We therefore generated stable transfectants from the 3T3 cell line which expressed human CD9, feline CD9 or feline CD4 (negative control). The anti-human CD9 antibody recognized human CD9 (Fig. 4i, 95.7%) and feline CD9 (Fig. 4f, 87.2%) but not feline CD4 (Fig. 4c, 0.3%) as predicted. However, although the vpgl 5 antibody reacted strongly with the feline CD9 transfectant (Fig. 4e, 86.3%) but not the feline CD4 transfectant (Fig. 4b, 0.2%), it also recognized the human CD9 transfectant (Fig. 4h, 49.7%) confirming our transient transfection studies. The anti-feline CD4 antibody, vpg33, reacted specifically with the feline CD4 transfectant (Fig. 4a, 44.9%) and did not react with either the feline CD9 transfectant (Fig. 4d, 1.9%) or the human CD9 transfectant (Fig. 4g, 1.7%). Therefore, the epitope recognized by the vpgl5 antibody appears to be present on human CD9 when expressed on murine cells but is either absent from or not recognized on human CD9 when expressed on human cells.