Fig 2: Immunocytochemical JAM-A tight junction staining in primary HBMEC is reduced under barrier-disturbing pro-inflammatory conditions.Cells were either left unstimulated or treated with TNF-a 10 ng/mL and IFN-? 100 U/mL for 24 h. A Immunocytochemical staining was performed with primary antibodies against JAM-A (M.Ab.F11), occludin and ZO-1. F-Actin was stained with Alexa Fluor 488 phalloidin. Representative for 5 independent experiments with different HBMEC preparations. B Analysis of paracellular barrier function by dextran 3000 transwell permeability assays. Mean and SD of 4 independent experiments in octuplicates with different HBMEC preparations. The two-tailed Mann-Whitney U test was used for statistical analysis. Bar = 25 µm.

Fig 3: HBMEC constitutively release sJAM-A, which is not altered under inflammatory or hypoxia/reoxygenation conditions. A Representative Western blot of a HBMEC whole cell protein extract, employing the goat anti-JAM-A antibody which was used as capture antibody for a self-developed sandwich ELISA (compare Figure 3). B Typical standard curve of the self-developed ELISA for JAM-A detection in culture supernatants and serum. C and D Concentrations of sJAM-A in culture supernatants of primary HBMEC (C) or hMCEC/D3 (D) which were either left unstimulated for 48 h, stimulated with TNF-α 10 ng/mL and IFN-γ 100 U/mL for 48 h or subjected to 1% O2 for 24 h followed by a reoxygenation period of 24 h. Please note: cell culture supernatants were concentrated by filter centrifugation before ELISA measurement, resulting in ∼10-fold increased JAM-A concentration. Mean and SD of 3 independent experiments. The Kruskal-Wallis test was used for statistical analysis. n.s., not significant. E Comparative Western blot analysis of a whole cell protein extract and a TCA-precipitated protein extract from cell culture supernatant, using a rb polyclonal antibody (Zymed) for JAM-A detection. Representative of 3 independent experiments with different HBMEC preparations. The asterisk represents full-length JAM-A, open circles represent soluble JAM-A. F Comparative Western blot analysis of JAM-A in TCA-precipitated cell culture supernatants, using the same primary antibody as in E, after stimulation as in C and D. Representative of 3 independent experiments with different HBMEC preparations. Abbreviations: WCE, whole cell protein extract; SN, supernatant.

Fig 4: Soluble JAM-A serum levels do not indicate BBB disturbance in multiple sclerosis and ischemic stroke. A Levels of sJAM-A in serum samples from 45 patients with stable relapsing-remitting MS not receiving secondary prophylactic therapy were measured in duplicates by ELISA and compared to those in 14 untreated MS patients during an acute relapse. B Levels of sJAM-A were measured in serum samples from 13 patients with non-small vessel complete ischemic stroke where a first sample was obtained within the first 3 h of symptom onset and a second sample 24 h after clinical onset. Serum samples were not concentrated before ELISA measurement. Dots represent single values, lines median values. Serum levels in stable and inflammatory-active MS patients were statistically compared by the Mann-Whitney U test, serum levels in stroke patients over time by the Wilcoxon matched pairs test. n.s., not significant.

Fig 5: Pro-inflammatory stimulation of HBMEC induces JAM-A dissociation from the actin cytoskeleton. A Primary HBMEC were either left unstimulated or treated with TNF-α 10 ng/mL or IFN-γ 100 IU/mL alone or in combination. Cell protein extracts were generated with a Nonidet-P40 based cell lysis buffer and subjected to Western blot analysis. JAM-A was stained with M.Ab.F11. Staining of the same, peroxidase-inactivated membranes with a rabbit antibody against β-actin served as a loading control. B N-deglycosylation of JAM-A with increasing concentrations of PNGase F for 2 h. A rabbit polyclonal antibody against JAM-A (Zymed) was used for the detection of JAM-A. The asterisk represents N-glycosylated, the open circle N-declycosylated JAM-A. Representative experiments out of at least 5 independent experiments with different EC preparations for each subpanel of the figure are shown.