In the present study, we showed that for a moderate contusive spinal cord injury, surviving cells that were transplanted 2 hours after injury were still detected after 4 weeks and that such acutely administered cell transplants may reduce the amount of secondary injury and partially restore the functionality of the ascending sensory pathways in the spinal cord. Takahashi et al.
demonstrated that BLI technology can successfully detect a minimum of 1,000 mouse cells in a model of contused SCI 
. Our results are similar to previous reports of stem cell-derived cell transplants in mouse models of SCI 
for which BLI intensity was high on the first day after transplantation but significantly decreased during the following two weeks. At 2 weeks post-transplantation, bioluminescence was significantly higher in the laminectomy sham group compared to the injured group; this may be due to inflammation and other secondary injury mechanisms that occur following spinal cord trauma that lead to a hostile microenvironment for the transplanted cells. However, there was no statistically significant difference in the observed BLI intensity between the injury and laminectomy groups at subsequent time points. This indicates that although cell survival at 2 weeks was reduced due to the SCI, cells nonetheless survived the secondary injury phase and were able to proliferate, reaching cell counts comparable to that which was found in a healthy sham control by week 3. Thus, contrary to conventional beliefs, the contusive injury microenvironment almost immediately after injury (2 hours) can still accept transplants without affecting long-term cell survival.
One study using iPS cell-derived neurospheres for transplant after contusive SCI has reported some evidence of MEP improvement after SCI 
. No study to date has evaluated the effect of stem cell derived transplants on the somatosensory pathways in SCI using electrophysiology, despite the fact that the somatosensory pathways receive the primary impact of the majority of contusive SCIs in humans. Somatosensory evoked potentials or SSEPs, are used to assess integrity of sensory tracts of the spinal cord. The improvement in the function of the sensory pathways can be seen by the increase in SSEP amplitude over the course of 6 weeks post-injury. Our results show for the first time the extent to which SSEP amplitudes were reduced in rats immediately following injury was less in rats receiving hES cell transplants compared to controls. This benefit in OPC treated rats is unlikely to be due to remyelination, as two weeks is not a sufficient time for immature OPCs to differentiate into mature OLs and produce sufficient and integrated myelin. However, others have shown the ability of progenitor cells to reduce acute inflammatory responses in damaged regions of the central nervous system (CNS
). These studies have included the transplantation of neural precursor cells and neural-induced mesenchymal stem cells (MSCs
) into the injured CNS, which result in reduced inflammation and fibrosis associated with secondary injury processes and improved neuronal regeneration 
. Although, the precise mechanisms through which these transplanted progenitors attenuate inflammation after SCI are still unclear, it is reasonable to infer that hES-derived OPCs may also possess some of the same anti-inflammatory properties that have been observed for other less differentiated progenitors.
Consistent with the changes in SSEP amplitude, the latencies of both groups were affected by their treatment. Initially, the SSEP latencies of both groups increased by approximately 40% at day 4 post-injury. This increase reflects the decrease in conductivity that results after injury due to damage to both the axons and their surrounding insulating myelin. Our results showed that following OPC transplantation, the latency of the OPC treatment group decreased to normal baseline values by week 6. Therefore, by 6 weeks after injury, the presence of transplanted OPCs may facilitate the remyelination of intact axons and restore their conductivity. This would be suggested by the increased conductivity via remyelination of spared axons that would then result in a decrease in the latencies of SSEPs, which are measured at the cortex due to stimulation rostral to the site of injury. Although SSEP latency recovered in OPC treated rats, SSEP amplitudes in these animals remained below baseline values. This is consistent with the persistence of the injury-induced cavity, which remains in treated animals such that any severed axons are unable to re-grow across the lesion.
In addition, it can be noted that the SSEPs of each animal did not always recover in a symmetrical fashion (). Although the contusion injury is induced at the midline, the natural progression of injury in all directions (laterally, proximodistally, or ventrodorsally) over time is unpredictable both in human and experimental models. The SSEP is an objective and sensitive tool to assess the integrity of the sensory pathways, which extends from periphery to the somatosensory cortex. It reflects the true electrical conductivity of the dorsal ascending pathways, which in turn reveals the progress of injury and not just the damage induced at the site of first impact. Therefore, it is reasonable to observe that a midline contusion in the spinal cord, which evolves into an asymmetrical injury over time, results in dissimilar SSEP recovery of the left and right sides.
It has been postulated that maximal amplitude improvement in SSEP is a function of the absolute number of axons that remain intact and that cavitation prevents the endogenous regrowth of axons across the lesion site 
. Thus, the recovery in SSEP latency in OPC treated rats may indicate that this fraction of axons, which were non-functional due to demyelination, were successfully remyelinated and their conductivity was restored to pre-injury levels. The general somatic afferents transduce signals of touch, pressure, movement and body position from the periphery to the cortex. They are two main sensory pathways: one from the lower trunk and legs called fasciculus gracilis located in the medial dorsal column and one from the upper trunk and arms called fasciculus cuneatus located in the lateral dorsal column. The neurons (first-order) of the fasciculus gracilis and fasciculus cuneatus ascend uninterrupted in the dorsal column and terminate in their respective nuclei in the medulla, the nucleus gracilis and the nucleus cuneatus. Here, second-order neurons of the internal arcuate fibers originate and cross to the contralateral brainstem in the medulla, eventually synapsing in the thalamus and activating third-order neurons that reach the cortex. The latency we reported herein reflects only changes in first-order neurons of the spinal cord. We report the latency recovery of the injured hindlimbs after transplantation with ES-derived OPCs. To verify that this recovery is due to repair at the injury site, we underscore that the latency of the forelimb SSEPs remains unchanged (). This indicates that the function of second- and third-order neurons did not change over time for either experimental group. Therefore, the latency recovery of the OP-treated group can only be due to recovery of the first-order hindlimb neurons of the fascilus gracilis. In addition, because the untreated control group did not exhibit improved latency, we can conclude that the reduction in latency is due to remyelination of neurons, allowing faster conduction through the site of injury.
This is consistent with the histological analysis performed on spinal cords from euthanized animals at the end of the longitudinal BLI and SSEP studies. The tumorigenicity of hES-derived cell lineages has been a point of concern over the past decade and has been observed in various studies 
. Our BLI results show that the transplanted OPCs did not migrate more than 1–2 mm from the locus of the injury or to other regions of the CNS, such as the brain. The pluripotent marker OCT4 was not detected in spinal cords treated with transplanted cells. Although this study incorporated a time in which tumor formation from ES cell transplantation can easily be detected it does not rule out the possibility of tumor formation had the study proceeded for a longer period of time. Thus, long-term experiments are required for this determination, which were not the intent of this study.
We also performed a series of immunostains for markers of oligodendrocytes, and myelin on slices of extracted spinal cord segments to verify the nature of our transplanted cells. ES-OPCs began to express MBP two weeks after transplantation. This suggests the presence of mature human OLs derived from the transplanted OPCs. Others have reported that stem cell-derived OPCs differentiate into other neural subtypes including astrocytes 
. However, the human transplanted OPCs identified by human specific antigen did not show the presence of GFAP+
cells, indicating that the transplanted OPCs did not give rise to astrocytes.
The injured spinal cord is known to express a host of factors that inhibit the outgrowth and differentiation of OLs at the site of injury. To maximize the remyelination ability of our transplanted OPCs, additional steps can include a combination of cell therapy with the manipulation of growth factors or various signaling molecules. For example, bone morphogenic proteins (BMPs) expressed by reactive astrocytes inhibit the differentiation of OPCs to OLs while promoting the formation of astrocytes 
, and the release of BMP is drastically increased following SCI. Blocking BMP activity by the antagonist noggin reversed the effect and promoted OPCs to follow the OL lineage 
. Reactive astrocytes in the glial scar may also inhibit OPCs via the release of tumor necrosis factor-α 
Finally, we were able to observe the formation of new thinly wrapped myelin sheaths around axons for an OPC treated rat using TEM. Because no reliable myelin associated antibodies that can distinguish human and rat exist for histological purposes, it is not possible to distinguish myelin that is produced by the injected OPCs from endogenously produced myelin. However, it is promising that newly formed myelin was only observed in the rat that received the OPC injection. Furthermore, because our immunohistochemical evaluation verified that our injected cells did form myelin-producing OLs, we can infer that at least a portion of these OLs functioned to remyelinate axons.
There remains much to be investigated regarding how stem cells mediate recovery in spinal cord injured patients. The ability to track cell survival and migration serves to augment current knowledge about the mechanisms of stem cell therapies. Here, we have shown for the first time improvements in functional electrophysiological behavior in conjunction with biological evidence of both OL differentiation and myelin formation in vivo. Combining this evidence with additional studies in order to elucidate the full mechanisms of stem cell integration and spinal cord repair will inevitably lead to un-matched treatment for patients with SCI in the near future.