Enhanced JAM-A shedding in response to PMA and proinflammatory mediators. (A) HUVECs were stimulated with PMA (200 ng/mL) or vehicle control (DMSO) for 2 hours, and subsequently conditioned media and cell lysates were subjected to Western blot analysis for sJAM-A and flJAM-A, respectively. A standard of serially diluted recombinant sJAM-A was run in parallel. (B) HUVECs were stimulated with PAF (100 nM) or TNF-α/IFN-γ (10 ng/mL) or treated with vehicle control (DMSO) for 16 hours, and conditioned media were analyzed by Western blotting using an anti–hJAM-A monoclonal antibody. Data are shown as representative Western blots of least 3 independent experiments. (C) HUVECs or HEK293 cells were left unstimulated or stimulated as described in panels A and B, and subsequently JAM-A surface expression was assessed by flow cytometry. Data were shown as representative histograms (top panel) and as mean plus or minus SD of the fluorescence signals obtained in 3 independent experiments (bottom panel). (D) Effect of IFN-γ and TNF-α on JAM-A mRNA expression. HUVECs were stimulated with IFN-γ and TNF-α or vehicle control (PBS) for 16 hours and analyzed for JAM-A mRNA expression by quantitative real-time RT-PCR. Expression of JAM-A mRNA was measured in 3 independent experiments and expressed as percentage of GAPDH mRNA expression. (E) Serum levels of murine JAM-A. Wt mice and jam-a^−/− mice were intravenously treated with IFN-γ and TNF-α (40 and 50 ng/g, respectively) or vehicle control (0.9% NaCl) or left untreated (n = 3 per group), and after 2.5 hours blood serum was investigated for released JAM-A by ELISA. *Significant increase in the JAM-A level induced by the cytokine treatment compared with the vehicle-treated control (P < .05).