not in the enzyme for too long as this can harm or even kill the cells. After the cells are in the PSC aggregate stage, it is important to keep the cells at a low density. The density is also quite important when the PSC aggregates are plated. These cells will expand on the plate over the next few days and precautions should be taken to ensure that they will not touch after they grow bigger. During the attachment step (2.2), it is imperative to ensure that the cells do not have a prolonged exposure to the FBS as this could affect the gene expression. Before isolating the neuroepithelial cells, it is also important to scratch off the non-neuronal clusters in order to yield a more pure population of cells. If one wanted to characterize whether or not their cells are going into the neuronal lineage, various factors can be looked at such as Pax6 or Sox1. Pax6 turns on around day 10 during neural differentiation, and Sox1 turns on by 2 weeks after differentiation12,26,30.
This protocol is at the forefront of neural differentiation as it recapitulates many steps that take place during the development of the human nervous system. Without the use of caudalizing factors (retinoic acid, basic FGF), neuroepithelial cells efficiently differentiate into telencephalic progenitors, which coincides with neuroectoderm cells first acquiring a rostral phenotype during in vivo development31. These telencephalic progenitors possess a dorsal phenotype due to endogenous Wnt signaling which dorsalizes the cells27. This system generates dorsal progenitors which can then be further differentiated into glutamatergic cells. While this cell type is very important, it is by no means the only cell type that this system is capable of forming. For example, the addition of SHH has been shown to ventralize the cells, allowing them to differentiate into GABAergic cells27,32. It was even demonstrated that these hESC derived GABAergic neurons can be transplanted into mice and that they are able to correct locomotor defects due to brain lesions32. Because the protocol that is demonstrated in this article goes through the stepwise checkpoints, it offers a tool to produce a wide array of cells within the human nervous system which is merely limited by our understanding of development and our imaginations.
The ability to form human neurons from PSCs opens a large number of doors from both a basic science as well as a clinical perspective. Postmortem tissue sources are limited and the quality can vary, whereas human PSC differentiation allows researchers to yield an unlimited and consistent supply of cells to work with. With the advent of induced pluripotent stem cell (iPSC) technology1,3,33,34, it is now feasible to get hESC-like cells from patient fibroblasts including those with various diseases35-38. As shown by our group28 as well as many others, the successful generation of neurons from iPSC has been and will continue to be a unique and useful tool for those who have long sought after human models for disease and development. In addition, several groups have shown that PSC-derived neurons can model certain aspects of the disease process36,37,39-42 and thus can be utilized to screen therapeutic compounds43. This is particularly encouraging because without a good model system to test candidate drugs for the central nervous system with, only about 8% of them have been shown to be clinically effective44. However, this method and others like it could help to dramatically improve this number.
The authors have nothing to disclose.
The authors would like to thank Dr. Y. Sasai for generously providing the FOXG1 antibody. This work was supported by Connecticut Stem Cell Research Grants (08-SCB-UCHC-022 and 11-SCB24) and Spastic Paraplegia Foundation.
Reagent | Supplier | Catalog # |
Dulbecco’s modified eagle medium with F12 nutrient mixture (DMEM/F12) | Gibco | 11330-032 |
Knockout Serum Replacer | Gibco | 10828-028 |
L-glutamine (200 mM) | Gibco | 25030 |
Non Essential Amino Acids | Gibco | 1140-050 |
2-Mercaptoethanol (14.3 M) | Sigma | M-7522 |
Neurobasal medium | Gibco | 21103-049 |
N2 | Gibco | 17502-048 |
B27 | Gibco | 12587-010 |
Heparin | Sigma | H3149 |
Poly-L-ornithine hydrobromide (polyornithine) | Sigma | 116K5103 |
Laminin (human) | Sigma | L-6274 |
Laminin (mouse) | Invitrogen | 23017-015 |
FBS | Gemini | 100-106 |
Bovine serum albumin (BSA) | Sigma | A-7906 |
Dispase | Gibco | 17105-041 |
Collagenase | Invitrogen | 17104-019 |
Accutase | Innovative Cell Technologies | AT104 |
ROCK Inhibitor | Stemgent | 04-0012 |
SB431542 | Stemgent | 04-0010 |
Dorsomorphin | Stemgent | 04-0024 |
Fibroblast growth factor 2 (FGF2, bFGF) | Invitrogen | 13256-029 |
Trypsin inhibitor | Gibco | 17075 |
0.1% gelatin | Millipore | ES-006-B |
Foxg1 antibody | Dr. Y. Sasai | |
Hoxb4 antibody (1:50) | Developmental Studies Hybridoma Bank | I12 |
Pax6 antibody (1:5000) | Developmental Studies Hybridoma Bank | PAX6 |
Nkx2.1 antibody (1:200) | Chemicon | MAB5460 |
Tbr1 antibody (1:2000) | Chemicon | AB9616 |
vGLUT1 antibody (1:100) | Synaptic Systems | 135302 |
Brain derived neurotrophic factor (BDNF) | PrepoTech Inc. | 450-02 |
Glial derived neurotrophic factor (GDNF) | PrepoTech Inc. | 450-10 |
Insulin growth factor 1 (IGF1) | PrepoTech Inc. | 100-11 |
Cyclic AMP (cAMP) | Sigma | D-0260 |
Sonic hedgehog (SHH) | R&D | 1845-SH |
50 ml tubes | Becton Dickinson (BD) | 352098 |
15 ml tubes | BD | 352097 |
6 well plates | BD | 353046 |
24 well plates | BD | 353047 |
T25 flasks (untreated) | BD | 353009 |
T75 flasks (untreated) | BD | 353133 |
Coverslips | Chemiglass Life Sciences | 1760-012 |
6 cm Petri dishes | BD | 353004 |
9” glass pipetes | Fisher | 13-678-20D |
Steriflip filters (0.22 μM) | Millipore | SCGP00525 |
Stericup filters 1,000 ml (0.22 μM) | Millipore | SCGPU10RE |
Phase contrast microscope (Observer A1) | Zeiss | R2625 |
Carbon dioxide incubator (Hera Cell 150) | Thermo Electron Corporation | |
Biosafety hood (Sterilgard III Advance) | The Baker Company | |
Centrifuge (5702 R) | Eppendorf |