Interaction of NG2(+) glial progenitors and microglia/macrophages from the injured spinal cord.

Spinal cord contusion produces a central lesion surrounded by a peripheral rim of residual white matter. Despite stimulation of NG2(+) progenitor cell proliferation, the lesion remains devoid of normal glia chronically after spinal cord injury (SCI). To investigate potential cell-cell interactions of the predominant cells in the lesion at 3 days after injury, we used magnetic activated cell sorting to purify NG2(+) progenitors and OX42(+) microglia/macrophages from contused rat spinal cord. Purified NG2(+) cells from the injured cord grew into spherical masses when cultured in defined medium with FGF2 plus GGF2. The purified OX42(+) cells did not form spheroids and significantly reduced sphere growth by NG2(+) cells in co-cultures. Conditioned medium from these OX42(+) cells, unlike that from normal peritoneal macrophages or astrocytes also inhibited growth of NG2(+) cells, suggesting inhibition by secreted factors. Expression analysis of freshly purified OX42(+) cells for a panel of six genes for secreted factors showed expression of several that could contribute to inhibition of NG2(+) cells. Further, the pattern of expression of four of these, TNFalpha, TSP1, TIMP1, MMP9, in sequential coronal tissue segments from a 2 cm length of cord centered on the injury epicenter correlated with the expression of Iba1, a marker gene for OX42(+) cells, strongly suggesting a potential regional influence by activated microglia/macrophages on NG2(+) cells in vivo after SCI. Thus, the nonreplacement of lost glial cells in the central lesion zone may involve, at least in part, inhibitory factors produced by microglia/macrophages that are concentrated within the lesion.

NG2 cell response in the CNP-EGFP mouse after contusive spinal cord injury.

NG2(+) cells in the adult CNS are a heterogeneous population. The extent to which the subpopulation of NG2(+) cells that function as oligodendrocyte progenitor cells (OPCs) respond to spinal cord injury (SCI) and recapitulate their normal developmental progression remains unclear. We used the CNP-EGFP mouse, in which oligodendrocyte lineage cells express EGFP, to study NG2(+) cells in the normal and injured spinal cord. In white matter of uninjured mice, bipolar EGFP(+)NG2(+) cells and multipolar EGFP(neg)NG2(+) cells were identified. After SCI, EGFP(+)NG2(+) cell proliferation in residual white matter peaked at 3 days post injury (DPI) rostral to the epicenter, while EGFP(neg)NG2(+) cell proliferation peaked at 7 DPI at the epicenter. The expression of transcription factors, Olig2, Sox10, and Sox17, and the basic electrophysiological membrane parameters and potassium current phenotype of the EGFP(+)NG2(+) population after injury were consistent with those of proliferative OPCs during development. EGFP(neg)NG2(+) cells did not express transcription factors involved in oligodendrogenesis. EGFP(+)CC1(+) oligodendrocytes at 6 weeks included cells that incorporated BrdU during the peak of EGFP(+)NG2(+) cell proliferation. EGFP(neg)CC1(+) oligodendrocytes were never observed. Treatment with glial growth factor 2 and fibroblast growth factor 2 enhanced oligodendrogenesis and increased the number of EGFP(neg)NG2(+) cells. Therefore, based on EGFP and transcription factor expression, spatiotemporal proliferation patterns, and response to growth factors, two populations of NG2(+) cells can be identified that react to SCI. The EGFP(+)NG2(+) cells undergo cellular and physiological changes in response to SCI that are similar to those that occur in early postnatal NG2(+) cells during developmental oligodendrogenesis.

Stem cells in spinal cord injury.

Traumatic injury to the adult spinal cord results in a massive loss of cells and permanent functional deficits. However, recent studies demonstrate that there is a proliferative response of endogenous glial precursors and progenitors and perhaps also pluripotent neural stem cells. These cells may prove to be an important new therapeutic target to improve recovery after injury to the spinal cord and brain.

Glial cell loss, proliferation and replacement in the contused murine spinal cord.

Studies in the rat have shown that contusive spinal cord injury (SCI) results in devastating pathology, including significant loss of mature oligodendrocytes and astrocytes even in spared white matter. Subsequently, there is increased proliferation of endogenous NG2(+) cells, postulated to contribute to replacement of mature glia chronically, which is important for functional recovery. Studies of mechanisms that stimulate endogenous progenitor cells would be facilitated by using mouse models with naturally occurring and genetically engineered mutations. To determine whether the murine response is similar to that in the rat, we performed contusive SCI on adult female C57Bl/6 mice at the T8-9 level. Animals received bromodeoxyuridine injections in the first week following injury and were killed at 1, 3, 4, 7 or 28 days postinjury (DPI). The overall loss of macroglia and the temporal-spatial response of NG2(+) cells after SCI in the (C57Bl/6) mouse was very similar to that in the (Sprague-Dawley) rat. By 24 h after SCI nearly half of the macroglia in spared ventral white matter had been lost. Cell proliferation was increased at 1-7 DPI, peaking at 3-4 DPI. Dividing cells included NG2(+) cells and Cd11b(+) macrophages and microglia. Furthermore, cells dividing in the first week expressed markers of mature glia at 28 DPI. The similarities in endogenous progenitor cell response to SCI in the mouse and rat suggest that this is a fundamental injury response, and that transgenic mouse models may be used to further probe how this cellular response to SCI might be enhanced to improve recovery after SCI.

Phenotypic changes in NG2+ cells after spinal cord injury.

Following contusive spinal cord injury (SCI), 50% of oligodendrocytes in the residual white matter are lost within 24 h. NG2-expressing cell proliferation is maximal 3 days after SCI, and may be the source of mature oligodendrocytes and astrocytes that chronically replace those that were lost. We studied NG2(+) cells dissociated from the 3-day injured spinal cord for comparison with those from uninjured adult and early postnatal cords. After 24 h in serum-containing medium, we performed patch clamp analysis and immunocytochemistry for NG2 in combination with nestin (progenitors), and A2B5, O4, and O1 (oligodendrocyte lineage markers). We observed an NG2(+)/A2B5-/O4-/O1- population in both adult preparations. More than double the normal number of NG2(+) cells was isolated from the injured cord, but OX42(+) microglia/macrophages were the predominant cell type after injury. Most cells isolated at P7 were NG2-/A2B5(+), whereas those from the normal adult were NG2(+)/A2B5-. NG2(+) cells after SCI displayed altered voltage-gated potassium current profiles compared to normal adult and P7 animals. Additionally, less than 25% of adult cells (normal and injured) responded to GABA and glutamate, compared to 100% of P7 cells. Our results indicate that the adult NG2(+) cell pool is antigenically and physiologically different than the early postnatal pool, and that contusive injury induces changes in adult NG2(+) cells.