In recent years, with the development of regenerative medicine, stem cell transplantation for the treatment of central nervous system disease is considered to be a way of great potential. NSCs derived from iPSCs seem to be the ideal seed cells for transplantation therapy for neurologic diseases. Because iPSCs share similar characteristics and potential with ESCs, we may study iPSCs by applying the same protocols that had been previously developed in ES cells.
Various culture protocols for the generation of NSCs from iPSCs have been reported in many studies, and the main methods are the following: RA-mediated induction , stromal cell coculture, conditioned medium induction, and serum-free-medium induction . RA has been demonstrated to induce significant neural differentiation of ESCs [33–35]. In 1995, Bain  used RA to induce mouse ESCs into nerve cells successfully, and these cells had the electrophysiological properties and functions of nerve cells. Later, Zhou and Bahrvand [36, 37] reported that the use of RA could promote human ESCs differentiation into neural stem cells, but efficiency was low, and the survival capacity of RA-induced neuronal cells was limited . Stromal cell coculture and conditioned-medium induction are also able to induce ESCs to differentiate into neural cells, but the two methods have a common shortcoming, in that the medium composition is complex, and this complicates the study of its mechanism.
In recent years, induction protocols using cytokines and serum-free medium have been established. Koch  successfully used serum-free medium to induce human ESCs into NSCs. The serum-free medium contains N2, insulin, FGF, and B27, and the data showed that FGF was able to stimulate proliferation and differentiation of a variety of cell types originated from multigerminal layers. A study  showed that neuroepithelial precursor cells differentiated from ESCs proliferate in the presence of basic fibroblast growth factor (bFGF), and differentiate into both neurons and glia after withdrawal of bFGF. Daadi  successfully induced ESCs into NSCs by using serum-free medium containing bFGF, leukemia inhibitory growth factor (LIF), and EGF. These factors are known to stimulate the proliferation of NSCs, and LIF has the potential to maintain the pluripotency of NSCs and to prevent NSCs from differentiating into neurons. Definite components of serum-free medium are beneficial to study the relation between cell growth and medium components.
In the present study, we established a modified four-stage culture system to induce iPSCs to NSCs by taking the advantages of previously developed protocols, by using RA combined with serum-free medium containing bFGF, EGF, LIF, B27, and heparin sodium to induce iPSCs to NSCs in adherent culture. The results showed that this system was stable and effective to induce neural differentiation of iPSCs. Induced cells expressed high levels of NSC markers, Nestin and Sox2, and have been amplified and passaged in vitro for more than 35 cell generations.
In addition, NSCs can be expanded and differentiated in both suspension and adherent cultures. However, it is difficult to control the quality and quantity of cells in the process of suspension culture, because ESC-derived neurosphere cells will lose the potential of self-renewal and differentiation gradually during the long suspension culture, and the types of neurosphere cells are very diverse [40–42]. The adherent culture may avoid these shortcomings.
After implantation into the brain in a rat model of stroke, iPSC-derived NSCs were able to survive and express NSC makers. One week after implantation, implanted cells expressed mainly β-tubulin in addition to Nestin; after 2 weeks, besides continuing to express Nestin and β-tubulin, a few implanted cells expressed GFAP (maker of astrocytes). These results show that grafted NSCs have the potential to differentiate into neurons and glia, and create the possibility of repairing brain injury or other neurologic defects through a neural cell-replacement strategy.
Tumorigenesis is a big challenge in stem cell-replacement therapy. Some animal studies demonstrated that direct implantation of ESCs or iPSCs led to tumor formation [21, 43]. In this study, two rats survived for more than 3 months, and no abnormal proliferation of cells was observed (data not shown). It implied that the implanted cells were not susceptible to tumorigenesis. Although we did not observe tumor formation in this study, it could be that a short period of observation lasting for only 2 weeks was not long enough for tumor development and formation. Therefore, the long-term safety of iPSC-derived NSCs implantation is still a serious problem.
Grafted stem cells have the potential to migrate in vivo. A number of studies [44, 45] observed migration of various implanted stem cells into lesions of the brain, and indicated that possibly stem cells are targeted by inflammatory chemotactic factors and cytokines extracted from ischemic brain tissue. This feature of chemotaxis is the pathophysiologic basis of repairing brain injury by cell-implantation therapy. In this study, implanted NSCs were observed migrating from the striatum toward the ischemic boundary at 2 weeks after implantation.
In our study, we observed a significant increase of neurologic scores 3 weeks after transplantation compared with the control groups, demonstrating a positive effect of repairing ischemic brain with iPSC-derived NSCs. Although 3 weeks is long enough for differentiation of implanted NSCs into neurons, it is not enough for implanted cells to replace dead neurons and function. As a result, some scholars [46, 47] believe that early recovery of function with stem cell transplantation is due to neurologic protection of growth factor release, not new neural cells replacement. Therefore, restoration of neurologic function in this study after cell implantation may be more relevant to a neuroprotective effect.
Note that PBS injection also improved the behavior of rats at 2 weeks (Figure 2B through D). One reason is that collateral circulation has been formed in brain tissue, and the function of impaired cells may be recovered in 2 weeks, which contributes to the improved behavior of rats. Moreover, the improvement of behavior in this rat model may be partly owing to functional compensation after brain injury.