Journal of Pathology J Pathol 2014; 232: 541–552 Published online 5 February 2014 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/path.4323
ORIGINAL PAPER
Flightless I over-expression impairs skin barrier development, function and recovery following skin blistering Zlatko Kopecki,1,2* Gink N Yang,1 Ruth M Arkell,3 Jessica E Jackson,1 Elizabeth Melville,1 Hiroaki Iwata,4 Ralf J Ludwig,4 Detlef Zillikens,4 Dedee F Murrell5 and Allison J Cowin1,2 1
Centre for Regenerative Medicine, Mawson Institute, University of South Australia, Adelaide, Australia Women’s and Children’s Health Research Institute, Adelaide, Australia 3 Research School of Biological Sciences, Australian National University, Canberra, Australia 4 Department of Dermatology, University of Lubeck, Germany 5 Department of Dermatology, St George Hospital, University of New South Wales, Sydney, Australia 2
*Correspondence to: Z Kopecki, Centre for Regenerative Medicine, Mawson Institute, University of South Australia. e-mail: [email protected]
Abstract Development of an intact epidermis is critical for maintaining the integrity of the skin. Patients with epidermolysis bullosa (EB) experience multiple erosions, which breach the epidermal barrier and lead to increased microbial colocalization of wounds, infections and sepsis. The cytoskeletal protein Flightless I (Flii) is a known regulator of both development and wound healing. Using Flii +/− , WT and Flii Tg/Tg mice, we investigated the effect of altering Flii levels in embryos and adult mice on the development of the epidermal barrier and, consequently, how this affects the integrity of the skin in EB. Flii over-expression resulted in delayed formation of the epidermal barrier in embryos and decreased expression of tight junction (TJ) proteins Claudin-1 and ZO-2. Increased intercellular space and transepidermal water loss was observed in FliiTg/Tg adult mouse skin, while FliiTg/Tg keratinocytes showed altered TJ protein localization and reduced transepithelial resistance. Flii is increased in the blistered skin of patients with EB, and over-expression of Flii in experimental EBA showed impaired Claudin-1 and -4 TJ protein expression and delayed recovery of functional barrier post-blistering. Immunoprecipitation confirmed Flii associated with TJ proteins and in vivo actin assays showed that the effect of Flii on actin polymerization underpinned the impaired barrier function observed in FliiTg/Tg mice. These results therefore demonstrate an important role for Flii in the development and regulation of the epidermal barrier, which may contribute to the impaired healing and skin fragility of EB patients. Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.
Keywords: Flii; epidermolysis bullosa; skin barrier; epidermis
Received 18 September 2013; Revised 10 December 2013; Accepted 20 December 2013
No conflicts of interest were declared.
Introduction An intact epidermis is critical for maintaining the integrity and function of healthy skin. This physical barrier consists of the stratum corneum, cell–cell tight junctions (TJ) and associated cytoskeletal proteins [1]. Patients with epidermolysis bullosa (EB) experience recurrent skin blistering, leading to disruption or loss of the protective barrier, increased microbial wound colocalization and the development of chronic infected wounds [2,3]. Accordingly, sepsis is the leading cause of death in infants with EB [4]. During development and epidermal wound healing, TJ proteins are expressed well before stratum corneum formation [5,6]. Similarly to wound healing, blistering results in TJ protein redistribution in marginal cells [7,8] and upregulation in the hyperproliferative zone surrounding the blister. An initial protective barrier is first formed by cells re-epithelializing the blister wound, well before Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
stratum corneum formation [8–10]. Establishment of an intact barrier contributes to skin strength, while the quick recovery of the barrier post-blistering is critical for blister healing. As the complex response of epidermal cells to barrier disruption may exacerbate EB, treatments aimed at aiding barrier re-establishment or dampening the epidermal stress response may improve the healing observed in EB patients. TJs are intercellular junctions which form the paracellular barrier and consist of occludin, claudins (Cldns) and junctional adhesion molecules linked to the actin cytoskeleton through scaffolding plaque proteins; zonula occludens (ZO)-1, -2, -3, MUPP-1 and cingulin [11]. Different studies have shown the importance of Cldns and ZO-1 and -2 in the formation of TJ strands and paracellular barrier functions [12,13]. Protein–protein interactions between Cldns and scaffolding plaque proteins provide the link to the actin cytoskeleton allowing the transduction of the J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
542
regulatory signals to and from TJs [14] and highlight the diverse role of TJs in regulating adhesion, signalling, differentiation, proliferation, migration and tissue permeability [5]. TJ proteins were initially implicated in regulating epithelial motility and polarity during wound healing [15], and more recent studies have shown a role for TJ proteins in regulating keratinocyte motility during wound healing and growth, motility and metastasis during tumorigenesis [16–18]. Actin remodelling is required for these cellular functions, the assembly and maturation of TJs and for establishing the epidermal barrier [19]. Flightless I (Flii) is a member of the Gelsolin family of actin remodelling proteins and a nuclear receptor co-activator with unique roles in regulating cell adhesions and intracellular signalling [20,21]. Homozygous knockout of Flii is embryonic lethal, highlighting the role of Flii as an essential regulator of fetal development [22]. Flii expression is increased in the epidermis during development and epidermal stratification [23]. Different studies have shown that Flii is capable of regulating cellular proliferation and migration [24], innate immunity through its downregulation of the Toll-like receptor pathway [25] and NF-κB signalling [26], and cytokine production via caspase activation and IL-1β maturation [27]. Flii is increased in the blistered skin of EB patients [28], and levels of Flii are significantly higher in human blister fluid compared to acute wound fluid [21]. Additionally, Flii affects TGFβ signalling, collagen assembly and fibroblast contraction, leading to increased dermal–epidermal blistering [21,29]. Blisterinduced barrier disruption leads to cytokine production [10] and keratinocyte proliferation and differentiation to restore the epidermal barrier and assemble new TJs. Impairment of these processes results in the development of the inflamed chronic non-healing wounds seen in EB patients. Previous studies have described the Flii effect on keratinocyte and fibroblast cellular adhesion structures [30,31]. Flii regulates transmembrane proteins, including CD151 [30] and Drebrin1, which are important in the cellular adhesion [32] and stabilization of TJs, respectively [33]. In addition, Flii has been identified as one of the 900 proteins present in TJ protein complexes [10]. In the present study we have investigated the effect of Flii on TJ protein expression and determined its effect on skin barrier development, function and recovery following blistering.
Materials and methods Animal studies Mice were maintained according to Australian Standards for Animal Care under protocols approved by The Australian National University Animal Ethics Committee and the Child, Youth and Women’s Health Service Animal Ethics Committee. All strains were BALB/c-congenic and were maintained as homozygous colonies or by continuous backcrossing to BALB/c animals. Wild-type controls were obtained Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
Z Kopecki et al
from BALB/c inbred litters. The murine alleles of Flii used in this study were: 1. A targeted null allele of the murine Flightless I gene (Flii ): Flii tm1Hdc (MGI:2179825) [22]. Animals homozygous null for Flii are embryonic lethal, but heterozygous carriers of the null allele are viable and fertile [22]. A heterozygous carrier of this allele is written as Flii+/− . 2. Tg(FLII)2Hdc (MGI:4939366) [24], a transgenic strain expressing exogenous human flightless I (FLII ), due to the insertion of a 17.8 kb fragment derived from human cosmid clone c110H8 containing all FLII introns and exons and the endogenous promoter (see Figure 1 [34] for extent of the 17.8 kb transgene). Mice homozygous for the transgene were used in this study and had two copies of the mouse Flii gene and two copies of the human FLII transgene (Flii+/+ ; FLIITg/Tg ) and had elevated levels of Flii protein relative to wild-type animals [34]. They are denoted throughout this article as FliiTg/Tg [34]. Embryonic (E15.5, E16.5, E17.5, E18.5 and E19.5) and adult (6 weeks) skin collected from Flii+/− , WT and FliiTg/Tg mice was assessed for expression and localization of TJ proteins, including Cldn1, -4, occludin, ZO-1 and -2, using confocal microscopy as described in Supplementary materials and methods (see supplementary material). Primary keratinocytes from adult Flii+/− , WT and FliiTg/Tg mice were isolated as previously described [30] and used in transepithelial resistance and paracellular tracer flux assays, while the HaCaT keratinocyte cell line was used in co-immunoprecipitation experiments, as detailed in Supplementary materials and methods (see supplementary material). Epidermolysis bullosa acquisita (EBA) was induced by injecting 3–4 week-old wild-type (WT), Flii+/− or FliiTg/Tg mice with rabbit anti-mouse ColVII antibody (0.3 mg/g body weight) subcutaneously every second day for 10 days, as previously described [28]. The mice were examined daily for evidence of cutaneous lesions and the extent of skin blistering was assessed daily. Mice progressively develop skin blistering, with the maximum number of blisters peaking at day 12 following the initial injection. The mice were euthanized and blistered skin collected at day 16 of the experiment, when most of the blisters were undergoing healing and repair. The effect of skin blistering on the localization and expression of TJ proteins during healing was analysed using immunohistochemistry, as described in Supplementary materials and methods (see supplementary material).
In vivo permeability assays In order to study the effect of differential Flii expression on the skin barrier during embryonic development and epidermal stratification, embryos of Flii+/− , wildtype and FliiTg/Tg mice at days 15.5, 16.5, 17.5, 18.5 J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
Flii regulation of skin barrier function
A
B
543
dye (Sigma Aldrich, Australia), following described protocols [35]. After extensive destaining in PBS, images of the embryos were taken, with the blue colour revealing non-functional skin barrier. Images of the embryos were used to quantify the intensity of blue staining, using Image ProPlus v 5.1 (Media Cybernetics Inc.). Materials and methods for reagents and antibodies, ultrastructural analysis, immunofluorescence analysis, cornified envelope extraction, calcium switch assay, western blotting, transepithelial resistance, paracellular tracer flux assay, co-immunoprecipitation, in vivo actin assay and statistical analysis may be found in Supplementary materials and methods (see supplementary material).
Results Over-expression of Flii results in delayed development of skin barrier C
Figure 1. Over-expression of Flii results in delayed development of the skin barrier and increased TEWL. (A) Epidermal permeability assay using toluidine blue staining on E15.5–E19.5 Flii+/− , WT and FliiTg/Tg mouse embryos revealed a defective outside–in permeability barrier at E17.5 in FliiTg/Tg embryos (n = 6). (B) The intensity of the blue staining was quantified using Image ProPlus software, revealing significantly increased permeability at E16.5 and E17.5 in FliiTg/Tg embryos (n = 6). (C) TEWL measurements on E15.5–E19.5 Flii+/− , WT and FliiTg/Tg mouse embryos revealed significantly increased water loss in FliiTg/Tg embryos during epidermal stratification (n = 6). Mean ± SEM; *p < 0.05
and 19.5 dpc were used for TEWL measurement and the ’outside–in’ permeability assay. Permeability of the skin barrier in the embryonic skin was initially assessed by measurements of trans-epidermal water loss (TEWL), using a hand-held vapometer (Delfin Technologies, Finland), following the manufacturer’s instructions. Embryos were subsequently killed and used in the ’outside–in’ permeability assay. Permeability of the skin barrier was assessed in killed embryos by 1 min dehydration in methanol/PBS baths (25%, 50% and 75% methanol/PBS), following by 1 min in 100% methanol. The embryos were then rehydrated in the same series of methanols for 1 min each and stained for 30 min in 0.1% toluidine blue O Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
Using embryonic and adult skin of Flii heterozygous (Flii+/− ), wild-type (WT) and Flii transgenic (FliiTg/Tg ) mice, we investigated the localization and expression of Flii during development. Flii expression increased during skin development, with predominant staining in the developing epidermis and minimal staining observed within fibroblasts of the dermis (see supplementary material, Figure S1A, B). FliiTg/Tg mice have increased Flii nuclear staining in the epidermis compared to Flii+/− and WT counterparts (Figure S1A). Increased Flii levels in the FliiTg/Tg mouse skin were confirmed using western blotting analysis (see supplementary material, Figure S1A–C). To investigate the effect of Flii on the development of the epidermal barrier, the toluidine blue permeability assay was performed on E15.5–E.19.5 Flii+/− , WT and FliiTg/Tg mouse embryos. WT and Flii+/− mouse embryos acquired a functional ’outside–in’ permeability barrier by E17.5 (Figure 1A). In contrast, Flii over-expressing embryos still show penetration of the dye at E17.5 but not at E18.5, indicating that the over-expression of Flii results in delayed development of the epidermal barrier (Figure 1A). Quantifying the intensity of blue staining revealed that FliiTg/Tg embryos have significantly increased permeability at E16.5 and E17.5 compared to WT and Flii+/− counterparts (Figure 1B). In addition, delayed development of an epidermal barrier in FliiTg/Tg mice was confirmed using TEWL measurements, which showed significantly increased water loss in FliiTg/Tg embryos during E16.5–E19.5 when compared to both Flii+/− and WT embryos (Figure 1B).
Flii affects the expression of Cldn1 and ZO-2 in embryonic skin during development In order to ascertain the effect of Flii on epithelial barrier development the expression of structural TJ J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
544
Z Kopecki et al
A
B
C
D
E
Figure 2. FliiTg/Tg embryos had significantly decreased expression of TJ proteins during development. (A) Immunolocalization of Cldn1, -4, occludin, ZO-1 and -2 in E17.5 epidermis of Flii+/− , WT and FliiTg/Tg embryos. (B, C) Graphical analysis of Cldn1 and ZO-2 expression in E17.5 epidermis of Flii+/− , WT and FliiTg/Tg embryos (n = 6). (D, E) Western blot analysis of total skin lysates from E17.5 and E19.5 Flii+/− , WT and FliiTg/Tg embryos revealed decreased expression of TJ proteins in FliiTg/Tg mouse embryos, namely ZO-2 (at E17.5), Cldn1 (at E17.5 and E19.5), Cldn4 (at E17.5) and occludin (at E19.5). Figure is representative of two independent experiments (n = 6). Magnification = ×20; scale bar = 500 µm; mean ± SEM; *p < 0.05.
proteins (Cldn1, Cldn4, occludin) and adaptor signalling TJ plaque proteins (ZO-1, -2) was examined in E17.5 and E19.5 Flii+/− , WT and FliiTg/Tg mouse skin (Figure 2A–E). No differences in the localization of TJ proteins were observed in the different mice genotypes; however, the effect of Flii on the expression of TJ proteins Cldn1 and ZO-2 was pronounced. While Flii+/− mouse embryonic skin showed increased expression of Cldn1at E17.5 compared to WT and FliiTg/Tg counterparts (Figure 2B), over-expression of Flii resulted in significantly decreased expression of both Cldn1 (Figure 2B) and ZO-2 (Figure 2C) compared to WT and Flii+/− mouse embryos. Western blot analysis confirmed the effect of Flii on Cldn1 and Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
ZO-2 expression and, interestingly, while Flii overexpression resulted in decrease of ZO-2 during development, Cldn1 expression was decreased at both E17.5 and E19.5 (Figure 2D, E). To investigate whether an alteration of cell differentiation was involved in the delayed skin barrier observed in FliiTg/Tg mice, expression of keratin 1, involucrin and filaggrin differentiation markers was examined in E17.5, E19.5 and adult Flii+/− , WT and FliiTg/Tg mouse skin (see supplementary material, Figures S2–S4). Reduction of Flii resulted in increased involucrin during early stratification (E17.5) and keratin 1 during development and in adult skin when compared to WT counterparts (see supplementary material, Figures S2, S3); in contrast, J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
Flii regulation of skin barrier function
545
Figure 3. Altered expression of Flii gene in adult mouse skin resulted in impaired formation of the skin barrier, with increased TJ intercellular space and altered cornfied envelope structure. (A) TEM showed impaired skin barrier formation in FliiTg/Tg adult mouse skin, with a significantly increased TJ intercellular space. No difference was observed in lamellar body extrusions, keratohyalin granules or the structure of the desmosome. Magnification = ×13 000; scale bar = 500 nm; n = 4; mean ± SEM; *p > 0.05.
over-expression of Flii led to decreased keratin 1 and increased filaggrin expression at E17.5 when compared to WT counterparts (Figures S3, S4).
Flii over-expression impairs tight junction formation in adult skin Using transmission electron microscopy, we investigated the effect of Flii on skin barrier formation using adult Flii+/− , WT and FliiTg/Tg mouse skin. No effect on lamellar body localization, keratin granules or desmosomes and TJ structure and numbers were observed in stratum granulosum between different genotypes; however, Flii Tg/Tg mouse skin showed significantly larger TJ intercellular space between contact sites of intramembrane fibrils of TJ, suggesting increased epidermal permeability (Figure 3A). Examining the effect of Flii over-expression on cornified envelope formation in adult skin showed no significant difference from WT counterparts (see supplementary material, Figure S5).
Over-expression of Flii alters expression and localization of TJ proteins in cultured adult keratinocytes To analyse the TJ protein cellular localization and expression in response to differential Flii levels, keratinocytes were isolated from adult Flii+/− , WT and FliiTg/Tg mouse skin. Cells were switched to a high-Ca2+ medium to induce TJ formation. Cellular TJ protein localization and expression were examined over a 6 day period. Immunostaining for the Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
transmembrane proteins Cldn1, Cldn4 and occludin and adaptor proteins ZO-1 and -2, using confocal microscopy, revealed no difference between Flii+/− and WT keratinocytes; however, differences in FliiTg/Tg keratinocytes were observed at day 6 following the Ca2+ switch (Figure 4A). Flii over-expression resulted in distorted cell junction outlines, decreased Cldn1 expression at cell contact points, diffuse pattern and decreased expression of ZO-1 and -2 adaptor proteins, reduced staining of occludin and membrane duplication indicative of intercellular spaces and impaired TJ formation (Figure 4A). The effect of Flii overexpression on impaired TJ formation was also confirmed, using western blot analysis of Ca2+ -treated cells, revealing decreased expression of Cldn1 at day 6 and occludin at days 3 and 6 following Ca2+ switch, and decreased Cldn4 and ZO-1 expression at day 6 following Ca2+ switch when compared to WT counterparts (Figure 4B).
Effect of Flii over-expression on TJ barrier function can be rescued following treatment with Flii-neutralizing antibodies Decreased expression and altered localization of TJ proteins has previously been reported to affect TJ barrier function in cultured keratinocytes [36,37]. The effect of Flii on complete transepithelial resistance (TER), including paracellular resistance of TJs and transcellular resistance of apical and basal membranes, was investigated using Flii+/− , WT and FliiTg/Tg keratinocytes in a transepithelial resistance assay, following Ca2+ switch. No significant difference was J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
546
Z Kopecki et al
A
B
Figure 4. Localization and expression of TJ proteins in adult Flii+/− , WT and FliiTg/Tg cells. (A) IHC localization (day 6 post-Ca2+ switch) and (B) protein expression (days 1– 6 post-Ca2+ switch) of TJ proteins in cultured keratinocytes. Impaired formation of TJ was observed in FliiTg/Tg keratinocytes, with decreased Cldn1 expression at cell contact points, diffuse pattern and decreased expression of ZO-1 and -2 adaptor proteins, reduced staining of occludin and membrane duplication at day 6 post-Ca2+ switch. In addition, FliiTg/Tg keratinocytes showed decreased expression of Cldn1 and occludin at days 3 and 6 post-Ca2+ switch and decreased Cldn4 and ZO-1 expression at day 6 post-Ca2+ switch. Figure is representative of two independent experiments. For (a, b) n = 3; magnification = ×60; scale bar = 20 µm.
observed in TER between Flii+/− and WT keratinocytes over a 14 day period (Figure 5A). In contrast, Flii over-expression delayed the formation of TER with significantly reduced TER at days 2–6 compared to both Flii+/− and WT keratinocytes, with the highest reductions of 63 ± 3% and 59 ± 7%, respectively, observed at day 4 (Figure 5A). In addition, FliiTg/Tg keratinocytes never reached the same TER levels of Flii+/− and WT keratinocytes (Figure 5A). Treatment of FliiTg/Tg keratinocytes with Flii-neutralizing antibody (FnAb; 50 µg/ml) but not control IgG antibody significantly increased TER formation in these cells, similar to levels observed in WT control counterparts (highest increase of 51 ± 2% in TER observed at day 6) (Figure 5B). In order to further characterize the effect of Flii on TJ and size selectivity of paracellular permeability, Flii+/− , WT and FliiTg/Tg keratinocytes were subjected to a paracellular tracer flux assay following Ca2+ switch. No significant difference Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
was observed in barrier function and paracellular permeability of 4 kDa (Figure 5C) and 10 kDa (Figure 5D) FITC dextrans in response to varying Flii levels. Treatment of WT and FliiTg/Tg keratinocytes with FnAb (50 µg/ml) or control IgG did not alter keratinocyte paracellular permeability to 4 or 10 kDa FITC dextrans (data not shown).
Flii affects the recovery of skin barrier following skin blistering Flii is increased in blistered skin and contributes to excessive epidermal–dermal separation during skin blistering, affecting both TEWL and healing of blisters [21,28]. In order to investigate the effect of differential Flii on skin barrier recovery follwing skin blistering, Flii+/− , WT and FliiTg/Tg mice were used in an established model of EBA previously described [21,38]. The effect of Flii on recovery of skin barrier J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
Flii regulation of skin barrier function
547
A
B
C
D
Figure 5. FliiTg/Tg keratinocytes had impaired TJ barrier function, which could be recovered using FnAb. (A) TER of cultured Flii+/− , WT and FliiTg/Tg keratinocytes following Ca2+ switch was determined over a period of 14 days. Flii+/− and WT keratinocytes established a TER of up to 15 000 /cm2 by day 6 of the experiment. In contrast, the TER in FliiTg/Tg cells was significantly delayed; however, it was significantly increased in response to FnAb treatment (B). Paracellular diffusion of 4 kDa (C) and 10 kDa (D) FITC–dextran through Flii+/− , WT and FliiTg/Tg keratinocyte monolayers was determined for 144 h after Ca2+ switch; data are expressed as a percentage of 0 h values. Flii+/− , WT and FliiTg/Tg keratinocytes formed similar permeability barriers for 4 and 10 kDa FITC–dextran. For (a, b) n = 6; mean ± SEM; * p < 0.05
function and re-establishment of new TJs was investigated post-blistering, using immunohistochemistry for structural TJ proteins Cldn1 and -4 in the stratum granulosum of the newly formed epidermis covering the healing blisters. Compared to Flii+/− EBA and WT EBA mouse skin, FliiTg/Tg EBA mouse skin revealed diffuse and significantly reduced expression of both Cldn1 and -4, suggesting that high levels of Flii affect TJ protein formation and may contribute to the impaired healing of blisters and increased skin fragility, with impaired skin barrier function in FliiTg/Tg EBA mice (Figure 6A, B).
The strongest association was with structural TJ proteins Cldn1 and -4 and adaptor protein ZO-2, while Flii was not observed to co-immunoprecipitate with occludin (Figure 7A). To investigate the effect of Flii on actin remodelling during TJ formation, Flii+/− , WT and FliiTg/Tg keratinocytes were used in an in vivo actin assay following Ca2+ differentiation. In contrast to Flii+/− and WT keratinocytes, which showed equal F-actin:G-actin distribution, FliiTg/Tg keratinocytes showed no G-actin (Figure 7B). Treatment of FliiTg/Tg keratinocytes with FnAb (50 µg/ml) or cytochalasin D (positive control) restored the Factin:G-actin distribution, favouring improved TJ barrier formation and function (Figure 7B).
Flii associates with TJ proteins and affects actin remodelling during TJ formation To gain a mechanistic insight into the role of Flii in TJ structure and function, immunoprecipitates and cell lysates from HaCaT keratinocytes were analysed, using western blotting following Ca2+ switch-induced formation of TJs in vitro. Flii, but not the family member Gelsolin, was found to co-immunoprecipitate with structural TJ proteins Cldn1, -4 and -6 and adaptor TJ plaque proteins ZO-1 and -2 (Figure 7A). Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
Discussion Flii colocalizes with actin and microtubule based structures and is an important conserved regulator of cellular adhesion with regulatory roles during development and wound healing [22,30,39–42]. Homozygous Flii knockdown in mice results in irregular actin organization and lethality early in embryogenesis, while J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
548
Z Kopecki et al
A
B
Figure 6. Flii affects the recovery of skin barrier formation and function post-blistering. (A, B) IHC localization of Cldn1 and -4 in the epidermis of healing blisters in Flii+/− , WT and FliiTg/Tg EBA mice. Cldn1 and -4 staining was localized to the stratum granulosum, highlighting newly formed TJ post-blistering. FliiTg/Tg mouse blistered skin had a diffuse and significantly reduced expression of both Cldn1 and -4. Magnification = ×40; scale bar = 200 nm; n = 4, mean ± SEM; * p > 0.05.
Flii over-expression results in phenotypically normal mice with thin fragile skin, altered cell adhesions and similar features to those observed in patients with skin-blistering diseases [23,30]. In addition, Flii overexpression in fetal and adult wounds results in a delayed wound-healing response and increased skin blistering in adult mice [23,28,29]. TJ regulation and function is a primary pathophysiological factor in many skin diseases and a contributing factor to skin regeneration [5,43]. Understanding the effect of the regulatory proteins involved in development and/or function of TJs may help in the design of new therapies aimed at ameliorating different clinical skin conditions, including blistering. This study describes the involvement of the cytoskeletal protein Flii during embryonic development, function and recovery of the protective skin barrier in adult mice. We demonstrate that skin barrier development, TJ protein level and localization, as well as TJ function and recovery, are impaired when Flii levels are increased, as is observed in blistered skin. The level of Flii increases during skin development and is significantly elevated in Flii overexpressing mice [23]. Flii over-expression leads to delayed formation of the protective barrier during Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
epidermal stratification, increased transepidermal water loss and decreased expression of structural TJ protein Cldn1, and signalling actin anchoring protein ZO-2 in the embryonic skin. Previous studies have described an important effect of complete knockout of TJ proteins Cldn1 and ZO-2 on skin barrier development and function, resulting in permeable TJs and mouse lethality early after birth and during embryogenesis, respectively [12,44]. Despite the decreased expression of Cldn1 and ZO-2 during embryogenesis, FliiTg/Tg mice show normal phenotype and life expectancy during adulthood, suggesting that other TJ proteins may be providing a compensatory mechanism. In agreement with these findings, complete loss of Cldn1 in humans with neonatal ichthyosis-sclerosing cholangitis syndrome results in a less severe phenotype compared to the phenotype observed in Cldn1−/− mice [45,46]. Furthermore, recent studies using mouse and human keratinocytes reveal that Cldn1 is dispensable for TJ water barrier function, also suggestive of a possible compensatory mechanism [36]. Examining the effect of Flii on cell differentiation revealed a possible keratin 1/filaggrin compensatory mechanism in Flii overexpressing mice during early stratification. No difference in lamellar body localization, keratin granules, J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
Flii regulation of skin barrier function
A
B
Figure 7. Flii effects on actin remodelling and TJ proteins may underpin impaired barrier function in FliiTg/Tg mice. (A) Anti-Cldn1, -4 and -6; ZO-1 and -2 and occludin IP were prepared from HaCaT keratinocytes following Ca2+ differentiation. Flii colocalized with structural (Cldn1, -4 and -6) and adaptor (ZO-1 and -2) TJ proteins. (B) Flii+/− , WT and FliiTg/Tg keratinocytes were used in in vivo actin assay, following Ca2+ differentiation. While Flii+/− and WT keratinocytes had an equal F-actin:G-actin distribution, FliiTg/Tg keratinocytes showed no G-actin. Treatment with Flii neutralizing antibody (FnAb) or cytochalasin D (control) restored the actin distribution in FliiTg/Tg keratinocytes. Data are representative of three independent experiments.
desmosome numbers or cornified envelope formation were observed between Flii over-expressing and wildtype mice, suggesting that the observed delay in skin barrier development, function and recovery is unlikely to be due to effects on cell differentiation, but more likely to relate to the decreased expression of TJ proteins. In addition, the effect of Flii on TJ integrity was further evident, with the significantly increased TJ intercellular space observed in FliiTg/Tg adult skin suggestive of permeable TJs and possible structural and functional effects on skin barrier permeability, which may contribute to the increased skin fragility and blistering observed in these mice [28]. Examining the effect of Flii on the expression and localization of TJ proteins in adult keratinocytes following Ca2+ switch revealed a similar effect of Flii over-expression observed in embryonic skin. Flii over-expression resulted in reduced expression, altered localization and diffuse pattern of both structural and actin-anchoring TJ proteins at cell–cell adherent points. Expression of Cldn1 and occludin was reduced at days 3 and 6 following Ca2+ switch, while expression of Cldn4 and ZO-1 was reduced only at day 6 of the experiment, suggesting a role for Flii in the assembly of TJs. While no effect of Flii on the expression of ZO-2 was observed in adult keratinocytes, Flii’s Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
549
effect on other TJ protein expression and localization during TJ assembly suggests a functional effect of Flii on skin barrier permeability and junctional integrity. Recent studies have described the correlation between TJ protein expression and barrier function, and detailed the distinct TJ protein contribution to both TJ-dependent paracellular resistance and transcellular resistance in cultured keratinocytes [11,36,47]. A direct functional effect of Flii on TJ formation and function, with significantly decreased transepithelial resistance in FliiTg/Tg keratinocytes, was also observed. This could be reversed following treatment with FnAb, suggesting a direct functional effect and a relationship between Flii levels and transepithelial resistance. Flii affected the transcellular resistance of apical and basolateral cell membranes, but no significant differences were observed in the paracellular permeability resistance of TJ. This suggests that Flii affects the flow of water and small ions, but not large molecules. Flii is increased in the blistered skin of patients with different EB subtypes [28], and is present in human blister fluid at higher levels than those observed in human acute wound fluid [21]. We have previously described the negative role of Flii as a contributor to increased dermal–epidermal blistering in a ColVII mouse model of epidermolysis bullosa acquisita (EBA) [28]. To examine the functional effect of Flii on TJ recovery following skin blistering, we examined the expression of structural TJ proteins Cldn1 and -4 during the healing of blisters in Flii+/− , WT and FliiTg/Tg mouse skin, using an EBA mouse model of skin blistering [38]. A decreased and fragmentary expression of structural TJ proteins Cldn1 and -4 at cell–cell borders in the stratum granulosum was observed in healing blistered skin in FliiTg/Tg EBA mice compared to WT and Flii+/− counterparts. These findings suggest that Flii, through its effects on TJ protein expression and localization, affects the recovery of the skin barrier following skin blistering. In agreement with previous studies describing increased skin blistering in FliiTg/Tg EBA mice [28] and decreased TEWL in WT EBA mice treated with FnAb during and following skin blistering [21], these findings suggest that the Flii effect on skin barrier function and recovery contributes to both the increased blistering and the delayed healing of blisters observed in FliiTg/Tg EBA mice [28]. Flii has previously been described as an important regulator of cellular adhesions and a modulator of both actin remodelling and cell signalling, with roles in actin polymerization and nuclear receptor coactivation, respectively [20,48]. Considering its identification as one of the proteins present in the TJ protein complexes [10], we investigated the mechanism behind its effects on TJ formation, function and recovery. Here, we show that Flii does not associate with occludin during TJ formation; however, its possible association with Cldn1, -4 and -6 structural TJ proteins and the actin-linking proteins ZO-1 and -2 may account for the impaired skin barrier function observed in Flii over-expressing mice. TJ permeability can be J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
550
regulated directly, through modification of TJ proteins, or indirectly, through effects on the actin cytoskeleton [49]. Inhibition of F-actin turnover has previously been shown to induce disassembly and internalization of TJs [50], highlighting the relationship between skin barrier function and actin cytoskeleton remodelling. Flii possesses calcium-independent G-actin binding activity as well as F-actin binding and severing activities, and has been shown to actively regulate actin polymerization, leading to altered functional cell outcomes, including adhesion and migration [51,52]. Previous studies have shown that Flii colocalizes with actin and macrotubular structures [40] and has the ability to regulate turnover of cell adhesion structures [31]. Surprisingly, we now demonstrate that Flii alters the Factin:G-actin ratio during TJ formation. Interestingly, the altered F-actin:G-actin ratio seen in FliiTg/Tg keratinocytes during TJ formation can be restored following treatment with FnAb, suggesting that modulation of Flii levels may be beneficial to improved healing and recovery of skin barrier following injury. While the mechanisms underpinning Flii ability to regulate the F-actin:G-actin ratio and/or interact with number of actin monomer-stabilizing or actin-sequestering proteins are yet to be identified, this finding is of potential importance in our understanding of the effects of Flii on cellular function. Flii may, therefore, directly interact with structural and adaptor TJ proteins affecting the formation of the TJ protein complex and association with the actin cytoskeleton, hence affecting skin barrier development, function and recovery following injury. The exact mechanism behind the Flii direct or indirect effect on TJs during skin barrier development, function and recovery are yet to be identified; however, our findings provide novel insight into the relationship between Flii and skin barrier function. Flii promotes keratinocyte permeability during development, in adult mice in vivo and keratinocyte monolayers in vitro, by regulating the expression and organization of TJ proteins and the actin cytoskeleton. These studies provide new insight into the relationship between Flii and skin development and repair following injury.
Acknowledgements This work was supported by grants from the Women’s and Children’s Hospital Foundation (to ZK and AJC) and the NHMRC (Grant Nos 626802 to AJC, DFM and RMA and 1003648 to RMA and AJC). AJC is supported by an NHMRC Senior Research Fellowship (Grant No. 1002009). ZK is supported by an NHMRC Early Career Fellowship (Grant No. 1036509). DZ and RJL are supported by the Excellence Cluster Inflammation at Interfaces (Grant No. DFG EXC 306/2) and the Research Training Group ’Modulation of Autoimmunity’ (Grant No. DFG GRK 1727/1). We acknowledge the technical assistance of Nicole Thomsen and Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
Z Kopecki et al
the Dystrophic Epidermolysis Bullosa Research Association of South Australia for their support.
Author contributions ZK and AJC conceived the experiments; ZK carried out all experiments with the assistance of GNY and JEJ; EM made the Flii neutralizing antibody; RMA provided the Flii genetic mice; DZ, RJL, HI and DFM provided ColVII antibody for the induction of skin blistering; ZK and AJC wrote the manuscript; and all authors contributed to manuscript preparation and approved the final submitted and published versions.
References 1. Koster MI, Roop DR. Mechanisms regulating epithelial stratification. Annu Rev Cell Dev Biol 2007; 23: 93–113. 2. Kopecki Z, Murrell DF, Cowin A. Raising the roof on epidermolysis bullosa (EB): a focus on new therapies. Wound Pract Res 2009; 17: 76–82. 3. van der Kooi-Pol MM, de Vogel CP, Westerhout-Pluister GN, et al. High anti-staphylococcal antibody titers in patients with epidermolysis bullosa relate to long-term colonization with alternating types of Staphylococcus aureus. J Invest Dermatol 2013; 133: 847–850. 4. Fine J-D, Johnson LB, Weiner M, et al. Cause-specific risks of childhood death in inherited epidermolysis bullosa. J Pediat 2008; 152: 276–280, e272. 5. O’Neill CA, Garrod D. Tight junction proteins and the epidermis. Exp Dermatol 2011; 20: 88–91. 6. Herard AL, Zahm JM, Pierrot D, et al. Epithelial barrier integrity during in vitro wound repair of the airway epithelium. Am J Respir Cell Mol Biol 1996; 15: 624–632. 7. Brandner JM, Kief S, Grund C, et al. Organization and formation of the tight junction system in human epidermis and cultured keratinocytes. Eur J Cell Biol 2002; 81: 253–263. 8. Malminen M, Koivukangas V, Peltonen J, et al. Immunohistological distribution of the tight junction components ZO-1 and occludin in regenerating human epidermis. Br J Dermatol 2003; 149: 255–260. 9. Brandner JM, Kief S, Wladykowski E, et al. Tight junction proteins in the skin. Skin Pharmacol Physiol 2006; 19: 71–77. 10. Tang VW. Proteomic and bioinformatic analysis of epithelial tight junction reveals an unexpected cluster of synaptic molecules. Biol Direct 2006; 1: 37. 11. Kirschner N, Haftek M, Niessen CM, et al. CD44 regulates tightjunction assembly and barrier function. J Invest Dermatol 2011; 131: 932–943. 12. Furuse M, Hata M, Furuse K, et al. Claudin-based tight junctions are crucial for the mammalian epidermal barrier: a lesson from claudin-1-deficient mice. J Cell Biol 2002; 156: 1099–1111. 13. Umeda K, Ikenouchi J, Katahira-Tayama S, et al. ZO-1 and ZO-2 independently determine where claudins are polymerized in tightjunction strand formation. Cell 2006; 126: 741–754. 14. Heiskala M, Peterson PA, Yang Y. The roles of claudin superfamily proteins in paracellular transport. Traffic 2001; 2: 93–98. 15. Bement WM, Forscher P, Mooseker MS. A novel cytoskeletal structure involved in purse string wound closure and cell polarity maintenance. J Cell Biol 1993; 121: 565–578. 16. Rachow S, Zorn-Kruppa M, Ohnemus U, et al. Occludin is involved in adhesion, apoptosis, differentiation and Ca2+ homeostasis of human keratinocytes: implications for tumorigenesis. PLoS One 2013; 8: e55116. J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
Flii regulation of skin barrier function
17. Shang X, Lin X, Alvarez E, et al. Tight junction proteins claudin3 and claudin-4 control tumor growth and metastases. Neoplasia 2012; 14: 974–985. 18. Webb PG, Spillman MA, Baumgartner HK. Claudins play a role in normal and tumor cell motility. BMC Cell Biol 2013; 14: 19. 19. Hartsock A, Nelson WJ. Adherens and tight junctions: structure, function and connections to the actin cytoskeleton. Biochim Biophys Acta 2008; 1778: 660–669. 20. Kopecki Z, Cowin AJ. Flightless I: an actin-remodelling protein and an important negative regulator of wound repair. Int J Biochem Cell Biol 2008; 40: 1415–1419. 21. Kopecki Z, Ruzehaji N, Turner C, et al. Topically applied flightless I neutralizing antibodies improve healing of blistered skin in a murine model of epidermolysis bullosa acquisita. J Invest Dermatol 2013; 133: 1008–1016. 22. Campbell HD, Fountain S, McLennan IS, et al. Flii, a gelsolinrelated cytoskeletal regulator essential for early mammalian embryonic development. Mol Cell Biol 2002; 22: 3518–3526. 23. Lin CH, Waters JM, Powell BC, et al. Decreased expression of Flightless I, a gelsolin family member and developmental regulator, in early-gestation fetal wounds improves healing. Mamm Genome 2011; 22: 341–352. 24. Cowin A, Adams D, Strudwick X, et al. Flightless I deficiency enhances wound repair by increasing cell migration and proliferation. J Pathol 2007; 211: 572–581. 25. Lee YH, Stallcup MR. Interplay of Fli-I and FLAP1 for regulation of β-catenin dependent transcription. Nucleic Acids Res 2006; 34: 5052–5059. 26. Wang T, Chuang TH, Ronni T, et al. Flightless I homolog negatively modulates the TLR pathway. J Immunol 2006; 176: 1355–1362. 27. Li J, Yin HL, Yuan J. Flightless-I regulates proinflammatory caspases by selectively modulating intracellular localization and caspase activity. J Cell Biol 2008; 181: 321–333. 28. Kopecki Z, Arkell RM, Strudwick XL, et al. Overexpression of the Flii gene increases dermal–epidermal blistering in an autoimmune ColVII mouse model of epidermolysis bullosa acquisita. J Pathol 2011; 225: 401–413. 29. Cowin AJ, Adams DH, Strudwick XL, et al. Flightless I deficiency enhances wound repair by increasing cell migration and proliferation. J Pathol 2007; 211: 572–581. 30. Kopecki Z, Arkell R, Powell BC, et al. Flightless I regulates hemidesmosome formation and integrin-mediated cellular adhesion and migration during wound repair. J Invest Dermatol 2009; 129: 2031–2045. 31. Kopecki Z, O’Neill GM, Arkell RM, et al. Regulation of focal adhesions by flightless I involves inhibition of paxillin phosphorylation via a Rac1-dependent pathway. J Invest Dermatol 2011; 131: 1450–1459. 32. Geary SM, Cowin AJ, Copeland B, et al. The role of the tetraspanin CD151 in primary keratinocyte and fibroblast functions: implications for wound healing. Exp Cell Res 2008; 314: 2165–2175. 33. Majoul I, Shirao T, Sekino Y, et al. Many faces of drebrin: from building dendritic spines and stabilizing gap junctions to shaping neurite-like cell processes. Histochem Cell Biol 2007; 127: 355–361. 34. Thomsen N, Chappell A, Ali RG, et al. Mouse strains for the ubiquitous or conditional overexpression of the Flii gene. Genesis 2011; 49: 681–688. 35. Hardman MJ, Sisi P, Banbury DN, et al. Patterned acquisition of skin barrier function during development. Development 1998; 125: 1541–1552.
Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
551
36. Kirschner N, Rosenthal R, Furuse M, et al. Contribution of tight junction proteins to ion, macromolecule, and water barrier in keratinocytes. J Invest Dermatol 2013; 133: 1161–1169. 37. Yuki T, Haratake A, Koishikawa H, et al. Tight junction proteins in keratinocytes: localization and contribution to barrier function. Exp Dermatol 2007; 16: 324–330. 38. Sitaru C, Mihai S, Otto C, et al. Induction of dermal–epidermal separation in mice by passive transfer of antibodies specific to type VII collagen. J Clin Invest 2005; 115: 870–878. 39. Davy DA, Ball EE, Matthaei KI, et al. The flightless I protein localizes to actin-based structures during embryonic development. Immunol Cell Biol 2000; 78: 423–429. 40. Davy DA, Campbell HD, Fountain S, et al. The flightless I protein colocalizes with actin- and microtubule-based structures in motile Swiss 3 T3 fibroblasts: evidence for the involvement of PI 3-kinase and Ras-related small GTPases. J Cell Sci 2001; 114: 549–562. 41. Kopecki Z, O’Neill GM, Arkell R, et al. Regulation of focal adhesions by Flightless I involves inhibition of paxillin phosphorylation via a Rac1-dependent pathway. J Invest Dermatol 2011; 131: 1450–1459. 42. Segre JA. Epidermal barrier formation and recovery in skin disorders. J Clin Invest 2006; 116: 1150–1158. 43. Yoshida K, Yokouchi M, Nagao K, et al. Functional tight junction barrier localizes in the second layer of the stratum granulosum of human epidermis. J Dermatol Sci 2013; 71: 89–99. 44. Xu J, Kausalya PJ, Phua DC, et al. Early embryonic lethality of mice lacking ZO-2, but not ZO-3, reveals critical and nonredundant roles for individual zonula occludens proteins in mammalian development. Mol Cell Biol 2008; 28: 1669–1678. 45. Feldmeyer L, Huber M, Fellmann F, et al. Confirmation of the origin of NISCH syndrome. Hum Mutat 2006; 27: 408–410. 46. Hadj-Rabia S, Baala L, Vabres P, et al. Claudin-1 gene mutations in neonatal sclerosing cholangitis associated with ichthyosis: a tight junction disease. Gastroenterology 2004; 127: 1386–1390. 47. Yamamoto T, Kurasawa M, Hattori T, et al. Relationship between expression of tight junction-related molecules and perturbed epidermal barrier function in UVB-irradiated hairless mice. Arch Dermatol Res 2008; 300: 61–68. 48. Claudianos C, Campbell HD. The novel flightless-I gene brings together two gene families, actin-binding proteins related to gelsolin and leucine-rich-repeat proteins involved in Ras signal transduction. Mol Biol Evol 1995; 12: 405–414. 49. Harhaj NS, Antonetti DA. Regulation of tight junctions and loss of barrier function in pathophysiology. Int J Biochem Cell Biol 2004; 36: 1206–1237. 50. Ivanov AI, Parkos CA, Nusrat A. Cytoskeletal regulation of epithelial barrier function during inflammation. Am J Pathol 2010; 177: 512–524. 51. Goshima M, Kariya K, Yamawaki-Kataoka Y, et al. Characterization of a novel Ras-binding protein Ce-FLI-1 comprising leucinerich repeats and gelsolin-like domains. Biochem Biophys Res Commun 1999; 257: 111–116. 52. Mohammad I, Arora PD, Naghibzadeh Y, et al. Flightless I is a focal adhesion-associated actin-capping protein that regulates cell migration. FASEB J 2012; 26: 3260–3272. 53. Enikanolaiye A, Lariviere N, Troy TC, et al. Involucrin–claudin-6 tail deletion mutant (C206) transgenic mice: a model of delayed epidermal permeability barrier formation and repair. Dis Models Mechan 2010; 3: 167–180.
J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
552
Z Kopecki et al
SUPPLEMENTARY MATERIAL ON THE INTERNET The following supplementary material may be found in the online version of this article: Supplementary materials and methods Figure S1. Flii levels are increased during skin development. Figure S2. Flii over-expression results in decreased keratin 1 levels during epidermal stratification. Figure S3. Decreasing Flii expression increases involucrin levels in embryonic mouse skin. Figure S4. Flii over-expression results in increased filaggrin levels early during embryonic skin development. Figure S5. Flii over-expression does not affect the formation of the cornified envelope.
Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
ORIGINAL PAPER
Flightless I over-expression impairs skin barrier development, function and recovery following skin blistering Zlatko Kopecki,1,2* Gink N Yang,1 Ruth M Arkell,3 Jessica E Jackson,1 Elizabeth Melville,1 Hiroaki Iwata,4 Ralf J Ludwig,4 Detlef Zillikens,4 Dedee F Murrell5 and Allison J Cowin1,2 1
Centre for Regenerative Medicine, Mawson Institute, University of South Australia, Adelaide, Australia Women’s and Children’s Health Research Institute, Adelaide, Australia 3 Research School of Biological Sciences, Australian National University, Canberra, Australia 4 Department of Dermatology, University of Lubeck, Germany 5 Department of Dermatology, St George Hospital, University of New South Wales, Sydney, Australia 2
*Correspondence to: Z Kopecki, Centre for Regenerative Medicine, Mawson Institute, University of South Australia. e-mail: [email protected]
Abstract Development of an intact epidermis is critical for maintaining the integrity of the skin. Patients with epidermolysis bullosa (EB) experience multiple erosions, which breach the epidermal barrier and lead to increased microbial colocalization of wounds, infections and sepsis. The cytoskeletal protein Flightless I (Flii) is a known regulator of both development and wound healing. Using Flii +/− , WT and Flii Tg/Tg mice, we investigated the effect of altering Flii levels in embryos and adult mice on the development of the epidermal barrier and, consequently, how this affects the integrity of the skin in EB. Flii over-expression resulted in delayed formation of the epidermal barrier in embryos and decreased expression of tight junction (TJ) proteins Claudin-1 and ZO-2. Increased intercellular space and transepidermal water loss was observed in FliiTg/Tg adult mouse skin, while FliiTg/Tg keratinocytes showed altered TJ protein localization and reduced transepithelial resistance. Flii is increased in the blistered skin of patients with EB, and over-expression of Flii in experimental EBA showed impaired Claudin-1 and -4 TJ protein expression and delayed recovery of functional barrier post-blistering. Immunoprecipitation confirmed Flii associated with TJ proteins and in vivo actin assays showed that the effect of Flii on actin polymerization underpinned the impaired barrier function observed in FliiTg/Tg mice. These results therefore demonstrate an important role for Flii in the development and regulation of the epidermal barrier, which may contribute to the impaired healing and skin fragility of EB patients. Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.
Keywords: Flii; epidermolysis bullosa; skin barrier; epidermis
Received 18 September 2013; Revised 10 December 2013; Accepted 20 December 2013
No conflicts of interest were declared.
Introduction An intact epidermis is critical for maintaining the integrity and function of healthy skin. This physical barrier consists of the stratum corneum, cell–cell tight junctions (TJ) and associated cytoskeletal proteins [1]. Patients with epidermolysis bullosa (EB) experience recurrent skin blistering, leading to disruption or loss of the protective barrier, increased microbial wound colocalization and the development of chronic infected wounds [2,3]. Accordingly, sepsis is the leading cause of death in infants with EB [4]. During development and epidermal wound healing, TJ proteins are expressed well before stratum corneum formation [5,6]. Similarly to wound healing, blistering results in TJ protein redistribution in marginal cells [7,8] and upregulation in the hyperproliferative zone surrounding the blister. An initial protective barrier is first formed by cells re-epithelializing the blister wound, well before Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
stratum corneum formation [8–10]. Establishment of an intact barrier contributes to skin strength, while the quick recovery of the barrier post-blistering is critical for blister healing. As the complex response of epidermal cells to barrier disruption may exacerbate EB, treatments aimed at aiding barrier re-establishment or dampening the epidermal stress response may improve the healing observed in EB patients. TJs are intercellular junctions which form the paracellular barrier and consist of occludin, claudins (Cldns) and junctional adhesion molecules linked to the actin cytoskeleton through scaffolding plaque proteins; zonula occludens (ZO)-1, -2, -3, MUPP-1 and cingulin [11]. Different studies have shown the importance of Cldns and ZO-1 and -2 in the formation of TJ strands and paracellular barrier functions [12,13]. Protein–protein interactions between Cldns and scaffolding plaque proteins provide the link to the actin cytoskeleton allowing the transduction of the J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
542
regulatory signals to and from TJs [14] and highlight the diverse role of TJs in regulating adhesion, signalling, differentiation, proliferation, migration and tissue permeability [5]. TJ proteins were initially implicated in regulating epithelial motility and polarity during wound healing [15], and more recent studies have shown a role for TJ proteins in regulating keratinocyte motility during wound healing and growth, motility and metastasis during tumorigenesis [16–18]. Actin remodelling is required for these cellular functions, the assembly and maturation of TJs and for establishing the epidermal barrier [19]. Flightless I (Flii) is a member of the Gelsolin family of actin remodelling proteins and a nuclear receptor co-activator with unique roles in regulating cell adhesions and intracellular signalling [20,21]. Homozygous knockout of Flii is embryonic lethal, highlighting the role of Flii as an essential regulator of fetal development [22]. Flii expression is increased in the epidermis during development and epidermal stratification [23]. Different studies have shown that Flii is capable of regulating cellular proliferation and migration [24], innate immunity through its downregulation of the Toll-like receptor pathway [25] and NF-κB signalling [26], and cytokine production via caspase activation and IL-1β maturation [27]. Flii is increased in the blistered skin of EB patients [28], and levels of Flii are significantly higher in human blister fluid compared to acute wound fluid [21]. Additionally, Flii affects TGFβ signalling, collagen assembly and fibroblast contraction, leading to increased dermal–epidermal blistering [21,29]. Blisterinduced barrier disruption leads to cytokine production [10] and keratinocyte proliferation and differentiation to restore the epidermal barrier and assemble new TJs. Impairment of these processes results in the development of the inflamed chronic non-healing wounds seen in EB patients. Previous studies have described the Flii effect on keratinocyte and fibroblast cellular adhesion structures [30,31]. Flii regulates transmembrane proteins, including CD151 [30] and Drebrin1, which are important in the cellular adhesion [32] and stabilization of TJs, respectively [33]. In addition, Flii has been identified as one of the 900 proteins present in TJ protein complexes [10]. In the present study we have investigated the effect of Flii on TJ protein expression and determined its effect on skin barrier development, function and recovery following blistering.
Materials and methods Animal studies Mice were maintained according to Australian Standards for Animal Care under protocols approved by The Australian National University Animal Ethics Committee and the Child, Youth and Women’s Health Service Animal Ethics Committee. All strains were BALB/c-congenic and were maintained as homozygous colonies or by continuous backcrossing to BALB/c animals. Wild-type controls were obtained Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
Z Kopecki et al
from BALB/c inbred litters. The murine alleles of Flii used in this study were: 1. A targeted null allele of the murine Flightless I gene (Flii ): Flii tm1Hdc (MGI:2179825) [22]. Animals homozygous null for Flii are embryonic lethal, but heterozygous carriers of the null allele are viable and fertile [22]. A heterozygous carrier of this allele is written as Flii+/− . 2. Tg(FLII)2Hdc (MGI:4939366) [24], a transgenic strain expressing exogenous human flightless I (FLII ), due to the insertion of a 17.8 kb fragment derived from human cosmid clone c110H8 containing all FLII introns and exons and the endogenous promoter (see Figure 1 [34] for extent of the 17.8 kb transgene). Mice homozygous for the transgene were used in this study and had two copies of the mouse Flii gene and two copies of the human FLII transgene (Flii+/+ ; FLIITg/Tg ) and had elevated levels of Flii protein relative to wild-type animals [34]. They are denoted throughout this article as FliiTg/Tg [34]. Embryonic (E15.5, E16.5, E17.5, E18.5 and E19.5) and adult (6 weeks) skin collected from Flii+/− , WT and FliiTg/Tg mice was assessed for expression and localization of TJ proteins, including Cldn1, -4, occludin, ZO-1 and -2, using confocal microscopy as described in Supplementary materials and methods (see supplementary material). Primary keratinocytes from adult Flii+/− , WT and FliiTg/Tg mice were isolated as previously described [30] and used in transepithelial resistance and paracellular tracer flux assays, while the HaCaT keratinocyte cell line was used in co-immunoprecipitation experiments, as detailed in Supplementary materials and methods (see supplementary material). Epidermolysis bullosa acquisita (EBA) was induced by injecting 3–4 week-old wild-type (WT), Flii+/− or FliiTg/Tg mice with rabbit anti-mouse ColVII antibody (0.3 mg/g body weight) subcutaneously every second day for 10 days, as previously described [28]. The mice were examined daily for evidence of cutaneous lesions and the extent of skin blistering was assessed daily. Mice progressively develop skin blistering, with the maximum number of blisters peaking at day 12 following the initial injection. The mice were euthanized and blistered skin collected at day 16 of the experiment, when most of the blisters were undergoing healing and repair. The effect of skin blistering on the localization and expression of TJ proteins during healing was analysed using immunohistochemistry, as described in Supplementary materials and methods (see supplementary material).
In vivo permeability assays In order to study the effect of differential Flii expression on the skin barrier during embryonic development and epidermal stratification, embryos of Flii+/− , wildtype and FliiTg/Tg mice at days 15.5, 16.5, 17.5, 18.5 J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
Flii regulation of skin barrier function
A
B
543
dye (Sigma Aldrich, Australia), following described protocols [35]. After extensive destaining in PBS, images of the embryos were taken, with the blue colour revealing non-functional skin barrier. Images of the embryos were used to quantify the intensity of blue staining, using Image ProPlus v 5.1 (Media Cybernetics Inc.). Materials and methods for reagents and antibodies, ultrastructural analysis, immunofluorescence analysis, cornified envelope extraction, calcium switch assay, western blotting, transepithelial resistance, paracellular tracer flux assay, co-immunoprecipitation, in vivo actin assay and statistical analysis may be found in Supplementary materials and methods (see supplementary material).
Results Over-expression of Flii results in delayed development of skin barrier C
Figure 1. Over-expression of Flii results in delayed development of the skin barrier and increased TEWL. (A) Epidermal permeability assay using toluidine blue staining on E15.5–E19.5 Flii+/− , WT and FliiTg/Tg mouse embryos revealed a defective outside–in permeability barrier at E17.5 in FliiTg/Tg embryos (n = 6). (B) The intensity of the blue staining was quantified using Image ProPlus software, revealing significantly increased permeability at E16.5 and E17.5 in FliiTg/Tg embryos (n = 6). (C) TEWL measurements on E15.5–E19.5 Flii+/− , WT and FliiTg/Tg mouse embryos revealed significantly increased water loss in FliiTg/Tg embryos during epidermal stratification (n = 6). Mean ± SEM; *p < 0.05
and 19.5 dpc were used for TEWL measurement and the ’outside–in’ permeability assay. Permeability of the skin barrier in the embryonic skin was initially assessed by measurements of trans-epidermal water loss (TEWL), using a hand-held vapometer (Delfin Technologies, Finland), following the manufacturer’s instructions. Embryos were subsequently killed and used in the ’outside–in’ permeability assay. Permeability of the skin barrier was assessed in killed embryos by 1 min dehydration in methanol/PBS baths (25%, 50% and 75% methanol/PBS), following by 1 min in 100% methanol. The embryos were then rehydrated in the same series of methanols for 1 min each and stained for 30 min in 0.1% toluidine blue O Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
Using embryonic and adult skin of Flii heterozygous (Flii+/− ), wild-type (WT) and Flii transgenic (FliiTg/Tg ) mice, we investigated the localization and expression of Flii during development. Flii expression increased during skin development, with predominant staining in the developing epidermis and minimal staining observed within fibroblasts of the dermis (see supplementary material, Figure S1A, B). FliiTg/Tg mice have increased Flii nuclear staining in the epidermis compared to Flii+/− and WT counterparts (Figure S1A). Increased Flii levels in the FliiTg/Tg mouse skin were confirmed using western blotting analysis (see supplementary material, Figure S1A–C). To investigate the effect of Flii on the development of the epidermal barrier, the toluidine blue permeability assay was performed on E15.5–E.19.5 Flii+/− , WT and FliiTg/Tg mouse embryos. WT and Flii+/− mouse embryos acquired a functional ’outside–in’ permeability barrier by E17.5 (Figure 1A). In contrast, Flii over-expressing embryos still show penetration of the dye at E17.5 but not at E18.5, indicating that the over-expression of Flii results in delayed development of the epidermal barrier (Figure 1A). Quantifying the intensity of blue staining revealed that FliiTg/Tg embryos have significantly increased permeability at E16.5 and E17.5 compared to WT and Flii+/− counterparts (Figure 1B). In addition, delayed development of an epidermal barrier in FliiTg/Tg mice was confirmed using TEWL measurements, which showed significantly increased water loss in FliiTg/Tg embryos during E16.5–E19.5 when compared to both Flii+/− and WT embryos (Figure 1B).
Flii affects the expression of Cldn1 and ZO-2 in embryonic skin during development In order to ascertain the effect of Flii on epithelial barrier development the expression of structural TJ J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
544
Z Kopecki et al
A
B
C
D
E
Figure 2. FliiTg/Tg embryos had significantly decreased expression of TJ proteins during development. (A) Immunolocalization of Cldn1, -4, occludin, ZO-1 and -2 in E17.5 epidermis of Flii+/− , WT and FliiTg/Tg embryos. (B, C) Graphical analysis of Cldn1 and ZO-2 expression in E17.5 epidermis of Flii+/− , WT and FliiTg/Tg embryos (n = 6). (D, E) Western blot analysis of total skin lysates from E17.5 and E19.5 Flii+/− , WT and FliiTg/Tg embryos revealed decreased expression of TJ proteins in FliiTg/Tg mouse embryos, namely ZO-2 (at E17.5), Cldn1 (at E17.5 and E19.5), Cldn4 (at E17.5) and occludin (at E19.5). Figure is representative of two independent experiments (n = 6). Magnification = ×20; scale bar = 500 µm; mean ± SEM; *p < 0.05.
proteins (Cldn1, Cldn4, occludin) and adaptor signalling TJ plaque proteins (ZO-1, -2) was examined in E17.5 and E19.5 Flii+/− , WT and FliiTg/Tg mouse skin (Figure 2A–E). No differences in the localization of TJ proteins were observed in the different mice genotypes; however, the effect of Flii on the expression of TJ proteins Cldn1 and ZO-2 was pronounced. While Flii+/− mouse embryonic skin showed increased expression of Cldn1at E17.5 compared to WT and FliiTg/Tg counterparts (Figure 2B), over-expression of Flii resulted in significantly decreased expression of both Cldn1 (Figure 2B) and ZO-2 (Figure 2C) compared to WT and Flii+/− mouse embryos. Western blot analysis confirmed the effect of Flii on Cldn1 and Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
ZO-2 expression and, interestingly, while Flii overexpression resulted in decrease of ZO-2 during development, Cldn1 expression was decreased at both E17.5 and E19.5 (Figure 2D, E). To investigate whether an alteration of cell differentiation was involved in the delayed skin barrier observed in FliiTg/Tg mice, expression of keratin 1, involucrin and filaggrin differentiation markers was examined in E17.5, E19.5 and adult Flii+/− , WT and FliiTg/Tg mouse skin (see supplementary material, Figures S2–S4). Reduction of Flii resulted in increased involucrin during early stratification (E17.5) and keratin 1 during development and in adult skin when compared to WT counterparts (see supplementary material, Figures S2, S3); in contrast, J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
Flii regulation of skin barrier function
545
Figure 3. Altered expression of Flii gene in adult mouse skin resulted in impaired formation of the skin barrier, with increased TJ intercellular space and altered cornfied envelope structure. (A) TEM showed impaired skin barrier formation in FliiTg/Tg adult mouse skin, with a significantly increased TJ intercellular space. No difference was observed in lamellar body extrusions, keratohyalin granules or the structure of the desmosome. Magnification = ×13 000; scale bar = 500 nm; n = 4; mean ± SEM; *p > 0.05.
over-expression of Flii led to decreased keratin 1 and increased filaggrin expression at E17.5 when compared to WT counterparts (Figures S3, S4).
Flii over-expression impairs tight junction formation in adult skin Using transmission electron microscopy, we investigated the effect of Flii on skin barrier formation using adult Flii+/− , WT and FliiTg/Tg mouse skin. No effect on lamellar body localization, keratin granules or desmosomes and TJ structure and numbers were observed in stratum granulosum between different genotypes; however, Flii Tg/Tg mouse skin showed significantly larger TJ intercellular space between contact sites of intramembrane fibrils of TJ, suggesting increased epidermal permeability (Figure 3A). Examining the effect of Flii over-expression on cornified envelope formation in adult skin showed no significant difference from WT counterparts (see supplementary material, Figure S5).
Over-expression of Flii alters expression and localization of TJ proteins in cultured adult keratinocytes To analyse the TJ protein cellular localization and expression in response to differential Flii levels, keratinocytes were isolated from adult Flii+/− , WT and FliiTg/Tg mouse skin. Cells were switched to a high-Ca2+ medium to induce TJ formation. Cellular TJ protein localization and expression were examined over a 6 day period. Immunostaining for the Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
transmembrane proteins Cldn1, Cldn4 and occludin and adaptor proteins ZO-1 and -2, using confocal microscopy, revealed no difference between Flii+/− and WT keratinocytes; however, differences in FliiTg/Tg keratinocytes were observed at day 6 following the Ca2+ switch (Figure 4A). Flii over-expression resulted in distorted cell junction outlines, decreased Cldn1 expression at cell contact points, diffuse pattern and decreased expression of ZO-1 and -2 adaptor proteins, reduced staining of occludin and membrane duplication indicative of intercellular spaces and impaired TJ formation (Figure 4A). The effect of Flii overexpression on impaired TJ formation was also confirmed, using western blot analysis of Ca2+ -treated cells, revealing decreased expression of Cldn1 at day 6 and occludin at days 3 and 6 following Ca2+ switch, and decreased Cldn4 and ZO-1 expression at day 6 following Ca2+ switch when compared to WT counterparts (Figure 4B).
Effect of Flii over-expression on TJ barrier function can be rescued following treatment with Flii-neutralizing antibodies Decreased expression and altered localization of TJ proteins has previously been reported to affect TJ barrier function in cultured keratinocytes [36,37]. The effect of Flii on complete transepithelial resistance (TER), including paracellular resistance of TJs and transcellular resistance of apical and basal membranes, was investigated using Flii+/− , WT and FliiTg/Tg keratinocytes in a transepithelial resistance assay, following Ca2+ switch. No significant difference was J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
546
Z Kopecki et al
A
B
Figure 4. Localization and expression of TJ proteins in adult Flii+/− , WT and FliiTg/Tg cells. (A) IHC localization (day 6 post-Ca2+ switch) and (B) protein expression (days 1– 6 post-Ca2+ switch) of TJ proteins in cultured keratinocytes. Impaired formation of TJ was observed in FliiTg/Tg keratinocytes, with decreased Cldn1 expression at cell contact points, diffuse pattern and decreased expression of ZO-1 and -2 adaptor proteins, reduced staining of occludin and membrane duplication at day 6 post-Ca2+ switch. In addition, FliiTg/Tg keratinocytes showed decreased expression of Cldn1 and occludin at days 3 and 6 post-Ca2+ switch and decreased Cldn4 and ZO-1 expression at day 6 post-Ca2+ switch. Figure is representative of two independent experiments. For (a, b) n = 3; magnification = ×60; scale bar = 20 µm.
observed in TER between Flii+/− and WT keratinocytes over a 14 day period (Figure 5A). In contrast, Flii over-expression delayed the formation of TER with significantly reduced TER at days 2–6 compared to both Flii+/− and WT keratinocytes, with the highest reductions of 63 ± 3% and 59 ± 7%, respectively, observed at day 4 (Figure 5A). In addition, FliiTg/Tg keratinocytes never reached the same TER levels of Flii+/− and WT keratinocytes (Figure 5A). Treatment of FliiTg/Tg keratinocytes with Flii-neutralizing antibody (FnAb; 50 µg/ml) but not control IgG antibody significantly increased TER formation in these cells, similar to levels observed in WT control counterparts (highest increase of 51 ± 2% in TER observed at day 6) (Figure 5B). In order to further characterize the effect of Flii on TJ and size selectivity of paracellular permeability, Flii+/− , WT and FliiTg/Tg keratinocytes were subjected to a paracellular tracer flux assay following Ca2+ switch. No significant difference Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
was observed in barrier function and paracellular permeability of 4 kDa (Figure 5C) and 10 kDa (Figure 5D) FITC dextrans in response to varying Flii levels. Treatment of WT and FliiTg/Tg keratinocytes with FnAb (50 µg/ml) or control IgG did not alter keratinocyte paracellular permeability to 4 or 10 kDa FITC dextrans (data not shown).
Flii affects the recovery of skin barrier following skin blistering Flii is increased in blistered skin and contributes to excessive epidermal–dermal separation during skin blistering, affecting both TEWL and healing of blisters [21,28]. In order to investigate the effect of differential Flii on skin barrier recovery follwing skin blistering, Flii+/− , WT and FliiTg/Tg mice were used in an established model of EBA previously described [21,38]. The effect of Flii on recovery of skin barrier J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
Flii regulation of skin barrier function
547
A
B
C
D
Figure 5. FliiTg/Tg keratinocytes had impaired TJ barrier function, which could be recovered using FnAb. (A) TER of cultured Flii+/− , WT and FliiTg/Tg keratinocytes following Ca2+ switch was determined over a period of 14 days. Flii+/− and WT keratinocytes established a TER of up to 15 000 /cm2 by day 6 of the experiment. In contrast, the TER in FliiTg/Tg cells was significantly delayed; however, it was significantly increased in response to FnAb treatment (B). Paracellular diffusion of 4 kDa (C) and 10 kDa (D) FITC–dextran through Flii+/− , WT and FliiTg/Tg keratinocyte monolayers was determined for 144 h after Ca2+ switch; data are expressed as a percentage of 0 h values. Flii+/− , WT and FliiTg/Tg keratinocytes formed similar permeability barriers for 4 and 10 kDa FITC–dextran. For (a, b) n = 6; mean ± SEM; * p < 0.05
function and re-establishment of new TJs was investigated post-blistering, using immunohistochemistry for structural TJ proteins Cldn1 and -4 in the stratum granulosum of the newly formed epidermis covering the healing blisters. Compared to Flii+/− EBA and WT EBA mouse skin, FliiTg/Tg EBA mouse skin revealed diffuse and significantly reduced expression of both Cldn1 and -4, suggesting that high levels of Flii affect TJ protein formation and may contribute to the impaired healing of blisters and increased skin fragility, with impaired skin barrier function in FliiTg/Tg EBA mice (Figure 6A, B).
The strongest association was with structural TJ proteins Cldn1 and -4 and adaptor protein ZO-2, while Flii was not observed to co-immunoprecipitate with occludin (Figure 7A). To investigate the effect of Flii on actin remodelling during TJ formation, Flii+/− , WT and FliiTg/Tg keratinocytes were used in an in vivo actin assay following Ca2+ differentiation. In contrast to Flii+/− and WT keratinocytes, which showed equal F-actin:G-actin distribution, FliiTg/Tg keratinocytes showed no G-actin (Figure 7B). Treatment of FliiTg/Tg keratinocytes with FnAb (50 µg/ml) or cytochalasin D (positive control) restored the Factin:G-actin distribution, favouring improved TJ barrier formation and function (Figure 7B).
Flii associates with TJ proteins and affects actin remodelling during TJ formation To gain a mechanistic insight into the role of Flii in TJ structure and function, immunoprecipitates and cell lysates from HaCaT keratinocytes were analysed, using western blotting following Ca2+ switch-induced formation of TJs in vitro. Flii, but not the family member Gelsolin, was found to co-immunoprecipitate with structural TJ proteins Cldn1, -4 and -6 and adaptor TJ plaque proteins ZO-1 and -2 (Figure 7A). Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
Discussion Flii colocalizes with actin and microtubule based structures and is an important conserved regulator of cellular adhesion with regulatory roles during development and wound healing [22,30,39–42]. Homozygous Flii knockdown in mice results in irregular actin organization and lethality early in embryogenesis, while J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
548
Z Kopecki et al
A
B
Figure 6. Flii affects the recovery of skin barrier formation and function post-blistering. (A, B) IHC localization of Cldn1 and -4 in the epidermis of healing blisters in Flii+/− , WT and FliiTg/Tg EBA mice. Cldn1 and -4 staining was localized to the stratum granulosum, highlighting newly formed TJ post-blistering. FliiTg/Tg mouse blistered skin had a diffuse and significantly reduced expression of both Cldn1 and -4. Magnification = ×40; scale bar = 200 nm; n = 4, mean ± SEM; * p > 0.05.
Flii over-expression results in phenotypically normal mice with thin fragile skin, altered cell adhesions and similar features to those observed in patients with skin-blistering diseases [23,30]. In addition, Flii overexpression in fetal and adult wounds results in a delayed wound-healing response and increased skin blistering in adult mice [23,28,29]. TJ regulation and function is a primary pathophysiological factor in many skin diseases and a contributing factor to skin regeneration [5,43]. Understanding the effect of the regulatory proteins involved in development and/or function of TJs may help in the design of new therapies aimed at ameliorating different clinical skin conditions, including blistering. This study describes the involvement of the cytoskeletal protein Flii during embryonic development, function and recovery of the protective skin barrier in adult mice. We demonstrate that skin barrier development, TJ protein level and localization, as well as TJ function and recovery, are impaired when Flii levels are increased, as is observed in blistered skin. The level of Flii increases during skin development and is significantly elevated in Flii overexpressing mice [23]. Flii over-expression leads to delayed formation of the protective barrier during Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
epidermal stratification, increased transepidermal water loss and decreased expression of structural TJ protein Cldn1, and signalling actin anchoring protein ZO-2 in the embryonic skin. Previous studies have described an important effect of complete knockout of TJ proteins Cldn1 and ZO-2 on skin barrier development and function, resulting in permeable TJs and mouse lethality early after birth and during embryogenesis, respectively [12,44]. Despite the decreased expression of Cldn1 and ZO-2 during embryogenesis, FliiTg/Tg mice show normal phenotype and life expectancy during adulthood, suggesting that other TJ proteins may be providing a compensatory mechanism. In agreement with these findings, complete loss of Cldn1 in humans with neonatal ichthyosis-sclerosing cholangitis syndrome results in a less severe phenotype compared to the phenotype observed in Cldn1−/− mice [45,46]. Furthermore, recent studies using mouse and human keratinocytes reveal that Cldn1 is dispensable for TJ water barrier function, also suggestive of a possible compensatory mechanism [36]. Examining the effect of Flii on cell differentiation revealed a possible keratin 1/filaggrin compensatory mechanism in Flii overexpressing mice during early stratification. No difference in lamellar body localization, keratin granules, J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
Flii regulation of skin barrier function
A
B
Figure 7. Flii effects on actin remodelling and TJ proteins may underpin impaired barrier function in FliiTg/Tg mice. (A) Anti-Cldn1, -4 and -6; ZO-1 and -2 and occludin IP were prepared from HaCaT keratinocytes following Ca2+ differentiation. Flii colocalized with structural (Cldn1, -4 and -6) and adaptor (ZO-1 and -2) TJ proteins. (B) Flii+/− , WT and FliiTg/Tg keratinocytes were used in in vivo actin assay, following Ca2+ differentiation. While Flii+/− and WT keratinocytes had an equal F-actin:G-actin distribution, FliiTg/Tg keratinocytes showed no G-actin. Treatment with Flii neutralizing antibody (FnAb) or cytochalasin D (control) restored the actin distribution in FliiTg/Tg keratinocytes. Data are representative of three independent experiments.
desmosome numbers or cornified envelope formation were observed between Flii over-expressing and wildtype mice, suggesting that the observed delay in skin barrier development, function and recovery is unlikely to be due to effects on cell differentiation, but more likely to relate to the decreased expression of TJ proteins. In addition, the effect of Flii on TJ integrity was further evident, with the significantly increased TJ intercellular space observed in FliiTg/Tg adult skin suggestive of permeable TJs and possible structural and functional effects on skin barrier permeability, which may contribute to the increased skin fragility and blistering observed in these mice [28]. Examining the effect of Flii on the expression and localization of TJ proteins in adult keratinocytes following Ca2+ switch revealed a similar effect of Flii over-expression observed in embryonic skin. Flii over-expression resulted in reduced expression, altered localization and diffuse pattern of both structural and actin-anchoring TJ proteins at cell–cell adherent points. Expression of Cldn1 and occludin was reduced at days 3 and 6 following Ca2+ switch, while expression of Cldn4 and ZO-1 was reduced only at day 6 of the experiment, suggesting a role for Flii in the assembly of TJs. While no effect of Flii on the expression of ZO-2 was observed in adult keratinocytes, Flii’s Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
549
effect on other TJ protein expression and localization during TJ assembly suggests a functional effect of Flii on skin barrier permeability and junctional integrity. Recent studies have described the correlation between TJ protein expression and barrier function, and detailed the distinct TJ protein contribution to both TJ-dependent paracellular resistance and transcellular resistance in cultured keratinocytes [11,36,47]. A direct functional effect of Flii on TJ formation and function, with significantly decreased transepithelial resistance in FliiTg/Tg keratinocytes, was also observed. This could be reversed following treatment with FnAb, suggesting a direct functional effect and a relationship between Flii levels and transepithelial resistance. Flii affected the transcellular resistance of apical and basolateral cell membranes, but no significant differences were observed in the paracellular permeability resistance of TJ. This suggests that Flii affects the flow of water and small ions, but not large molecules. Flii is increased in the blistered skin of patients with different EB subtypes [28], and is present in human blister fluid at higher levels than those observed in human acute wound fluid [21]. We have previously described the negative role of Flii as a contributor to increased dermal–epidermal blistering in a ColVII mouse model of epidermolysis bullosa acquisita (EBA) [28]. To examine the functional effect of Flii on TJ recovery following skin blistering, we examined the expression of structural TJ proteins Cldn1 and -4 during the healing of blisters in Flii+/− , WT and FliiTg/Tg mouse skin, using an EBA mouse model of skin blistering [38]. A decreased and fragmentary expression of structural TJ proteins Cldn1 and -4 at cell–cell borders in the stratum granulosum was observed in healing blistered skin in FliiTg/Tg EBA mice compared to WT and Flii+/− counterparts. These findings suggest that Flii, through its effects on TJ protein expression and localization, affects the recovery of the skin barrier following skin blistering. In agreement with previous studies describing increased skin blistering in FliiTg/Tg EBA mice [28] and decreased TEWL in WT EBA mice treated with FnAb during and following skin blistering [21], these findings suggest that the Flii effect on skin barrier function and recovery contributes to both the increased blistering and the delayed healing of blisters observed in FliiTg/Tg EBA mice [28]. Flii has previously been described as an important regulator of cellular adhesions and a modulator of both actin remodelling and cell signalling, with roles in actin polymerization and nuclear receptor coactivation, respectively [20,48]. Considering its identification as one of the proteins present in the TJ protein complexes [10], we investigated the mechanism behind its effects on TJ formation, function and recovery. Here, we show that Flii does not associate with occludin during TJ formation; however, its possible association with Cldn1, -4 and -6 structural TJ proteins and the actin-linking proteins ZO-1 and -2 may account for the impaired skin barrier function observed in Flii over-expressing mice. TJ permeability can be J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
550
regulated directly, through modification of TJ proteins, or indirectly, through effects on the actin cytoskeleton [49]. Inhibition of F-actin turnover has previously been shown to induce disassembly and internalization of TJs [50], highlighting the relationship between skin barrier function and actin cytoskeleton remodelling. Flii possesses calcium-independent G-actin binding activity as well as F-actin binding and severing activities, and has been shown to actively regulate actin polymerization, leading to altered functional cell outcomes, including adhesion and migration [51,52]. Previous studies have shown that Flii colocalizes with actin and macrotubular structures [40] and has the ability to regulate turnover of cell adhesion structures [31]. Surprisingly, we now demonstrate that Flii alters the Factin:G-actin ratio during TJ formation. Interestingly, the altered F-actin:G-actin ratio seen in FliiTg/Tg keratinocytes during TJ formation can be restored following treatment with FnAb, suggesting that modulation of Flii levels may be beneficial to improved healing and recovery of skin barrier following injury. While the mechanisms underpinning Flii ability to regulate the F-actin:G-actin ratio and/or interact with number of actin monomer-stabilizing or actin-sequestering proteins are yet to be identified, this finding is of potential importance in our understanding of the effects of Flii on cellular function. Flii may, therefore, directly interact with structural and adaptor TJ proteins affecting the formation of the TJ protein complex and association with the actin cytoskeleton, hence affecting skin barrier development, function and recovery following injury. The exact mechanism behind the Flii direct or indirect effect on TJs during skin barrier development, function and recovery are yet to be identified; however, our findings provide novel insight into the relationship between Flii and skin barrier function. Flii promotes keratinocyte permeability during development, in adult mice in vivo and keratinocyte monolayers in vitro, by regulating the expression and organization of TJ proteins and the actin cytoskeleton. These studies provide new insight into the relationship between Flii and skin development and repair following injury.
Acknowledgements This work was supported by grants from the Women’s and Children’s Hospital Foundation (to ZK and AJC) and the NHMRC (Grant Nos 626802 to AJC, DFM and RMA and 1003648 to RMA and AJC). AJC is supported by an NHMRC Senior Research Fellowship (Grant No. 1002009). ZK is supported by an NHMRC Early Career Fellowship (Grant No. 1036509). DZ and RJL are supported by the Excellence Cluster Inflammation at Interfaces (Grant No. DFG EXC 306/2) and the Research Training Group ’Modulation of Autoimmunity’ (Grant No. DFG GRK 1727/1). We acknowledge the technical assistance of Nicole Thomsen and Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
Z Kopecki et al
the Dystrophic Epidermolysis Bullosa Research Association of South Australia for their support.
Author contributions ZK and AJC conceived the experiments; ZK carried out all experiments with the assistance of GNY and JEJ; EM made the Flii neutralizing antibody; RMA provided the Flii genetic mice; DZ, RJL, HI and DFM provided ColVII antibody for the induction of skin blistering; ZK and AJC wrote the manuscript; and all authors contributed to manuscript preparation and approved the final submitted and published versions.
References 1. Koster MI, Roop DR. Mechanisms regulating epithelial stratification. Annu Rev Cell Dev Biol 2007; 23: 93–113. 2. Kopecki Z, Murrell DF, Cowin A. Raising the roof on epidermolysis bullosa (EB): a focus on new therapies. Wound Pract Res 2009; 17: 76–82. 3. van der Kooi-Pol MM, de Vogel CP, Westerhout-Pluister GN, et al. High anti-staphylococcal antibody titers in patients with epidermolysis bullosa relate to long-term colonization with alternating types of Staphylococcus aureus. J Invest Dermatol 2013; 133: 847–850. 4. Fine J-D, Johnson LB, Weiner M, et al. Cause-specific risks of childhood death in inherited epidermolysis bullosa. J Pediat 2008; 152: 276–280, e272. 5. O’Neill CA, Garrod D. Tight junction proteins and the epidermis. Exp Dermatol 2011; 20: 88–91. 6. Herard AL, Zahm JM, Pierrot D, et al. Epithelial barrier integrity during in vitro wound repair of the airway epithelium. Am J Respir Cell Mol Biol 1996; 15: 624–632. 7. Brandner JM, Kief S, Grund C, et al. Organization and formation of the tight junction system in human epidermis and cultured keratinocytes. Eur J Cell Biol 2002; 81: 253–263. 8. Malminen M, Koivukangas V, Peltonen J, et al. Immunohistological distribution of the tight junction components ZO-1 and occludin in regenerating human epidermis. Br J Dermatol 2003; 149: 255–260. 9. Brandner JM, Kief S, Wladykowski E, et al. Tight junction proteins in the skin. Skin Pharmacol Physiol 2006; 19: 71–77. 10. Tang VW. Proteomic and bioinformatic analysis of epithelial tight junction reveals an unexpected cluster of synaptic molecules. Biol Direct 2006; 1: 37. 11. Kirschner N, Haftek M, Niessen CM, et al. CD44 regulates tightjunction assembly and barrier function. J Invest Dermatol 2011; 131: 932–943. 12. Furuse M, Hata M, Furuse K, et al. Claudin-based tight junctions are crucial for the mammalian epidermal barrier: a lesson from claudin-1-deficient mice. J Cell Biol 2002; 156: 1099–1111. 13. Umeda K, Ikenouchi J, Katahira-Tayama S, et al. ZO-1 and ZO-2 independently determine where claudins are polymerized in tightjunction strand formation. Cell 2006; 126: 741–754. 14. Heiskala M, Peterson PA, Yang Y. The roles of claudin superfamily proteins in paracellular transport. Traffic 2001; 2: 93–98. 15. Bement WM, Forscher P, Mooseker MS. A novel cytoskeletal structure involved in purse string wound closure and cell polarity maintenance. J Cell Biol 1993; 121: 565–578. 16. Rachow S, Zorn-Kruppa M, Ohnemus U, et al. Occludin is involved in adhesion, apoptosis, differentiation and Ca2+ homeostasis of human keratinocytes: implications for tumorigenesis. PLoS One 2013; 8: e55116. J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
Flii regulation of skin barrier function
17. Shang X, Lin X, Alvarez E, et al. Tight junction proteins claudin3 and claudin-4 control tumor growth and metastases. Neoplasia 2012; 14: 974–985. 18. Webb PG, Spillman MA, Baumgartner HK. Claudins play a role in normal and tumor cell motility. BMC Cell Biol 2013; 14: 19. 19. Hartsock A, Nelson WJ. Adherens and tight junctions: structure, function and connections to the actin cytoskeleton. Biochim Biophys Acta 2008; 1778: 660–669. 20. Kopecki Z, Cowin AJ. Flightless I: an actin-remodelling protein and an important negative regulator of wound repair. Int J Biochem Cell Biol 2008; 40: 1415–1419. 21. Kopecki Z, Ruzehaji N, Turner C, et al. Topically applied flightless I neutralizing antibodies improve healing of blistered skin in a murine model of epidermolysis bullosa acquisita. J Invest Dermatol 2013; 133: 1008–1016. 22. Campbell HD, Fountain S, McLennan IS, et al. Flii, a gelsolinrelated cytoskeletal regulator essential for early mammalian embryonic development. Mol Cell Biol 2002; 22: 3518–3526. 23. Lin CH, Waters JM, Powell BC, et al. Decreased expression of Flightless I, a gelsolin family member and developmental regulator, in early-gestation fetal wounds improves healing. Mamm Genome 2011; 22: 341–352. 24. Cowin A, Adams D, Strudwick X, et al. Flightless I deficiency enhances wound repair by increasing cell migration and proliferation. J Pathol 2007; 211: 572–581. 25. Lee YH, Stallcup MR. Interplay of Fli-I and FLAP1 for regulation of β-catenin dependent transcription. Nucleic Acids Res 2006; 34: 5052–5059. 26. Wang T, Chuang TH, Ronni T, et al. Flightless I homolog negatively modulates the TLR pathway. J Immunol 2006; 176: 1355–1362. 27. Li J, Yin HL, Yuan J. Flightless-I regulates proinflammatory caspases by selectively modulating intracellular localization and caspase activity. J Cell Biol 2008; 181: 321–333. 28. Kopecki Z, Arkell RM, Strudwick XL, et al. Overexpression of the Flii gene increases dermal–epidermal blistering in an autoimmune ColVII mouse model of epidermolysis bullosa acquisita. J Pathol 2011; 225: 401–413. 29. Cowin AJ, Adams DH, Strudwick XL, et al. Flightless I deficiency enhances wound repair by increasing cell migration and proliferation. J Pathol 2007; 211: 572–581. 30. Kopecki Z, Arkell R, Powell BC, et al. Flightless I regulates hemidesmosome formation and integrin-mediated cellular adhesion and migration during wound repair. J Invest Dermatol 2009; 129: 2031–2045. 31. Kopecki Z, O’Neill GM, Arkell RM, et al. Regulation of focal adhesions by flightless I involves inhibition of paxillin phosphorylation via a Rac1-dependent pathway. J Invest Dermatol 2011; 131: 1450–1459. 32. Geary SM, Cowin AJ, Copeland B, et al. The role of the tetraspanin CD151 in primary keratinocyte and fibroblast functions: implications for wound healing. Exp Cell Res 2008; 314: 2165–2175. 33. Majoul I, Shirao T, Sekino Y, et al. Many faces of drebrin: from building dendritic spines and stabilizing gap junctions to shaping neurite-like cell processes. Histochem Cell Biol 2007; 127: 355–361. 34. Thomsen N, Chappell A, Ali RG, et al. Mouse strains for the ubiquitous or conditional overexpression of the Flii gene. Genesis 2011; 49: 681–688. 35. Hardman MJ, Sisi P, Banbury DN, et al. Patterned acquisition of skin barrier function during development. Development 1998; 125: 1541–1552.
Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
551
36. Kirschner N, Rosenthal R, Furuse M, et al. Contribution of tight junction proteins to ion, macromolecule, and water barrier in keratinocytes. J Invest Dermatol 2013; 133: 1161–1169. 37. Yuki T, Haratake A, Koishikawa H, et al. Tight junction proteins in keratinocytes: localization and contribution to barrier function. Exp Dermatol 2007; 16: 324–330. 38. Sitaru C, Mihai S, Otto C, et al. Induction of dermal–epidermal separation in mice by passive transfer of antibodies specific to type VII collagen. J Clin Invest 2005; 115: 870–878. 39. Davy DA, Ball EE, Matthaei KI, et al. The flightless I protein localizes to actin-based structures during embryonic development. Immunol Cell Biol 2000; 78: 423–429. 40. Davy DA, Campbell HD, Fountain S, et al. The flightless I protein colocalizes with actin- and microtubule-based structures in motile Swiss 3 T3 fibroblasts: evidence for the involvement of PI 3-kinase and Ras-related small GTPases. J Cell Sci 2001; 114: 549–562. 41. Kopecki Z, O’Neill GM, Arkell R, et al. Regulation of focal adhesions by Flightless I involves inhibition of paxillin phosphorylation via a Rac1-dependent pathway. J Invest Dermatol 2011; 131: 1450–1459. 42. Segre JA. Epidermal barrier formation and recovery in skin disorders. J Clin Invest 2006; 116: 1150–1158. 43. Yoshida K, Yokouchi M, Nagao K, et al. Functional tight junction barrier localizes in the second layer of the stratum granulosum of human epidermis. J Dermatol Sci 2013; 71: 89–99. 44. Xu J, Kausalya PJ, Phua DC, et al. Early embryonic lethality of mice lacking ZO-2, but not ZO-3, reveals critical and nonredundant roles for individual zonula occludens proteins in mammalian development. Mol Cell Biol 2008; 28: 1669–1678. 45. Feldmeyer L, Huber M, Fellmann F, et al. Confirmation of the origin of NISCH syndrome. Hum Mutat 2006; 27: 408–410. 46. Hadj-Rabia S, Baala L, Vabres P, et al. Claudin-1 gene mutations in neonatal sclerosing cholangitis associated with ichthyosis: a tight junction disease. Gastroenterology 2004; 127: 1386–1390. 47. Yamamoto T, Kurasawa M, Hattori T, et al. Relationship between expression of tight junction-related molecules and perturbed epidermal barrier function in UVB-irradiated hairless mice. Arch Dermatol Res 2008; 300: 61–68. 48. Claudianos C, Campbell HD. The novel flightless-I gene brings together two gene families, actin-binding proteins related to gelsolin and leucine-rich-repeat proteins involved in Ras signal transduction. Mol Biol Evol 1995; 12: 405–414. 49. Harhaj NS, Antonetti DA. Regulation of tight junctions and loss of barrier function in pathophysiology. Int J Biochem Cell Biol 2004; 36: 1206–1237. 50. Ivanov AI, Parkos CA, Nusrat A. Cytoskeletal regulation of epithelial barrier function during inflammation. Am J Pathol 2010; 177: 512–524. 51. Goshima M, Kariya K, Yamawaki-Kataoka Y, et al. Characterization of a novel Ras-binding protein Ce-FLI-1 comprising leucinerich repeats and gelsolin-like domains. Biochem Biophys Res Commun 1999; 257: 111–116. 52. Mohammad I, Arora PD, Naghibzadeh Y, et al. Flightless I is a focal adhesion-associated actin-capping protein that regulates cell migration. FASEB J 2012; 26: 3260–3272. 53. Enikanolaiye A, Lariviere N, Troy TC, et al. Involucrin–claudin-6 tail deletion mutant (C206) transgenic mice: a model of delayed epidermal permeability barrier formation and repair. Dis Models Mechan 2010; 3: 167–180.
J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
552
Z Kopecki et al
SUPPLEMENTARY MATERIAL ON THE INTERNET The following supplementary material may be found in the online version of this article: Supplementary materials and methods Figure S1. Flii levels are increased during skin development. Figure S2. Flii over-expression results in decreased keratin 1 levels during epidermal stratification. Figure S3. Decreasing Flii expression increases involucrin levels in embryonic mouse skin. Figure S4. Flii over-expression results in increased filaggrin levels early during embryonic skin development. Figure S5. Flii over-expression does not affect the formation of the cornified envelope.
Copyright 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk
J Pathol 2014; 232: 541–552 www.thejournalofpathology.com
