Transduction of SC has been used as a mean to act as gene delivery vehicles. Boosting SC expression of neurotrophins such as BDNF and NT3 has profound effects on neural protection, plasticity, and regeneration after spinal cord injury in the developing and adult CNS (Kromer and Cornbrooks 1985; Martin et al. 1991; Montero-Menei et al. 1992; Levi and Bunge 1994). Since NT3 also promotes oligodendrocyte proliferation, survival and differentiation (McTigue et al. 1998), and SC migration (Yamauchi et al. 2005) in vitro, we explored the benefit of this strategy on remyelination after transplantation into the demyelinated spinal cord. In doing so, we demonstrated that forced expression of BDNF and NT3 in macaque SC promoted functional motor recovery of the demyelinated mouse spinal cord (Girard et al. 2005). This beneficial effect resulted from multiple events including enhanced remyelination by endogenous oligodendrocytes and SC, enhanced neuroprotection, and reduced scar formation, thus highlighting the issue of ex vivo neurotrophin delivery based on SC therapy to enhance global repair of demyelinated lesions. A similar strategy combining overexpression of NT3 and BDNF was also used in spinal cord trauma and found to improve efficiently axonal length and the number of myelinated axons in the lesion (Golden et al. 2007). However this strategy did not distinguish a direct or indirect effect of neurotrophins on myelination, through either creating an increase in axon number and/or acceleration of SC differentiation.

An alternative strategy to enhance their intra-CNS integration/migration has been to modify SC membrane properties. As mentioned above, one major limitation of SC-based therapy is the poor potential of their integration/migration within white and gray matter of the CNS. By contrast, neural precursors and OEC interact more efficiently with astrocytes. Interestingly, both of these cells express the polysialylated form of NCAM (PSA-NCAM), a cell adhesion molecule highly expressed during development and involved in CNS plasticity, remodeling, and repair (reviewed in Rutishauser 2008). Since SC express NCAM but not PSA, it was speculated that forcing NCAM polysialylation in SC would improve their integration/migration in the CNS. Indeed, ectopic expression of PSA by rodent SC enhanced their migratory properties in vitro and functional repair in a trauma model in vivo without altering their myelination potential (Lavdas et al. 2006; Papastefanaki et al. 2007). This strategy was further explored with adult macaque SC grafted in the demyelinated rodent spinal cord. Forced expression of PSA-NCAM by the grafted SC enhanced their recruitment by distant lesions and promoted their ability to remyelinate the demyelinated axons while wild-type SC were unable to do so (Bachelin et al. 2010). This pro-migratory effect resulted in increased SC fluidity and exit from the graft most likely by reduced adhesion among sister SC and SC–astrocytes. Alternatively, it may also have decreased SC sensitivity to inhibitory molecules and/or increased their response to chemotactic cues as demonstrated for PSA-NCAM+ embryonic-stem-cell-derived neural precursors (Glaser et al. 2007).

Other strategies have targeted specific molecular cues of the glial scar. Indeed, reducing levels of chondroitin sulfate proteoglycans (CSPG) (Grimpe and Silver 2004; Santos-Silva et al. 2007) and N-cadherin–N-cadherin (Fairless et al. 2005) interaction promoted SC–astrocyte mixing in vitro and in vivo. Interestingly, another inhibitor of axonal regeneration, semaphorin 3A (Sema 3A), was found to be repulsive for SC in vitro and in vivo in a model of spinal trauma. Treatment with a Sema 3A inhibitor was able to reverse this effect and increase SC-mediated myelination and axonal regeneration (Kaneko et al. 2006). Data also indicate that ephrins may play a role in SC–astrocyte non-mixing (Afshari et al. 2010). However, whether neutralizing these axonal growth inhibitors would also improve exogenous SC migration through CNS parenchyma has not been reported so far. Several inhibitors of axonal regeneration are also contained in myelin. However, their role in Schwann cell migration inhibition remains to be elucidated. Interestingly, SC, unlike OEC and OPC, do not secrete metalloendoproteases, enzymes known to degrade inhibitors as extracellular matrix proteoglycans and myelin axonal inhibitors Nogo (Amberger et al. 1997; Gueye et al. 2011).

Another issue with SC-based therapy concerns the control of SC myelination by axon diameter. Injections of SC into areas of small-diameter axons do not result in remyelination (Brook et al. 1993). However, using lentiviral-driven overexpression of NRG 1 type III, Taveggia et al. (2005) induced SC to myelinate small-diameter axons. Furthermore, overexpression of NRG1 type III in transgenic mice increased the thickness of endogenous newly formed PNS but not CNS myelin in response to demyelination (Brinkmann et al. 2008). Interestingly, several members of the ADAM (a disintegrin and metalloprotease) secretase have been implicated in the regulation of NRG 1 type III activity notably by regulating its processing. While some secretases such as BACE (beta secretase) positively regulate the myelination process (Shirakabe et al. 2001; Hu et al. 2006), recent data showed that another secretase TACE (tumor necrosis factor alpha-converting enzyme) is a negative regulator of PNS myelination (La Marca et al. 2011). These last observations underline the therapeutic potential of TACE modulation since several inhibitors of TACE are already under investigation for other disease treatment. The possibility to direct selectively exogenous SC remyelination towards CNS axons irrespectively of their size may thus be useful in enhancing their contribution to CNS remyelination.

OEC (initially named olfactory SC) are differentiated glial cells very similar to SC that belong to the peripheral olfactory system where they ensheath but do not myelinate the axons belonging to the first cranial nerve (Franklin and Barnett 2000; Vincent et al. 2005).

They share also features with astrocytes expressing “fibrillar” GFAP and forming part of the glial limitans of the olfactory bulb (Doucette 1991). The origin of OEC has long been controversial. During early development, they originate from the nasal placodes and provide essential growth and guidance to sensory axons which they do not myelinate. Although they were believed to be ectodermal derivatives, recent evidences based on the Wnt1Cre fate mapping technology show that at early embryonic stages OEC within the placode derive from neural crest cells (NCC) (Barraud et al. 2010; Forni et al. 2011). These findings could explain the high level of similarities existing between OEC and SC. In view of these recent data, we will consider OEC as an alternative source of neural crest-derived cells rather than cells deriving from another lineage.

In vitro, OEC and SC have remarkable similarities. Both cell types express the low-affinity NGF receptor (p75ntr), S100, GFAP, and cell adhesion molecules such as L1 and NCAM. They also express extracellular matrix proteins such as laminin and fibronectin (Wewetzer et al. 2002) and secrete neurotrophins including NGF, BDNF, FGF, and vascular endothelial growth factor (VEGF) (reviewed in Sasaki et al. 2011). Although OEC do not normally myelinate the very small axons of the first cranial nerve, studies have shown that OEC can (re)myelinate larger axons with myelin that is indistinguishable morphologically and biochemically from that made by SC when transplanted in the CNS. This was first observed in cocultures of OEC with DRG neurons (Devon and Doucette 1992) and further confirmed in vivo with an OEC cell line (Franceschini and Barnett 1996) and primary OEC (Franklin et al. 1996; Smith et al. 2001; Lakatos et al. 2003b). Cell suspensions of acutely dissociated OEC from neonatal rats remyelinated and enhanced axonal conduction, when focally injected into the ethidium-bromide-demyelinated areas of the dorsal columns of the spinal cord (Imaizumi et al. 1998). Moreover, xenotransplanted canine, human, or porcine OEC isolated from the adult olfactory bulb are capable of extensive functional remyelination following transplantation into demyelinated rat CNS (Kato et al. 2000; Smith et al. 2001; Deng et al. 2006). The myelin formed contains the peripheral myelin protein P0, and the cell’s association with axons resembles that of SC. Furthermore, when faced with demyelinated axons, transplanted OEC express SC transcription factors Krox-20 and SCIP mRNA, suggesting that similar mechanisms are involved in the manner OEC and SC form myelin (Smith et al. 2001). In addition to these observations, OEC transplantation in trauma models have suggested that OEC have multiple beneficial effects, reducing glial scarring and cavitation, increasing angiogenesis, and providing a cellular scaffold for regenerating axons (reviewed in Toft et al. 2011). Improved regeneration and reduced scarring was also observed after autologous transplantation in a canine model of spinal cord injury (Jeffery et al. 2005). However, the repair capacities of OEC have been challenged given that their ability to support axonal growth was not fully confirmed (Tetzlaff et al. 2011). It was also reported that OEC are less efficient than SC at forming myelin in vitro and in vivo (Plant et al. 2003; Boyd et al. 2004). In addition, it was proposed that a SC contamination could be responsible for all the observed effects (Boyd et al. 2004). However, transplantation of eGFP-OEC in a contusive SCI lesion (Sasaki et al. 2004), in the injured nerve (Dombrowski et al. 2006; Radtke et al. 2009), or in the X-EB lesions (Sasaki et al. 2011) provided proof that OEC have the ability to remyelinate axons with, in some circumstances, recovery of locomotor functions. Moreover, co-transplantation of purified OEC and meningeal cells indicated that this potential can be enhanced by meningeal cells which contaminate OEC preparations (Lakatos et al. 2003a). So far, the behavior of OEC in a context of immune inflammation comparable to MS has not been addressed.

Several studies suggested that OEC migrate better than SC when confronted with CNS components. Indeed, OEC mix better when cocultured with astrocytes (Lakatos et al. 2000), and transplanted OEC elicit a diminished expression of GFAP and CSPG expression in astrocytes compared to SC (Lakatos et al. 2003a; Garcia-Alias et al. 2004; Andrews and Stelzner 2007). This is explained via decreased N-cadherin-mediated adhesion with astrocytes (Fairless et al. 2005) and induction of a weaker astrocyte hypertrophy (Lakatos et al. 2000). Moreover, as mentioned before, OEC secrete matrix metalloproteinases (MMPs), which modulate their migration in vitro (Gueye et al. 2011). Since MMPs also induce changes in astrocyte morphology (Ogier et al. 2006), these enzymes could also play an important role in OEC mixing with astrocytes.

However, the greater migration potential of OEC in the injured spinal cord remains a subject of debate (Deng et al. 2006; Lu et al. 2006; Pearse et al. 2007). While several studies show that OEC, SC, and OPC survive and migrate poorly after transplantation in normal white matter (Iwashita et al. 2000; Hinks et al. 2001; Lankford et al. 2008), only one report highlights that transplanted OEC (and OPC), but not SC, migrate extensively in gray and white matter of the X-irradiated spinal cord (Lankford et al. 2008). Although these data stress potential differences between transplanted SC and OEC with respect to their sensitivity to CNS inhibitory cues. However the specific environment of the X-irradiated spinal cord besides NG2+OPC a depletion in NG2+ OPC remains to be identified.

Since OEC myelinate in a manner very similar to SC but mix better with astrocytes in vitro and in vivo, it has been proposed that from a therapeutic point of view, OEC transplantation would have advantages over SC. However, immature stages of the SC lineage migrated profusely on transplantation into the CNS (Fig. 6.2 and 6.3) and elicited similar reduced changes in astrocyte reactivity including CSPG production (Fig. 6.4) (Woodhoo and Sommer 2008; Zujovic et al. 2010). This suggests that OEC represent a stage of the neural crest lineage with plasticity more similar to SC precursors (SCp) and boundary cap cells (BC, see further) than mature SC.

Whether OEC and SC display species differences in terms of growth factor requirements for expansion has also been investigated. Similar to rodent OEC, OEC from large animal species share many shared features with SC. In fact a recent review by Wewetzer et al. (2011) concluded that rodent, but not canine, human, and monkey, OEC and SC require mitogens for expansion and undergo spontaneous immortalization after prolonged growth factor treatment. These observations highlight that both canine and monkey OEC and SC may serve as valuable translational models for addressing their regenerative capacities at a preclinical stage.

Autologous SC grafts in monkeys and autologous OEC grafts in dogs have been proven feasible and successful in terms of repair. Whether human OEC and SC have equivalent regenerative/remyelinating potentials remain to be defined. Considering autologous grafts studies in humans, lack of adverse effects but no proof of benefit were reported for autologous SC grafts in the injured spinal cord injury 1 year after injury (Saberi et al. 2008). Likewise no conclusions except for safety could be drawn from the suspended MS trial. Autologous olfactory mucosa engraftment (Lima et al. 2006) and autologous olfactory ensheathing cell transplantation (Feron et al. 2005) appeared to be safe and/or provide potential benefit. However, their application to promote remyelination/regeneration in MS patients has not been reported. Clearly more studies on MS-like preclinical models are necessary to evaluate the outcome of grafted cells in an inflammatory demyelinating milieu.

An alternative to the use of committed SC or OEC for cell therapy would be the use of more immature stages of the neural crest (NC) lineage. While embryonic and adult precursors have been identified, few reports have addressed the repair potential of these PNS stem/precursor cells in the CNS. These include SC precursors (SCp), boundary cap cells (BC), and olfactory epithelial progenitors (OEp).

While the main asset of embryonic stem cells is their capacity to give rise to all the cells of the body, their foremost limitation is the source of these cells. However, one of the major issues of autologous cell transplantation therapy resides in the accessibility of the cell sources. This is the reason why many studies focused on skin as the most accessible organ. Based on the modification of adult fibroblasts transduced into induced pluripotent stem cells (iPS) or directly reprogrammed into differentiated cells, the skin offers great future opportunities for cell replacement and regenerative medicine.