Targeting ependymal stem cells in vivo as a non-invasive therapy for spinal cord injury

Spinal cord injury (SCI) is a debilitating and devastating condition, and there are approximately 12,000 new cases in the USA each year and an estimated number of sufferers reaching 250,000–300,000 in the USA alone. Although important advances have been made in the medical treatment of SCI recently, it is not yet possible to completely restore neuronal function after SCI. In rodents and primates, SCI causes irreversible loss of function distal to the lesion as a result of axonal damage, demyelination, and death of oligodendrocytes, astrocytes and neurons (including both spinal cord interneurons and motor neurons) (Grossman et al., 2001). Replacing the lost cell types, and integrating newly transplanted or generated cells into the spinal cord circuitry, are key aims in designing potential therapies for patients that suffer from SCI. Recent advances in our understanding of stem cell biology are revealing important potential applications for these cells in treating patients with SCI.

The majority of research into developing treatments for SCI patients currently focuses on the use of stem cells to regenerate damaged tissue. The plasticity and ability to self-renew of three different types of stem cells suggests that they hold potential for regeneration following SCI: human embryonic stem cells (hESCs), neural stem cells (NSCs) and induced pluripotent stem cells (iPSCs). hESCs are pluripotent cells that can differentiate into nearly all cell types, including oligodendrocytes and motor neurons. NSCs are multi-potent cells that can give rise to neurons, oligodendrocytes and astrocytes. Patient-derived iPSCs, which can be created by reprogramming adult somatic cells using a variety of methods, have recently been proposed for use in SCI therapy. iPSCs can differentiate into many cell types, including glia and motor neurons (for a review, see Ronaghi et al., 2010). In addition, unlike transplantation of foreign cells, patient-derived iPSCs should not be rejected by the immune system. However, it should be noted that all forms of stem cell-based therapy that involve transplantation require labor-intensive in vitro propagation and manipulation, followed by transplantation and the establishment of the cells into appropriate sites in injured patients. In addition, iPSCs bring other disadvantages, including the potential for teratoma formation, aberrant reprogramming and the presence of transgenes in iPSC populations (Yamanaka, 2009; Ronaghi et al., 2010).

Ependymal stem cells (EpSCs) are multipotent stem cells found in the adult tissue surrounding the ependymal canal of the spinal cord (reviewed by Ronaghi et al., 2010). Direct reprogramming of this resident spinal cord stem cell population might be a promising way to avoid the need for stem cell transplantation to treat SCI. Work from Meletis et al. suggests that such an approach is possible: they reported that the EpSCs of the spinal cord proliferate in response to SCI in adult rodents (Meletis et al., 2008). After SCI, the EpSC progeny is recruited to the injury site (even when the injury does not affect the EpSCs or their processes) to give rise to scar-forming glial cells and, to a lesser degree, oligodendrocytes. Thus, the cells of an injured host can give rise to some of the cell types that are necessary to regenerate healthy spinal cord tissue.

Meletis et al. generated two transgenic mouse lines expressing tamoxifen-dependent Cre recombinase under the control of FoxJ1 or nestin regulatory sequences, which allowed genetic labeling of cells of the ependymal layer in the adult spinal cord (Meletis et al., 2008). FoxJ1 is a specific marker of cells with motile cilia or flagella, whereas NSCs and neural progenitor cells express nestin. The authors used cell culture, immunohistochemistry and electron microscopy to characterize these populations and their progeny before and after SCI in mice. The EpSCs respond to SCI by proliferating and differentiating, and their progeny migrate towards the injury site; increasing numbers of cells accumulate in the forming glial scar over several weeks (Meletis et al., 2008). By using molecular markers, the authors showed that most of the EpSC progeny have an astrocyte-like morphology and were negative for GFAP (the major intermediate filament protein of mature astrocytes), and that a smaller sub-population expressed GFAP and nestin. None of the recombined cells in the scar tissue had neuronal morphology or was immunoreactive to NeuN (neuron-specific phenotype), whereas a population of cells expressed Olig2 (an oligodendrocyte lineage transcription factor). The first month after injury, the Olig2-expressing cells displayed an ultrastructural morphology that corresponded to immature oligodendrocytes. At 10 months after SCI, most of the ependymal-derived progeny were located in the scar tissue, but a substantial number of these cells were dispersed in the intact gray and white matter bordering the lesion. Most of these cells expressed Olig2 and displayed a mature oligodendrocyte morphology with processes that enwrapped myelin-basic-protein-immunoreactive myelin-ensheathing axons. Thus, EpSCs generate both astrocytes and myelinating oligodendrocytes (Meletis et al., 2008). The intrinsic potential of EpSCs to replace some of the cells in the spinal cord following injury opens up the opportunity for developing non-invasive therapies for patients with SCI, through activating the differentiation of EpSCs into various cell types. The transgenic animals generated by Meletis et al. could be used in further experiments to determine whether EpSCs can also function as NSCs.

At this point, additional work is needed to increase our understanding of stem cell differentiation pathways to move forward in developing treatments for SCI. The use of animal models in which functional regeneration occurs after SCI will be of great help. The spontaneous regeneration of neuronal cells from stem cell progenitors after SCI has been reported in adult fish [e.g. serotonergic interneurons in goldfish (Takeda et al., 2008) and motor neurons in zebrafish (Reimer et al., 2008)], which is in marked contrast to the absence of neuronal replacement observed in mammals (Meletis et al., 2008). Reimer et al. showed that regeneration of motor neurons occurs following SCI in adult zebrafish, and that the new motor neurons are integrated into the spinal cord circuits. The plasticity of Olig2-expressing progenitor cells seems to allow them to generate motor neurons through activation of the transcription factors HB9, islet-1 (ISL1) and ISL2, which are also found in developing motor neurons of mammals (Tsuchida et al., 1994; William et al., 2003) and zebrafish (Cheesman et al., 2004; Park et al., 2007). An in vitro study in mice has shown that, with a specific differentiation protocol, 90% of differentiated cultures of EpSCs obtained after SCI stain positive for the motor-neuron-specific marker HB9, with 32% of these motor neurons displaying electrophysiological properties that resemble those of functional spinal motor neurons (Moreno-Manzano et al., 2009), which shows that the manipulation of EpSCs after SCI might be a viable strategy for restoring neuronal dysfunction in humans.

The next step in studies of rodent or non-human primate models should focus on the in vivo manipulation of the EpSC population. Clues for moving forward are provided by zebrafish studies: in zebrafish, the Olig2-expressing progenitors respond to a Sonic hedgehog signal to regenerate motor neurons (Reimer et al., 2009). Intraperitoneal injection of cyclopamine inhibits Sonic hedgehog signaling and reduces ventricular proliferation and motor neuron regeneration in zebrafish (Reimer et al., 2009). This finding indicates the possibility that Olig2-expressing progenitors in mammals could be manipulated in vivo to stimulate generation of specific neuronal cells. Indeed, it has recently been shown in rats that application of an intravenous hedgehog agonist increases the size of the population of neural precursor cells after SCI (Bambakidis et al., 2009). Further research should not only focus on the pharmacological manipulation of the EpSC population, but also on other aspects of EpSCs: for example, enhanced physical activity in adult rats induces an endogenous response that leads to increased proliferation and differentiation of EpSCs, mainly into macroglia or cells that express nestin (Cizkova et al., 2009). Recent research has shown that physical exercise maintains nestin expression in the EpSCs and improves functional recovery in rats following SCI (Foret et al., 2010).

Therefore, the study by Meletis et al. opens up the possibility that combined pharmacological and physiotherapeutic treatments could be used to manipulate the resident EpSC population in patients with SCI to improve functional recovery. This approach could bypass the significant risks associated with therapies involving transplantation of non-patient-derived donor cells. Since the report by Meletis et al. was published (Meletis et al., 2008), other studies have shown that it is indeed possible to manipulate activated EpSCs to improve neurological recovery after SCI (e.g. Moreno-Manzano et al., 2009; Reimer et al., 2009; Fortet et al., 2010). Potential use of EpSCs in spinal cord regenerative medicine will depend on the development of strategies for directed in vivo differentiation into different functional cell types (see Zhang et al., 2010). To acheive this aim, it will be important to advance our understanding of stem cell differentiation pathways in the mature nervous system. The use of animal models such as zebrafish, in which functional regeneration of neurons from EpSCs occurs after the SCI event, should facilitate progress towards this goal.