Fig 2: Actin protrusive network is upregulated by AMPK-Rac1-PAK1-Cofilin signaling cascade in cells lacking CAV-1. (A) Immunofluorescence microscopy analysis demonstrating that more pronounced endogenous ARPC2 colocalizes with F-actin (visualized by Alexa 568 phalloidin) on the lamellipodia protrusions in CAV-1 deficient cells. Magnified regions of cell edges on the right show the distribution of ARPC2 in WT and CAV-1 KO cells. Bars, 10 µm (in cell images) and 5 µm (in the magnified box). Quantification of width of lamellipodial protrusions are shown on the right. n = 16 regions from 16 cells for each group. (B) Representative images of membrane ruffling by time-lapse microscope. 1-pixel-wide areas were cut out to generate a 300 frame 2 s interval kymograph. Yellow dashed lines indicate the track of cell movement. An enlarged region is displayed on the right, vertical dashed lines show the membrane protrusion distance, while horizontal dashed lines mark the duration of protrusion. Bars, 10 µm (in cell images) and 2 µm (in the magnified box). Quantification of protrusion rate are shown on the right. n = 16 regions from 16 cells for each group. (C) P-AMPK (Thr172) and total AMPK were detected from the lysates of each group by western blotting. Please note that CAV-1(Y14F)-mEGFP can’t be detected by using phospho-CAV-1(Tyr14) antibody. Asterisk denotes the non-specific band. Quantification of P-AMPK (Thr172) levels (compared to total AMPK) from each group was shown on the right panel. n = 3. (D) Pull-down assays were performed for WT, CAV-1 KO, and CAV-1 KO; CAV-1-mEGFP re-expressed cells. Proteins bound to GST-PAK binding domain were analyzed by western blots and further quantified (compared to total Rac1) based on the band’s intensity. n = 3. (E) Western blot analysis and quantification (compared to total PAK1 and Cofilin) of the levels of phosphorylated PAK1 (Thr423) and Cofilin (Ser3) in WT and CAV-1 KO cell lysates. n = 3. (F) Immunostaining and quantification of endogenous P-Cofilin (Ser3) and F-actin distribution in WT and CAV-1 KO cells. A 16 µm length line was used to generate a line profile to illustrate the co-localization of P-Cofilin (Ser3) and F-actin. The lamellipodia region was enlarged on the right, and 2 µm width region was chosen to analyze the mean intensity of P-Cofilin (Ser3) on the leading edge. Bars, 10 µm (in cell images) and 5 µm (in the magnified box). n = 18 regions from 18 cells for each group. (G) Immunostaining and quantification of endogenous P-cofilin (Ser3) distribution upon compound C treatment in WT and CAV-1 KO cells. Magnified regions represent the lamellipodia region. Bars, 10 µm (in cell images) and 5 µm (in the magnified box). n = 18 regions from 18 cells for each group. (H) Western blot analysis and quantification of the phosphorylated levels of AMPK (Thr172), PAK1 (Thr423), Cofilin (Ser3) and activity of Rac1 and RhoA upon compound C treatment. n = 3. In panel (C–E,H), the obtained intensity value from wild-type cells was set to 1. All the data are presented as mean ± SD. In (C,G,H), ***P < 0.001; *P < 0.05; N.A., not significant (one-way ANOVA). In (D–F), ***P < 0.001; **P < 0.01 (unpaired t-test). All the data are from three independent experiments.

Fig 3: Increases in extracellular matrix (ECM) stiffness promoted RhoA and YAP/TAZ activity, which led to decreased expression of regenerative markers. (A–E) Western blot shows that the expression of key factors can be regulated by ECM stiffness in Schwann cells (SCs). Four unique cell lysates were created for Western blots. (F,G) Nuclear YAP/TAZ (green) is upregulated as ECM stiffness increases. A total of 98 cells were quantified per well, with 6 wells quantified in each of 2 unique trials. Scale bar = 50 µm. (H–J) Expression of c-Jun and p75NTR are enhanced by RhoA and YAP/TAZ inhibition, with RhoA inhibition resulting in higher expression compared to YAP/TAZ inhibition. Three lysates for c-Jun and two lysates for p75NTR were created for Western blots. (K) Heatmap generated by microarray analysis shows up- and downregulated genes that are associated with either mechanotransduction or SC regenerative phenotypes. Data are presented as the mean ± SEM. * p < 0.05, ** p < 0.005, *** p < 0.0005, p-values found in Supplemental Table S1.

Fig 4: Cellular localization of RhoA in Cav1-competent and Cav1-deficient human TM cells. ((A), Top) Under basal conditions, Cav1 protein localized almost exclusively to membrane fraction in TM cells (250-fold, ****p < 0.0001, paired t-test, n = 6 experiments using five different TM cell strains). ((A), Bottom) Representative blots from a hTM cell strain depicting Cav1 signal in cytosolic verses membrane fractions, normalized to total protein. ((B), Top) For the membrane fraction only, Cav1 protein levels were significantly decreased by approximately 53% in cells treated with CAV1-silencing shRNA compared to cells treated with scrambled shRNA (**p = 0.001, paired t-test, n = 6 experiments using five different human TM cell strains). ((B), Bottom) Representative blots from an hTM cell strain depicting membrane-bound Cav1 protein abundance, normalized to total protein. ((C), Top) Under basal conditions, a-tubulin protein localized preferentially to the cytosolic fraction in TM cells (5.5-fold, *p = 0.012, paired t-test, n = 6 experiments using five different TM cell strains). ((C), Bottom) Representative blots from a hTM cell strain depicting a-tubulin signal in cytosolic verses membrane fractions, normalized to total protein. ((D), Top) Under basal conditions, RhoA protein localized preferentially to the membrane fraction (5-fold) as compared to cytosolic fraction (****p < 0.0001, paired t-test, n= 6 experiments using five different TM cell strains). ((D), Bottom) Representative blots from a hTM cell strain depicting RhoA signal in cytosolic verses membrane fractions, normalized to total protein. ((E), Top) For the membrane fraction only, RhoA protein levels were not significantly different in cells treated with CAV1-silencing shRNA compared to cells treated with scrambled shRNA (p = 0.13, paired t-test, n = 6 experiments using five different human TM cell strains). ((E), Bottom) Representative blots from a hTM cell strain depicting membrane-bound RhoA protein abundance, normalized to total protein. ((F), Top) For the cytosolic fraction only, RhoA protein levels were significantly decreased by approximately 38% in cells treated with CAV1-silencing shRNA compared to cells treated with scrambled shRNA (***p = 0.0004, paired t-test, n = 6 experiments using five different human TM cell strains). ((F), Bottom) Representative blots from an hTM cell strain depicting cytosolic RhoA protein abundance, normalized to total protein.

Fig 5: Schematic of hypothesized regulation of Rho/ROCK activity by Cav1 in TM cells. (A) Cartoon showing components of the irideocorneal angle, highlighting aqueous humor flow (blue arrows) into and through the conventional outflow pathway. A higher magnification illustration of the conventional outflow pathway, depicting the TM, juxtacanalicular tissue (JCT) and SC on the right. (B) Diagram demonstrating results of the present study, showing the status of Rho/ROCK activity and MLC phosphorylation in Cav1 competent, Cav1 deficient, and cavtratin-treated TM cells. For example, we propose that in the presence of Cav1, RhoA localizes to the plasma membrane, where it is sequestered and inhibited via physical interaction with the Cav1 scaffolding domain, leading to a low basal level of Rho/ROCK activity and MLC phosphorylation. In the absence of Cav1, we propose that RhoA is liberated and basal Rho/ROCK activity and MLC phosphorylation is elevated. Following treatment with the Cav1 scaffolding domain peptide, cavtratin, we hypothesize that RhoA is inhibited, leading to reduced ROCK activity and subsequent MLC dephosphorylation.