br Experimental Procedures br Author Contributions br Acknow

Experimental Procedures
Author Contributions
Acknowledgments
Introduction Multiple sclerosis (MS) is a chronic, inflammatory, demyelinating disease of the CNS that is distinguished by recurrent episodes of focal inflammatory demyelination and consequent neurological symptoms (relapsing remitting MS [RRMS]). Although relapses usually resolve in spontaneous remission, RRMS can evolve with time into a secondary progressive form characterized by irreversible accumulation of disabilities. Furthermore, patients affected by the most severe primary progressive form (PPMS) experience a steady neurological decline from the onset of the disease (Antel et al., 2012). Currently available treatments targeting the immune system are highly effective at reducing or even stopping the intermittent episodes of inflammation, but they do not influence the course of progressive MS. Therapeutic options for PPMS patients are limited to symptomatic treatments and the long-term prognosis is generally poor (Rice et al., 2013). Clearly, the unsolved challenge in the MS field is to develop neuroprotective and remyelinating strategies for the treatment of progressive MS patients (Hauser et al., 2013). The generation of patient-specific cells from induced pluripotent stem cells (iPSCs) or somatic cell nuclear transfer has recently emerged as a promising strategy for the development of autologous cell therapies (Goldman et al., 2012; Yamada et al., 2014). iPSC-derived oligodendrocyte progenitor cells (OPCs) were shown to successfully remyelinate and rescue a hypomyelinated mouse model, raising the possibility of future clinical trials (Wang et al., 2013). However, oligodendrocyte differentiation protocols are still inefficient and require over 120 days in culture. Therefore, an improved protocol that can generate large numbers of purified OPCs in a relatively short time is highly desirable. Moreover, this protocol should be reproducible and highly efficient among different iPSC lines, including those derived from MS patients. We have pioneered the efficient and robust generation of iPSC-derived OPCs from PPMS patients. Our protocol recapitulates the major steps of oligodendrocyte differentiation from neural stem cells to OLIG2+ progenitors and finally to O4+ OPCs in a significantly shorter time than the 120–150 days required by the most recently published protocols (Wang et al., 2013; Stacpoole et al., 2013). Furthermore, O4+ OPCs were able to differentiate into MBP+ mature oligodendrocytes in vitro and to myelinate mglur in vivo when injected into immunocompromised shiverer (shi/shi) mice. No abnormal growths were observed. Our results provide a proof of concept that transplantation of iPSC-derived, patient-specific cells for remyelination is technically feasible.
Results
Discussion In this work, using a fast and highly reproducible protocol, we demonstrated efficient in vivo myelination of neurons by iPSC-derived OPCs from PPMS patients. A previous report on MS-derived iPSCs showed that oligodendrocytes could be differentiated in vitro from an integrating, retrovirally reprogrammed iPSC line from one 35-year-old RRMS patient (Song et al., 2012); here, we generated four integration-free iPSC lines from PPMS patients of both sexes and with ages ranging from 50 to 62 years. Since most protocols for oligodendrocyte differentiation have been optimized using only one or two hESC lines and their reproducibility with iPSC lines is controversial (Alsanie et al., 2013), we tested our protocol with two hESC and four hiPSC lines derived from PPMS patients. Previous work has elegantly shown that iPSC-derived OPCs from healthy controls were able to rescue a mouse model of hypomyelination, although differentiation to the O4 stage was rather inefficient and required more than 120 days (Wang et al., 2013). We provide an improved differentiation protocol and further proof that patient-specific iPSC lines can be successfully used to generate oligodendrocytes. We obtained 44%–70% O4+ cells in all lines after 75 days of differentiation, compared with the minimum of 120 days required according to previous reports (Wang et al., 2013; Stacpoole et al., 2013). There are several critical differences between our approach and previously published protocols. First, we began neural induction with dual SMAD inhibition in adherent as opposed to suspension cultures (Nistor et al., 2005; Hu et al., 2009b; Wang et al., 2013). Using this approach, we started with only 10,000 cells/cm2 of iPSCs at day 0 and achieved a great expansion of neural progenitors, ultimately generating an abundance of human OPCs. The use of RA and SHH as caudalizing and ventralizing patterning agents recapitulates the signals that are present around the pMN domain of the spinal cord, from which motor neurons and oligodendrocytes are believed to arise (Hu et al., 2009a). The high efficiency in the generation of OLIG2+ cells at d12 can be explained by the synergistic effect of activin/nodal receptor kinase inhibition, BMP4 inhibition, and RA and SHH signaling (Patani et al., 2011; Miller et al., 2004). The optimal concentration of RA in our hands is 100 times less than the concentration commonly used by other groups (Nistor et al., 2005; Izrael et al., 2007; Gil et al., 2009; Hu et al., 2009b). Surprisingly, induction with RA alone (without exogenous SHH) generated a large population of OLIG2+ cells. While the combination of RA and fibroblast growth factor (FGF) signaling is known to promote OLIG2 expression during chicken development and has been used for in vitro differentiation of both hESCs and hiPSCs (Novitch et al., 2003; Nistor et al., 2005; Pouya et al., 2011), we achieved OLIG2 induction in the absence of any exogenous FGF in our culture conditions. We show that RA, synergistically with the dual inhibition of SMAD proteins, upregulates OLIG2 (Figure S1C), possibly by stimulating the endogenous expression of SHH (Figure S1A). We confirmed that SAG is an efficient replacement for SHH and indeed showed superior efficacy in our hands (Stacpoole et al., 2013). The addition of HGF to the medium, although not essential, appeared to slightly improve the differentiation efficiency (data not shown; Hu et al., 2009c). Finally, the transition from adherent cultures to spheres proved to be a critical step to enrich the OLIG2+ population and possibly restrict differentiation to the oligodendrocyte lineage.