Traumatic spinal cord injury results in immediate functional impairments below the level of injury caused by the loss of neural cells and axons due to the initial impact. Following this primary injury, a secondary pathophysiological cascade causes progressive tissue loss for weeks to months after the insult, leading to the formation of fluid-filled cysts surrounded by scar tissue [1]. The endogenous response within the injured spinal cord fails to reorganize spinal cord tissue in a way that leads to functional repair. Currently no treatments exist that can effectively restore lost motor, sensory and autonomous function after spinal cord injury.

Mesenchymal stem cells (MSCs), derived mostly from bone marrow, but also from adipose tissue and umbilical cord, are being studied as a potential repair strategy for spinal cord injury. Typically, MSCs can be easily isolated, cultured and prepared for transplantation into a spinal cord lesion. MSCs secrete numerous molecules that are known to exert paracrine effects resulting in repair. After spinal cord injury, MSCs have the potential to decrease secondary tissue loss after spinal cord injury and this neuroprotective effect has been shown to be correlated with moderate functional improvements [2]. MSCs secrete neurotrophic factors, such as brain-derived neurotrophic factor, glialderived growth factor and nerve growth factor that have the potential to decrease neuronal apoptosis and/or promote axonal regeneration [3]. In addition, MSCs secrete factors that have proliferative and stabilizing effects on blood vessels, including vascular endothelial growth factor and angiopoietin-1, respectively [4]. However, survival of MSCs in the injured spinal cord is poor, limiting the availability of these trophic factors to the nearby nervous tissue and thus their effects that lead to repair. Because it has been shown that improved survival of MSCs is associated with improved anatomical and/or functional repair [5-13], it is important to understand mechanisms of transplanted cell death and to develop strategies to improve MSC survival. This review provides an overview of studies that have investigated MSC survival in the injured spinal cord and summarizes current MSC survival promoting strategies, specifically focusing on bone marrow- derived MSCs.

Traumatic spinal cord injury results in immediate functional impairments below the level of injury caused by the loss of neural cells and axons due to the initial impact. Following this primary injury, a secondary pathophysiological cascade causes progressive tissue loss for weeks to months after the insult, leading to the formation of fluid-filled cysts surrounded by scar tissue [1]. The endogenous response within the injured spinal cord fails to reorganize spinal cord tissue in a way that leads to functional repair. Currently no treatments exist that can effectively restore lost motor, sensory and autonomous function after spinal cord injury.

Mesenchymal stem cells (MSCs), derived mostly from bone marrow, but also from adipose tissue and umbilical cord, are being studied as a potential repair strategy for spinal cord injury. Typically, MSCs can be easily isolated, cultured and prepared for transplantation into a spinal cord lesion. MSCs secrete numerous molecules that are known to exert paracrine effects resulting in repair. After spinal cord injury, MSCs have the potential to decrease secondary tissue loss after spinal cord injury and this neuroprotective effect has been shown to be correlated with moderate functional improvements [2]. MSCs secrete neurotrophic factors, such as brain-derived neurotrophic factor, glialderived growth factor and nerve growth factor that have the potential to decrease neuronal apoptosis and/or promote axonal regeneration [3]. In addition, MSCs secrete factors that have proliferative and stabilizing effects on blood vessels, including vascular endothelial growth factor and angiopoietin-1, respectively [4]. However, survival of MSCs in the injured spinal cord is poor, limiting the availability of these trophic factors to the nearby nervous tissue and thus their effects that lead to repair. Because it has been shown that improved survival of MSCs is associated with improved anatomical and/or functional repair [5-13], it is important to understand mechanisms of transplanted cell death and to develop strategies to improve MSC survival. This review provides an overview of studies that have investigated MSC survival in the injured spinal cord and summarizes current MSC survival promoting strategies, specifically focusing on bone marrow- derived MSCs.

MSC Survival Rates

MSCs can be tracked after transplantation by virally transducing the cells to express green-fluorescent protein (GFP) or isolating cells from a GFP-transgenic donor. Most reports on MSC transplantation after spinal cord injury provide only qualitative or semi-quantitative data on MSC transplant survival. Table 1 provides an overview of studies that have reported on bone marrow-derived MSC survival after transplantation into animal models of spinal cord injury. Studies that provide quantitative data on MSC survival after SCI report survival rates between 0 [14,15] and 52% [8] one week after transplantation and between 0 [6,16-20] and 8% [21] one month after transplantation without survival enhancing therapies. In some cases, presence of MSCs up to two months [2,13,22-25] and even three months [11,26,27] was reported after transplantation into models of spinal cord injury, but usually no or very few cells survive at these time points. The large variation in survival rates can in part be explained by the model system used. Interestingly, in spinal cord transection models, cells are usually reported to be present at the end point of the study, with reported survival rates up to 7% at eight weeks after transplantation [28]. In the transected spinal cord, a piece of gelfoam is often used to fill the injury gap and/or to provide a scaffold for the MSCs. Alternatively, cells are injected directly in the lesion penumbra, i.e., the nervous tissue adjacent to the actual transection. The environment in the lesion penumbra is different from the lesion epicenter in terms of immune cell presence, scar tissue and blood supply [29]. Together, this may explain why survival seems to be better in transection models than in contusion models in which the cells are mostly injected into the lesion environment. Because contusion models are clinically more relevant, in more than 70% of the cases a contusion is the mechanism of injury, it is imperative to understand the low and variable survival rates in contusion models of spinal cord injury.

Table 1: Overview of studies that have investigated bone marrow-derived MSC survival after transplantation into the injured spinal cord.

Effect of Timing, Dose and Delivery Method on MSC Transplant Survival

One important factor that should be accounted for is the timing of cell transplantation. Table 2 gives an overview of studies that have studied the effect of timing of transplantation on MSC survival in spinal cord injury models. Of the four studies that so far looked at survival after intralesional transplantation, three report better or similar survival after acute or 3-day-delayed transplantation than after 7-, 14- or 21-days-delayed transplantation [8,12,16]. The fourth study described better survival after 7-day-delayed transplantation than after acute transplantation. However, here the survival rate was reported at 2 8 days after the injury, implicating that the transplant was in fact quantified after 21 days, which could account for the discrepancy. MSCs seem to be able to migrate to the injury site and survive to some extent both when injected in the spinal cord away from the injury site [30], intrathecally [7,15,19,20,31,32] and intravenously [9,33,34]. A direct comparison of cell survival after an intralesional injection or intravenous injection of MSCs revealed that the former approach resulted in better survival [33]. Studies of direct comparisons between the other delivery modes are absent. After intrathecal delivery up to 5% of MSCs can survive 28 days after transplantation after repeated weekly injections starting 7 days post-injury [31]. With intrathecal delivery, three-day-delayed [19,31] or late injection (≥ 14-day-delayed) [7,15] seems less beneficial for cell survival than 7-day-delayed injection. However, a 5-day-delayed injection was shown to lead to better survival than a 7-day-delayed injection [19]. Given the current data, it seems reasonable to conclude that the optimal time point for intralesional transplantation is three days post-lesion. Although less data is available about the beneficial effects of intrathecal MSC injections and differences in methodology between studies make direct comparisons difficult, current data seem to suggest the use of intrathecal injections as a delivery method for MSC therapy for spinal cord injury [7,15,31]. For intrathecal injection the optimal time point seems to be five to seven days after injection. These time points are also favorable for clinical translation. Repeated deliveries (three deliveries at weekly intervals) seem to have beneficial effects on cell survival and associated anatomical and functional recovery after intrathecal delivery of MSCs [7,31]. A repeated injection regimen, however, seems problematic for intralesional injections where multiple surgeries will be necessary. Studies using single MSC injections did not find a dose-effect on MSC survival or associated recovery [7,28].

Table 2: Overview of studies that have investigated the effect of timing of MSC transplantation on survival of MSCs in animal models of spinal cord injury.

The Role of the Immune System in MSC Transplant Survival

There are a number of plausible causes of death of MSCs after transplantation into the injured spinal cord. The overwhelming presence of neutrophils in the first days after injury, followed by activation of resident microglia and a massive influx of macrophages may cause MSC death by direct phagocytosis. However, simply reducing the presence of macrophages does not increase MSC survival [35]. Neutrophils and macrophages may also cause MSC death by the formation of reactive oxygen species that cause membrane damage leading to death. Indeed, transplanting the cells within a polyurethane-based biogel with antioxidative properties increases short term (one week) survival of MSCs [6]. Longer term (4 week) survival however, is unaffected by this gel, probably due to biodegradation of the gel. The adaptive immune response may also play a role in MSC death after transplantation, the extent of which, however, is debated. MSCs have low expression of MHC class I molecules and absence of co-stimulatory molecules and have been reported to suppress the function of T-cells, B-cells, natural killer cells and dendritic cells [36]. The effect of immunosuppressants on MSC survival has been studied by different groups with conflicting results. In syngeneic transplantation, where cells are taken from an inbred strain and transplanted into a genetically similar individual from the same inbred strain, immunosuppressants are deemed unnecessary because the contribution of the immune system to MSC death is thought to be low [24]. In allogeneic transplantation, i.e. cells from an outbred strains and transplanted into individuals from another strain, immunosuppressants were shown to be effective by some investigators [21,24] but not by others [20,35]. Cyclosporine A (CsA) dose may partly explain these differences. Indeed, Parr et al. [21] studied the effect of CsA dose on MSC survival after 28 days and showed no effect of low dose CsA treatment (10 mg/kg/day) but a small statistically significant effect of high dose CsA treatment (20 mg/kg/day) on MSC survival 28 days after injection (8% vs 11%) in an allogeneic model system. Similarly, Swanger et al. [24] showed increased graft volumes, which were interpreted as larger cell numbers, with high dose CsA treatment (30 mg/kg/day for three days prior to transplantation, followed by 15 mg/kg/day for the duration of the study), compared to low dose (10 mg/kg/day) CsA treatment at 4 and 8 weeks post-transplantation. This difference was seen in an allogeneic model using MSCs from transgenic Fischer (inbred) rats transplanted into Sprague-Dawley (SD; outbred) rats without a spinal injury. In another study, in which MSCs from SD rats were transplanted into SD rats, this same dose regimen did not result in improved MSC transplant survival [35]. Aside from the strain difference, the fact that the latter study used spinal cord injured rats whereas the former study used uninjured rats likely explains the difference, since MSC death mechanisms in the injured spinal cord are different from those in the uninjured spinal cord. Xenotransplantation of MSCs, in which cells from one species (typically humans) are transplanted into another species (typically rats), usually results in poor survival even when using nude rats or CsA after injection into a spinal cord contusion [17]. This kind of rejection, however, is less relevant from a translational point of view, since a projected MSC therapy for human spinal cord injury would optimally be autologous and at the least allogeneic.

Deprivation from Oxygen, Nutrients and Growth Substrates

After spinal cord injury, there is a shortage of oxygen and nutrients resulting from rupturing of blood vessels and death of endothelial cells caused by the initial impact and subsequent inflammatory processes. The ischemic environment likely contributes to MSC death after transplantation, although there is some evidence that MSCs are relatively resistant to hypoxia [37]. An increasing body of research is focusing on strategies to restore vascularization after spinal cord injury, the effect of which on MSC survival has yet to be determined. Another mechanism of MSC death in an injury site may be anoikis, which is defined as apoptosis induced by the lack of a substrate for attachment. Indeed, MSCs express integrin receptors, providing ligands for attachment to extracellular matrix molecules including collagen, fibronectin and laminin [38]. MSCs are anchorage-dependent cells and lack of a substrate to adhere to may induce apoptotic pathways. Transplantation of MSCs within scaffolds, in addition to possibly providing protection against macrophages, may prevent anoikis. Indeed, transplantation within a fibrin matrix was shown to increase MSC survival compared to controls [39].

In addition to the survival enhancing strategies described above, including optimizing timing, immunosuppression and transplantation within a scaffold for protection against anoikis, reactive oxygen species and macrophages, a number of other treatments have been shown to be successful at increasing MSC survival. Table 1 summarizes these studies. Therapies include blockade of the IL-6/IL-6receptor [12], silencing p75 receptors in MSCs [10], culturing MSCs under microgravity conditions [9], electroacupunture [40,41], electrical stimulation [42], co-treatment with granulocyte-colony stimulating factor [13], Salvianolic acid B [43] or a NOGO-66R antagonist [22] and BDNF-overexpression in MSCs [11]. Strategies that failed to improve MSC survival include neural induction of MSCs [44] and co-culture with Schwann cells prior to transplantation [45]. Most of these studies however, did not provide quantitative data from which percentages of surviving cells could be derived. Instead, increases in MSC staining intensity, cell number per unit area, graft volume or only qualitative data are reported, making these studies very difficult to interpret and to compare.

There have been some early clinical trials assessing the safety and primary efficacy of MSC transplantation into spinal cord injured patients, concluding that MSC transplantation is feasible and safe [62,63]. Larger randomized, controlled, blinded clinical trials are needed to assess efficacy of MSCs in humans. Moreover, to our knowledge, survival rates of MSCs in humans have not been studied, but it is plausible that MSC survival enhancing strategies developed in rodents will benefit survival rates and efficacy of MSCs in humans.

MSC survival after transplantation in the injured spinal cord is poor, especially in the clinically relevant contusion models, which limits their therapeutic efficacy. Optimal timing, route of delivery, pretreatment of MSCs and co-treatment strategies may enhance survival to an extent, but quantitative data is scarce and when provided shows only small or short-term improvements in survival. It is clear from the current literature that more research is needed to elucidate mechanisms of MSC transplant death, so that rational survival enhancing strategies may be developed that can further develop MSC transplantation as a clinically relevant spinal cord injury therapeutic.

Financial support was provided by Wings for Life Foundation (grant WFLUS- 015/13) and Craig H. Nielsen Foundation (grant 284621)