Fig 2: Effect of anti-PAK siRNAs on JURKAT cell adhesion.Cells were nucleofected with 2 µl esiRNA against PAK1 or PAK2, with the combination of both esiRNAs (PAK1+PAK2) or with non-targeting siRNA (ctrl) and the cell adhesion to fibronectin (ACF- adherent cell fraction) was tested 22 h post nucleofection. The level of PAK1, PAK2 and pPAK2 (Ser141) was assessed at the same time by western-bloting. (A) The experiment with the highest silencing efficacy obtained. The graph of cell adhesivity shows means and standard deviations of sample quadruplicates. (B) Summary of results for all adhesion experiments (N = 7). The adherent cell fraction was expressed as relative to the adhesivity of controls and the resulting values were compared using paired Student’s t-test.

Fig 3: PAK2 localizes to the leading edge of activated neutrophils.Replicate samples of total neutrophil (PMN) cell lysates (50 µg per lane) were analyzed for (A) the expression of PAK1 (68 kD), PAK2 (61 kD), PAK3 (65 kD) and PAK4 (72 kD) or (D) phosphorylation of PAK1/2 Thr432/402 at indicated time points after the addition of fMLP (10 nM) by Western blot. Human neutrophils adherent on fibronectin surfaces were treated in the presence or absence of fMLP (10 nM) for 3 min and stained for (B) PAK1, PAK2 or PAK4 (green) and F-actin (red), (E) phospho-PAK2 Ser20 (green) and Rac1-GTP (red), (F) phospho-PAK2 Ser20 (green) and Cdc42-GTP (red), or (G) phospho-PAK1/2 Thr423/402 (green) and F-actin (red). In selected experiments, neutrophils were pretreated with vehicle (0.1% DMSO), EHT1864 (Rac1/2 inhibitor, 50 µM) or wortmannin (PI3K inhibitor, 100 nM). Results are quantified from at least 60 cells and presented as the mean percentage±SEM of (C) neutrophils displaying PAK immunofluorescence at actin-rich leading edge, or (H) neutrophils displaying phospho-PAK1/2 immunofluorescence in the fields of view. Representative images obtained from 3 independent experiments are shown. * P<0.05, compared to the basal level. Scale ba = 10 µm.

Fig 4: A retained ER function is a requisite for anti-inflammatory action of vildagliptin under diabetes(A) Representative immunoblot images and quantification of PAK2 activation and CHOP expression from isolated ARCM in the fact of vildagliptin (Vil) (20 µM) with diabetes-mimicking stress (N = 4).(B) Strategy for Vildagliptin administration (5 mg/kg/day in drinking water) with or without application of AAV9-Pak2 in Pak2cKO mice fed with HFHSD.(C and D) (C) BW (D) intraperitoneal glucose tolerance test.(E and F) (E) IVRT from pulse-wave Doppler analysis and (F) calculated FS from M-mode echocardiography (N = 6 in C–F).(G) Representative immunofluorescence staining (scale bar: 25 µm) and quantification of left: Mac3- (red) and right: CD68- (green)-positive non-cardiomyocytes in heart sections. Cardiomyocytes are counterstained with a-actinin (green) and troponin (red). Nuclei are stained with DAPI (blue) (N = 4/group).(H) Relative mRNA expression (normalized to chow, AU) of pro- and anti-inflammatory markers in Pak2cKO hearts (N = 5–6/group).(I) Immunoprecipitation of acetylated HMGB1 from protein lysate. Immunoblot images and quantification of CHOP and HMGB1 (N = 4/group). Gß is loading control.Inset: key for data points; each data point represents one animal/experiment. All data are presented as mean ± SEM. p values (shown in each panel) versus corresponding controls, determined by ANOVA with Tukey's post hoc tests in (A and C–I).

Fig 5: ER dysfunction facilitates cardiac HMGB1 expression and subsequent release(A) Correlation graph of relative Hmgb1 expression to PAK2 (N = 16 pairs) expression in C576L/BJ hearts. (B-C) From Pak2fl/fl and Pak2cKO mice, (B) immunoprecipitation from hearts (N = 5–8/group, AU) for HMGB1 acetylation.(C) serum analyzed by ELISA (N = 4–6/group) and immunoblot (N = 4/group).(D) Representative immunohistochemical images (scale bar: 45 µm) (black square scale bar: 25 µm) of CHOP-stained nuclei (black arrowheads), with quantification graph depicting nuclear H-score from 10 images from PAK2 mice hearts.(E) Immunofluorescence staining (scale bar: 50 µm) showing cellular localization of CHOP (red) and HMGB1 (green) following HFHG in H9C2. Lysosomes and nuclei are stained with LysoTracker (red) and DAPI (blue), respectively, and yellow puncta denote HMGB1 and lysosome colocalization. Relative CHOP nuclear intensity (AU) from 100 nuclei/data point (N = 4).(F) Schematic figure showing CHOP-binding sites in the promoter region (-8000 bp from transcription start site, TSS) in rat, human, and mouse Hmgb1. ChIP performed on H9C2s using anti-CHOP antibody (normalized to the input chromatin, N = 4). Primer pair 1 product sequence is highlighted gray; forward (F) and reverse (R) primer sequences flanking the corresponding region are annotated as Red = Primer one; Orange = Primer 2 and Green = Primer 3.(G) Strategy for macrophages stimulation using the conditional medium culturing H9C2s.(H–L) Immunoblot from H9C2s and conditional medium from (H and I) CHOP overexpression (N = 5) and (K and L) HFHG stimulation following PAK2 knockdown using siPAK2 (N = 3-5 experiments). Conditioned medium was analyzed for release of acetyl-HMGB1. Gß and Coomassie stain are loading controls.(M and N) Representative flow contour plots and quantification of %CD206+ and %CD86+ macrophages with boxed mean M1/M2 ratio for BMM (chow-fed mice) treated with culture medium of H9C2 with (J) CHOP overexpression, (M) siScramble/PAK2 with HFHG (8 h) stimulation, and (N) Ad-Control/Ad-PAK2 with HFHG (12 h) stimulation. All M1/M2 ratios are calculated from mean %CD206+ and %CD86+ cells gated from parent (Ly6G-F4/80+) macrophages. Each data point represents one animal/experiment. All IP experiments (N = 1/group/cohort).All data are presented as mean ± SEM. p values (shown in each panel) versus corresponding controls determined by two-tailed Student's t test in (C, D, F, H, and I) and by ANOVA with Tukey's/Sidak's post hoc tests in (B, C, E, and J–N). Pearson R coefficient analysis in (A).