NSCs have been shown to express trophic factors (proteins and other molecules) that promote survival, proliferation, and differentiation of host brain cells after stroke, and this has led to the general hypothesis of NSC-mediated recovery through trophic support [53]. In vitro profiling of NSCs has shown that they secrete such key factors as brain-derived neurotrophic factor (BDNF), glial-derived neurotrophic factor (GDNF), vascular endothelial growth factor (VEGF), and several others [29, 54–56]. However, relatively little is known about which trophic factors mediate efficacy in vivo, due in part to the need for improved methods to profile transplanted NSC expression/secretion in vivo, as well as the need for additional studies that selectively upregulate or suppress trophic factors within transplanted NSCs. For example, IC delivery of VEGF-transfected NSCs into stroke-injured rats resulted in higher levels of vascular repair and functional recovery compared to unmodified NSCs [57]. Further, a recent study by our group showed that in stroke-injured rats, the recovery enhancement associated with IC delivery of human NSCs could be suppressed by systemic infusion of bevacizumab (Avastin), an antibody specific for human VEGF [30], thus enabling us to tease apart the role of the NSC-secreted VEGF from that of endogenous rodent VEGF. Results of this study further indicated that recovery enhancement by NSCs coincided with increased neovascularization and blood–brain barrier integrity, along with decreased microglial/immune cell activation within the peri-infarct region, and that both of these changes could be suppressed by cotreatment with Avastin. These results highlight the concept that NSC-associated functional improvements are unlikely to occur via a single mechanism and instead may involve simultaneous modulation of multiple endogenous repair pathways.

Regeneration of the vascular system by localized activation of angiogenesis is one form of endogenous repair that is correlated with improved stroke outcome [58, 59] and has the potential for modulation by NSCs and other cell therapies. Different types of NSCs have been shown to express trophic factors that can stimulate neovascularization (angiogenesis or vasculogenesis), including VEGF, epidermal growth factor (EGF), insulin-like growth factor 1 (IGF1), and basic fibroblast growth factor (bFGF) [30, 60]. IC delivery of NSCs into stroke-injured rats was correlated with increased angiogenesis or neovascularization [30, 61], and poststroke stimulation of neovascularization by transplanted human fetal-derived NSCs could be suppressed by selective blockade of human VEGF [30]. Further evidence of VEGF as a key mediator of these effects was shown by intravenous coadministration of NSCs and recombinant VEGF [62] or IC delivery of VEGF-transfected NSCs [57] into stroke-injured rats; both approaches led to increased vascular repair and functional recovery beyond that of NSCs alone or unmodified NSCs. The trophic factors placental growth factor and angiopoietin-1 have been identified as candidate mediators of neovascularization by bone marrow-derived stromal cells [63–65], but their potential involvement in NSC-based therapies has not yet been elucidated.

The immune response after stroke involves both an acute phase that is associated with injury progression and a delayed phase that can aid subsequent repair [66]. Preclinical studies suggest that IC delivery of NSCs can modulate this response in favor of repair by reducing the number of activated microglia and macrophages that infiltrate surviving peri-infarct tissue [28, 30, 67]. Similarly, systemic delivery of NSCs has been shown to reduce infiltration of activated monocytes into the brain of stroke-injured rodents, and this was correlated with reduced levels of proinflammatory cytokines in the spleen and brain, including tumor necrosis factor, interferon gamma, and interleukin-6 [68, 69]. Additional in vitro studies have shown that NSCs can suppress T cell activation and dendritic cell maturation in a coculture setting, and that these effects are mediated in part by NSC secretion of such immunomodulatory factors as nitric oxide and prostaglandin E2 [70–72].

Activation of brain plasticity after stroke in the form of outgrowth and remapping of neuronal projections is correlated with improved behavioral outcome, and thought to be a driving factor behind spontaneous recovery after stroke [73–75]. Based on studies from our group, IC delivery of human fetal-derived NSCs augmented several aspects of brain plasticity in stroke-injured adult rats, including remapping and increased axonal sprouting of contralesional neurons into the ischemic hemisphere and through the cortical thalamic and cortical spinal tracts, increased dendritic branching in both cortical hemispheres, and increased axonal transport [29, 30]. Similarly, IC delivery of human embryonic-derived NSCs enhanced axonal sprouting of contralesional neurons in a rodent model of neonatal hypoxia–ischemia [76]. Moreover, the plasticity-enhancing effects of fetal-derived NSCs could be recapitulated in vitro with a primary neuron coculture system, and the use of neutralizing antibodies implicated SLIT and thrombospondin 1 and 2 as NSC-secreted factors potentially driving these effects [29]. In addition, IC-transplanted human embryonic-derived NSCs were shown to promote recruitment and/or proliferation of endogenous oligodendrocytes in the ischemic hemisphere [77], as well as differentiate into oligodendrocytes themselves [33, 77, 78], thus demonstrating the potential to support or even directly participate in the myelination of newly formed circuits.

In response to a stroke injury, endogenous NSCs residing in the subventricular zone of the lateral ventricles can undergo proliferation and migrate to the lesion’s boundary, where a small proportion become mature neurons and integrate into surviving circuits [79–81]. In rats subject to striatal ischemic stroke, subsequent IC delivery of human fetal-derived NSCs was correlated with increased migration and survival of immature neurons in the infarct area [67]. However, given that the majority of recruited cells either die or remain in an immature state upon arrival to the infarct site [79, 82], the exact mechanisms by which injury- and cell therapy-induced neurogenesis might lead to improved stroke recovery still need to be resolved and may be more akin to various types of therapeutic paracrine signaling attributed to transplanted NSCs.

As we continue to explore the potential of NSCs and IC administration as therapeutic strategies for stroke, several challenges remain. First, to meet the demand of the increasingly high number of stroke cases that occur each year (estimated 795,000 new or recurrent stroke cases annually in the USA [97]), improved methods for large-scale production of clinical-grade NSCs must be established. While embryonic-derived NSCs or immortalized cell lines may be the best option for this, given the virtually unlimited scalability of the starting material, they may also pose the greatest safety risk due to the potential for undifferentiated cells to persist in the final product and undergo tumorigenesis or ectopic tissue formation in vivo, particularly when transplanted into the stroke-injured brain [98, 99]. Similarly, though iPSC technologies may ultimately allow reasonable scalability of patient-specific NSCs from a more ethically acceptable source, they will also require strategies to address any undifferentiated cells that remain in the final product [98, 100], as well as the carryover of aberrant genetic or epigenetic changes that can accumulate in aged somatic cells. Alternatively, fetal-derived NSCs pose less tumorigenic risk, but their scalability is more limited, both by the relative availability of appropriate fetal starting material and the extent to which intermediate precursors or the final NSC product can be expanded in vitro. Because extended time in culture can alter a cell’s phenotype and even lead to chromosomal abnormalities, any efforts to scale up NSC production will require careful monitoring of karyotypic stability, as well as the specific molecular signatures that determine an NSC product’s clinical release criteria (California Institute for Regenerative Medicine (CIRM) Disease Team Grant DR1 01480) [40].

A second major challenge faced by NSC and non-NSC cell therapies is the need to adequately characterize the final cell product, including quantitative measures of cell identity, purity, and any potential impurities (California Institute for Regenerative Medicine (CIRM) Disease Team Grant DR1 01480) [40]. Together, these measures define key elements of the release criteria that determine the consistency and clinical eligibility of each manufactured lot for a given cell-therapy product. Establishing appropriate quantitative measures of cell identity and purity remains difficult for NSCs, given that surface proteins and intracellular markers typically associated with NSCs either are expressed only transiently during differentiation or are also expressed by other cell types. Use of multiple identity/purity markers can help overcome these limitations, but also can increase the risk of a failed manufacturing run if any of the predetermined release criteria are not met. Furthermore, to ensure safety of an NSC product, its release criteria must also include appropriate impurity markers indicative of unwanted cell types, with acceptable impurity levels ideally based on preclinical studies where the known impurity is spiked into the NSC product at increasing levels and tested for its relative impact on safety. In the case of embryonic- and iPSC-derived NSCs, residual pluripotent cells represent the most concerning type of impurity given their known tumorigenic potential [98, 99], and pluripotency markers such as Oct4, Tra-1-60, and SSEA-4 often are used to detect these cells.

In addition, a detailed understanding of the production/derivation process is necessary to identify any other unwanted cell types that might arise and persist in the final product. An example of this was seen with the first ever clinical trial to test a human embryonic-derived cell therapy, which consisted of embryonic-derived oligodendrocyte progenitor cells (OPCs) developed by Geron Corporation as a direct intraparenchymal spinal treatment for patients with subacute thoracic spinal cord injury. Prior to its initiation, the trial was put on temporary hold after preclinical safety studies indicated the presence of microscopic epithelial cysts in some rats treated with the OPCs [101], thus indicating that this type of NSC product may require detection or removal of epithelial impurities as part of the manufacturing process. Based on a favorable safety profile with the five patients treated to date, it was recently announced that this OPC product, since acquired by Asterias Biotherapeutics, Inc., will advance into a phase I/phase IIa dose escalation study for patients with subacute cervical spinal cord injury (www.​asteriasbiothera​peutics.​com).

Based on guidelines provided by the US Food and Drug Administration (FDA), the development of candidate NSC products must include extensive preclinical testing to sufficiently characterize key functional properties posttransplantation, including their proliferative capacity, differentiation potential, and putative mechanism(s) of action [40, 102]. The proliferative capacity of an NSC product for stroke is an important aspect of its safety profile, as unrestricted growth could lead to transplanted cells spreading beyond the intended graft site or even tumorigenesis. However, most preclinical stroke studies with NSCs indicate very low posttransplant survival and proliferation [21, 31, 51, 52]. Efforts to enhance this, such as hypoxia preconditioning [37, 103] or coadministration of NSCs with some kind of structural support matrix [104, 105], may improve efficacy of NSCs, particularly for large stroke lesions, but whether this will also increase the tumorigenic risk remains to be determined.

As discussed in “Mechanisms of Action for Transplanted NSCs in Stroke,” maximizing the efficacy of NSCs in stroke requires an in-depth understanding of how transplanted NSCs act in the host brain to promote recovery, including their capacity to survive, differentiate, integrate into the host brain, and stimulate endogenous repair through paracrine signaling. For example, the in vivo secretion profile of transplanted NSCs remains largely unknown, but such knowledge could guide development of optimized NSCs that express higher levels of active paracrine factors or enrichment strategies used to purify active subpopulation(s) from an NSC product. It is widely accepted that NSCs are heterogeneous, and this may be advantageous especially if multiple cell types (e.g., neurons, astrocytes, and oligodendrocytes) are required for efficacy. However, if only a certain subpopulation of cells is required (e.g., VEGF-expressing cells or neuronal precursors) then enrichment for this active suppopulation could increase the “effective dose” of the cells.