Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Jun;14(6):621-628.
doi: 10.1038/nmeth.4291. Epub 2017 May 15.

Generation of Pure GABAergic Neurons by Transcription Factor Programming

Affiliations
Free PMC article

Generation of Pure GABAergic Neurons by Transcription Factor Programming

Nan Yang et al. Nat Methods. .
Free PMC article

Abstract

Approaches to differentiating pluripotent stem cells (PSCs) into neurons currently face two major challenges-(i) generated cells are immature, with limited functional properties; and (ii) cultures exhibit heterogeneous neuronal subtypes and maturation stages. Using lineage-determining transcription factors, we previously developed a single-step method to generate glutamatergic neurons from human PSCs. Here, we show that transient expression of the transcription factors Ascl1 and Dlx2 (AD) induces the generation of exclusively GABAergic neurons from human PSCs with a high degree of synaptic maturation. These AD-induced neuronal (iN) cells represent largely nonoverlapping populations of GABAergic neurons that express various subtype-specific markers. We further used AD-iN cells to establish that human collybistin, the loss of gene function of which causes severe encephalopathy, is required for inhibitory synaptic function. The generation of defined populations of functionally mature human GABAergic neurons represents an important step toward enabling the study of diseases affecting inhibitory synaptic transmission.

Figures

Figure 1
Figure 1
Rapid and efficient generation of human inhibitory iN cells with Ascl1, Dlx2 and Myt1l. (a) Identification of transcription factor (TF) combinations to generate GABAergic neurons. TF combinations give rise to different fractions of iN cells (pie charts) with spontaneous EPSCs (magenta), IPSCs (blue) or mixed postsynaptic currents (green) and varying frequencies (mean ± s.e.m.; n = 5–7 cells) of EPSCs and IPSCs (histograms). Filled circles represent values measured from individual cells. (b) Representative images of AMD-iN cells 5 weeks after infection (4 weeks after coculture with mouse glia). iN cells were detected with antibodies against MAP2 and Human Nuclear Antigen (HuNu). Bar graphs show percentage of HuNu-positive cellsthat express neuronal marker MAP2 (n = 3 independent experiments). (cg) MAP2-expressing AMD-iN cells colabel with additional GABAergic markers including DLX proteins, GAD1/2 (GAD67/GAD65), GABA and vGAT. (h) Repetitive series of action potentials (APs) in response to step-current injections in AMD-iN cells. (i) Intrinsic and active membrane properties of AMD-iN cells observed as AP generation in response to amplitude of step-current stimulation (left); resting membrane potential (Vrest), AP threshold (APt), AP height (APh) and after hyperpolarization potentials (AHP) (second left); membrane capacitance (Cm, second right) and input resistance (Rm, right). Values are mean ± s.e.m. (n = 8–18 cells). (j) Spontaneous IPSCs recorded from AMD-iN cells. (k) Evoked IPSCs in AMD-iN cells as elicited by single (top left) or a train (top right) of stimulation (triangles). Inset, expanded view of asynchronous delayed IPSCs. Evoked IPSC amplitude (bar graph) and synaptic depression (filled circles) are presented as mean ± s.e.m. (l,m) AMD-iN cells received spontaneous excitatory (red asterisk and example trace) and inhibitory (blue asterisk and example trace) synaptic inputs when cocultured with Ngn2-iN cells. Pie chart indicates percentages of synaptic event types (IPSC, blue; EPSC, red) recorded from an AMD-iN cell cocultured with Ngn2-induced neurons. Scale bars, 50 um (b,cf,l).
Figure 2
Figure 2
Transient expression of Ascl1 and Dlx2 (AD) generates GABAergic iN cells from human ES cells. (a) Omission of Myt1l slows the formation of neuronal morphologies, but exogenous Myt1l is not necessary for neuronal induction mediated by Ascl1 and Dlx2. (b) Endogenous MYT1 family members are induced in AD-transduced human ES cells (qRT-PCR results expressed as log2 fold change between AD-transduced cells and human ES cells. Mean ± s.e.m., n = 3 biological replicates). (c) MYT1L protein is present in AD-iN cells 5 weeks postinfection. (d) Lentiviral vectors and timeline for GABAergic iN cell induction. Cells are transduced with three viruses expressing rtTA, a fused Ascl1-T2A-puromycin resistance gene, and a fused Dlx2-hygromycin resistance gene. (e) AD-iN cells express telencephalic marker FOXG1 and GABAergic neuron markers (GABA, DLX proteins, GAD1/2 (GAD67/65)). (f) MAP2-positive cells coexpress GABA, DLX proteins or GAD1/2 (GAD65/67) after 5 weeks of conversion. (g) qRT-PCR analysis in AMD and AD-iN cells (5 WPI) cultured on mouse glia (three biological replicates each), mouse glia, and Ngn2-iN cells (i) or EGFP-sorted AD-iN cells (ii and iii) after coinfection of hES cells with a constitutive EGFP virus. (h) AD- iN cells cocultured with mouse glia for 4 weeks show highly branched MAP2-positive neurons that coimmunostain for CB, CR and SST. Expression of CB and SST or CR and SST is largely nonoverlapping. (i) Quantification of marker overlap. (mean ± s.e.m., n = 3 biological replicates). (j) PV-expressing AD-iN cells were detected. (k) Single-cell qRT-PCR analysis of 64 AD-iN cells for genes indicated on the left. Ct, crossing threshold (g,k). Scale bars, 50um (a,e,h,j).
Figure 3
Figure 3
Functional maturation and synaptic integration of AD-iN cells in vitro and long-term stability of GABAergic fate in vivo. (a) Current injection-induced action potentials recorded from AD-iN cells as replated on mouse glia for 4 to 6 weeks (red and blue, respectively) or when cocultured with Ngn2-induced excitatory neurons for 6 weeks (brown). (b) AP-generation properties (i) measured as number of APs generated with current-pulse amplitude (left) and as resting membrane potential (Vrest), AP threshold (APt), AP height (APh) and after-hyperpolarization potentials (AHP) (right); and intrinsic properties (ii) measured as membrane capacitance (Cm, left) and input resistance (Rm, right). All values are mean ± s.e.m. (n = 14 cells, 5 WPI; n = 12 cells, 7 WPI;n = 32 cells, 7 WPI Ngn2-iN coculture; Student’s t-test; *, P > 0.05). (c) Sample traces (left) of spontaneous IPSCs and event amplitude and frequency (mean ± s.e.m., right), as recorded from AD-iN cells cocultured with mouse glia for 4 weeks (red) or 6 weeks (blue). (d) Sample traces of train-stimulation-induced evoked IPSCs (left) and first IPSC amplitude and total charge transfer from IPSC trains (mean ± s.e.m., right) recorded from AD-iN cells when coculturedwith mouse glia for 4 weeks (red) or 6 weeks (blue). (c,d) Filled circles represent individual cells (n = 15 cells per condition; Student’s t-test; *, P > 0.05). (e) Representative traces (left) of spontaneous EPSCs and IPSCs recorded from GABAergic AD-iN cell when cocultured with Ngn2-induced glutamatergic neurons for 7 weeks. Arrowheads, network activity. Inset, expanded view of boxed area. Asterisks, EPSC-like events with fast kinetics (magenta) and IPSC-like events with slow kinetics (blue). Pie charts indicate cell fraction with EPSCs (magenta), IPSCs (blue) or both types of responses (mixed, green; n = 54 cells). (f) Representative image of grafted AD-iN cells identified by human nuclei antibodies (magenta) and expressed GABAergic neuron markers including GABA, CR, CB, SST, NPY and PV (green). Histogram shows percentage of human cells that express different markers.
Figure 4
Figure 4
Induced GABAergic neurons for human neurological disease modeling. (a) Collybistin expression is reduced by shRNA knockdown (hairpin numbers 1–5) compared with control (Ctrl) in iN cells as measured by qRT-PCR (mean ± s.e.m.). (b) Representative traces of postsynaptic currents induced by exogenous application of GABA (1 mM). (c,d) Cumulative (cum.) plots (i) and summary graphs (ii) show reduction in average peak amplitude (c) and total charge-transfer (d) of GABA puff-induced IPSCs in human Ngn2-iN cells subjected to collybistin knockdown. (e) Patch–clamp configuration for postsynaptic recordings performed on day 28–30 human neurons expressing collybistin shRNAs. Collybistin KD iN cells (EGFP positive) were cocultured with Ascl1, Dlx2, Myt1l-generated iN cells (EGFP negative, black arrowheads). Rec, recording electrode. Scale bars, 15 μm. (f) Sample traces of GABAAR- mediated spontaneous IPSCs recorded from control (top) or collybistin shRNA no. 4 (Sh# 4)-infected neurons (bottom). (g) Cumulative plots (left) and average graphs (right) representing mean ± s.e.m. values of sIPSC amplitude (amp, i) and event frequency (freq, ii), recorded from control versus collybistin shRNA no. 4 – infected neurons in (e). Numbers inside bars indicate total number of independent batches (for batch-wise comparisons) or total number of cells recorded/number of batches. Two-tailed, unpaired Student’s t-test; ***, P < 0.005; **, P < 0.01; *, P < 0.05; n.s., not significant. P > 0.05 was used for all comparisons except batch-wise comparisons, where paired t-tests were performed. For cumulative plots, circles represent average values recorded from individual cells.

Similar articles

See all similar articles

Cited by 49 articles

See all "Cited by" articles

Substances

Feedback