Integrin beta 1 inhibition alleviates the chronic hyperproliferative dermatitis phenotype of SHARPIN-deficient mice

SHARPIN (Shank-Associated RH Domain-Interacting Protein) is a component of the linear ubiquitin chain assembly complex (LUBAC), which enhances TNF-induced NF-κB activity. SHARPIN-deficient (Sharpincpdm/cpdm) mice display multi-organ inflammation and chronic proliferative dermatitis (cpdm) due to TNF-induced keratinocyte apoptosis. In cells, SHARPIN also inhibits integrins independently of LUBAC, but it has remained enigmatic whether elevated integrin activity levels in the dermis of Sharpincpdm/cpdm mice is due to increased integrin activity or is secondary to inflammation. In addition, the functional contribution of increased integrin activation to the Sharpincpdm/cpdm phenotype has not been investigated. Here, we find increased integrin activity in keratinocytes from Tnfr1-/- Sharpincpdm/cpdm double knockout mice, which do not display chronic inflammation or proliferative dermatitis, thus suggesting that SHARPIN indeed acts as an integrin inhibitor in vivo. In addition, we present evidence for a functional contribution of integrin activity to the Sharpincpdm/cpdm skin phenotype. Treatment with an integrin beta 1 function blocking antibody reduced epidermal hyperproliferation and epidermal thickness in Sharpincpdm/cpdm mice. Our data indicate that, while TNF-induced cell death triggers the chronic inflammation and proliferative dermatitis, absence of SHARPIN-dependent integrin inhibition exacerbates the epidermal hyperproliferation in Sharpincpdm/cpdm mice.

Integrins are the major cell adhesion receptors that mediate the interaction of a cell with the surrounding extracellular matrix, including the basement membrane. Binding of integrins to extracellular ligands triggers a conformational change in integrin structure that allows recruitment of a plethora of cellular factors resulting in activation of several signaling pathways [18]. SHARPIN acts as an integrin inhibitor through binding to the integrin cytoplasmic domain, preventing the recruitment of integrin activating proteins and supporting the integrin inactive conformation [19]. Functionally, SHARPIN-mediated integrin inhibition regulates cell adhesion and migration [19], as well as lymphocyte detachment during transmigration [20]. Importantly, SHARPIN plays mutually exclusive roles in regulating integrins and LUBAC such that SHARPIN inhibits integrins independent of LUBAC [21].
Deregulated integrin activity is implicated in many human pathological conditions, including immune diseases, skin blistering, bleeding disorders, and cancer [18]. In the skin, integrin expression is predominantly confined to the basal keratinocytes that anchor the epidermis to the basal lamina [22,23]. Transgenic mice overexpressing integrin beta 1 (Itgb1) in the suprabasal layer of the epidermis (Tg(Itgb1)0869Fmw) exhibit epidermal hyperproliferation, perturbed keratinocyte differentiation and skin inflammation [24], which resembles the Sharpin cpdm/cpdm phenotype. Furthermore, integrin blocking therapies have been shown to alleviate psoriasis in mice [25] and in human patients [26]. These data suggest that increased integrin activation in Sharpin cpdm/cpdm mice may contribute to the dermatitis phenotype. We have previously demonstrated increased Itgb1 activity in the basal layer of the epidermis of Sharpin cpdm/cpdm mice [19]. However, whether this was due to the absence of SHARPIN-mediated integrin inhibition or secondary to the chronic inflammation-driven proliferative dermatitis has remained unclear.
Using Tnfr1 -/-Sharpin cpdm/cpdm double knockout mice we now present data suggesting that SHARPIN acts as an integrin inhibitor in vivo also in the absence of chronic inflammation. In addition, we demonstrate that Itgb1 inhibition with a function blocking antibody alleviates the excessive proliferation and apoptosis observed in Sharpin cpdm/cpdm epidermis, but does not ameliorate the chronic and systemic inflammation in Sharpin cpdm/cpdm mice, suggesting that increased integrin activity in the absence of SHARPIN exacerbates the hyperproliferative skin phenotype in Sharpin cpdm/cpdm mice.
To analyze this observation quantitatively and without the interference of high dermal Itgb1 activity (Fig 1A), we isolated keratinocytes from the epidermal layer of Tnfr1 +/+ Sharpin cpdm/cpdm , Tnfr1 -/-Sharpin +/? and Tnfr1 -/-Sharpin cpdm/cpdm mice, and investigated Itgb1 activity specifically in basal cell layer keratinocytes that also express Integrin alpha 6 (Itga6, also known as CD49f; Fig 1B) (see Methods for details). In order to distinguish basal keratinocytes by flow cytometry, any residual cells expressing lineage markers for leukocytes [CD45, Protein Tyrosine Phosphatase, Receptor Type C (PTPRC)] or endothelial cells [CD31, Platelet And Endothelial Cell Adhesion Molecule 1 (PECAM1)] were gated out by flow cytometry, and cells expressing Itga6 were further analyzed for binding active Itgb1 (9EG7) or total Itgb1 (HMβ1-1) antibodies ( Fig 1C). Quantification of these data showed that the amount of active Itgb1 on the surface of the double knockout Tnfr1 -/-Sharpin cpdm/cpdm basal keratinocytes was indeed significantly increased compared to Tnfr1 -/-Sharpin +/? keratinocytes, although it was lower than in the hyperproliferative Tnfr1 +/+ Sharpin cpdm/cpdm cells (Fig 1D). The total Itgb1 levels were not significantly altered even though total Itgb1 levels seem elevated in Tnfr1 +/+ Sharpin cpdm/cpdm basal keratinocytes ( Fig 1E). Importantly, FACS experiments with primary keratinocytes from four individual mice showed that staining with the 9EG7 antibody results in approximately 6 fold higher signals than isotype control (Panel A and B in S1 Fig). Altogether, these data suggest that increased Itgb1 activity in Sharpin cpdm/cpdm mice is not solely due to chronic inflammation and proliferative dermatitis but that SHARPIN functions as an integrin inhibitor in vivo.
Itgb1 inhibition ameliorates the epidermal hyperproliferation phenotype in Sharpin cpdm/cpdm mice Given that SHARPIN indeed inhibits integrin activity in mouse epidermis (Fig 1), we hypothesized that the increased Itgb1 activity might contribute to the Sharpin cpdm/cpdm dermatitis phenotype. Since epidermal Itgb1 depletion is lethal [28], Sharpin cpdm/cpdm and age-and gendermatched control (Sharpin +/+ or Sharpin +/cpdm ) mice were systemically treated with an Itgb1 function-blocking antibody or with PBS. Tnfr1 -/-Sharpin +/? and Tnfr1 -/-Sharpin cpdm/cpdm mice were not included in this experiment as they display a normal skin phenotype. As expected, the epidermis of Sharpin cpdm/cpdm mice was much thicker than in control treated mice (Fig  2A-2C). Interestingly, blocking Itgb1 function significantly reduced epidermal thickness in Sharpin cpdm/cpdm , but not in control mice (Fig 2A-2C), without significantly affecting dermal thickness (Panel C in S1 Fig). From each animal the epidermal thickness was measured from five individual tissue sections with two measurements per skin section, thus minimizing the chances of recording experimental outliers. Importantly, plotting all data showed equal  Staining for the proliferation marker Ki67 indicated that inhibition of Itgb1 in Sharpin cpdm/cpdm mice reduced the number of proliferating epidermal cells (Fig 2D and 2E), suggesting that Itgb1 inhibition reduces keratinocyte proliferation. To determine if the reduced keratinocyte proliferation is an autonomous effect of Itgb1 inhibition, we cultured Sharpin cpdm/cpdm and Sharpin +/? primary mouse keratinocytes, isolated from 6 weeks old mice ( Fig  2F and 2G; Panel E in S1 Fig), and evaluated the effect of either Itgb1 function blocking antibody or isotype control antibody on cell proliferation (Fig 2H). While Sharpin cpdm/cpdm keratinocytes demonstrated comparable or even higher proliferation rate than Sharpin +/? cells in vitro (Fig 2G), integrin inhibition with the Itgb1 function blocking antibody significantly reduced both Sharpin cpdm/cpdm and Sharpin +/? primary keratinocyte proliferation (Fig 2H), showing that the Itgb1 blocking antibody inhibits keratinocyte proliferation autonomously and suggesting that amelioration of the chronic proliferative dermatitis (Fig 2A-2C) is, at least partly, linked to decreased keratinocyte proliferation.

Itgb1 inhibition does not reduce the chronic inflammation in
Sharpin cpdm/cpdm mice Chronic inflammation in Sharpin cpdm/cpdm mice is characterized by enhanced infiltration of immune cells to many tissues including the skin [1,29], as well as increased numbers of white blood cells in the peripheral blood [30]. Interestingly, leukocyte (CD45 + ; Fig 3A and 3B), macrophage (F4/80 + ; Fig 3C and 3D) and mast cell (Toluidine blue + ; Fig 3E and 3F) infiltration into the dermis of Sharpin cpdm/cpdm mice was not affected by the Itgb1 function-blocking antibody, indicating that inflammation was not reduced. Also, the increased numbers of white blood cells in the peripheral blood of Sharpin cpdm/cpdm mice remained elevated or even increased further upon Itgb1 inhibition (Panel A in S2 Fig), further suggesting that the chronic inflammation in Sharpin cpdm/cpdm mice does not depend on Itgb1.

Elevated keratinocyte apoptosis in Sharpin cpdm/cpdm mouse skin is partially rescued by Itgb1 inhibition
While the chronic proliferative dermatitis of Sharpin cpdm/cpdm mice fully depends on TNFinduced keratinocyte apoptosis [6,16,17], Itgb1 antibody treatment did not affect the previously reported [31] elevation of Tnf expression levels in the Sharpin cpdm/cpdm skin (Panel B and C in S2 Fig), suggesting that altered Tnf expression levels are not mediating the effects of the Itgb1 antibody treatment. However, systemic Itgb1 antibody treatment reduced the increased apoptosis in Sharpin cpdm/cpdm epidermis (Fig 4A and 4B). Since Tnf levels remained high despite Itgb1 antibody treatment (Panel B and C in S2 Fig), the reduction in apoptotic cell numbers could rather be related to the decreased proliferation as keratinocyte apoptosis balances cell proliferation to control epidermal thickness [32]. These data suggest that inhibition of integrin activity in Sharpin cpdm/cpdm mice with an anti-Itgb1 antibody ameliorates the chain reaction leading to hyperproliferative dermatitis in the absence of SHARPIN.
Through direct interaction with the integrin cytoplasmic domain SHARPIN inhibits integrin activity, which plays a role in cell adhesion and migration [19,20]. Consistently, increased levels of active Itgb1 were observed in the epidermis of Sharpin cpdm/cpdm mice [19]. However, as Sharpin cpdm/cpdm mice suffer from chronic proliferative dermatitis, it has remained unclear whether increased integrin activity in the Sharpin cpdm/cpdm epidermis is due to lack of SHAR-PIN or merely a side effect of the dermatitis. Using Tnfr1 -/-Sharpin cpdm/cpdm double knock out mice [6] we now report that Itgb1 activity is also increased in basal Tnfr1 -/-Sharpin cpdm/cpdm keratinocytes in the absence of proliferative dermatitis, supportive of a direct role for SHAR-PIN in integrin inhibition in vivo. The Tnfr1 -/-Sharpin cpdm/cpdm mice and other double knock out mice that rescue the Sharpin cpdm/cpdm inflammatory phenotype [6,16,17] are excellent tools to dissect which in vivo functions of SHARPIN are a side effect of the inflammatory phenotype and which phenotypes are independent of the inflammation. In addition, as Tnfr1 -/-Sharpin cpdm/cpdm mice have a considerably longer life span than Sharpin cpdm/cpdm mice, which need to be sacrificed by 8 weeks due to severe skin inflammation, these double knock out mice provide means to address whether SHARPIN has important functions later in development and in the development of diseases, such as cancer.
Sharpin cpdm/cpdm mice display skin thickening (hyperkeratosis), increased keratinocyte proliferation and apoptosis, as well as increased inflammatory cell infiltration [1,2,29]. The phenotype has been shown to originate from defective NF-kB signaling and increased apoptosis induction in response to Tnfr1 activation in keratinocytes, as crosses between Sharpin cpdm/cpdm mice and Tnf -/or Tnf1r -/mice fully rescued the chronic proliferative dermatitis [6,16,17]. Transgenic mice with suprabasal expression of Itgb1 displayed similar chronic proliferative dermatitis with deregulated keratinocyte proliferation and skin inflammation, although apoptosis was not significantly affected [24]. In Sharpin cpdm/cpdm mice, however, the increased Itgb1 activity was primarily observed in basal keratinocytes negative for Keratin-10 or involucrin expression [19]. Our study now shows that inhibition of Itgb1 in Sharpin cpdm/cpdm mice decreases both keratinocyte proliferation and apoptosis. Importantly, inflammation was not affected, suggesting that the cause of the alleviated phenotype is different from that in the aforementioned Tnf -/-Sharpin cpdm/cpdm and Tnfr1 -/-Sharpin cpdm/cpdm double knockout mice. Thus, while Tnf-induced keratinocyte death drives chronic proliferative dermatitis in Sharpin cpdm/cpdm mice, increased Itgb1 activity appears to exacerbate the hyperproliferative skin phenotype, consistent with the important role of Itgb1 in keratinocyte proliferation [34].
Recently, SHARPIN was also shown to bind and inhibit the T cell receptor (TCR) in a LUBAC-independent fashion, thereby playing an intrinsic role in the generation of regulatory T cells (Treg cells) [35]. Importantly, transfer of wild type Treg cells into Sharpin cpdm/cpdm mice considerably alleviated their systemic inflammation [35]. These and our data together show that, though chronic inflammation in Sharpin cpdm/cpdm mice is triggered by Tnf-mediated apoptosis [6,16,17], other regulatory roles of SHARPIN contribute to different aspects of the complex Sharpin cpdm/cpdm phenotype.
In summary, our data suggest that Sharpin functions as an integrin inhibitor in vivo and that integrins contribute to the chronic proliferative dermatitis phenotype in Sharpin cpdm/cpdm mice. We therefore postulate that different aspects of the complex Sharpin cpdm/cpdm phenotype, also in tissues other than skin, could be a result of SHARPIN's ability to regulate several proteins.

Mice
The C57BL/KaLawRij-Sharpin cpdm /RijSunJ mouse strain (Stock No: 007599 |) with a spontaneous mutation leading to the complete loss of SHARPIN protein [1,2] was acquired from The Jackson Laboratory (Bar Harbor, ME). The colony was maintained in heterozygote breeding and genotyped for the Sharpin cpdm/cpdm mutation to obtain Sharpin cpdm/cdpm homozygous mice and littermate control mice (Sharpin +/+ or Sharpin +/cpdm ) for experiments. DNA was extracted with KAPA Mouse Genotyping Kit (KK7302) and the Sharpin cpdm mutation detected using 40x genotyping assay mix (TaqMan SNP Genotyping Assays, 5793982, Applied Biosystems) and TaqMan Universal PCR Master Mix. Tnfrsf1a tm1Imx Sharpin cpdm/cdpm mice [16] (Tnfrsf1a tm1Imx Stock No: 003242; hereafter called Tnfr1 -/-), obtained from Prof. H. Walczak, were also maintained in heterozygote breeding. PCR amplification was used for detection of the deletion in Tnfr1 gene in somatic DNA. Age-matched mice were used in all the experiments described here. As both male and female Sharpin cpdm/cdpm mice develop the same symptoms and we have to perform heterozygous breedings, both genders were used in the experiments.
Mice were housed in standard conditions (12-h light/dark cycle) with food and water available ad libitum. The viability, clinical signs and behaviour of the mice were monitored daily.

Antibody treatment of mice
Five pairs of 5-week old gender-matched Sharpin cpdm/cpdm and Sharpin +/? littermates were injected intraperitoneally with 0.1 ml of 0.05 mg/ml function-blocking anti-mouse ITGB1 antibody (LEAF™ purified anti-mouse CD29 Armenian hamster IgG (clone HMB1-1, Biolegend) or PBS twice a week. Mice were injected for a total of 3 weeks (6 injections) and then euthanized by CO 2 asphyxiation. Complete necropsies were performed as described [36]. Skin was collected in Fekete's acid-alcohol-formalin. Dorsal skin was also collected in 4% paraformaldehyde for immunofluorescence or fresh frozen for IHC. Blood was collected in EDTA tubes for analysis using the ADVIA 120 Hematology System (Siemens Healthcare Global). Four pairs of 5-week old littermates injected with 0.1 ml of 0.5 mg/ml anti-mouse CD29 antibody or Armenian hamster IgG negative control antibody (Biolegend) showed highly similar results.

Antibodies
The antibodies used in this study are LEAF™ purified Anti-mouse CD29 Armenian hamster IgG Immunofluorescence O.C.T. (Sakura) embedded frozen skin sections were fixed (4% paraformaldehyde), quenched (100mM glycine in PBS), and blocked/permeabilized (2% BSA, 0.1% Triton X100 in PBS). Samples were labeled with primary antibodies (1 hour at RT in 2% BSA in PBS), washed three times and labeled with fluorescent secondary antibodies. After washes, samples were mounted in Vectashield mounting medium (Vector Laboratories) and viewed using a 20x objective on either a 3i Marianas Spinning disk confocal microscope (Intelligent Imaging Innovations) or Zeiss Axiovert 200M with Yokogawa CSU22 spinning disk confocal microscope unit with Hamamatsu Orca ER CCD camera (Hamamatsu Photonics K.K.). Images were processed and quantified using Fiji image analysis software [37].
Isolation and culture of mouse primary keratinocytes 6-8 weeks old littermate mice were sacrificed, dorsal skin was shaved and a piece of dorsal skin (1.5 cm x 1.5 cm) was harvested in cold PBS. Tissue was cut further to smaller pieces, rinsed with Hank's Balanced Salt Solution without Ca 2+ and Mg 2+ (HBSS; Sigma), and incubated overnight in 0,25% porcine trypsin (Sigma) in HBSS at +4˚C on a shaker. Next day, the dermis was removed with forceps, the remaining epidermis was chopped with a scalpel and incubated in 0.2% Collagenase XI (Sigma) at 37˚C incubator for 30 min. Keratinocytes were dissociated from the tissue by pipetting the suspension every 5-10 minutes. Keratinocytes were filtered through a cell strainer (70 μm; BD Biosciences). Finally, cells were incubated with 20 U/ml DNase I (Roche) for 5 min on ice and spun down.

Keratinocyte proliferation assay
Equal amounts of primary mouse keratinocytes were plated on 96-well plates coated with 20 μg/ml rat tail Collagen Type I. One day after seeding, the keratinocytes were treated with 10 μg/ml anti-Itgb1(Anti-mouse CD29 Armenian hamster IgG) or isotype IgG control antibody. Theamount of cells) in each well was quantified with Cell Proliferation Reagent WST-1 (Roche, Sigma Aldrich) according to the manufacturer's instructions and a BioTek Synergy H1 Hybrid Multi-Mode Microplate Reader. The relative proliferation was calculated by normalizing the amount of cells in each time point to day zero (Fig 2G) or by subtracting the amount of cells at day zero from the value at day two (corrected cell amount), followed by normalizing to the corrected cell amount of IgG control. A minimum of two wells per keratinocyte primary cell line were analysed per time point.

Western blotting
Protein content in primary mouse keratinocytes was analysed by immunoblotting using standard western blotting techniques and the Odyssey LICOR imaging system.

Statistical analysis
All statistical analyses were performed using GraphPad Prism (GraphPad Software, San Diego, USA). Normal distribution was assumed for the data and unpaired Student's t-test was used for all experiments. A p < 0.05 was considered significant.