Contributions of Donor and Host Blood Vessels in CNS Allografts

Abstract

The contributions of blood vessels in various transplantation paradigms of solid CNS tissue or cell suspension allografts placed into adult host brains were investigated immunohistochemically using the PVG-RT1Cand PVG-RT1Uinbred rat strains and a panel of highly specific monoclonal antibodies. The monoclonal antibodies included OX-27 and U9F4 against major histocompatibility complex (MHC) class I antigens of the PVG-RT1Cand PVG-RT1Urats, respectively; OX-26 against the rat transferrin receptor located on blood–brain barrier (BBB) endothelia; and OX-7 against rat neuronal Thy 1.1 for evaluating graft survival. Our study is the first to address the immunogenicity of blood vessels in surviving CNS allografts. Solid fetal or neonatal PVG-RT1Ccortex was grafted into the third or lateral cerebral ventricle or caudate/putamen of PVG-RT1Uadult hosts for 30 days to 7 months. All allografts expressed demonstrable Thy 1.1 immunoreactivity with OX-7 antibody and appeared well-vascularized with blood vessels that immunostained with the OX-26 antibody against the transferrin receptor. For the most part, the allografts were supplied sparsely with donor (PVG-RT1C) MHC class I-positive (OX-27) blood vessels clustered in pockets. Donor MHC class I-positive vessels entered the host brain only from allografts in the third ventricle; these vessels were restricted to the host median eminence and no longer immunostained with OX-26 for the transferrin receptor (normally the median eminence is supplied with non-BBB vessels that do not possess the transferrin receptor and do not stain with OX-26). In host brains harboring a third ventricle allograft, host MHC class I-positive vessels immunostained with the U9F4 antibody were evident throughout the host CNS, including the median eminence, and throughout the allografts excluding sites inhabited by donor PVG-RT1Cvessels. Cell suspension neural allografts (donor PVG-RT1C) placed within the brain parenchyma of PVG-RT1Uhosts revealed no significant differences in vascular contributions between donor and host when compared to results obtained from solid CNS allografts. A unique immunohistochemical approach of introducing ascites fluid OX-27 as the primary antibody intravenously to the PVG-RT1Uhost demonstrated that in donor PVG-RT1Cposterior pituitary allografts, donor and not host vessels predominate and are restricted to the graft. Finally, blood vessels isolated from adult PVG-RT1Cbrains were mixed with solid fetal PVG-RT1Ucortical tissue and grafted into the brain parenchyma of adult PVG-RT1Uhosts. Immunostaining with OX-27 antibody against MHC class I of the PVG-RT1Crat strain disclosed that the PVG-RT1Cblood vessels survived and were confined to the PVG-RT1Usyngeneic graft. The results suggest that blood vessels supplying CNS allografts placed within the host brain are predominantly of host origin; surviving donor vessels are restricted to the allograft with rare exceptions, which may be dictated by the type of neural allograft and the host CNS site receiving the allograft. The survival of isolated allogeneic CNS blood vessels grafted into the host brain suggests that such blood vessels can present an endothelial genotype and phenotype different from those of host vessels indigenous to the CNS site receiving the allogeneic vessel graft. This finding may have implications in the circumvention of the blood–brain fluid barriers for the CNS delivery of blood-borne therapeutics.