Objective To facilitate the identification of transplanted retinal pigment epithelial (RPE) cells, we sought to double-label the cells with 5-bromodeoxyuridine (BrdU) and with natural pigment. The BrdU is not lost during cell division but does require immunohistochemical methods for visualization; the pigment on the other hand, allows immediate, obvious identification, but is gradually lost with cell division. Together they provide a convenient, long-term double label.Methods Non-confluent RPE cells at the second to the fifth passages were labelled with 5-BrdU and pigment. The double-labelled RPE cells were transplanted onto Bruch's membrane of 72 eyes of New Zealand albino rabbits. The labelled cells were localized by anti-BrdU antibody and the avidin biotin-alkaline phosphatase complex (ABC-AP) method, and by visible inclusions of pigment. Results The transplanted RPE cells had distinct basal and apical morphology, and were in close contact with the photoreceptor outer segments of the host. The BrdU label was restricted to the nuclei of the RPE cells, which were stained blue. The pigment was located in the cytoplasm of the apical portion of these RPE cells. No evidence of severe rejection was seen.Conclusions Using this double-label method, transplanted RPE cells could be readily and reliably identified till one year after transplantation. The transplanted RPE cells revealed normal morphology with some function: they had distinct basal and apical morphology as seen in close contact with the outer segment of photoreceptor cells. The junctional complexes were well formed with neighboring RPE cells. The transplanted RPE cells phagocytose shed outer segments of the host. No evidence of rejection was observed, suggesting that the subretinal space of the rabbit may possess some degree of immunologic privilege. This experiment provides reliable evidence for the clinical research of allotransplantation of RPE cells.

Objective.To detemine optimal conditions by using 5-bromodeoxyuridine (BrdU) as a marker of transplanted retinal pigment epithelial (RPE) cells in the subretinal space of albino rabbits.Methods.Pigmented rabbit RPE cells at second to fifth passage were fed with 20 μmol/L BrdU in Eagle's minimal essential medium (MEM) for five days. After extensive wash with phosphate buffered saline (PBS),the cells were detached by trypsin and used for transplantation onto Bruch's membrane of albino rabbits. Eyes were enucleated at various times post transplantation. Acetone, 4% paraformaldehyde, periodate-lysine-paraformaldehyde (PLP),or half-strength Karnovsky's fixatives were individually used to fix the tissue in order to find optimal condition for detecting BrdU marker.The fixation was followed by embedding in OCT compound, glycol plastic, or paraffin. The transplanted area was then sectioned, pepsin digested, and used for immunohistochemical staining with monoclonal antibody against BrdU and avidin biotin-alkaline phosphatase complex (ABC-AP).Results.Frozen sections of acetone or paraformaldehyde fixed tissue gave strong immunostaining of BrdU but the overall morphology was poor. Karnovsky's fixed tissue offered strong staining but this was buried by strong background. When using PLP as a fixative, we obtained strongly positive blue staining with very low background, and also excellent morphologic preservation.Conclusion.In combination with immunohistochemical detection method, BrdU labeling is an excellent long term marker for RPE transplantation one year after surgery. To use BrdU as a marker necessitates the use of pepsin digestion to make the BrdU antigen in the nuclei accessible to antibody. But the pepsin digestion may damage other tissues and result in overall poor morphology. Among the fixatives tested in this study, PLP fixed tissues offered both strong BrdU staining and good preservation of structural integrity, particularly the fine structure of photoreceptor and RPE cells.