Fig 2: Immunofluorescence of serial sections through the back skin of adult mice. Sections were stained with Abs against NFAT5 (ab110995, Abcam) and Mmp3 (ab53015 (Abcam) and counterstained with DAPI as indicated. Length of the bars: 100 μm.

Fig 3: NFAT5 affects the expression of Mmp3, Klk5 and Asprv1 protease genes in KCs. (A) Left panels, results of qRT-PCR assays showing the effect of NFAT5 on the expression of Mmp3, Klk7 and Asprv1 genes upon transduction of KCs obtained from the tails of adult ice (T) or from the skin of newborn (NB) mice. Right panels, effect of NFAT5 ablation on Mmp3 and Klk7 expression. Results of 3 independent NGS assays are shown. (B) Expression of protease genes in KCs cultured for 1 or 3 weeks (ws). NGS data of three independent batches of KCs cultured for 1 or 3 ws are shown. (C–E) Detection of NFAT5 expression upon transduction of KCs. Cultured KCs were transduced with retroviruses expressing NFAT5, or with an ‘empty’ EGZ virus. (C) Detection of human NFAT5 RNA by RT-PCR assays in tail (T) and NB KCs upon transduction. (D) Western blots showing the expression of NFAT5 upon transduction of KCs from newborn and tails of adult mice. (E) Fluorescence microscopy of NFAT5-GFP expression in transduced adult KCs. (*p<0.05).

Fig 4: The expression of MMP-3 associates with the expression of IL-1α and IL-1β in human atheroma.(A) Representative immunohistochemical stainings of IL-1 isoforms, MMP-3, SMC α-actin (α-SMA), and CD68 in fibrous cap (top panels, original magnification 100x); tunica media and tunica intima (middle panels, 100x); and macrophage-rich shoulder area with lipid core (bottom panels). On the first column, black rectangles designate areas magnified on the following panels. Controls for antibody specificity using rabbit, mouse, or goat IgG were performed on serial sections (right panels) (B) Linear regressions demonstrate the relationship between MMP-3+ area and IL-1α (upper left), IL-1β (upper right), αSMA (lower left), and CD68+ (lower right) areas in the tunica intima (n = 17 for MMP-3, IL-1α and IL1-β; n = 16 for CD68). (C) A principal component regression demonstrates that IL-1α, IL-1β, and CD68+ areas associate similarly and significantly with the MMP-3+ area. Left panel: bar graphs represent β-coefficient and 95% confidence intervals. Right panel: regression of predicted vs. observed MMP-3+ area demonstrates fitness of the model and global R2 = 0.77. Dashed lines represent 95% confidence intervals.

Fig 5: IL-1α and IL-1β elicit similar levels of protease expression in ex-vivo tissue cultures of human atheromata.Cultures were stimulated by 1040 pM IL-1α or IL-1β for 48 hours and the levels of the indicated proteases were quantified in conditioned media. The results are normalized to protein concentration in tissue lysates. (A) MMP-3, determined by ELISA. The groups (n = 19) were compared with Friedman’s test and Dunn’s post hoc test; **P<0.01, n.s., not significant. (B) Immunoblotting of MMP-1 containing a 53kDa band corresponding to the pro-enzyme and a 45kDa band corresponding to the cleaved form. (C) Normalized densitometric units of MMP-1 immunoblots (n = 3, groups did not differ significantly). (D) Gelatin zymography demonstrating gelatinolytic activity corresponding to pro-MMP9 at 92kDa and cleaved MMP-2 at 63kDa. (E) Normalized densitometric analyses of zymograms (n = 7, *P<0.05).