A rat assay for CNS repair

Rah, Berend (2007). A rat assay for CNS repair. PhD thesis The Open University.

DOI: https://doi.org/10.21954/ou.ro.0000fb4a

Abstract

Olfactory ensheathing cells from the adult olfactory bulb (OECs) have been shown at our laboratory to facilitate axon regeneration in the spinal cord. They share similarities with Schwann cells at the histological level but appear to be more efficient at promoting axon regeneration. In particular, they enable axons to leave the transplant and re-enter the host CNS and establish lost connections with the target area. OECs and Schwann cells have been used at different stages of maturity and at varying purities. Besides these two cell types there are other candidate repair cells, most notably neural stem cells and OECs from the olfactory mucosa which may also be of use in reconstruction of severed fibre tracts. For screening the reparative abilities of these different cell types experiments in the spinal cord are time-consuming and technically challenging. The following experiments were therefore designed to address the need for an assay system to allow comparison of various candidate repair cells and deliver reliable, time-efficient and reproducible results.

Based on previous work from this laboratory, stereotaxically guided microinjection of cell suspensions into the lateral ventroposterior thalamic nucleus (VPL) was used to compare the behaviour of different cell types in terms of cell survival, cell migration, interaction with the host tissue and axon recruitment. OECs, adult Schwann cells (ASCs), neonatal Schwann cells (NSCs), neural crest stem cells (NCSSs) and OECs derived from either the rat or human olfactory mucosa were investigated. The sham group only received injections of medium into the thalamus.

Approximately 3-5 \i\ of cell suspension at a concentration of 100x106 cells/ml was injected into the thalamus via a micropipette that was slowly withdrawn so as to lay an artificial tract of the grafted cells. Animals were culled after survival times of one, two and three weeks and their brains processed for immunohistochemistry. Donor cells were identified by expression of the low affinity nerve growth factor receptor (p75), fibronectin and nestin. OECs from surgical biopsy samples of the human olfactory mucosa were detected by an antibody directed against human mitochondria (MIT) or by a human-specific p75 antibody. Axons were identified by immunoreactivity for neurofilament (NF). Host astrocytes and their processes were identified by glial fibrillary acidic protein (GFAP). An antibody against peripherin was used to identify NCSSs.

Samples were processed for immunohistochemistry based on the avidin-biotin complex method or for immunofluorescence studies using confocal microscopy. Immunohistochemistry was primarily employed to obtain tissue slides for quantitative assessment and to examine the immune response to human OECs transplanted into the thalamus.

Histological sections were then subjected to both qualitative and quantitative analysis. For qualitative analysis cell alignment and survival, interactions between host and donor tissue, and the recruitment of axons were examined.

The qualitative analysis showed all cell types survived. The cells formed a vertically orientated column aligned in parallel strands along the line of the pipette penetration of the thalamus. Differences in the shape and alignment of cells were noted. Neonatal Schwann cells and OECs from the rat mucosa displayed the highest degree of alignment and cells had long and slender cell nuclei in contrast to OECs from the bulb that had nuclei of plump and more oval morphology. NSC, ASC and OECs from bulb and mucosa could all be identified by p75. Fibronectin staining identified fibroblasts in OECs and Schwann cell transplants. Columns derived from all cell types were capable of recruiting axons. The astrocytic processes appeared to provide a guidance structure for axons. Astrocytes in close vicinity of the graft were hypertrophied but no significant differences were observed in their response to the different transplanted cell types employed.

NCSSs behaved in much the same way in vivo as they did in vitro. They differentiated into either neurons of immunohistochemically peripheral type or Schwann cell-like glial cells. They also appeared to recruit axons but it was impossible to ascertain the origin or destination of delicate axons seen inside the graft.

Human xenografts were investigated for immune rejection using antibodies directed against CD4, CD8, MHC I and MHC II. OECs derived from the human olfactory mucosa were invariably rejected by the rat immune system if only oral immunosuppression with cyclosporine was employed. The immune response was significantly reduced in animals receiving systemic cyclosporine with few lymphocytes infiltrating the graft and better cell survival. However, the immune response was not completely abolished. The grafts were surrounded by MHC ll-positive cells of microglial origin, but some grafted OECs could clearly be identified and survived well. The immune response measured by the degree of perivascular lymphocyte cuffing was notably stronger in animals that had only received oral cyclosporine. Perivascular cuffing of lymphocytes, a hallmark or intense rejection, was clearly visible in the vicinity of human grafts in orally treated animals. These cells also often displayed a honey-comb like arrangement and focal areas of necrosis.

Quantitative analysis was carried out for all cell types except NCSSs. It revealed that all cell types could recruit axons at a higher density than the sham group. NSC recruited the longest axons and at the highest density. There were unexpected differences between OECs obtained from the rat olfactory bulb and those from the olfactory mucosa, with the latter cell type being almost as successful at axon recruitment as the NSC in terms of mean axon density and mean axon length. OECs from the bulb recruited axons at far lesser density and axons of slightly shorter length than their counterparts from the olfactory mucosa. Axon density and mean axon length for adult Schwann cells fell between that for OECs from the bulb and NSCs. OECs from the human olfactory mucosa attracted the least number of axons which did not differ much from the sham group, although they achieved a slightly higher axon density than the sham group after 2 weeks. The mean axon length was comparable to the sham group. These measurements probably reflect persistent immunorejection of the graft despite the fact that these animals had received systemic immunosuppression with cyclosporine.

The thalamus as a model did not lend itself to the study of axon re-entry into the host parenchyma given the lack of a suitable labelling method and a well-defined pathway. Hence, the postcommissural fornix was chosen to investigate host re-entry of fibres. The fornix is a very accessible white matter tract with a well-defined unilateral projection to the mammillary body. Radiofrequency heat lesions were generated with an electrode that was stereotaxically lowered to a point crossing the path of the post commissural fornix. Following removal of the electrode, a micropipette was then lowered to the lesion site and the cell suspension injected. In the final step the fornix was labelled by injecting the subiculum with 2 jj,I of a 10 % solution of the tracer biotin dextran. For the main experiments, biotin dextran (BD)-labelling was performed shortly after the lesion had been generated, and the cell suspension had been injected. In some experiments, animals were labelled at the time of delayed grafting.

OECs and NSCs were transplanted into the fornix. The cells were identified with p75. Typically, a small column had formed at the dorsal site of the lesion, which ended in a somewhat spherical bolus of cells filling the lesion cavity. The bolus often had a central, necrotic core. In all cases, fornix fibres failed to advance into or past the lesion site. Regeneration into the bolus also failed when cells were transplanted with a 2 weeks delay. BD-labelled axons in close proximity to the lesion became thickened, exhibited many varicosities and often ended in growth-cone-like expansions. These axons resembled axons in pure lesions without grafting of cell suspensions, and just like them sprouted at the point were they came into contact with the lesion site.

Although axons were unable to advance into the lesion site, cells were seen to igrate from the graft into the proximal and distal ends of the severed fornix. At the proximal end cells clearly intermingled with the BD-labelled fibres. Interestingly, NF-immunoreactive axons were seen in both column and bolus but their origin could not be determined.

Embryonic rat brain cells obtained from whole brain preparations of 14 day old embryos failed to enable regeneration past the lesion site just like in the other cell types. These cells had a heterogeneous distribution and differentiated into various cell types. They sometimes appeared to be organised in subpopulations. They did neither express p75 nor GFAP with the exception of cells adjacent to surviving fibres of the fornix which expressed p75 weakly and diffusely.