Mechanosensing by the α6-integrin confers an invasive fibroblast phenotype and mediates lung fibrosis

Matrix stiffening is a prominent feature of pulmonary fibrosis. In this study, we demonstrate that matrix stiffness regulates the ability of fibrotic lung myofibroblasts to invade the basement membrane (BM). We identify α6-integrin as a mechanosensing integrin subunit that mediates matrix stiffness-regulated myofibroblast invasion. Increasing α6-expression, specifically the B isoform (α6B), couples β1-integrin to mediate MMP-2-dependent pericellular proteolysis of BM collagen IV, leading to myofibroblast invasion. Human idiopathic pulmonary fibrosis lung myofibroblasts express high levels of α6-integrin in vitro and in vivo. Genetic ablation of α6 in collagen-expressing mesenchymal cells or pharmacological blockade of matrix stiffness-regulated α6-expression protects mice against bleomycin injury-induced experimental lung fibrosis. These findings suggest that α6-integrin is a matrix stiffness-regulated mechanosensitive molecule which confers an invasive fibroblast phenotype and mediates experimental lung fibrosis. Targeting this mechanosensing α6(β1)-integrin offers a novel anti-fibrotic strategy against lung fibrosis.

TREs. The TRE binding activities of c-fos, c-jun, fosB, fra1, junB, junD and fra2 were quantified by colorimetric ELISA. G, The binding of c-fos/c-jun complex to mouse  6 promoter under soft vs. stiff matrix conditions was measured by quantitative ChIP. H, Effects of c-fos/cjun inhibitors (T-5224 and decoy ODNs) on matrix stiffness-regulated mouse  6 expression were evaluated by immunoblot. Results are the means + SD of at least three separate experiments; *p < 0.05; **p < 0.01.

Supplementary Figure 3: Evaluation of matrix stiffness-regulated lung myofibroblast
invasion into the BM by PDMS hydrogel system, a self-developed "Sandwich" system and rat mesentery. A, IPF lung myofibroblasts were adapted to soft (2 kPa) and stiff (30 kPa) PDMS gels. Cells were detached by trypsinization. The ability of IPF myofibroblasts cultured on soft vs. stiff PDMS gels to invade the BM was evaluated by Matrigel invasion assay. B, Schematic shows a "sandwich" invasion assay system. C, Invasion indexes in the "sandwich" invasion assay were calculated as the ratio of the percent invasion of test cells (lung MFBs cultured on soft or stiff gels) over the percent invasion of control cells (lung MFBs cultured on the regular tissue culture plates). D, Rat mesentery (white arrow) was isolated from euthanized rats, decellularized by 0.5 N ammonium hydroxide and mounted on 24-well invasion inserts using Super Glue. E, An equal number of living IPF lung myofibroblasts pre-cultured on soft or stiff PA gels were plated on rat mesentery for 24 hours. Invading cells in the lower chambers Supplementary Figure 3 were detached by trypsinization and the number of cells was counted. Invasion indexes were calculated as described in C. Results are the means + SD of at least three independent arrays; *p < 0.05. **p < 0.01.

Quantitative real-time PCR
Total RNA was isolated using TRIzol reagent (Invitrogen). 1 μg total RNA was reversely transcribed into cDNA with a cDNA Synthesis Kit (Thermo Scientific). Quantitative PCR reactions were carried out in a Bio-Rad iCycler. Relative quantification was calculated using the comparative C T method 3 .

Quantitative chromatin immunoprecipitation (ChIP) assay
Lung fibroblasts were treated with 1% formaldehyde for 10 minutes at 37°C to cross-link histones to DNA. The cross-linked chromatin was sonicated to shear into chromatin fragments of 200-1,000 base pairs. A portion of sheared chromatin was reversed at 65°C for 4hr and crosslinked DNA was purified by phenol/chloroform extraction. The DNA was saved and used for internal reference control in the following Real-time PCR reactions. The rest of sonicated chromatin was immunoprecipitated with anti-phospho c-Fos antibody at a concentration recommended by the manufacturer, while negative control was immunoprecipitated with IgG.
Immunocomplexes were recovered with Protein A agarose beads. Crosslinks were reversed and treated by Proteinase K to remove protein from the DNA. DNA was purified by phenol/chloroform extraction. Real-time PCR was performed to quantify c-Fos/c-Jun-binding  6 promoter fragments using the following primers: forward 5'-TGTACTCCCTCCCCAGGTCT-

Construction of  6 -expressing lentiviruses and cell infection
The full-length cDNA of human  6 corresponding to NCBI Reference Sequence:

Assessment of lung inflammation and injury
Mice were sacrificed and lungs were flushed with three 1mL aliquots of complete media.
Recovered BAL fluid samples were centrifuged (300 g for 8 min). The cell pellets were resuspended in 0.5 ml complete media. Total cells were counted with a hemocytometer. For differential cell counts, cells were stained with Diff-Quik (Baxter Healthcare, Miami, FL) according to manufacturer's recommendation. BAL cell differential counts were determined based on standard morphological criteria and staining properties under a light microscope with evaluation of ≥500 cells/slide.

Subcellular fractionation, co-immunoprecipitation, immunoblot and densitometry analysis
Nuclear proteins and cytoplasmic proteins were isolated using NE-PER Nuclear and Cytoplasmic Extraction Kit (Thermo Scientific) as described in our previous studies 5 .
Cell lysates were precleared by adding 30 l of protein G-Sepharose (Amersham Pharmacia Biotech). Supernatants were incubated with rabbit polyclonal antibodies against  6 at a concentration recommended by the manufacturer or an equal amount of non-immune rabbit IgG (SouthernBiotech) at 4°C for 1 h. Precipitation of the immunocomplexes was achieved by adding goat anti-rabbit IgG (SouthernBiotech) at a final concentration of 1 g/ml.
Immunocomplexes were pelleted with 100 l of protein G-Sepharose. After 1 h of incubation at 4°C, the resins were washed three times and  6 -binding immunocomplexes were eluted by boiling in Laemmli buffer (Bio-Rad).
Immunoblot was performed as described in our previous studies 6 . Blot images were scanned.

Flow cytometry
Single

Lung histology and hydroxyproline content assay
Masson's trichrome stain for collagen and H&E stain were performed using a kit from Poly

Micro-computed tomography (Micro-CT) and 3D volume reconstructions
Mouse lungs were inflated with 4% paraformaldehyde in PBS and fixed in fresh fixative for 4 h. Fixed lungs were transferred to 15% sucrose in PBS and left overnight at 4°C. Lungs were dehydrated for 2 h each in 70%, 80% and 90% ethanol and then overnight in 100% ethanol.
Dehydrated lungs were incubated in 100% hexamethyldisilazane for 2 h. After air dry, lungs were placed in a 16 mm diameter sample holder and scanned using a Scanco analyses were performed on the outlined lungs. A fixed threshold of 32 (of maximal grey-scale value) was used to determine the whole lung volumes. A further analysis using a fixed threshold of 120 was used to determine the lung tissue volumes. Aerated lung volumes were calculated by subtraction of the lung tissue volumes from the whole lung volumes.