Identification and characterization of cells infiltrating the graft and aqueous humour in rat corneal allograft rejection

In a rat model of corneal transplantation, Fischer 344 (RT1(lv1)) rats received orthotopic corneal isografts or Wistar-Furth (RT1(u)) donor allografts. Rejection was observed in 25 of 26 allograft recipients, at a median time of 18 days, with all isografts surviving > 100 days. Flow cytometric analysis of aqueous humour identified cellular infiltration of the aqueous at the time of allograft rejection, in contrast to the acellular aqueous found in isografts at corresponding times following transplantation. A higher proportion of CD8(+) than CD4(+) cells was found at days 1-3 following rejection, whereas there was a higher proportion of CD4(+) cells at days 5-8. No changes in peripheral blood T cell subsets were found at the time of rejection. Immunohistochemical analysis of cells infiltrating recipient iris and grafted cornea undertaken at days 1-2, 4 and 7-10 following onset of rejection, demonstrated inflammatory cells in the graft epithelium, stroma and aggregated on the endothelium. Large numbers of macrophages, T cells (CD4(+) > CD8(+) at all time points), natural killer (NK) cells and neutrophils were detected in graft tissue at days 1-2 and 4, diminishing after that time. Most infiltrating cells expressed MHC class II antigen, and a smaller number expressed IL-2R. Expression of the co-stimulatory marker B7 was identified in a few cells at day 4 in the region of the graft-host wound. The immune response in graft rejection was characterized at day 4 also by expression of intercellular adhesion molecule-1 (ICAM-1) on endothelial cells of iris and corneal vessels, demonstration of interferon-gamma on mononuclear cells in the peripheral (recipient) cornea, and tumour necrosis factor-alpha on aggregated mononuclear cells on the graft, but not recipient, endothelium. Only sparse cellular infiltrates were found in isograft controls, with inflammation located at the graft-host wound. These findings suggest that inflammatory cells reach a corneal allograft by two routes-from vessels in the peripheral recipient cornea, and from vessels in the recipient iris via the aqueous humour. Different aqueous and intragraft T cell subset proportions were seen early in rejection, although a preponderance of CD4(+) cells was found in both aqueous and graft at later times.