We describe the procedure for the in vitro differentiation of mouse embryonic stem cells into neuronal cells using the hanging drop method. Furthermore, we perform a comprehensive phenotypic analysis through RT-qPCR, immunofluorescence, RNA-seq, and flow cytometry.
We describe the step-by-step procedure for culturing and differentiating mouse embryonic stem cells into neuronal lineages, followed by a series of assays to characterize the differentiated cells. The E14 mouse embryonic stem cells were used to form embryoid bodies through the hanging drop method, and then induced to differentiate into neural progenitor cells by retinoic acid, and finally differentiated into neurons. Quantitative reverse transcription polymerase chain reaction (RT-qPCR) and immunofluorescence experiments revealed that the neural progenitors and neurons exhibit corresponding markers (nestin for neural progenitors and neurofilament for neurons) at day 8 and 12 post-differentiation, respectively. Flow cytometry experiments on an E14 line expressing a Sox1 promoter-driven GFP reporter showed that about 60% of cells at day 8 are GFP positive, indicating the successful differentiation of neural progenitor cells at this stage. Finally, RNA-seq analysis was used to profile the global transcriptomic changes. These methods are useful for analyzing the involvement of specific genes and pathways in regulating the cell identity transition during neuronal differentiation.
Since their first derivation from the inner cell mass of the developing mouse blastocysts1,2, mouse embryonic stem cells (mESC) have been used as powerful tools to study stem cell self-renewal and differentiation3. Furthermore, studying mESC differentiation leads to tremendous understanding of molecular mechanisms that may improve efficiency and safety in stem cell-based therapy in treating diseases such as neurodegenerative disorders4. Compared to animal models, this in vitro system provides many advantages including simplicity in practice and assessment, low cost in maintaining cell lines in contrast to animals, and relative ease in genetic manipulations. However, the efficiency and quality of differentiated cell types are often affected by different lines of mESCs as well as the differentiation methods5,6. Also, the traditional assays to evaluate differentiation efficiency rely on qualitative examination of selected marker genes which lack robustness and they therefore fail to grasp global changes in gene expression.
Here we aim to use a battery of assays for systematic assessment of the neuronal differentiation. Using both traditional in vitro analyses on selected markers and RNA-seq, we establish a platform for measurement of the differentiation efficiency as well as the transcriptomic changes during this process. Based on a previously established protocol7, we generated embryoid bodies (EBs) through the hanging drop technique, followed by induction using supraphysiologic amount of retinoic acid (RA) to generate neural progenitor cells (NPCs), which were subsequently differentiated to neurons with neural induction medium. To examine the efficiency of the differentiation, in addition to traditional RT-qPCR and immunofluorescence (IF) assays, we performed RNA-seq and flow cytometry. These analyses provide comprehensive measurement of the progression of the stage-specific differentiation.
1. mESC culture
2. EB, NPC, and neuron differentiation
3. Characterization of mESCs and differentiated cells
As a representation of our method, we performed an EB, NPC, and neuron differentiation experiment on E14 cells. E14 cells were cultured on γ-irradiated MEFs (Figure 1A) until the γ-irradiated MEF population diluted out. We confirmed the pluripotency of the E14 cells by performing Alkaline Phosphatase (AP) staining (Figure 1B) and later RT-qPCR (see below) for Nanog and Oct4 markers. The γ-irradiated MEF-free E14 cells were then induced for differentiation using the protocol outlined in Figure 2A. Briefly, differentiation media droplets of 20 µL containing 500 cells were seeded on the lid of the culture plate (see protocol section 2 for details). The EBs formed were then collected and placed in suspension in fresh differentiation media. From day 4 to day 8 of differentiation, 5 μM RA was added to the culture plates to induce NPCs. Differentiated EBs showed round shape and their size continued to increase during differentiation (Figure 2B). At day 8, the NPCs were harvested and trypsinized, and then the resulting single cell suspension was plated in a tissue culture chamber in DMEM/F12 medium with N2 supplement and later in B27 supplement. By day 10, NPCs differentiating into neurons appear to have elongated-cell shape (Figure 2B).
To further evaluate our differentiation experiment, we performed immunofluorescence (IF) experiments on E14 NPCs at day 8 and E14 neurons at day 12. We observed positive staining for nestin in NPCs and neurofilament (NF) signal for neurons (Figure 3A). Alternatively, RT-qPCR and RNA-seq confirmed the induction of NPC marker genes and loss of pluripotency genes in NPCs (Figure 3B,D,F). As a quantitative method to test the success of ESC differentiation, we differentiated a mouse ESC line expressing a Sox1 promoter-driven GFP reporter9, followed by flow cytometry analysis on ESCs and NPCs. We found that 58.7% of total cells at the NPC stage are GFP-positive while the GFP signal is 0.0% at the ESC stage (Figure 3C). To profile the transcriptomic changes during differentiation, RNA-seq experiments for E14 ESCs, EB day 3, and NPC day 8 were performed and revealed gene clusters associated with the respective stages (Figure 3D). The genes in the RNA-seq heat-map were sorted based on their expression levels to identify differentially expressed genes in the different stages during differentiation. Gene Ontology (GO) analysis for the four gene clusters showed that these clusters correspond to distinct cellular functions or pathways indicating that the three cell stages of mESC neuronal differentiation each have a group of genes that are highly expressed in their respective stage but not others (Figure 3E). For example, genes in Cluster 3 are highly expressed in E14 NPCs compared to other stages and correspond to pathways related to neuronal development. Clusters 1, 2, and 4 do not contain highly expressed genes related to any germ layer lineage specifications but they are related to cellular growth and proliferation. Thus, the RNA-seq and accompanying GO analysis showed that the E14 cells have differentiated into the neuronal lineage by day 8 of differentiation.
Figure 1: E14 ESCs in culture. (A) The light microscope images show E14 cells (black arrow) growing in colonies atop the γ-irradiated MEFs. E14 colonies continue to proliferate as seen in the colony size difference between day 1 and day 3 cultures. (B) Confirmation of the pluripotency of E14 ESCs by the alkaline phosphatase (AP) stain. Purple arrows indicate the mESCs that were positive for AP stain. Please click here to view a larger version of this figure.
Figure 2: E14 differentiation into EBs, NPCs, and neurons. (A) The schematic summarizes the major steps for differentiating E14 cells into EBs, NPCs, and neurons. (B) E14 ESCs cultured in medium without LIF and 2i in a suspension plate form individual spheres of EBs visible at day 2 where they continue to grow and expand in size in subsequent days. RA is added at day 4 of differentiation to induce the differentiation into NPCs. After 4 days of induction, these NPCs are plated for differentiation into neurons, which are shown in the bottom panel. Please click here to view a larger version of this figure.
Figure 3: Characterization of differentiated cells. (A) Immunofluorescence images in the top panel show NPC-containing EBs at day 8 probed for nestin (green) and nuclei (DAPI, blue). The bottom panel shows immunofluorescence images for neurons at day 12 probed for neurofilament (green). The red box in the merged images are zoomed in 3x for better view. (B) RT-qPCR analysis showing the pluripotency markers (Nanog and Oct4) and NPC markers (Pax6, NeuroD1, and Nes) of E14 ESCs and NPCs. Error bars are mean ± SD. (C) E14 cells expressing a Sox1 promoter-driven GFP reporter were differentiated into NPCs. Both the ESCs and NPCs at day 8 were quantified for positive GFP fluorescent signal with flow cytometry. (D) The heatmap shows the z-scores for the determined FPKM of the genes expressed in the ESC, EB, and NPC stages. Four distinct gene clusters were identified signifying groups of genes that are differentially expressed in either the ESC, EB, or NPC stage. (E) GO analysis was performed using the R package, clusterProfiler, on the four clusters identified in the RNA-seq. (F) The graph shows the FPKM values for three other pluripotency markers, Sox2, Klf4, and Myc as well as neural markers, NeuN, Map2, and Tubb3 for E14 cells at the ESC, EB day 3, and NPC day 8 stages. Please click here to view a larger version of this figure.
The method for neural differentiation of mouse embryonic stem cells has been established for decades and researchers have continued to modify the previous protocols or create new ones for various purposes7,10,11. We utilized a series of assays to comprehensively analyze the efficiency and progress of the differentiation stages of mESCs to neurons, which may be used in analysis of other lineage differentiation of mouse or human ESCs. Furthermore, our approaches have proved to be useful tools to evaluate the impact of specific genes or pathways on neuronal differentiation in vitro8.
With our methods, pluripotent un-committed ESCs treated with retinoic acid (RA) commit to the neural lineage with a high efficiency and are further induced to generate neurons7. To improve the successful differentiation of ES cells into neural cells and reduce the heterogeneity, it is important to keep the ES cells in an undifferentiated state12. Non-proliferative MEFs, treated with γ-irradiation or mitomycin C, function to maintain the pluripotency of the ESCs and provide a scaffold for their growth13,14. To obtain consistent results, we start each differentiation experiment with mESCs cultured on γ-irradiated MEFs. After a few passages, the γ-irradiated MEFs die out and the culture eventually becomes homogenous for mESC cells. Alternatively, the mESCs can be pre-plated on gelatin for about 45 min before γ-irradiated MEFs are seeded to better remove them on the next passage. Leukemia inhibitory factor (LIF) has long been used to maintain the pluripotency state of cultured mouse ES cells by activating the JAK/STAT pathway15,16,17. More recently, PD0325901 (PD, a MEK inhibitor) and CHIR99021 (CH, a GSK inhibitor) were found to provide additional pluripotency maintenance of the ES cells3,18. In our protocol, we culture mESCs with these inhibitors together with LIF to maintain high pluripotency of mESCs.
Another critical factor to achieve successful differentiation is the quality of EBs. We perform the differentiation of E14 cells by the hanging drop method, which has been applied by other investigators5,19,20. With this method, single ES cells are allowed to suspend in the differentiation medium droplet for 2 days where they spontaneously aggregate and form EBs. The resulting EBs are typically more well-defined in terms of their morphology (Figure 2B) compared to the method of suspending isolated mESCs in medium, which results in EB sizes in a much wider range in our experience (data not shown). To prevent EBs from attaching to the plates, it is important to do a low speed rotation starting from day 3 continuing through the differentiation process. The EBs are induced to differentiate into NPCs by treating them with RA. The resulting NPCs from RA treatment at EB day 4 are typically heterogenous for neural cells such as oligodendrocytes and astrocytes21. Using RT-qPCR or immunofluorescence, the NPC population can be probed for neuron and other neural lineage markers such as Gfap for astrocytes and Olig2 for oligodendrocytes. To further induce the differentiation of NPCs into neurons, the NPCs are cultured in optimal neuron medium where the most important components are the N2 and B27 supplements. N2 supplement mainly functions to help the NPCs to commit to the neuronal lineage while the B27 supplement functions to maintain the longevity of the neurons.
The samples can be collected across the differentiation period (e.g., ESC stage, EB day 2, EB day 4, NPC day 6, NPC day 8, neurons day 10, and neurons day 12) to track the differentiation process by performing RT-qPCR for pluripotency and ectoderm markers. Comparing the pluripotency markers such as Oct4, Nanog, Sox2, Klf4, and Myc between the different cell stages will verify the pluripotency of the mESCs (Figure 3B,F). To investigate the efficiency of the neural differentiation, markers for the mesoderm layer such as Hand1, Snai1, and Tbxt; and endoderm layer such as Eomes and Gata4 can also be probed for (data not shown). Further verification can be performed with immunofluorescence (IF) probing for NPC or neuronal markers (Figure 3A). However, these methods are not quantitative and biased toward the selected markers. To overcome these limitations, we incorporate flow cytometry and RNA-seq analyses (Figure 3C‒F). The cell line used in the flow cytometry experiment is a Sox1–GFP E14 cell line, which was used specifically in this experiment to assess the quality of the NPC differentiation procedure. Sox1 is one of the earliest specific neuronal marker during neuroectoderm development22 hence making it an excellent marker for NPC lineage. Sox1 can be probed for using RT-qPCR or Western blot to evaluate the NPC population. These analyses are particularly beneficial to investigate the differentiation defect caused by gene manipulation or chemical treatment.
It is important to note that there are a few limitations to our protocol presented here. First of all, we are only presenting the comprehensive analysis for one wild-type mESC cell line. Other ESC lines originating from mice or humans might require changes and further optimization in the protocol to ensure successful and efficient neuron differentiation. Secondly, we present an in vitro neuron differentiation method, which naturally bears its own set of limitations. As mentioned before, EBs are treated with a supraphysiological level of RA to drive them towards the NPC lineage. The resulting NPCs are then placed in neuron-optimum media to mimic the physiological conditions and encourage neuron lineage commitment, growth, and longevity. Here, N2 and B27 supplements are used to culture neurons but other supplements are also available such as NS2123 for similar purposes, which may alter the success and efficiency of neuron differentiation. These conditions are synthetically reconstituted in the cell culture assays, which may not fully represent physiological conditions. The quality of the EBs, NPCs, and neurons highly depend on the starting mESCs. mESCs that have been passaged for too many times and kept in culture for more than 1 week typically start to lose pluripotency and may not successfully undergo differentiation. Thus, maintaining the mESCs in an optimal condition is key in ensuring that they can effectively differentiate into EBs, NPCs, and neurons. Other neuron culture methods such as 3D models have also been proposed to better mimic physiological conditions24,25,26 sometimes at the expense of throughput and feasibility27,28. We believe our protocols are useful to characterize these 3D culture models.
The authors have nothing to disclose.
This work was supported by a grant from the NIH (1R35GM133496-01) to Z. Gao. We would like to thank Dr. Ryan Hobbs for the assistance in sectioning. We thank Penn State College of Medicine's core facilities, including the Genome Sciences and Bioinformatics, the Advanced Light Microscopy Imaging, and the Flow Cytometry. We also thank Dr. Yuka Imamura for the assistance in RNA-seq analysis.
0.05% Trypsin + 0.53mM EDTA 1X | Corning | 25-052-CV | |
0.1% Gelatin | Sigma | G1890-100G | Prepared in de-ionized water |
16% Paraformaldehyde | Thermo Scientific | 28908 | Diluted in 1X PBS |
40-μm cell strainer | Falcon | 352340 | |
Albumax | Thermo Fisher Scientific | 11020021 | |
AlexaFluor 488 goat anti-mouse IgG (H+L) | Invitrogen | A11001 | Antibody was diluted at 1:500 for IF |
Alkaline Phosphatase Staining Kit II | Stemgent | 00-0055 | |
AzuraQuant Green Fast qPCR Mix LoRox | Azura Genomics | AZ-2105 | |
B27 supplement | Thermo Fisher Scientific | 17504044 | |
BD FACSCanto | BD | 657338 | |
bFGF | Sigma | 11123149001 | |
BioAnalyzer High Sensitivity DNA Kit | Agilent | 5067-4626 | |
Chir99021 | Cayman Chemicals | 13122 | |
Chloroform | C298-500 | Fisher Chemical | |
DAPI | Invitrogen | R37606 | |
DMEM | Corning | 10-017-CM | |
DMEM/F12 medium | Thermo Fisher Scientific | 11320033 | |
EB buffer | Qiagen | 19086 | |
Ethanol | 111000200 | Pharmco | Diluted in de-ionized water |
Fetal bovine serum | Atlanta Biologicals | S10250 | |
Fisherbrand Superfrost Plus Microscope Slides | Fisher Scientific | 12-550-15 | |
HiSeq 2500 Sequencing System | Illumina | SY-401-2501 | |
Isopropanol | BDH1133-4LG | BDH VWR Analytical | Diluted in de-ionized water |
L-glutamine | Thermo Fisher Scientific | 25030024 | |
LIF | N/A | N/A | Collected from MEF supernatant |
m18srRNA primers | IDTDNA | N/A | 5'-GCAATTATTCCCCATGAACG-3' 5'-GGCCTCACTAAACCATCCAA-3' |
MEM Non-essential amino acids | Corning | 25-025-Cl | |
mNanog primers | IDTDNA | N/A | 5'-AGGCTTTGGAGACAGTGAGGTG-3' 5'-TGGGTAAGGGTGTTCAAGCACT-3' |
mNes primers | IDTDNA | N/A | 5'-AGTGCCCAGTTCTAGTGGTGTCC-3' 5'-CCTCTAAAATAGAGTGGTGAGGGTTG-3' |
mNeuroD1 primers | IDTDNA | N/A | 5'-CGAGTCATGAGTGCCCAGCTTA-3' 5'-CCGGGAATAGTGAAACTGACGTG-3' |
mOct4 primers | IDTDNA | N/A | 5'-AGATCACTCACATCGCCAATCA-3' 5'-CGCCGGTTACAGAACCATACTC-3' |
mPax6 primers | IDTDNA | N/A | 5'-CTTGGGAAATCCGAGACAGA-3' 5'-CTAGCCAGGTTGCGAAGAAC-3' |
N2 supplement | Thermo Fisher Scientific | 17502048 | |
Nestin primary antibody | Millipore | MAB5326 | Antibody was diluted at 1:200 for IF |
Neural basal | Thermo Fisher Scientific | 21103049 | |
Neurofilament primary antibody | DSHB | 2H3 | |
NEXTflex Illumina Rapid Directional RNA-Seq Library Prep Kit | BioO Scientific | NOVA-5138-07 | |
PD0325901 | Cayman Chemicals | 13034 | |
Penicillin/streptomycin | Corning | 30-002-Cl | |
Phosphate-buffered saline (PBS) | N/A | N/A | Prepared in de-ionized water |
– Potassium chloride | P217-500G | VWR | |
– Potassium phosphate monobasic anhydrous | 0781-500G | VWR | |
– Sodium chloride | BP358-10 | Fisher Bioreagents | |
– Sodium phosphate, dibasic, heptahydrate | SX0715-1 | Milipore | |
Random hexamer primer | Thermo Scientific | SO142 | |
Retinoic acid | Sigma | R2625 | Prepared in DMSO |
Sodium pyruvate | Corning | 25-000-Cl | |
Sucrose | Sigma | 84097 | Diluted in 1X PBS |
SuperScript III Reverse Transcriptase | Invitrogen | 18064022 | |
Tissue-Tek O.C.T. compound | Sakura | 4583 | |
TriPure Isolation Reagent | Sigma-Aldrich | 11667165001 | |
TruSeq Rapid | Illumina | 20020616 | |
β-mercaptoethanol | Fisher BioReagents | BP176-100 |