Rapid and robust generation of functional oligodendrocyte progenitor cells from epiblast stem cells

Abstract

Myelin-related disorders such as multiple sclerosis and leukodystrophies, for which restoration of oligodendrocyte function would be an effective treatment, are poised to benefit greatly from stem cell biology. Progress in myelin repair has been constrained by difficulties in generating pure populations of oligodendrocyte progenitor cells (OPCs) in sufficient quantities. Pluripotent stem cells theoretically provide an unlimited source of OPCs, but current differentiation strategies are poorly reproducible and generate heterogenous populations of cells. Here we provide a platform for the directed differentiation of pluripotent mouse epiblast stem cells (EpiSCs) through defined developmental transitions into a pure population of highly expandable OPCs in 10 d. These OPCs robustly differentiate into myelinating oligodendrocytes in vitro and in vivo. Our results demonstrate that mouse pluripotent stem cells provide a pure population of myelinogenic oligodendrocytes and offer a tractable platform for defining the molecular regulation of oligodendrocyte development and drug screening.

Figure 1. Efficient differentiation of epiblast stem cells into region-specific neuroepithelial cells in five days

Treatment of (a) pluripotent EpiSCs which express (b) Oct3/4 (green) with factors to block both BMP and activin/Nodal signaling results in their progression to (c) neural rosettes expressing (d) Sox1 (red) and Pax6 (green). The five day treatment regimen results in patterned neural rosettes expressing the (e) region-specific transcription factors Olig2 (red) and Nkx2.2 (green) which are (f) normally expressed in the ventral, ventricular zone of the mouse embryonic spinal cord. (g) Genome wide transcriptional profiling during the transition of EpiSCs to patterned neural rosettes confirmed a rapid down-regulation of pluripotency genes and up-regulation of genes specific to the developing neuroectoderm. Scale bars, 100 µm (a–e) and 50 µm (f).

Figure 2. Highly expandable OPCs derived from patterned EpiSC-derived neuroectoderm

Patterned neural rosettes derived from EpiSCs were dissociated and grown on a laminin substrate in the presence of FGF2, PDGF-AA, and SHH resulting in the formation of OPCs. (a) Phase contrast image of EpiSC-derived OPCs showing characteristic bi-polar morphology. Immunostaining revealed that the OPCs were highly pure as (b) 89.9% co-expressed transcription factors Olig2 (red) and Nkx2.2 (green) and (c) 89.2% expressed Sox10 (green) without sorting or selection (n>600 cells from random fields were manually counted for each marker). (d) EpiSC-derived OPCs could be extensively expanded at least 8 passages. Shown is a graph depicting cumulative cell number counts over multiple passages (passage denoted ‘p’). (e) Flow cytometric analyses revealed that both low passage (p2) and high passage (p12) EpiSC-derived OPCs robustly co-expressed OPC surface markers NG2 and PDGFRα. (f) Genome wide transcriptional profiling during the transition of EpiSC-derived neural rosettes to OPCs revealed a rapid down-regulation of rosette-specific genes and up-regulation of genes specific to OPCs. Scale bars, 50 µm.

Figure 3. EpiSC-derived OPCs differentiate into oligodendrocytes in vitro

(a–b) EpiSC-derived OPCs cultured in differentiation-inducing media for three days (a) adopt a mature oligodendrocyte (‘oligo’) morphology by phase contrast and (b) co-express O4 (green) and MBP (red). (c) Genome wide transcriptional profiling during the transition of EpiSC-derived OPCs to oligodendrocytes revealed a rapid down-regulation of OPC-specific genes and up-regulation of genes specific to oligodendrocytes. (d) EpiSC-derived OPCs seeded on in vitro cultured embryonic mouse cortical neurons showed (see arrows) extension of MBP+ (green) processes along the βIII-tubulin+ (red) nerve axons, with DAPI (blue). Scale bars, 50 µm.

Figure 4. EpiSC-derived OPCs are myelinogenic

(a) Coronal sections of early postnatal (P2-4) mouse forebrain cultured for 13 days show extensive MBP+ (black) segments in sections from wild type mice and the complete lack of MBP+ (black) segments in shiverer (Mbp^shi/shi) mutants (lethally hypomyelinated due to lack of Mbp). Images were taken within the region outlined by the dotted box. (b) Injection of EpiSC-derived OPCs into shiverer forebrain sections reveal extensive MBP+ (black) myelin sheaths after 10 days. No differences were noted in the myelinogenic capacity of low passage (p4) and high passage (p10) EpiSC-derived OPCs. (c) Toluidine blue stained sections (1µm) as well as electron microscopic (EM) analysis showed that EpiSC-derived OPCs produce compact myelin. (d) Injection of EpiSC-derived OPCs into early post-natal shiverer mutant mice reveals clear myelination both in the corpus callosum as well as the contralateral striatum (day 21 after injection). CC = corpus callosum. Scale bars, 100 µm (a,b), 25 µm (c), and 50 µm (d).

Figure 5. Screening for extrinsic signals that regulate the fate of EpiSC-derived OPCs

(a) Stimulation of EpiSC-derived OPCs (Olig2+ and O4-negative) with a cocktail of defined factors results in their robust differentiation into Olig2+ and O4+ oligodendrocytes in 2 days. (b) Activation of the Notch pathway (with Jag1) signaling cascade during differentiation does not drastically alter the normal transition of OPCs to oligodendrocytes. (c) Inhibition of GSK3β (with CHIR99021), a negative regulator of the canonical Wnt/β-catenin signaling pathway, blocks the transition of OPCs into oligodendrocytes. (d) Stimulation with BMP4 and LIF results in the respecification of OPCs into GFAP+ astrocytes. (e) Quantification of the effect of GSK3β inhibition using CHIR99021 to block the differentiation of OPCs into O4+ oligodendrocytes (>2.4×10³ cells were manually scored from random fields of triplicate wells from two independent experiments and shown +/− s.d.). Scale bars, 50 µm.