Fig 2: ADAM15 binds directly to the proline-rich linker of PABP.Mammalian-two hybrid was used for the identification of the PABP domain interacting with ADAM15. (A) PABP fragments covering the complete sequence were cloned into a bait vector, the cytoplasmic domain of ADAM15 into a prey vector and co-transfected together with a firefly luciferase reporter into HEK293 cells, and transcribed luciferase signals measured after 48 hours. (B) An exclusive binding of ADAM15 to the proline-rich domain of PABP (392–470), but not the RRM domains 1 and 2 (1–192) as compared to an included negative protein binding control (p53 and TRAF). A representative experiment is shown. (C) depicts the mean (± S.D.) of five repeated assays of ADAM15-binding with each distinct PABP fragment. Measured luciferase values were normalized to the values obtained from a co-transfected Renilla luciferase control plasmid.

Fig 3: Impact of ADAM15 and SIRT1 on histone acetylation, ROS and NAD+ in mechanically strained SF. (A–C) ADAM15 was downregulated by siRNA and a non-silencing siRNA (neg) as control. (A) Immunoblots from nuclear and cytoplasmic lysates of SF, showing decreased deacetylated histone after 6 and 9 h strain in ADAM15-expressing cells only. Tubulin and histone deacetylase (HDAC1) served as loading controls. (B,C) ROS and NAD+ assays in ADAM15-expressing SF (neg control) as compared to ADAM15-silenced cells. Each symbol represents the mean value of one individual donor, the horizontal bar (-) the median of 6 different donors. * p < 0.05, by Wilcoxon signed-rank test for comparison of ADAM15-expressing versus non-expressing SF. (D) ROS and NAD+ assays from mechanically strained SF with prior downregulation of SIRT1 by siRNA (I) and a non-silencing siRNA (N) from one representative donor. *** p < 0.0005, by Student’s t-test for SIRT1-expressing versus non-expressing SF. (E) ROS and NAD+ assays from strained SF in the presence of SIRT1 inhibitor selisistat (0 µM; solid line, 50 µM; double line and 100 µM; dashed line) from one representative donor. *** p < 0.0005, by Student’s t-test, comparing DMEM with the inhibitor. Representative results of at least three independent experiments are shown.

Fig 4: Strain-induced ATP release is dependent on activated pannexin-1 (PANX1). (A) Immunoblots from SF, strained for 0–9 h, with prior downregulation of ADAM15 by siRNA (I) or non-silencing siRNA (N), showing increased phosphorylations of PANX1 and Src in ADAM15-expressing SF. (B) Immunoblots of strained SF in the presence of dasatinib. (C) ATP release and (D) total ATP of strained SF in the presence of dasatinib (1 µM) and the PANX1 channel inhibitor carbenoxolone (carbenoxo, 100 µM). Data show the mean ±SD from one representative experiment out of at least 3 independent experiments. *** p < 0.0005, by Student’s t-test, for comparison of DMEM with inhibitor-treated SF. (E) ATP release and (F) total ATP from SF with downregulated ADAM15 (siA15), double knockdown of ADAM15/HOTAIR (siA15+Hot), single knockdown of HOTAIR (siHot), and negative siRNA (Neg). ** p < 0.005; *** p < 0.0005, using Student’s t-test.

Fig 5: ADAM15-dependent downregulation of HOTAIR under mechanical strain, resulting in the upregulation of SIRT1. (A–D) SF with prior downregulation of ADAM15 were strained for 0–9 h, and HOTAIR and SIRT mRNA and protein expression was quantified by qPCR and immunoblotting. (A) GAPDH-normalized Ct values for HOTAIR from one representative donor, showing higher Ct values, i.e., reduced HOTAIR levels, in ADAM15-expressing SF (dashed line), as compared to SF treated with ADAM15 siRNA (solid line). (B,C) Fold change (mean ± SD from 7 donors) of HOTAIR and SIRT1 mRNA in ADAM15-expressing versus non-expressing SF. *** p < 0.0005, by Student’s t-test, when comparing stimulated versus unstimulated HOTAIR/SIRT levels. (D) Immunoblots from SF silenced with ADAM15 siRNA (I) and negative control siRNA (N), showing increased SIRT1 expression in ADAM15-expressing cells after 6 h strain. (E,F) Fold change of SIRT1 after HOTAIR downregulation in SF (n = 6) when unexposed to mechanical strain by siRNA and negative control siRNA (N), showing increased SIRT1 mRNA (E) and protein levels (F). Tubulin served as a loading control.