Fig. 5

Opioid actions in normal and CD8⁺-depleted arthritic mice. a The opioid antagonist NLXM was injected ipsilaterally in the early (left panel, d16) or late chronic phase (right panel, >d60) of arthritis. Paw withdrawal latencies are represented as means ± SEM. Paw withdrawal latency in mice that were not depleted of CD8⁺ cells was significantly reduced ipsilaterally at 15 min after NLXM injection as compared to baseline thresholds (one-way RM ANOVA and Bonferroni’s multiple comparison test followed by Bonferroni correction,^§§§ P < 0.001), N = 8 mice. No effect was observed in the ipsilateral paw of CD8⁺-depleted animals (early phase: one-way RM ANOVA and Bonferroni’s multiple comparison test followed by Bonferroni correction, P > 0.05; chronic phase: Friedman and Dunn’s test, P > 0.05) or in the contralateral paws of all groups (one-way RM ANOVA and Bonferroni’s multiple comparison test, P > 0.05). b, c The joint cells of ipsilateral arthritic knee joints was analyzed for opioid peptide release. Met-enkephalin immunoreactivities (ir-MENK, b left panel) and beta-endorphin immunoreactivities (ir-END, c) were assessed in the supernatant of cells, N = 9–10 and in the cell pellet (ir-MENK, b right panel). Statistics were performed using the Mann-Whitney U test to compare the effect of anti-CD8 vs. isotype matched antibody, *P > 0.05