Hemidesmosomes modulate force generation via focal adhesions

Abstract

Hemidesmosomes are specialized cell-matrix adhesion structures that are associated with the keratin cytoskeleton. Although the adhesion function of hemidesmosomes has been extensively studied, their role in mechanosignaling and transduction remains largely unexplored. Here, we show that keratinocytes lacking hemidesmosomal integrin α6β4 exhibit increased focal adhesion formation, cell spreading, and traction-force generation. Moreover, disruption of the interaction between α6β4 and intermediate filaments or laminin-332 results in similar phenotypical changes. We further demonstrate that integrin α6β4 regulates the activity of the mechanosensitive transcriptional regulator YAP through inhibition of Rho-ROCK-MLC- and FAK-PI3K-dependent signaling pathways. Additionally, increased tension caused by impaired hemidesmosome assembly leads to a redistribution of integrin αVβ5 from clathrin lattices to focal adhesions. Our results reveal a novel role for hemidesmosomes as regulators of cellular mechanical forces and establish the existence of a mechanical coupling between adhesion complexes.

Figure 1.

Inverse correlation between integrin β4 expression and FA maturation. (A) Immunofluorescence confocal images of migrating PA-JEB/β4 keratinocytes stained for BP230 (green), integrin β4 (Itg. β4; red), phosphorylated paxillin (PAX-pY31; blue), phosphorylated MLC (MLC-pS19; green), vinculin (Vinc; red), and actin (Act; blue) 72 h after creating the gap. Nuclei were counterstained with DAPI (cyan). T means the trailing area, while L means the leading area of migrating cell monolayer. Yellow arrowheads indicate the enrichment of hemidesmosomal structure at the leading border, and red arrowheads indicate the higher level of MLC-pS19 and larger FAs in the leader cells. Scale bars: 20 µm. For quantification, type I HD (based on β4 and BP230 colocalization), MLC-pS19, and vinculin-positive areas are calculated as a percentage of the total ROI area. The values represent the mean (± SD) of three independent experiments, with ∼18 images per experiment. ***, P < 0. 001; ****, P < 0.0001. (B) Representative confocal fluorescence microscopy images of PA-JEB (β4 −) and PA-JEB/β4 (β4 +) keratinocytes cultured for 1 d in complete KGM medium and then switched to DMEM (10% FCS) for 16 h. Cells were immunostained for β4 (green), plectin (Plec; red), and laminin-332 (Ln332; blue). Colocalization of β4, plectin, and laminin-332 is visualized in the overlay images. Nuclei were counterstained with DAPI (cyan). (C) Cells were immunostained for β4 (green), vinculin (Vinc; red), and actin (Act; blue). Nuclei were counterstained with DAPI (cyan). Scale bars: 10 µm. (D) Cell area and FA size probed by vinculin were quantified with ImageJ. The values represent the mean (± SD) of three independent experiments, with ∼20 images per experiment. ****, P < 0.0001.

Figure 2.

Close connection between HDs and FA components. (A) Proximity biotinylation assays were performed with PA-JEB/β4 keratinocytes expressing integrin β4 fused to the biotin ligase BirA*. An IL-2R-BirA* fusion protein, which is dispersed over the cell membrane, was used as negative control to identify the specific proximity interactors of integrin β4. The volcano plot shows the results from three independent experiments (threshold false discovery rate: 0.01 and S0: 2). Significant proximity interactors of β4 and IL-2R are indicated in light blue (IL-2R interactors, left, and integrin β4 interactors, right), red (FA components), or green (HD proteins). (B) Representative superresolution microscope image of β4 (+) cells showing vinculin (red) and integrin β4 (cyan). (C) Representative superresolution microscope image showing vinculin (red) and plectin (green) in β4 (+) and β4 (−) PA-JEB cells. (D) Representative superresolution microscope image showing vinculin (red) and keratin 14 (green) in β4 (+) and β4 (−) PA-JEB cells. Scale bars: 1 µm.

Figure S1.

Confirmation of BirA*-β4 interactions by streptavidin precipitation and Western blot. (A) PA-JEB/β4 keratinocytes (cultured in KGM or DMEM) expressing integrin β4 fused to the biotin ligase BirA* were used to perform proximity biotinylation assays. The proximity interactors of integrin β4 were pulled down by streptavidin beads and detected by Western blotting with the indicated antibodies. (B) Western blot analysis of whole-cell lysates (WCL) of plectin (Ple) knockout (KO) clones, probed with antibodies against plectin and integrin β4. β4 (+) PA-JEB cells were included as a positive control.

Figure 3.

Intact laminin-integrin β4-plectin linkage reduces FA size and cell spreading. (A) Domain structure of integrin β4 and plectin. Dots indicate the relative locations of the R1281W mutation in β4-R1281W and of the D230A, P232A and E233A mutations in β4-AD. ABD, actin-binding domain. (B) Representative confocal fluorescence microscopy images of β4 (−), β4 (+), β4-R1281W, and β4-AD PA-JEB keratinocytes. Cells were cultured for 24 h in DMEM (10% FCS) and then fixed and stained for β4 (green), plectin (Plec; red) or BP230 (red), and keratin-14 (K14; blue) or laminin-332 (Ln322; blue). Nuclei were counterstained with DAPI (cyan). Scale bars: 10 µm. (C) Inverse black-and-white images of confocal micrographs of β4 (−), β4 (+), β4-R1281W, and β4-AD PA-JEB keratinocytes showing cell morphology and vinculin-stained FAs (black). Scale bars: 10 µm. (D) Quantification of cell area and FA size with ImageJ. The data are presented as the mean (± SD) from three independent experiments, with ∼20 images per experiment. ****, P < 0.0001. (E) Confocal microscopy images of vinculin-stained FAs (red) and plectin-stained HDs (green) in β4 (+) and plectin-deficient β4 (+) keratinocytes (Plec KO). Nuclei (blue) were visualized with DAPI staining. Scale bars: 10 µm. (F) Quantification of cell area and FA size from three independent experiments, with ∼20 images per experiment. ****, P < 0.0001. (G) PA-JEB/β4 keratinocytes were cultured for 24 h in DMEM (10% FCS) with or without the β4 blocking antibody ASC-8 (supernatant diluted 1:5). Shown are quantification of cell area and FA size from three independent experiments, with ∼20 images per experiment. **, P < 0.01; ****, P < 0.0001.

Figure S2.

Focal contact area varies depending on the expression and function of integrin β4. (A) Cells were treated with or without integrin α3–blocking mAb (J143; 20 µg/ml) in suspension before a short-term (45 min) adhesion assay was performed on a laminin-332–rich matrix substrate. Data are presented as the mean (± SD) from three independent experiments. *, P < 0.05; ****, P < 0.0001. (B) Quantification of FA length probed by vinculin with ImageJ. Data are presented as the mean (± SD) from three independent experiments, with ∼20 images per experiment. **, P < 0.01; ****, P < 0.0001. (C) Inverse black-and-white images of confocal micrographs of β4 (−), β4 (+), β4-R1281W, and β4-AD PA-JEB keratinocytes showing cell morphology and paxillin-stained FAs (black). Scale bars: 10 µm. (D and E) Quantification of FA size and length probed by paxillin with ImageJ. Data are presented as the mean (± SD) from two independent experiments, with ∼20 images per experiment. **, P < 0.01; ****, P < 0.0001. (F) Integrin β4 (+) PA-JEB keratinocytes were treated with integrin β4–blocking mAb (ASC-8; supernatant diluted 1:5) alone or together with integrin α3-blocking mAb (J143; 20 µg/ml). Data are presented as the mean (± SD) from three independent experiments. *, P < 0.05; ****, P < 0.0001.

Figure 4.

Integrin α6β4 and plectin reduce cellular traction force and cellular tension. (A) Representative confocal image of β4 (+) cells exerting forces on stiff pillars. The cell seeded on the pillar top was stained for actin (green), vinculin (blue), and integrin β4 (red). The secondary fluorescence antibody against vinculin has a strong nuclear background (blue). The pillar tops were coated by fibronectin tagged by fluorescence (gray). Deflection of pillars is shown by arrows, whose length is relative to force magnitude. The cyan line represents the cell periphery. Scale bars: 10 µm. (B) Intensity profile of vinculin, integrin β4 (Itg. β4), and the pillar tops from the green line drawn following the deflection of pillars in A. (C and D) Cell spreading area and total cellular force of β4 (−), β4 (+), β4-R1281W, and β4-AD keratinocytes adhering on soft (29.5 kPa) and stiff (137.1 kPa) pillars. The data are presented as the mean (± SD) from three independent experiments, with ∼70 cells in total. **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. (E) Representative images from confocal microscopy comparing the phosphorylation of MLC at S19 (red; MLC-pS19) in PA-JEB cells stably expressing β4-GFP (green) with that in β4-deficient PA-JEB cells (left), PA-JEB cells expressing β4-R1281W (middle), or cells expressing β4-AD (right). Scale bars: 10 µm. Graph showing the quantification of the intensity of MLC-pS19 in different cell lines, expressed as a ratio of the total pixel intensity of MLC-pS19 staining divided by cell area. (F) Western blot analysis of whole-cell lysates from β4 (−), β4 (+), β4-R1281W, and β4-AD keratinocytes probed with antibodies against MLC-pS19 and GAPDH. PA-JEB cells treated with Y27632, a ROCK inhibitor, served as a negative control for validation of the MLC-pS19 antibody. A representative Western blot is shown (n = 3). (G) Western blot analysis of whole-cell lysates from integrin β4 (+), β4 (−) keratinocytes, and two plectin-knockout clones probed with antibodies against MLC-pS19 and GAPDH. A representative Western blot is shown (n = 2).

Figure S3.

Integrin α6β4 reduces traction force generation through modulation of RhoA activity. (A) Confocal images showing the deposition of laminin-332 (red) by β4 (−) and β4 (+) PA-JEB keratinocytes seeded on micropillars. Cells were visualized by actin staining (green), and pillar tops were coated with fibronectin (blue). Scale bars: 10 µm. (B) Quantifications of percentage of deflected pillars and force per pillar of the indicated cell lines seeded on pillars. (C) Representative immunoblot of RhoA activity in integrin β4 (−), β4 (+), β4-R1281W, and β4-AD PA-JEB keratinocytes. Active RhoA was pulled down from the lysates, using GST–rhotekin–binding domain fusion proteins with the RhoA-binding region of rhotekin. (D) RhoA activity is shown as active RhoA/total RhoA, and values were normalized to the β4 (−) group. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; n.s., not significant.

Figure 5.

Mechanotransduction-related signaling pathways are inactivated by integrin α6β4 and plectin. (A) Representative images from confocal microscopy showing the distribution of YAP (red in merge) and cell nuclei (blue in merge) in β4 (−), β4 (+), β4-R1281W, and β4-AD PA-JEB keratinocytes. Scale bars: 10 µm. (B) Quantification of the percentage of nuclear YAP. The percentage of nuclear YAP was calculated by dividing the YAP staining overlapped with DAPI with the total YAP staining intensity. The data are presented as the mean (± SD) from three independent experiments, with ∼60 cells in total. ****, P < 0.0001. (C) Western blot analysis showing the levels of phosphorylated FAK (Y397) and paxillin (PAX-pY31) in β4 (−), β4 (+), β4-R1281W, and β4-AD PA-JEB keratinocytes. A representative Western blot is shown (n = 2). (D) Western blot analysis showing the levels of MLC-pS19, FAK (at Y397), AKT (at S473), mTOR (at S2448), and YAP (S127) of β4 (−) and plectin (−) keratinocytes treated with ROCK (Y27632; 10 µM), myosin (blebbistatin; 10 µM), FAK (VS-4718; 1 µM), PI3K (GDC-0941; 1 µM), AKT (MK-2206; 1 µM), or mTOR (AZD-8055; 100 nM) inhibitors. Inhibitors were added 45 min before cell lysate. β4 (+) cells are used as a control group. DMSO was used as vehicle control. (E) The relative ratios of phosphorylated FAK/total FAK (pFAK/tFAK) and phosphorylated YAP/total YAP (pYAP/tYAP) were calculated and normalized to β4 (+) group treated by DMSO. Data are presented as the mean (± SD) from two independent experiments.

Figure 6.

HD assembly controls the localization of integrin αVβ5. (A) Confocal images showing the distribution of β5 (green) together with either clathrin (Cla; red) or vinculin (Vinc; red) in β4 (−) and β4 (+) PA-JEB keratinocytes. Nuclei were counterstained with DAPI (blue). Scale bars: 10 µm. (B) Quantification of the colocalization of β5 with clathrin or vinculin. (C) Representative images showing integrin β5 clustering in clathrin lattices in β4 (−) cells, untreated or treated with the myosin inhibitor blebbistatin (20 µM) for 30 min. (D) Quantification of the colocalization of β5 with clathrin. (E) Confocal images showing the distribution and colocalization of β5 (green) and vinculin (red) in β4 (−), β4 (+), β4-R1281W, and β4-AD PA-JEB keratinocytes. Nuclei are shown in blue. Scale bars: 10 µm. (F) Quantification of the colocalization of β5 with vinculin. (G) Confocal images showing the distribution of β5 (green) together with either vinculin (red) or clathrin (red) in PA-JEB/β4 wild-type (close-up in the middle panel) and plectin knockout PA-JEB/β4 (close-up in the left panel) keratinocytes. Plectin is shown in blue, and nuclei are shown in cyan. Scale bars: 10 µm. (H) Quantification of the colocalization of β5 with vinculin. Data are presented as the mean (± SD) from three independent experiments, with ∼60 cells in total. ****, P < 0.0001.

Figure 7.

HDs counteract actomyosin contractility in carcinoma cell lines. (A) Confocal images showing the distribution of β4 (red) together with plectin (green) and laminin (blue) in HaCaT, A431, A549, and MCF10A cells. Font in red indicates cells are skin derived, green indicates lung-derived cells, and blue indicates breast-derived cells. Scale bars: 10 µm. (B) Colocalization efficiency between the integrin β4 (Itg. β4) channel and the plectin channel, shown by Pearson’s R value. Data are presented as box-and-whisker plots in which the box extends the 25th to 75th percentiles, the middle line indicates the median, and whiskers go down to the smallest value and up to the largest (∼60 cells in total). The statistical analysis compares the PA-JEB/β4 group with each of the other groups. ****, P < 0.0001. (C) Western blot analysis of whole-cell lysates from PA-JEB, HaCaT, A431, A549, and MCF10A cells either expressing integrin β4 or lacking integrin β4 probed with antibodies against integrin β4, MLC-pS19, and GAPDH.

Figure 8.

Proposed model for the role of HDs in cellular tension regulation. Contractile forces generated by the actomyosin cytoskeleton can prestress IFs, which can counterbalance a further increase in force generation. Binding of the prestressed IFs to β4 in HDs and F-actin–anchored FAs is mediated by plectin (situation on the left). When the HD constraint on actomyosin generated force is eliminated, force transduction across integrins will increase and FAs can further grow and reorganize into larger and more complex structures (situation on the right). Upon increased cellular tension, more FAK can be phosphorylated and subsequently trigger the activation of the PI3K–YAP signaling pathway.