PAK proteins and YAP-1 signalling downstream of integrin beta-1 in myofibroblasts promote liver fibrosis

Abstract

Fibrosis due to extracellular matrix (ECM) secretion from myofibroblasts complicates many chronic liver diseases causing scarring and organ failure. Integrin-dependent interaction with scar ECM promotes pro-fibrotic features. However, the pathological intracellular mechanism in liver myofibroblasts is not completely understood, and further insight could enable therapeutic efforts to reverse fibrosis. Here, we show that integrin beta-1, capable of binding integrin alpha-11, regulates the pro-fibrotic phenotype of myofibroblasts. Integrin beta-1 expression is upregulated in pro-fibrotic myofibroblasts in vivo and is required in vitro for production of fibrotic ECM components, myofibroblast proliferation, migration and contraction. Serine/threonine-protein kinase proteins, also known as P21-activated kinase (PAK), and the mechanosensitive factor, Yes-associated protein 1 (YAP-1) are core mediators of pro-fibrotic integrin beta-1 signalling, with YAP-1 capable of perpetuating integrin beta-1 expression. Pharmacological inhibition of either pathway in vivo attenuates liver fibrosis. PAK protein inhibition, in particular, markedly inactivates the pro-fibrotic myofibroblast phenotype, limits scarring from different hepatic insults and represents a new tractable therapeutic target for treating liver fibrosis.

Figure 1. Integrin beta-1 is increased in activated HSCs and required for phenotype and function.

(a) integrin beta-1 is increased in activated (A) compared with quiescent (Q) rat HSCs by qRT–PCR. (b) Immunofluorescence of rat AHSCs shows integrin beta-1 costained in cells with pro-fibrotic markers α-SMA (green; DAPI as blue nuclear counterstain) and SOX9 (red). (c,d) Activated mouse HSCs (‘Control’) show decreased protein levels for α-SMA, COL1 and SOX9 by immunoblotting following the loss of integrin beta-1 (‘Itgb1-null’). Quantification from n≥3 experiments in c and example immunoblots shown in d. (e) α-SMA is lost by immunofluorescence from activated mouse HSCs (‘Control’) following integrin beta-1 inactivation (‘Itgb1-null’). DAPI (blue) used as nuclear counterstain. (f,g) Proliferation measured by BrdU incorporation of activated mouse HSCs (‘Control’) declines following integrin beta-1 inactivation (‘Itgb1-null’). Quantification from n≥3 experiments in f with example immunofluorescence in g. DAPI, blue nuclear counterstain. (h,j) Migration of activated mouse HSCs (‘Control’) over 24 h is almost entirely attenuated following the loss of integrin beta-1 (‘Itgb1-null’). Quantification from n=3 biological replicates with 30–77 cells in each experiment in h with an individual example of migratory tracks (in μm) shown in i,j. (k,l) Contractile properties of activated mouse HSCs (‘Control’) are markedly attenuated after inactivation of integrin beta-1 (‘Itgb1-null’). Quantification of gel contraction from n=3 experiments is shown in k in the presence or absence of mitomycin-C (Mito-C) and example images shown in l. Scale bars, 50 μm. Two-tailed unpaired t-test was used for statistical analysis. Data are shown as means±s.e.m. *P<0.05, **P<0.01, ^‡P<0.001. C, control-activated HSCs; N, integrin beta-1-null HSCs.

Figure 2. Integrin beta-1 is required for multiple pro-fibrotic features in activated HSCs.

(a) Heatmap and cluster analysis showing gene expression changes (1.5-fold; P<0.05) in biological duplicates of activated (A) control (‘Cnt’) HSCs and following integrin beta-1 inactivation (‘Null’). Quiescent (Q) HSCs are also included for comparison. Seven clusters were identified based on upregulated (red), downregulated (blue) and intermediate (yellow) gene expression. (b,c) Functional annotation by gene ontology for enrichment in Cluster 3. Proportions are shown in b. Individual categories and the genes underlying them are shown in c.

Figure 3. YAP-1 and MYL9 mediate pro-fibrotic aspects of liver fibrosis in activated HSCs.

(a,b) Quantified increase in total MYL9 and YAP-1 protein levels on activation of rat HSCs from n≥3 experiments (a) with representative immunoblot shown in b. (c–e) Total YAP-1 and MYL9 are diminished in activated mouse HSCs (‘Control’) following integrin beta-1 loss (‘Itgb1-null’). Quantification from n=3 experiments shown in c with representative immunoblot in d. In the immunofluorescence in e, note the rounded inactivated appearance of the Itgb1-null cells. The remaining total YAP-1 signal is more cytoplasmic (see h,i). Scale bar, 50 μm. (f,g) Itga11 knockdown in activated rat HSCs using two different siRNA oligos. Data for each oligo are shown relative to its own scrambled control in either black or grey (n=3 for each). (f) Detection of Myl9 transcripts was diminished to an almost identical extent as for Itga11. (g) Protein detection of MYL9 and total YAP-1 was also diminished following Itga11 knockdown. (h,i) The proportion of phosphorylated YAP (PYAP, inactive form), is increased following integrin beta-1 loss (‘Itgb1-null’) from activated mouse HSCs (‘Control’; n=3 experiments) and localises more predominantly to the cytoplasm (i). DAPI, blue nuclear counterstain, is shown. Scale bar, 50 μm. (j) Luciferase activity (in relative light units; RLU) following co-transfection of constructs containing the wild-type (MYL9-TEAD) or mutated (MYL9-ΔTEAD) TEAD motif from the 3′-untranslated region of the MYL9 gene with empty vector (Control) or YAP expression vector. Results are normalized to a Renilla vector and expressed relative to the control MYL9 luciferase construct without YAP. (k) Transcript levels by qRT–PCR following inhibition of YAP-TEAD interaction using VP in activated rat HSCs expressed relative to DMSO control. Two-tailed unpaired t-test was used for statistical analysis. Data are shown as means±s.e.m. *P<0.05, **P<0.01, ^†P<0.005, ^‡P<0.001.

Figure 4. Pharmacological inhibition of YAP-1 with VP improves liver fibrosis in vivo.

(a) PSR staining (collagen deposition in red) counterstained with fast green in olive oil control (Oil; top) or chronic CCl₄-induced fibrosis (bottom) in mice treated with DMSO (n=4) or VP (VP; n=3). (b) Quantification of surface area covered by the PSR staining in a. (c) PSR staining (collagen deposition in red) in sham-operated mice (Sham; top) or BDL to induce peribiliary fibrosis (bottom) with control DMSO (n=7) or VP (n=5) treatment. (d) Quantification of surface area covered by the PSR staining in c. Scale bar, 500 μm. Liver weight was unaffected in both models treated with VP. Two-tailed unpaired t-test was used for statistical analysis. Data are shown as means±s.e.m. *P<0.05, ^†P<0.005.

Figure 5. Integrin beta-1 inactivation identifies PAK signalling as a requirement for liver fibrosis.

(a–f) Quantification of immunoblots showing: (a) increased protein levels of PAK-1 and PAK-3 on activation of rat HSCs compared with quiescent (Q) cells; (b) decreases in both PAK-1 and PAK-3 following the loss of integrin beta-1 (‘Itgb1-null’) in activated mouse HSCs (AHSCs). Levels in mouse QHSCs are shown for comparison. (c) Expression of PAK-1 and PAK-3 proteins are similarly diminished following Itga11 knockdown using two independent siRNAs in activated mouse HSCs relative to control (scrambled siRNA). (d,e) Decreases in the levels of activated HSC markers, SOX9, COL1 and phosphoMyl9 (PMYL9) following PAK-1 abrogation by siRNA1 in activated rat HSCs relative to their respective scrambled control levels (‘Control’). Similar decreases were observed with a second independent siRNA (siRNA2). Representative immunoblot is shown in e. (f) Decreases in COL1 and SOX9 following the inhibition of group I PAKs using IPA3 treatment in rat and human-activated HSCs compared with DMSO control. (g) IPA3 treatment of activated rat HSCs disrupts the actin cytoskeleton (phalloidin staining in green). Note, the rounded cell appearance following IPA3 treatment similar to Fig. 3e and Supplementary Fig. 1f. Scale bar, 50 μm. (h) Relative expression levels by qRT–PCR following VP or IPA3 treatment of activate rat HSCs for three genes identified as markers of HSC inactivation. As expected, Col1a1 levels were decreased in response to both treatments. For experiments in a–h, n=3 or 4. Two-tailed unpaired t-test was used for statistical analysis. Data are shown as means±s.e.m. *P<0.05, **P≤0.01, ^†P<0.005, ^‡P≤0.001.

Figure 6. Pharmacological inhibition of PAK-1 improves liver fibrosis in rodents.

(a) PSR staining (collagen deposition in red) in olive oil control (Oil; top) or chronic CCl₄-induced fibrosis (bottom) with DMSO (n=5) or IPA3 (n=4) treatment. (b) Quantification of surface area covered by the PSR staining in a. (c) PSR staining (collagen deposition in red) in sham-operated mice (Sham; top) or BDL to induce peribiliary fibrosis (bottom) with control DMSO (n=6) or IPA3 (n=7) treatment. (d) Quantification of surface area covered by the PSR staining in c. (e) Immunohistochemistry for α-SMA (brown; activated HSC/myofibroblast marker) in mouse livers following CCl₄ (top) or BDL (bottom) induced fibrosis. (f) Quantification of surface area covered by α-SMA staining was reduced following IPA3 treatment in both models (n=5 for all animal groups). Scale bar, 500 μm. Two-tailed unpaired t-test was used for statistical analysis. Data are shown as means±s.e.m. *P<0.05, **P≤0.01.

Figure 7. Analysis of in vivo activated HSCs.

(a) Expression analysis by qRT–PCR of in vivo activated HSCs extracted from wild-type mice following CCl₄ injections compared with olive oil control. (b,c) FACS analysis for integrin beta-1 on in vivo activated HSCs extracted from Pdgfrb-BAC-eGFP mice following CCl₄-induced fibrosis compared with olive oil control. Fluorescence intensity is shown using a phycoerythrin (PE)-conjugated antibody to integrin beta-1 (integrin beta-1-PE) and isotype control (Iso-PE) (b). Graphical representation of integrin beta-1 fluorescence demonstrating increased integrin beta-1 on unpermeabilized live HSCs the context of liver fibrosis (c). All experiments are n=4. Two-tailed unpaired t-test was used for statistical analysis. Data in bar charts show means±s.e.m. *P<0.05, **P<0.01.