Human neural precursor cells in spinal cord repair

Abstract

Traumatic spinal cord injury (SCI) often results in substantial neural cell death, axonal degeneration and demyelination, accompanied by loss of sensory and motor functions. Cell therapy may have neuroprotective effects, substitute for lost neural cells, promote regeneration and enhance remyelination in the lesioned spinal cord. The availability of neural precursor cells (NPCs) suitable for both experimental and clinical purposes may be significantly improved by expansion of these cells in vitro. The primary objectives of the present thesis have been to evaluate conditions for culturing and in vitro expansion of human NPCs and to characterize the suitability of these cells for therapeutic treatment of SCI.

NPCs derived from the human first trimester forebrain and spinal cord were cultured as neurospheres in the presence of epidermal growth factor (EGF) and basic fibroblast growth factor (bFGF), with or without ciliary neurotrophic factor (CNTF). First trimester forebrain up to 12 weeks of gestation can serve as source for neurospheres, while spinal cord NPCs only expand successfully when the originating tissue is ≤ 9.5 weeks of gestation. Within these developmental time frames, gestational age of the donor tissue did not affect the proportion of cells with different phenotypes in culture. In vitro propagation enhanced the size of the cell population immunoreactive (IR) for CD133 without altering the numbers of cells expressing Tra-1-60-, Tra-1-81-, stage-specific embryonic antigen-4 (SSEA-4)-, nestin- and proliferating cell nuclear antigen (PCNA). The fraction of glial fibrillary acidic protein (GFAP)-IR cells increased, while the proportion of β-tubulin III-IR cells was reduced following five or more passages in the case of cultures originating from the spinal cord, but not from forebrain. Thus, even under the same culture conditions, the cellular composition of human neurospheres was dependent on the length of propagation in vitro and the regional heterogeneity of the tissue from which they were derived.

CNTF enhanced the rate of cell proliferation and expansion of human NPCs derived from the embryonic spinal cord without significantly altering the proportions of cells that expressed nestin, GFAP,β-tubulin III or O4. In contrast, this factor exerted no such influence on human NPCs derived from agematched forebrain. This difference might reflect, at least in part, differences in the expression of the CNTF receptor-α protein in these two regions of the central nervous system during development.

In vitro cultures of human NPCs expressed human leukocyte antigen (HLA) class I and II, but almost no co-stimulatory proteins (CD40, CD80 and CD86). The level of HLA increased during expansion of both spinal cord and forebrain NPCs. This phenomenon could, at least in theory, enhance the immunogenicity of NPCs and the risk for rejection after transplantation, but human peripheral lymphocytes exhibited no response when co-cultured with NPCs expanded in vitro. Together, these observations indicated that expanded NPCs demonstrate a low degree of immunogenicity, despite their pronounced expression of HLA, and the incompatibility of this antigen with that of a potential recipient.

In order to study the patterns of migration of human NPCs implanted into the injured spinal cord, these cells were transplanted close to traumatic spinal cord lesions performed in rats by clip compression or partial transection. A small fraction of the grafted NPCs had migrated, always confined to the white matter, several millimeters away from the site of implantation. There was no evidence of directed migration of grafted NPCs towards any of the two spinal cord lesions at six weeks after transplantation. The proportion of cells expressing nestin, GFAP or β-tubulin III, was similar at the graft center and among the cells that had migrated furthest, suggesting that their motility was not influenced by the degree of differentiation. In agreement, pre-differentiation of NPCs in vitro, achieved by inhibiting γ-secretase, did not affect migration of the grafted cells. Reduction of the extracellular chondroitin sulfate proteoglycans by chondroitinase ABC promoted migration of NPCs, demonstrating that although no evidence for homing of grafted NPCs was detected, extrinsic factors can affect their migration. In conclusion, implanted human NPCs mainly remained at the site of transplantation, a feature that may be beneficial for purposes of local neuroprotection and cell replacement following focal SCI.