Delayed-type hypersensitivity response in the skin is inhibited by anti-CD99 antibodies. (A) Blocking of CD99 partially inhibits immigration of T cells into inflamed skin. Radiolabeled in vivo–activated T cells were injected together with 70 μg control IgG from preimmune serum (co-IgG), affinity-purified antibodies against CD99 (anti-CD99), and affinity-purified antibodies against ESAM (anti-ESAM), 70 μg F(ab′)2 fragments (anti-CD99 F(ab′)2, or 70 μg Fab fragments (anti-CD99 Fab) of affinity-purified antibodies against CD99. Immigration of cells into the noninflamed control ears is depicted by the first bar (noninflamed). Results in the upper 2 panels are representative of 5 similar experiments; those in the bottom panel, of 3 similar experiments. For each determination, 4 mice were analyzed. Experiments shown by the 2 graphs were performed with 2 different preparations of T cells. Numbers on the left refer to the percentage of injected cells that were found in the analyzed ear. P < .001, anti-CD99 versus co-IgG (upper panel); P < .005, co-IgG versus anti-CD99 or F(ab′)2 (middle panel); P < .001, anti-CD99 versus co-IgG (bottom panel). Results are presented as mean ± SEM and are representative of at least 3 separate experiments. (B) CD99 and ESAM are accessible for antibodies from within the blood vessel lumen. DNFB-sensitized and -challenged mice were injected either with affinity-purified antibodies against CD99 or against ESAM, or with control IgG from rabbit preimmune serum (as indicated). Then, 15 minutes later anesthetized mice were perfused with PBS to remove unbound antibody and then with PFA to fix bound antibody. Cryostat sections of the inflamed ear were incubated with secondary antibodies and stained as described in “Materials and methods.” Bar represents 200 μm.