Optimization of Transcription Factor-Driven Neuronal Differentiation from Human Induced Pluripotent Stem Cells for Disease Modelling and Drug Screening

Abstract

Progress of human brain in vitro models stands as a keystone in neurological and psychiatric research, addressing the limitations posed by species-specific differences in animal models. The generation of human neurons from induced pluripotent stem cells (iPSCs) using transcription factor reprogramming protocols has been shown to reduce heterogeneity and improve consistency across different stem cell lines. Despite notable advancements, the current protocols still exhibit several shortcomings. This study focuses on standardizing and optimizing the procedure for iPSC-derived glutamatergic neurons generation through the inducible overexpression of Neurogenin-2. Noteworthy refinements include stringent scrutiny of genomic rearrangements post-fibroblast reprogramming, selection of a homogeneously integrated NGN2-cassettes population, and the incorporation of an intermediate step during neuronal differentiation to store neuronal progenitors. The neural culture showed a high degree of neuronal maturation and consistency, as shown by single-cell and network electrophysiological recordings. These advancements aim to provide more reliable tools for disease modelling and drug screening in neurological disorders.

Keywords: Electrophysiological recordings; FACS sorting; NGN2-mediated neuronal differentiation; Transcription factor–driven differentiation; iGluNeurons; iPSCs.

Fig. 1

Generation and characterization of iPSCs from fibroblasts. (A) Timeline of the generation of iPSCs, starting from the manual picking of emerging colonies after virus infection, further expansion, and full characterization. (B) Representative immunostaining of the undifferentiated state markers NANOG, SOX2, OCT3/4, and SSEA4 (scale bar 50 µM). (C, D) Images from 4 cell lines were quantified using ImageJ for colocalization of nuclear and surface markers (expressed as percentages of DAPI-positive nuclei; means ± SEM). (E) Comparison of SNP array results between fibroblasts and iPSCs reprogrammed from the same cell line

Fig. 2

Lentiviral infection with “all-in-one Tet-on”-vector NGN2 and FACS sorting of iPSC-NGN2 positive cells. (A) Timeline of the lentiviral infection with "all-in-one Tet-on” vector NGN2 and subsequent FACS sorting of cells. Briefly, after infection and selection, NGN2 and GFP expression were transiently induced and, the next day, GFP-positive cells were sorted and plated either as a pool or as single clones. (B) Representative dot plots of a single cell line showing the various steps of FACS sorting, starting from non-induced cells to single-cell seeding. These plots highlight how the selected cell region corresponds to the largest cell population exhibiting the same fluorescence intensity (middle panels). (C) Histograms of the percentage of GFP-positive cells before and after sorting, comparing pool and single clone strategies for all 7 IPSC lines (4 iPSC lines from healthy controls and 3 iPSC from patients carrying a mutation in the PRRT2 gene) tested along the study. Data are represented as means ± SEM

Fig. 3

IPSCs NGN2 pre-differentiation and differentiation steps. (A) Schematic protocol to differentiate iPSCs into iGluNeurons. Cells were pre-differentiated for 3 days, then cryopreserved. iGluNeurons were generated by plating pre-differentiated cells with rat astrocytes. (B, C) Representative immunostaining of neuroprogenitor cells using Nestin (B) and SOX2 (C) markers. (D) SOX2- and Nestin-positive cells expressed in percent of DAPI-positive nuclei for all IPSC lines tested. (E) Percentage of viability of neuroprogenitors after thawing and trypsinization. (F-L) Representative immunostaining of iGluNeurons performed between DIV 35 and DIV 49 for mature neuronal markers: NeuN (F), Tubulin beta III (G), SMI312 (H), MAP2 and NaV channels (I), synaptophysin and MAP2 (J), AnkyrinG and MAP2 (K), Synapsins I/II and MAP2 (L) (scale bar 10 µM)

Fig. 4

Action potential firing activity of individual iGluNeurons. Action potential firing activity in iGluNeurons. (A) iGluNeuron probed by patch pipette. (B-F) In vitro whole cell current clamp recording of the action potential firing activity elicited with the injection of depolarizing currents at five timepoints of in vitro differentiation: DIV 21 (B), 28 (C), 35 (D), 42 (E) and 49 (F). (G) Plot of the number of action potentials evoked (y axis) vs the amount of depolarizing current linearly injected (x axis). (H-L) Quantification of the main action potential parameters evaluated at each timepoint of differentiation. Mean firing frequency (Hz; H), Instantaneous firing frequency (Hz; I), rheobase (pA; J), action potential amplitude (mV; K), action potential width (ms; L). (M-Q) In vitro whole cell current clamp recording of the action potential spontaneous activity recorded at the subthreshold potential of −40 mV at each timepoint. (R) Bar plot showing the mean spontaneous firing frequency of iGluNeurons at each timepoint (Hz). N = 3 iPSC cultures from 3 healthy subjects, in 3 independent experiments (an average of n = 15 across all time points). Data are expressed as Mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001

Fig. 5

Voltage-gated sodium currents expression in iGluNeurons. (A) Exemplificative ramp protocol recorded from −100 mV to + 100 mV, (Vholding = −70 mV, duration of 500 ms) by iGluNeuron at 42 (green) and 49 (orange) DIV (B-F) Exemplificative I-V relationships recorded from −100 to + 80 mV (Duration of 500 ms, Vholding = −70 mV, Δ = 10 mV) by iGluNeurons at different timepoints of in vitro differentiation. (G-K) Representative whole cell current families recorded from iGluNeurons at each timepoint of in vitro differentiation previously described. As before, cells were clamped at −70 mV. (L-M) Mean I-V relationship of the inward (L) and outward (M) current densities J (pA/pF) at each voltage step obtained in each experimental timepoint. (N-O) Bar plot showing the current density measured at −30 (N) and + 80 mV (O) in each experimental condition. N = 3 iPSC cultures from 3 healthy subjects, in 3 independent experiments (an average of n = 15 for each time points). Data are expressed as Mean ± SEM. *P < 0.05; **P < 0.01; **P < 0.001; ****P < 0.0001

Fig. 6

Spontaneous electrophysiological activity of the iGluNeuron network. iGluNeurons were plated on MEA chips. (A) Raster plots of 60 s spontaneous activity of a representative neuronal culture at DIV 49 overlapped with the cumulative instantaneous firing rate profile (bin = 10 ms). Spikes are represented by black bars; vertical bands indicate network burst events overlapped with the cumulative instantaneous firing rate profile (bin = 10 ms). (B-F) Scatter plots of Mean Firing Rate (MFR; B), percentage of Active Electrodes (AE; inset in B), Mean Bursting Rate (MBR; C), percentage of Random Spikes (RS; D), Burst Duration (BD; E), and Network Bursting Rate (NBR; F). Data are shown as means ± SEM. (G) Aligned network bursts with respect to the peak of the longest network burst of a representative network at DIV 49, overlapped with the respective cumulative instantaneous firing rate profile (bin = 10 ms). N = 3 iPSC cultures from 3 healthy subjects, in 3 independent experiments (total n = 13). The mean values and statistical data are provided in Supplementary Tables 1-3