Human Oligodendrogenic Neural Progenitor Cells Delivered with Chondroitinase ABC Facilitate Functional Repair of Chronic Spinal Cord Injury

摘要

class="head no_bottom_margin" id="sec1title">IntroductionNeural progenitor cells (NPCs) represent a promising regenerative strategy to target several CNS disorders (). Neurobehavioral recovery following NPC transplantation into the injured spinal cord has been reported in both rodent and non-human primate models (, , ). Although NPCs have therapeutic value, their translational potential is hindered by limited availability, immunologic complications, and ethical concerns. Recently, transplantation of induced pluripotent stem cell (iPSC)-derived NPCs and iPSC-derived oligodendrocyte progenitor cell-enriched NPCs has demonstrated therapeutic promise (, , ). However, previous studies have identified the potential for tumorigenicity, immunogenicity, and genetic and epigenetic abnormalities following iPSC-based cell transplantation (). To avoid the risks associated with NPCs and iPSCs, we focused on human directly reprogrammed NPCs (drNPCs), which were directly generated from somatic cells, thus avoiding the pluripotent state ().Within the injured spinal cord environment, oligodendrocytes are highly susceptible to the cytotoxic conditions found both local and distant to the lesion epicenter, leading to demyelination of preserved axons (). investigated the temporal pattern of conduction failure in individual fibers across a contusion injury and examined changes in their conduction properties from acute to chronic stages of injury. Acutely (1–7 days) after spinal cord injury (SCI), complete conduction block was observed in ascending dorsal column axons, followed by a period of improved conduction during the subacute (2–4 weeks) phase, without further improvement at the chronic (3–6 months) phase. At 6 months after SCI, 16% of sampled fibers were capable of conducting across the lesion. Furthermore, this study demonstrated a population of axons (20% of the fibers tested) which are chronically demyelinated and viable but unable to conduct under normal physiological conditions (). Using detailed histological analyses, observed that the number of demyelinated axons progressively increased up to 450 days after injury. Although remyelination of axons was observed from 14 to 450 days post-SCI, it was found to be incomplete. The results of the previous studies indicated that chronic and progressive demyelination represents an important target for cell transplantation therapy. Previous studies have revealed the biological importance of remyelination and tissue sparing by graft-derived cells for neurobehavioral recovery following the transplantation of NPCs into the injured spinal cord (, ). Thus, remyelinating spared axons and promotion of neural plasticity and/or tissue sparing by graft-derived oligodendrocytes represents a promising therapeutic strategy following SCI (href="#bib43" rid="bib43" class=" bibr popnode">Trounson and McDonald, 2015). Building on these findings, we recently developed a novel method to generate human iPSC-derived NPCs or drNPCs biased toward an oligodendrogenic fate (oNPCs) for the treatment of SCI (href="#bib24" rid="bib24" class=" bibr popnode">Khazaei et al., 2017, href="#bib27" rid="bib27" class=" bibr popnode">Nagoshi et al., 2018). Transplantation of oNPCs in the subacute phase of SCI showed greater differentiation into oligodendrocytes, axonal remyelination, and tissue sparing, which ultimately resulted in the recovery of motor function (href="#bib27" rid="bib27" class=" bibr popnode">Nagoshi et al., 2018).The majority of studies have reported that NPCs transplanted in the subacute period following SCI have therapeutic effects (href="#bib7" rid="bib7" class=" bibr popnode">Cummings et al., 2005, href="#bib20" rid="bib20" class=" bibr popnode">Iwanami et al., 2005). Unfortunately, most patients with SCI have progressed past the subacute period, and effective therapies to target the chronic phase of SCI are greatly lacking (href="#bib28" rid="bib28" class=" bibr popnode">Nishimura et al., 2013, href="#bib40" rid="bib40" class=" bibr popnode">Suzuki et al., 2017). Although a number of animal model studies have aimed to achieve neurobehavioral recovery in the chronic phase of SCI with NPC transplantation, functional recovery has not been observed (href="#bib8" rid="bib8" class=" bibr popnode">Cusimano et al., 2012, href="#bib35" rid="bib35" class=" bibr popnode">Parr et al., 2007). Due to the inhibitory microenvironment of the chronically injured spinal cord, which includes a cystic cavity with surrounding glial scar as well as neuronal and glial cellular loss, repair and regeneration of the injured cord has proven challenging (href="#bib38" rid="bib38" class=" bibr popnode">Silver and Miller, 2004). The glial scar consists of a non-neural lesion core (fibrotic scar) and an astrocytic scar border (href="#bib39" rid="bib39" class=" bibr popnode">Sofroniew, 2018). The non-neural lesion core is comprised of stromal cells and extracellular matrix molecules including fibronectin, collagen, proteoglycans, and laminin (href="#bib31" rid="bib31" class=" bibr popnode">O'Shea et al., 2017). The astrocyte scar forms a border between the non-neural lesion core and adjacent spared neural tissue and restricts the spread of inflammation (href="#bib4" rid="bib4" class=" bibr popnode">Burda and Sofroniew, 2014). After SCI, a phenotypic change to astrocytes occurs, known as reactive astrogliosis. Naive astrocytes undergo a change including increased glial fibrillary acidic protein (GFAP) expression, hypertrophy, and process extension, resulting in a characteristic phenotype of reactive astrocytes within several days after SCI (href="#bib16" rid="bib16" class=" bibr popnode">Hara et al., 2017, href="#bib32" rid="bib32" class=" bibr popnode">Okada et al., 2006). Subsequently, reactive astrocytes overlap their processes and transform into scar-forming astrocytes (href="#bib16" rid="bib16" class=" bibr popnode">Hara et al., 2017, href="#bib38" rid="bib38" class=" bibr popnode">Silver and Miller, 2004). In SCI lesions, chondroitin sulfate proteoglycans (CSPGs) are produced not only by astrocytes, but also pericytes, fibroblast lineage cells, and inflammatory cells. CSPGs comprise the main inhibitory component of the glial scar (href="#bib2" rid="bib2" class=" bibr popnode">Anderson et al., 2016, href="#bib39" rid="bib39" class=" bibr popnode">Sofroniew, 2018). Importantly, a recent study demonstrated that genetically targeted astrocyte ablation did not reduce CSPGs, suggesting that astrocytes are not the primary producers of CSPGs in SCI lesions (href="#bib2" rid="bib2" class=" bibr popnode">Anderson et al., 2016). The inhibitory CSPGs limit integration and migration of grafted cells, which prevents remyelination of the spared axons and regeneration of neural circuits (href="#bib15" rid="bib15" class=" bibr popnode">Grijalva et al., 1996, href="#bib28" rid="bib28" class=" bibr popnode">Nishimura et al., 2013). To counter the CSPGs, the enzyme chondroitinase ABC (ChABC) has been shown to degrade the sulfated glycosaminoglycan chains on the CSPGs (href="#bib26" rid="bib26" class=" bibr popnode">Moon et al., 2001) and promote neural plasticity and functional recovery after acute SCI (href="#bib3" rid="bib3" class=" bibr popnode">Bradbury et al., 2002). Moreover, we have previously shown that a combinatorial therapy of ChABC and NPCs or iPSC-derived NPCs increased the long-term survival of NPCs and optimized their integration and migration in chronic SCI, resulting in neurobehavioral recovery (href="#bib12" rid="bib12" class=" bibr popnode">Fuhrmann et al., 2018, href="#bib22" rid="bib22" class=" bibr popnode">Karimi-Abdolrezaee et al., 2010, href="#bib40" rid="bib40" class=" bibr popnode">Suzuki et al., 2017).One of the challenges of providing sustained delivery of ChABC is the thermal instability of the enzyme (href="#bib41" rid="bib41" class=" bibr popnode">Tester et al., 2007), which limits the activity to less than 4 days in vivo (href="#bib6" rid="bib6" class=" bibr popnode">Crespo et al., 2007). In previous SCI studies, ChABC has been delivered continuously using an osmotic mini-pump with intrathecal catheters (href="#bib22" rid="bib22" class=" bibr popnode">Karimi-Abdolrezaee et al., 2010, href="#bib40" rid="bib40" class=" bibr popnode">Suzuki et al., 2017). However, the mini-pump delivery system is prone to dislodgement, can be associated with off-target delivery, is invasive, and is associated with complications such as infection and arachnoiditis (href="#bib11" rid="bib11" class=" bibr popnode">Follett et al., 2004). Previously, we demonstrated that an affinity-based release of ChABC from a crosslinked methylcellulose (XMC) hydrogel could reduce CSPG levels at the injury site in vivo for 2 weeks and promote functional repair following a single intrathecal injection (href="#bib34" rid="bib34" class=" bibr popnode">Pakulska et al., 2013, href="#bib33" rid="bib33" class=" bibr popnode">Pakulska et al., 2017). To achieve affinity release, ChABC is delivered as a fusion protein with Src homology 3 (SH3-ChABC), and XMC is modified with an SH3 binding peptide.The present study aimed to assess the efficacy of a combinatorial strategy employing an XMC hydrogel containing ChABC and human directly reprogrammed oNPCs to treat chronic SCI. Here, in a clip-contusion chronic SCI model using immunodeficient rats, we degraded CSPGs with a single intrathecal injection of an innovative XMC hydrogel containing ChABC. Next, we transplanted clinically relevant oNPCs 1 week after the intrathecal injection to assess the therapeutic potential of this combinatorial therapy in the chronically injured spinal cord.