Flightless I Regulates Hemidesmosome Formation and Integrin-Mediated Cellular Adhesion and Migration during Wound Repair  Zlatko Kopecki, Ruth Arkell, Barry C. Powell, Allison J. Cowin  Journal of Investigative Dermatology  Volume 129, Issue 8, Pages 2031-2045 (August 2009) DOI: 10.1038/jid.2008.461 Copyright © 2009 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 1 Flii overexpressing mice have thinner skin and reduced hemidesmosome formation. (a–d) Representative images of haematoxylin and eosin stained mice skin in Flii+/-, WT, FliiTg/+, and FliiTg/Tg illustrating significantly decreased dermal thickness in FliiTg/Tg unwounded skin compared to other mice groups. Scale bar in D refers to A–D and =500μm. n=9 *P<0.05. (e–h) Electron transmission microscopy reveals that skin from Flii+/- mice has significantly higher numbers of hemidesmosomes at the dermal–epidermal junction compared to WT, FliiTg/+, and FliiTg/Tg mice. Hemidesmosomes=solid black arrow. Region without hemidesmosomes=clear arrow. Extracellular matrix=star. Basal keratinocyte=triangle. Magnification × 7900, scale bar in H refers to E–H and=2μm. (i) Graph showing effect of Flii gene expression on dermal skin thickness. (j) Tensiometer analysis of Flii+/-, WT, FliiTg/+, and FliiTg/Tg mice skin. FliiTg/Tg mice skin had significantly weaker tensile strength compared to Flii+/-, WT, and FliiTg/+ counterparts, (*a-c) respectively (n=12). (k) Graph showing number of hemidesmosomes in unwounded skin with different levels of Flii gene expression. Flii+/- (*a) have significantly higher number of hemidesmosomes whereas both FliiTg/+ (*b) and FliiTg/Tg (*c) have significantly decreased number of hemidesmosomes compared to WT counterparts (n=2). All figures represent means and standard error of means *P<0.05. The color reproduction of this figure is available on the html full text version of the article. Journal of Investigative Dermatology 2009 129, 2031-2045DOI: (10.1038/jid.2008.461) Copyright © 2009 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 2 Flii overexpressing mice have impaired hemidesmosome structure. (a–d) FliiTg/+ and FliiTg/Tg show reduced inner plaque (white arrow) and significantly decreased number of hemidesmosomes with sub-basal dense plate (black arrow). Magnification × 64,000, scale bar=200nm. (e–h) FliiTg/+ and FliiTg/Tg mice have shorter hemidesmosomes adhesion sites (clear arrow). Magnification × 25,000, scale bar=500nm. (i–l) Flii+/- and WT mice show intracellular networks of dense tonofibrils (white arrow) and anchoring fibrils (thin black arrow) whereas FliiTg/+ and FliiTg/Tg mice have diffuse tonofilaments (thick black arrow) and decreased network of anchoring fibrils (thin black arrow). Black star=extracellular matrix. White triangle=basal keratinocyte. Magnification × 64,000, scale bar=200nm. (m) Graphical analysis of percentage of hemidesmosomes with sub-basal dense plate. Both FliiTg/+ and FliiTg/Tg mice have significantly reduced number of hemidesmosomes with sub-basal plate when compared to WT and Flii+/- mice respectively (*P<0.05). (n) Graphical analysis of hemidesmosome length. FliiTg/+ mice have significantly shorter hemidesmosome length compared to Flii+/- mice (*P<0.05) whereas FliiTg/Tg mice have significantly shorter hemidesmosome length compared to Flii+/-, WT, and FliiTg/+ mice. Figure represents means and standard error of means. One hundred and fifty hemidesmosomes measured per group, *P<0.05. Journal of Investigative Dermatology 2009 129, 2031-2045DOI: (10.1038/jid.2008.461) Copyright © 2009 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 3 Flii regulates keratinocyte and fibroblast cell adhesion. (a) Graph showing keratinocyte adhesion on laminin, collagen I and fibronectin. FliiTg/+ keratinocytes have significantly impaired adhesion on all three substrates (*a–h), respectively. (b) Graphical analysis of fibroblast cell adhesion on different substrates. Flii+/- fibroblasts have significantly improved adhesion compared to wild-type and FliiTgTg counterparts on all substrates (*a–i), respectively. Figures represent a mean and standard error of means, n=6 *P<0.05. Journal of Investigative Dermatology 2009 129, 2031-2045DOI: (10.1038/jid.2008.461) Copyright © 2009 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 4 Flii affects cellular spreading on different extracellular matrix substrates. Keratinocyte (a–l) and fibroblast (m–x) spreading on laminin, collagen I, fibronectin and glass was examined by staining cells with phalloidin-FITC. Representative images of cell spreading on different substrates are shown in a–x. Magnification × 100. Scale bar in a and m=10μm and refers to all images. Journal of Investigative Dermatology 2009 129, 2031-2045DOI: (10.1038/jid.2008.461) Copyright © 2009 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 5 Flii regulates keratinocyte cell migration. (a–c) Representative images of scratch wound areas in Flii+/-, wild-type, and FliiTg/Tg keratinocytes 48hours post-wounding. (d) Graph showing change in wound area over 48hours of keratinocyte migration. FliiTg/Tg and wild-type keratinocytes have significantly decreased wound area compared to Flii+/- counterparts (*a–t) respectively. Results represent mean and standard error of mean, n=6, *P<0/05. Scale bar in a=100μm and refers to all images. Journal of Investigative Dermatology 2009 129, 2031-2045DOI: (10.1038/jid.2008.461) Copyright © 2009 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 6 Differential effects of Flii on fibroblast outgrowth from skin explants. Here, 2mm punch biopsies from Flii+/-, WT, FliiTg/+, and FliiTg/Tg were allowed to adhere to collagen I, laminin, and fibronectin-coated coverslips prior to culturing over 7 days. Images were taken daily and cellular outgrowth measured as proximal distance. (a) Representative images of cellular outgrowth on collagen I taken at day 6 post planting. Fib=fibroblasts. Scale bar=100μm. (b) Flii+/- cells have significantly increased outgrowth on laminin compared to FliiTg/Tg fibroblasts from day 4 post-seeding, and compared to both WT and Flii+/- cells from day 5 post-seeding. (*a–r respectively). (c) Cellular outgrowth on collagen I shows a similar pattern to laminin, where Flii+/- cells have increased outgrowth compared to WT, FliiTg/+, and FliiTg/Tg fibroblasts (*a–z), respectively. (d) Cellular outgrowth on fibronectin is similar to those on laminin and collagen I, except no significant difference is observed between WT and FliiTg/+ fibroblasts outgrowth (*a–y, respectively). Figure represents means and standard error of means. *P<0.05. Journal of Investigative Dermatology 2009 129, 2031-2045DOI: (10.1038/jid.2008.461) Copyright © 2009 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 7 CD151 expression is increased in Flii-deficient mice during wound repair. (a) CD151 expression detected as a white fluorescence in the epidermis and dermis of wounded Flii+/-, WT, and FliiTg/+ skin at day 3 and 7 post-wounding. Epidermis=e. Wound matrix=w. Magnification × 40, scale bar=100μm. (b) Western Blot analysis of CD151 and β-tubulin (control) in protein extracted from WT and Flii+/- skin at day 3 and 7 post-wounding. (c) Both WT (*a) and Flii+/- (*b) wounds have significantly increased CD151 expression in epidermis at day 3 post-wounding when compared to FliiTg/+ wounds. Flii+/- wounds also have significantly increased CD151 expression compared to WT (*c) and FliiTg/+ (*d) wounds at day 7 post-wounding. Figure represents means and standard error of means. n=6 per group per time point. *P<0.05. Journal of Investigative Dermatology 2009 129, 2031-2045DOI: (10.1038/jid.2008.461) Copyright © 2009 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 8 Manipulation of Flii results in altered expression of laminin and laminin-binding integrin α3, β1, α6, and β4 subunits. (a) Western blot analysis of laminin and β-tubulin (control) in protein extracted from Flii+/-, wild-type and FliiTg/Tg skin at day 3 and 7 post-wounding. Flii+/- have increased laminin expression compared to wild-type and Flii overexpressing mice. (b) α3, β1, α6, and β4 integrin expression detected as white fluorescence in the epidermis and dermis of wounded skin at day 3 post-wounding. Epidermis=e. Wound matrix=w. Scale bar in (a) refers to all images=100μm. (c) Graph showing quantitative analysis of integrin β1 expression in the epidermis of Flii+/-, WT, and FliiTg/+ wounds. Flii+/- wounds have significantly increased β1 integrin expression compared to WT wounds at day 3(*a) and 14 (*d) post-wounding and compared to FliiTg/+ wounds at day 3(*b), 7 (*c), 14 (*e), and 21 (*f) post-wounding. (d) Western blot analysis of β1 integrin in protein extracted from Flii+/-, WT, and FliiTg/+ skin at day 3 and 7 post-wounding. Figure represents means and standard error of means. n=6 per group per time point. *P<0.05. Journal of Investigative Dermatology 2009 129, 2031-2045DOI: (10.1038/jid.2008.461) Copyright © 2009 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 9 Graphical representation of α3, α6, and β4 integrin optical density in Flii+/-, WT, and FliiTg/+ wounds. (a) Flii+/- wounds have significantly increased α3 integrin expression compared to FliiTg/+ (*a) wounds at day 14 post-wounding and compared to both WT (*b) and FliiTg/+ (*c) wounds at day 21 post-wounding. (b) Flii+/- wounds have significantly decreased α6 integrin expression compared to FliiTg/+ (*a) and WT (*b) wounds at day 3 post-wounding. At day 7 post-wounding both WT (*c) and Flii+/- (*d) wounds have significantly increased α6 integrin expression compared to FliiTg/+ wounds. Flii+/- wounds have significantly higher α6 integrin expression compared to WT and FliiTg/+ mice wounds at day 14 (*e, f) and 21 (*g, h) post-wounding, respectively. (c) In the epidermis, both Flii+/- (*a) and FliiTg/+ (*b) wounds have significantly higher β4 integrin expression at day 3 post-wounding when compared to WT counterparts. At day 7 post-wounding, WT wounds have significantly higher expression of β4 integrin compared to FliiTg/+ (*c) wounds whereas Flii+/- wounds have significantly decreased β4 integrin expression compared to both WT(*d) and FliiTg/+ (*e) wounds. Figure represents means and standard error of means. n=6 per group per time point. *P<0.05. Journal of Investigative Dermatology 2009 129, 2031-2045DOI: (10.1038/jid.2008.461) Copyright © 2009 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 10 Flii co-localization with integrin-binding proteins involved in adhesion-dependent signaling pathways. (a) Anti-integrin β1, β4, and tetraspanin CD151 immunoprecipitates (IP) were prepared from confluent HaCaTs keratinocytes and immunoblotted with Flii or Gelsolin (positive control) antibodies. No co-localization was observed with Flii and integrins β1, β4, or tetraspanin CD151 and specific bands were only observed in HaCaTs cell lysate controls. (b) Anti-talin, paxillin, vinculin, and α-actinin immunoprecipitates (IP) were prepared from confluent HaCaTs keratinocytes and immunoblotted with Flii and Gelsolin (positive control) antibodies. Flii colocalized with all actin-binding cytoskeletal proteins except α-actinin. Data are representative of three independent experiments. (c) Representative confocal images of dual immunofluorescence on HaCaTs keratinocytes illustrating Flii (green) and different actin-binding protein (red) staining. Merged images show Flii co-localization with talin, vinculin, and paxillin but not α-actinin. Magnification × 100. Scale bar=10μm and refers to all images. Data are representative of three independent experiments. Journal of Investigative Dermatology 2009 129, 2031-2045DOI: (10.1038/jid.2008.461) Copyright © 2009 The Society for Investigative Dermatology, Inc Terms and Conditions