Topically Applied Hsp90 Blocker 17AAG Inhibits Autoantibody-Mediated Blister-Inducing Cutaneous Inflammation

Accepted Manuscript Topically Applied Hsp90 Blocker 17AAG Inhibits Autoantibody-Mediated BlisterInducing Cutaneous Inflammation Stefan Tukaj, Katja Bieber, Konrad Kleszczyński, Mareike Witte, Rebecca Cames, Kathrin Kalies, Detlef Zillikens, Ralf J. Ludwig, Tobias. W. Fischer, Michael Kasperkiewicz PII:

S0022-202X(16)32453-8

DOI:

10.1016/j.jid.2016.08.032

Reference:

JID 533

To appear in:

The Journal of Investigative Dermatology

Received Date: 19 April 2016 Revised Date:

28 August 2016

Accepted Date: 31 August 2016

Please cite this article as: Tukaj S, Bieber K, Kleszczyński K, Witte M, Cames R, Kalies K, Zillikens D, Ludwig RJ, Fischer TW, Kasperkiewicz M, Topically Applied Hsp90 Blocker 17AAG Inhibits Autoantibody-Mediated Blister-Inducing Cutaneous Inflammation, The Journal of Investigative Dermatology (2016), doi: 10.1016/j.jid.2016.08.032. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Topically Applied Hsp90 Blocker 17AAG Inhibits AutoantibodyMediated Blister-Inducing Cutaneous Inflammation

RI PT

Topical Hsp90 Inhibition in Autoimmunity

Stefan Tukaj1,2,5, Katja Bieber3,5, Konrad Kleszczyński1, Mareike Witte1, Rebecca Cames1,

SC

Kathrin Kalies4, Detlef Zillikens1,3, Ralf J. Ludwig1,3, Tobias. W. Fischer1,

M AN U

and Michael Kasperkiewicz1

Department of Dermatology, University of Lübeck, Germany

2

Department of Molecular Biology, University of Gdańsk, Poland

3

Lübeck Institute of Experimental Dermatology, University of Lübeck, Germany

4

Institute of Anatomy, University of Lübeck, Germany

5

These authors contributed equally to this work

EP

TE D

1

Abbreviations: COL7, type VII collagen; EBA, epidermolysis bullosa acquisita; Flii,

AC C

Flightless I; Hsp, heat shock protein; MMP, matrix metalloproteinase

Keywords: autoantibody; autoimmunity; epidermolysis bullosa acquisita; heat shock protein; inflammation

Corresponding author: Michael Kasperkiewicz, Department of Dermatology, University of Lübeck, Ratzeburger Allee 160, D-23538 Lübeck, Germany; phone: +49 451 500 3836; fax: +49 451 500 5163, e-mail: [email protected]

1

ACCEPTED MANUSCRIPT

ABSTRACT Cell stress-inducible heat shock protein 90 (Hsp90) has been recognized as key player in mediating inflammatory responses. While its systemic blockade was successfully used to treat autoimmune diseases in preclinical models, efficacy of a topical route of Hsp90 inhibitor

mediated

dermatoses.

Here,

effects

of

the

Hsp90

RI PT

administration has so far not been evaluated in chronic inflammatory and autoimmuneblocker

17-allylamino-

demethoxygeldanamycin (17AAG) applied topically to the skin were determined in

SC

experimental inflammatory epidermolysis bullosa acquisita (EBA), an anti-type VII collagen autoantibody-induced blistering skin disease. Topical 17AAG ameliorated clinical disease

M AN U

severity when given before or during occurrence of skin lesions without causing cutaneous or systemic toxicity in mice with antibody transfer- and immunization-induced EBA. In both EBA models and in the setting of locally induced inflammation, topical 17AAG treatment was associated with (i) reduced neutrophilic infiltrates, (ii) decreased NFκB activation, (iii)

TE D

lowered expression of matrix metalloproteinases and Flightless I, and (iv) induction of antiinflammatory Hsp70 in the skin. Our results suggest that topical delivery of Hsp90 antagonists, offering the benefit of a reduced risk of systemic side effects of Hsp90 inhibition,

AC C

EP

may be useful for the control of EBA and possibly other related inflammatory skin disorders.

2

ACCEPTED MANUSCRIPT

INTRODUCTION Heat shock protein 90 (Hsp90) is a highly conserved and constitutively expressed molecular chaperone that can be up-regulated by cell stress factors for example in the context of cancer or inflammation. It is a central player in the folding, stabilization, maturation, and activation

RI PT

of many proteins. More than 200 Hsp90-dependent clients have been identified, including those considerably implicated in both cancer and inflammation such as transcription factors of

al., 2015; Tukaj and Węgrzyn, 2016; Wu et al., 2012).

SC

the NFκB protein family and tissue degrading matrix metalloproteinases (MMPs) (Sevin et

Consequently, several Hsp90 blockers that inhibit the ATPase activity of the chaperone like

M AN U

the geldanamycin derivative 17-allylamino-demethoxygeldanamycin (17AAG) are currently being evaluated in clinical phase I-III trials for the treatment of various malignancies (GarciaCarbonero et al., 2013). Due to selective blockade of innate and adaptive cells of the immune system, we and others have shown in experimental models of autoimmune encephalomyelitis,

TE D

rheumatoid arthritis, systemic lupus erythematosus, and autoimmune bullous disorders that pharmacological inhibition of Hsp90 also appears to be an effective treatment strategy for inflammatory autoimmune diseases (Tukaj et al., 2015b; Tukaj and Węgrzyn, 2016). Anti-

EP

inflammatory effects of Hsp90 inhibitors may not only result from delivery of Hsp90 client

AC C

proteins to proteasomal degradation but also from concurrent induction of anti-inflammatory genes including Hsp70, which, in contrast, may rather limit efficacy of Hsp90 inhibitors in cancer treatment due to its tumor-promoting capacity (Sevin et al., 2015; Tukaj and Węgrzyn, 2016).

Mechanobullous or inflammatory epidermolysis bullosa acquisita (EBA) is an organ-specific autoimmune bullous disease induced by autoantibodies to type VII collagen of the dermalepidermal junction. The inflammatory variant of EBA, which is clinically and pathophysiologically related to the most common subepidermal autoimmune blistering disease bullous pemphigoid, can be reproduced experimentally in mice. In this EBA type, 3

ACCEPTED MANUSCRIPT autoantibody binding is followed by a lesional inflammatory cell infiltration leading to proteolytic degradation of the basal membrane zone (Kasperkiewicz et al., 2016). We have previously demonstrated that systemically applied Hsp90 inhibitors exhibit activity in mice with experimental EBA by potently affecting inflammatory disease pathways (Kasperkiewicz

RI PT

et al., 2011; Tukaj et al., 2014).

While successful treatment with topical 17AAG has been recently reported in a murine model of cutaneous squamous cell carcinoma (Singh et al., 2015), efficacy of this route of Hsp90

SC

inhibitor administration has so far neither been tested in experimental EBA nor in any other in vivo models of chronic inflammatory skin disorders. Using three murine models of EBA (a

M AN U

systemic antibody transfer- and immunization-induced model as well as a local antibody transfer model) we show in this current study that topical 17AAG treatment is an effective

AC C

EP

TE D

and safe approach to inhibit autoantibody-mediated blister-inducing cutaneous inflammation.

4

ACCEPTED MANUSCRIPT

RESULTS Topical 17AAG treatment reduces disease activity in a systemic antibody transfer- and immunization-induced EBA mouse model In a first set of experiments, EBA was induced in mice by repetitive administration of

RI PT

pathogenic anti-COL7 IgG (Sitaru et al., 2005). In a prophylactic setting, when applied topically onto the shaved back before disease onset (daily on days 0-12), 17AAG (500 nmol) significantly reduced the development of characteristic clinical symptoms (i.e., erythema,

SC

blisters, erosions, and crusts) compared with vehicle (Figure 1a and c). In a therapeutic setting, when administered during ongoing disease (daily on days 8-15), mice treated with

M AN U

topical 17AAG (500 nmol) also showed a significantly lower extent of skin lesions on the back compared to control animals at the end of the treatment (Figure 1b and d). These clinical findings were accompanied by significantly reduced dermal neutrophil infiltrations in mice treated prophylactically (Figure 1e and g) or therapeutically (Figure 1f and h) with 17AAG.

TE D

Likewise, a combined prophylactic-therapeutic topical treatment with 17AAG (1000 nmol) on ears (prophylactic application) and body lesions (therapeutic application) on alternate days over several weeks significantly ameliorated the clinical and histological course of disease in

EP

COL7-immunized mice compared with vehicle (Figure 2a-d). In this immunization-induced

AC C

model (Iwata et al., 2013), some degree of non-significant trends was observed for differences between detectable serum concentrations of IL-2 (18.52±17.07 vs. 9.78±4.15, P=0.29), IL-6 (58.37±30.05 vs. 31.01±24.46, P=0.074), IL-10 (9.05±5.23 vs. 20.37±13.38, P=0.084), IL17A (48.66±46.34 vs. 19.98±27.68, P=0.25), IL-17F (36.47±35.07 vs. 5.74±1.56, P=0.17), and sIL-2R (2.45± 4.22 vs. 0.26±0.60, P=0.27) in vehicle- (n=9) and 17AAG-treated mice (n=8), respectively, at the end of the treatment period. For the remaining analyzed serum factors related to the T cell activation status (IL-4, IL-5, IL-9, IL-13, IL-21, IL-22, IFNγ, TNFα, and sCD4), no obvious differences were noted or a comparative analysis was not feasible due to serum levels being predominantly or completely below the detection limit of 5

ACCEPTED MANUSCRIPT the applied assays in either mouse group (data not shown). In addition, autoantibody production was not affected, as evidences by comparable levels of circulating autoantibodies in vehicle- and 17AAG-treated mice (Figure 2e). In both models, topical 17AAG treatments elicited neither skin irritation nor signs of systemic

RI PT

toxicity such as weight loss or death.

17AAG-induced amelioration of disease severity is associated with reduced NFκB

SC

activation as well as altered MMP2, MMP9, MMP12, Flii, and Hsp70 expression levels in the skin

M AN U

Activation of the transcription factor NFκB, which requires phosphorylation and consecutive degradation of its upstream inhibitor IκBα, is known to up-regulate various inflammatory genes including those that play a pathophysiological role in experimental EBA (Herrington et al., 2016; Kasperkiewicz et al., 2016). Compared with vehicle, topical 17AAG significantly

TE D

decreased expression of p-IκBα in antibody transfer- and immunization-induced experimental EBA, suggesting that NFκB activation is inhibited by this treatment (Figure 3a and b). In addition, 17AAG-treated animals had a significantly lower protein and mRNA content of

EP

MMP2 (Figure 3c-e), MMP9 (Figure 3f-h), MMP12 (Figure 3i-k), and Flii (Figure 3l-n), which are implicated in the disease process of EBA (Kopecki et al., 2011, 2013; Shimanovich

AC C

et al., 2004; Tukaj et al., 2015a). Finally, Hsp70, a marker of effective Hsp90 inhibition and negative regulator of inflammatory responses (Sevin et al., 2015; Tukaj and Węgrzyn, 2016), was significantly increased at the protein and transcriptional level after 17AAG treatment compared with vehicle (Figure 3o-q).

6

ACCEPTED MANUSCRIPT Effects of topical 17AAG treatment on cellular and molecular skin pathology are replicated in a local EBA mouse model In a local model of EBA in the ear, where the locally restricted pathology is induced by a single application of anti-COL7 IgG (Kasprick et al., 2015), we could confirm the effects of

RI PT

topical 17AAG treatment on the cellular and molecular factors observed in the systemic EBA models. Short-term topical 17AAG treatment significantly affected dermal neutrophil infiltration (Figure 4a), NFκB activation (Figure 4b), as well as MMP2 (Figure 4c), MMP9

SC

(Figure 4d), MMP12 (Figure 4e), Flii (Figure 4f), and Hsp70 (Figure 4g) expression in the skin, although no significant differences in the clinical disease phenotype were observed

AC C

EP

TE D

M AN U

between vehicle- and 17AAG-treated mice (data not shown).

7

ACCEPTED MANUSCRIPT

DISCUSSION We have previously shown that Hsp90 represents an important pathophysiological factor and novel treatment target in autoimmune bullous diseases (Tukaj et al., 2015b). In these previous studies, systemic Hsp90 inhibition by the intraperitoneally applied geldanamycin derivative

RI PT

17-DMAG or the short peptide derivative TCBL-145 ameliorated disease activity in mice with experimental EBA (Kasperkiewicz et al., 2011; Tukaj et al., 2014). Analysis of the underlying mechanistic effects in in vitro, ex vivo, and in vivo models of EBA revealed that

SC

Hsp90 inhibition interfered with the afferent disease stage of autoantibody production (by inhibition of autoreactive T cells and B cells as well as promotion of regulatory B cells) and

M AN U

the efferent phase of autoantibody-induced tissue damage (by inhibition of MMP- and reactive oxygen species-producing neutrophil effector cells) (Tukaj et al., 2015b). Topical Hsp90 inhibitor therapy is an attractive choice for the treatment of skin diseases due to the advantageous targeting of the drug to the site of the pathologic condition and reduction

TE D

of the risk of systemic side effects. Singh et al. (2015) have demonstrated that 17AAG in a DMSO-acetone solution shows good penetration into murine skin and can potently inhibit ultraviolet radiation-induced squamous cell carcinomas in mice. Using this topical 17AAG

EP

formulation, we now present that this treatment can also suppress the development of

AC C

experimentally induced EBA lesions in mice when administered before the appearance of clinical signs and/or during ongoing disease. In addition to reduced clinical disease activity, animals treated with topical 17AAG showed a less intense neutrophilic inflammatory cell infiltrate at the dermal-epidermal junction compared with vehicle. These findings, which correlated with the clinical and histological results obtained in our previous EBA studies of systemic treatment with Hsp90 inhibitors (Kasperkiewicz et al., 2011; Tukaj et al., 2014), were equally observed in two different systemic models of EBA, an antibody transfer- and immunization-induced mouse model, reproducing the efferent or both the efferent and

8

ACCEPTED MANUSCRIPT afferent autoimmune response of the disease, respectively (Iwata et al., 2013; Sitaru et al., 2005). Disease development may be diverse in mice with antibody transfer-induced EBA depending on various factors such as antibody preparations from different serum batches of COL7-

RI PT

immunized rabbits, which may explain the different degree of induced clinical manifestations between our independent experiments. In addition, the relatively low dose of injected antibodies compared to originally reported protocols (3 mg vs. up to 15 mg per injection;

SC

Sitaru et al., 2005) could have contributed to the relatively short disease duration with decrease of skin lesions during the observation period independently of vehicle or verum

M AN U

application. This effect, however, was more pronounced in mice treated with 17AAG, and the Hsp90 inhibitor was also capable of suppressing a gradual rise in disease extent in the immunization-induced EBA model, making it unlikely that the efficacy of 17AAG could have been triggered by treatment-independent disease improvement.

TE D

In the immunization-induced model, topical treatment with 17AAG did not significantly affect serological markers associated with the T cell immune response and had no impact on levels of circulating autoantibodies, suggesting that the achieved outcomes resulted from local

EP

effects within the skin tissue rather than from potential systemic adsorption of this drug. In fact, topical 17AAG treatment affected the activation status of NFκB in the skin as well as

AC C

cutaneous expression levels of MMPs, Flii, and Hsp70 at the protein and transcriptional level in both systemic EBA models as well as in a corresponding local in vivo model (Kasprick et al., 2015). In this latter model, however, no difference in the clinical disease phenotype was found between mice treated with 17AAG and vehicle, which may have resulted from the short follow-up period (48 h after anti-COL7 IgG injection) being not long enough to allow for the observation of potential clinical effects. This assumption is supported by the observation of a delayed onset of clinical differences between mice treated with previously described agents (e.g., EndoS, roflumilast, dimethylfumarate, anti-GM-CSF, intravenous immunoglobulins) 9

ACCEPTED MANUSCRIPT and controls in the systemic antibody transfer-induced EBA model (Hirose et al., 2012; Koga et al., 2016; Müller et al., 2016; Samavedam et al., 2014; Schwab et al., 2014). Nevertheless, the local EBA model is primarily suited for investigations of initial effector mechanisms of the disease at defined body regions and may, in contrast to the systemic EBA models, rather

RI PT

not represent a reliable or representative tool for general clinical outcome measurements of pharmacological treatments.

The NFκB family of transcription factors, which is known to be tightly regulated by the

SC

Hsp90 chaperone machinery, has an essential role in inflammation and innate immunity. Being conserved in all multicellular animals, inflammation-induced activation of the NFκB

M AN U

signaling pathway is associated with phosphorylation of IκBα by IKK, with the latter existing in complexes with Hsp90, and consecutive degradation of p-IκBα by the proteasome (Herrington et al., 2015; Sevin et al., 2015; Tukaj and Węgrzyn, 2016). This allows NFκB to enter the nucleus for gene expression regulation of a plethora of pro-inflammatory genes

TE D

including those that have been found to be over-expressed in experimental EBA and partly validated as novel drug targets in EBA mice such as IL-1, CXCR1/2, and GM-CSF (Hirose et al., 2013; Kasperkiewicz et al., 2016; Sadeghi et al., 2015; Samavedam et al., 2014). Herein,

EP

we found that topical 17AAG decreased NFκB activation in experimental EBA, as evidenced by down-regulated skin expression levels of p-IκBα compared with vehicle-treated mice.

AC C

However, since the functional significance of NFκB has not yet been specifically addressed in this disease, further studies are needed to define the relative contribution of the NFκB protein clients of Hsp90 in 17AAG-caused inhibition of experimentally induced EBA development. Proteolytic MMPs comprising MMP2, MMP9, and MMP12 are capable of degrading basement membrane components and have been suggested to be involved in autoantibodyinduced dermal-epidermal separation (Niimi et al., 2006; Shimanovich et al., 2004; Tukaj et al., 2015a). A close interaction between Hsp90 and MMPs was described in the context of tissue destruction by metastatic cancer cells (Eustace et al., 2004; Garcia-Carbonero et al., 10

ACCEPTED MANUSCRIPT 2013), and such Hsp90-MMP complexes have been recently also found in sera of EBA patients (Tukaj et al., 2015a). In our current experiments, a decrease in the levels of all 3 MMPs was demonstrated in 17AAG-treated skin of mice with experimental EBA compared with vehicle.

RI PT

The actin remodeling cytosceletal protein Flightless I (Flii) has been implicated in the development of autoantibody-induced blistering and impaired wound healing. Reduced Flii expression using Flii+/- mice or topical application of Flii neutralizing antibodies was shown

SC

to impair blister formation and improve healing of blistered skin in mice with experimental EBA (Kopecki et al., 2011, 2013). Since Flii was recently validated as a new client of Hsp90

M AN U

(Wu et al., 2012), investigation of this protein was of further particular interest to our study. In fact, along with the decrease in MMPs, the amount of Flii expression in the skin was also reduced after topical 17AAG treatment compared with vehicle.

Hsp90 inhibition by classical Hsp90 blockers is associated with the activation of the heat

TE D

shock transcription factor 1 which drives the expression of Hsp70. Hsp70 is not only regarded as indicator of effective pharmacological Hsp90 blockade, but also as blocker of NFκB activity, inactivator of antigen presenting cells, and inductor of immunoregulatory

EP

mechanisms (e.g., expansion of regulatory T cells) (Sevin et al., 2015; Tukaj and Węgrzyn, 2016). In accordance, Hsp70 levels were up-regulated following topical 17AAG treatment in

AC C

comparison with control in the skin of EBA mice. It is interesting to note that a pronounced immunohistochemical expression of the investigated proteins was observed in the epidermis of EBA mice including the upper layers. Similar findings were previously described in the skin of patients with bullous pemphigoid, in which MMPs and Hsp90 were strongly detected throughout epidermal cells, being not limited to an expression adjacent to the epidermal basement membrane zone (Niimi et al., 2006; Tukaj et al., 2013). This implies that besides inflammatory cells of the dermal-epidermal junction,

11

ACCEPTED MANUSCRIPT epidermal keratinocytes participate in the pathophysiology of bullous pemphigoid as well as bullous pemphigoid-like EBA and are prone to Hsp90 inhibition. In conclusion, our results suggest that topical delivery of Hsp90 antagonists, offering the benefit of a reduced risk of systemic side effects of Hsp90 inhibition, may be useful for the

AC C

EP

TE D

M AN U

SC

RI PT

control of EBA and possibly other related inflammatory skin disorders.

12

ACCEPTED MANUSCRIPT

MATERIALS AND METHODS Mice BALB/c and B6.SJL-H2s C3c/1CyJ (B6.s) mice aged 6-12 weeks were purchased from Charles River Laboratories (Sulzfeld, Germany). Animal experiments were approved by local

Systemic and local models of experimental EBA

SC

certified personnel.

RI PT

authorities of the Animal Care and Use Committee (Kiel, Germany) and performed by

Systemic induction of EBA by antibody transfer in mice followed published protocols with

M AN U

minor modifications (Sitaru et al., 2005). Briefly, IgG from rabbits immunized with recombinant fragments of murine COL7 was purified by protein G sepharose affinity chromatography (Amersham Biosciences, Freiburg, Germany), as previously reported (Sitaru et al., 2005). Reactivity of IgG fractions was analysed by immunofluorescence microscopy on

TE D

murine skin. BALB/c mice received a total of seven subcutaneous injections of 3 mg rabbit anti-mouse COL7 IgG in the neck region every second day (days 0, 2, 4, 6, 8, 10, and 12). Systemic induction of EBA by autoantigen immunization was performed as described

EP

elsewhere (Iwata et al., 2013). Briefly, B6.s mice were injected subcutaneously in the hind

AC C

footpads with a single injection of 120 µg recombinant murine COL7 von Willebrand Factor domain (vWFA2) emulsified in the nonionic block copolymer adjuvant TiterMax (Alexis Biochemicals, Norcross, GA, USA). Local induction of EBA in mice was carried out following published protocols with minor modifications (Kasprick et al., 2015). Briefly, BALB/c mice received a single intradermal injection of 0.6 mg rabbit anti-mouse COL7 IgG in the base of each ear.

13

ACCEPTED MANUSCRIPT Topical treatment and clinical outcome analysis 17AAG stock (50 mM; Selleck Chemicals, Munich, Germany) was prepared in DMSO and freshly reconstituted in acetone (DMSO:acetone 1:40 v/v) to a concentration of 500 nmol or 1000 nmol at the time of treatment, as previously described (Singh et al., 2015).

RI PT

In the case of systemic antibody transfer-induced EBA model, 17AAG (500 nmol) or vehicle solution was applied topically to back skin, which was shaved two days before the treatment, daily on days 0-12 (prophylactic setting) or days 8-15 (therapeutic setting). Disease severity

SC

was expressed as the percentage of surface area of the shaved back affected by skin lesions and determined at three time points (days 4, 8, and 12 or days 8, 11, and 15 in the setting of

M AN U

prophylactic or therapeutic treatment, respectively).

In the case of systemic immunization-induced EBA model, 17AAG (1000 nmol) or vehicle solution was applied topically to ear skin (both ears per agent) every other day starting one day before immunization (prophylactic treatment) as well as concurrently to any newly

TE D

developed skin lesion of the body every other day (therapeutic treatment) until week 10. Disease severity was expressed as the percentage of surface area of the ears or the total body affected by skin lesions and determined at six time points (weeks 5-10).

EP

In the case of local EBA model, 17AAG (500 nmol) or vehicle solution was applied topically to ear skin (both ears per agent) daily for three days starting 1 day before antibody transfer.

AC C

Disease severity was expressed as the percentage of surface area of the ears affected by skin lesions and determined 2 days following antibody transfer.

Histopathology Lesional skin specimens were prepared for examination by histopathology, as described previously (Sitaru et al., 2005). Briefly, biopsies from experimental animals were fixed in 4% buffered formalin, and paraffin-embedded tissue sections were stained with hematoxylin and eosin. Dermal neutrophil infiltration was analysed semiquantitatively in a blinded fashion 14

ACCEPTED MANUSCRIPT using the following scoring system: 0 = no infiltration, 1 = faint infiltration, 2 = moderate infiltration, and 3 = severe infiltration.

Enzyme-linked immunosorbent assay

RI PT

Circulating mouse COL7 vWFA2-specific antibodies were detected by enzyme-linked immunosorbent assay (ELISA), as described previously (Iwata et al., 2013). In brief, 96-well microtiter plates (Maxisorb; Nunc, Roskilde, Denmark) were coated with 250 ng murine

SC

vWFA2. To reduce nonspecific binding, plates were blocked with 1% BSA in PBS-T at room temperature (RT) for 60 min, followed by incubation with mouse sera (1:1,000 to 1:100,000

M AN U

in 1 x PBS with 1 x BSA) at RT for 60 min. Subsequently, plates were incubated with antimouse IgG conjugated with HRP (1:10,000; Bethyl Laboratories, Montgomery, TX, USA) at RT for 60 min. One-step Turbo TM-ELISA (Thermo Scientific, Rockford, IL, USA) substrate was added to visualize enzymatic reaction. The cutoff was defined as average of negative

Multiplex cytokine assay

TE D

controls + 3 SDs of negative controls.

EP

Serum concentrations of various cytokines (IL-2, IL-4, IL-5, IL-6, IL-9, IL-10, IL-13, IL17A, IL-17F, IL-21, IL-22, IFNγ, and TNFα) were determined by the LEGENDplex™ mouse

AC C

Th cytokine panel (13-plex) array (Biolegend, Fell, Germany) according to the manufacturer’s protocol. The analysis was performed with the flow cytometer MACSQuant Analyzer 10 (Miltenyi Biotec).

Soluble IL-2 receptor and CD4 measurement Serum levels of murine soluble IL-2 receptor (sIL-2R) and CD4 (sCD4) were analyzed by ELISA (MyBioSource, San Diego, CA, USA) according to the manufacturer’s protocol.

15

ACCEPTED MANUSCRIPT Fluorescent immunohistochemistry 6-µm-thick cryosections of mouse skin were dried at RT for 10 min, fixed with ice-cold acetone at -20oC for 10 min, and washed three times with 1 x PBS (pH 7.5) for 5 min. Blocking was performed with mouse Ig blocking reagent (Vector Laboratories, Inc.

RI PT

Burlingame, CA, USA) at RT for 30 min. Subsequently, pre-incubation was carried out using 5% normal goat serum (NGS) at RT for 1 h, and sections were incubated with the following mouse monoclonal primary antibodies at RT overnight: anti-p-IκBα(Ser32), anti-MMP2, anti-

SC

MMP9, anti-MMP12, anti-Flightless I (Flii), and anti-Hsp70 (all 1:50; all from Santa Cruz, Inc., Santa Cruz, CA, USA). Next, slides were washed three times with 1 x PBS for 5 min and

M AN U

labeled in presence of 2% NGS with pre-adsorbed goat anti-mouse polyclonal secondary IgG conjugated with Cy3 (1:500; Abcam plc, Cambridge, UK) at RT for 90 min. Slides were washed three times with 1 x PBS for 5 min, counterstained for cell nucleus detection with DAPI solution (1 µl/ml) at RT for 1 min, washed again three times with 1 x PBS for 5 min,

TE D

and mounted in Fluoromount-G mounting medium (SouthernBiotech, Birmingham, AL, USA). All fluorescent images were obtained using the Keyence BZ-9000 inverted fluorescence microscope (Keyence GmbH, Neu-Isenburg, Germany). The fluorescence

EP

intensity of the epidermal and dermal layers of stained skin sections was determined by densitometry measurements using Image J 1.38d software (National Institute of Health,

AC C

USA).

Real-time quantitative reverse transcriptase-polymerase chain reaction Total RNA was extracted from skin tissue using TRIZOL (Invitrogen, Carlsbad, CA, USA). 500 ng total RNA was used to perform the reverse transcription with SuperScript III FirstStrand Synthesis System (Invitrogen, Carlsbad, CA, USA). Resultant cDNA samples were used for real-time quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR) carried out by the LightCycler 480 (Roche, Mannheim, Germany) in a mixture containing 16

ACCEPTED MANUSCRIPT Maxima SYBR Green qPCR Master Mix (Fermentas GmbH, St. Leon-Rot, Germany), specific primers, and 10 ng of cDNA in a total volume of 20 µl in LightCycler borosilicate glass capillaries (Roche, Mannheim, Germany). qRT-PCR was performed using a 10 min initial denaturation at 95°C followed by 50 three-step cycles of 15 sec at 95°C (denaturation),

RI PT

30 sec at 59°C (annealing), and 30 sec at 72°C (extension). Relative expression of the genes was calculated with the 2–∆∆CT method, with the elongation of GAPDH gene used as a housekeeping gene. The following primers were used: GAPDH: forward primer, 5’and

CCTTCCACAATGCCAAAGTT-3’;

MMP2:

CACATCCTTCACCTGGTGTG-3’; TCTCTCTCCCGAGAGTCCAA-3’ CCTGAACCATAACGCACAGA-3’;

reverse

MMP9: and

forward

reverse

MMP12: and

TE D

TTTGGAGCTCACGGAGACTT-3’

and

forward

M AN U

AGCGTGAAGTTTGGAAGCAT-3’

reverse

SC

GGGTGTGAACCACGAGAAAT-3’

forward

primer,

primer, primer, primer, primer, primer,

5’5’5’5’5’5’-

reverse

primer,

5’-

TCTGCCTCATCAAAATGTGC-3’;

Flii:

forward

primer,

5’-

TTGACCATCTGGCAAATTGA-3’

and

reverse

primer,

5’-

EP

CATCCAGGAAGGTCTTGAGC-3’;

and

Hsp70:

forward

primer,

5’-

3’.

AC C

GCTCGAGTCCTATGCCTTCA-3’ and reverse primer, 5’-ATGACCTCCTGGCACTTGTC-

Statistics

Data was analyzed by Student’s t-test and Mann-Whitney U test using GraphPad prism 6 (San Diego, California, USA). A p-value <0.05 was considered statistically significant.

CONFLICT OF INTEREST The authors state no conflict of interest. 17

ACCEPTED MANUSCRIPT

ACKNOWLEDGMENTS This study was supported by the Else Kröner-Fresenius-Stiftung (2013_A117) and Deutsche Forschungsgemeinschaft (EXC 306/2 and KA3438/1-1). We are grateful to Dr. D. T. Woodley, Dr. M. Chen, Dr. W. Li, Dr. Y. Hou, A. Cames, and K. Döbber (Department of

RI PT

Dermatology, University of Southern California) who provided advice and/or technical

AC C

EP

TE D

M AN U

SC

assistance.

18

ACCEPTED MANUSCRIPT

REFERENCES Eustace BK, Sakurai T, Stewart JK, Yimlamai D, Unger C, Zehetmeier C et al. Functional proteomic screens reveal an essential extracellular role for hsp90 alpha in cancer cell

RI PT

invasiveness. Nat Cell Biol 2004;6:507-14

Garcia-Carbonero R, Carnero A, Paz-Ares L. Inhibition of HSP90 molecular chaperones:

SC

moving into the clinic. Lancet Oncol 2013;14:e358-69

Herrington FD, Carmody RJ, Goodyear CS. Modulation of NF-κB signaling as a therapeutic

M AN U

target in autoimmunity. J Biomol Screen 2016;21:223-42

Hirose M, Vafia K, Kalies K, Groth S, Westermann J, Zillikens D et al. Enzymatic autoantibody glycan hydrolysis alleviates autoimmunity against type VII collagen. J

TE D

Autoimmun 2012;39:304-14

Hirose M, Götz J, Recke A, Zillikens D, Ludwig RJ, Brandolini L et al. The allosteric

EP

CXCR1/2 inhibitor DF2156A improves experimental epidermolysis bullosa acquisita. J Genet

AC C

Syndr Gene Ther 2013;S3:005

Iwata H, Bieber K, Tiburzy B, Chrobok N, Kalies K, Shimizu A et al. B cells, dendritic cells, and macrophages are required to induce an autoreactive CD4 helper T cell response in experimental epidermolysis bullosa acquisita. J Immunol 2013;191:2978-88

Kasperkiewicz M, Müller R, Manz R, Magens M, Hammers CM, Somlai C et al. Heat-shock protein 90 inhibition in autoimmunity to type VII collagen: evidence that nonmalignant plasma cells are not therapeutic targets. Blood 2011;117:6135-42 19

ACCEPTED MANUSCRIPT

Kasperkiewicz M, Sadik CD, Bieber K, Ibrahim SM, Manz RA, Schmidt E et al. Epidermolysis bullosa acquisita: from pathophysiology to novel therapeutic options. J Invest

RI PT

Dermatol 2016;136:24-33

Kasprick A, Yu X, Scholten J, Hartmann K, Pas HH, Zillikens D et al. Conditional depletion of mast cells has no impact on the severity of experimental epidermolysis bullosa acquisita.

SC

Eur J Immunol 2015;45:1462-70

M AN U

Koga H, Recke A, Vidarsson G, Pas HH, Jonkman MF, Hashimoto T et al. PDE4 inhibition as potential treatment of epidermolysis bullosa acquisita. J Invest Dermatol 2016; pii: S0022202X(16)32095-4. doi: 10.1016/j.jid.2016.06.619

TE D

Kopecki Z, Arkell RM, Strudwick XL, Hirose M, Ludwig RJ, Kern JS et al. Overexpression of the Flii gene increases dermal-epidermal blistering in an autoimmune ColVII mouse model

EP

of epidermolysis bullosa acquisita. J Pathol 2011;225:401-13

Kopecki Z, Ruzehaji N, Turner C, Iwata H, Ludwig RJ, Zillikens D et al. Topically applied

AC C

flightless I neutralizing antibodies improve healing of blistered skin in a murine model of epidermolysis bullosa acquisita. J Invest Dermatol 2013;133:1008-16

Müller S, Behnen M, Bieber K, Möller S, Hellberg L, Witte M et al. Dimethylfumarate impairs neutrophil functions. J Invest Dermatol 2016;136:117-26

20

ACCEPTED MANUSCRIPT Niimi Y, Pawankar R, Kawana S. Increased expression of matrix metalloproteinase-2, matrix metalloproteinase-9 and matrix metalloproteinase-13 in lesional skin of bullous pemphigoid. Int Arch Allergy Immunol 2006;139:104-13

RI PT

Sadeghi H, Lockmann A, Hund AC, Samavedam UK, Pipi E, Vafia K et al. Caspase-1independent IL-1 release mediates blister formation in autoantibody-induced tissue injury

SC

through modulation of endothelial adhesion molecules. J Immunol 2015;194:3656-63

Samavedam UK, Iwata H, Müller S, Schulze FS, Recke A, Schmidt E et al. GM-CSF

acquisita. J Immunol 2014;192:559-71

M AN U

modulates autoantibody production and skin blistering in experimental epidermolysis bullosa

Schwab I, Mihai S, Seeling M, Kasperkiewicz M, Ludwig RJ, Nimmerjahn F. Broad

TE D

requirement for terminal sialic acid residues and FcγRIIB for the preventive and therapeutic activity of intravenous immunoglobulins in vivo. Eur J Immunol 2014;44:1444-53

EP

Sevin M, Girodon F, Garrido C, de Thonel A. HSP90 and HSP70: Implication in inflammation processes and therapeutic approaches for myeloproliferative neoplasms.

AC C

Mediators Inflamm 2015;2015:970242

Shimanovich I, Mihai S, Oostingh GJ, Ilenchuk TT, Bröcker EB, Opdenakker G et al. Granulocyte-derived elastase and gelatinase B are required for dermal-epidermal separation induced by autoantibodies from patients with epidermolysis bullosa acquisita and bullous pemphigoid. J Pathol 2004;204:519-27

21

ACCEPTED MANUSCRIPT Singh A, Singh A, Sand JM, Bauer SJ, Hafeez BB, Meske L et al. Topically applied Hsp90 inhibitor 17AAG inhibits UVR-induced cutaneous squamous cell carcinomas. J Invest Dermatol 2015;135:1098-107

RI PT

Sitaru C, Mihai S, Otto C, Chiriac MT, Hausser I, Dotterweich B et al. Induction of dermalepidermal separation in mice by passive transfer of antibodies specific to type VII collagen. J

SC

Clin Invest 2005;115:870-8

Tukaj S, Kleszczyński K, Vafia K, Groth S, Meyersburg D, Trzonkowski P et al. Aberrant

M AN U

expression and secretion of heat shock protein 90 in patients with bullous pemphigoid. PLoS One 2013;8:e70496

Tukaj S, Tiburzy B, Manz R, de Castro Marques A, Orosz A, Ludwig RJ et al.

TE D

Immunomodulatory effects of heat shock protein 90 inhibition on humoral immune responses. Exp Dermatol 2014;23:585-90

EP

Tukaj S, Hellberg L, Ueck C, Hänsel M, Samavedam U, Zillikens D et al. Heat shock protein 90 is required for ex vivo neutrophil-driven autoantibody-induced tissue damage in

AC C

experimental epidermolysis bullosa acquisita. Exp Dermatol 2015a;24:471-3

Tukaj S, Zillikens D, Kasperkiewicz M. Heat shock protein 90: a pathophysiological factor and novel treatment target in autoimmune bullous skin diseases. Exp Dermatol 2015b;24:56771

Tukaj S, Węgrzyn G. Anti-Hsp90 therapy in autoimmune and inflammatory diseases: a review of preclinical studies. Cell Stress Chaperones 2016;21:213-8 22

ACCEPTED MANUSCRIPT Wu Z, Moghaddas Gholami A, Kuster B. Systematic identification of the HSP90 candidate

AC C

EP

TE D

M AN U

SC

RI PT

regulated proteome. Mol Cell Proteomics 2012;11:M111.016675

23

ACCEPTED MANUSCRIPT

FIGURE LEGENDS Figure 1. Prophylactic or therapeutic treatment with topical 17AAG reduces clinical and histological disease severity in mice with antibody transfer-induced EBA. Clinical scores, calculated as the percentage of surface area of the shaved back covered by EBA lesions, were

RI PT

lower in mice treated with topical 17AAG (500 nmol) on the shaved back either (a) starting before development of skin lesion (prophylactic setting; daily on days 0-12) or (b) when skin lesions had already developed (therapeutic setting; daily on days 8-15) compared with

SC

vehicle. Representative clinical presentations of vehicle- and 17AAG-treated mice at the end of the (c) prophylactic (day 12) and (d) therapeutic treatment (day 15). Semiquantitative

M AN U

histological evaluation of dermal neutrophil infiltration revealed lower scores (0-3: no, mild, moderate, and severe infiltration, respectively) in 17AAG-treated mice at the end of the (e) prophylactic and (f) therapeutic treatment. Representative images of hematoxylin-eosinstained back skin sections of vehicle- and 17AAG-treated mice at the end of the (g)

TE D

prophylactic and (h) therapeutic treatment; scale bars represent 100 µm. Significances were calculated with t-test and expressed as mean ± SEM (5-6 mice per group). *P<0.05,

EP

**P<0.01.

AC C

Figure 2. Topical application of 17AAG in a mixed prophylactic-therapeutic setting ameliorates clinical and histological disease severity without affecting autoantibody serum levels in mice with immunization-induced EBA. (a) Clinical scores, calculated as percentage of surface area of the ears or the total body covered by EBA lesions, were lower in mice treated with topical 17AAG (1000 nmol) on both ears (prophylactic setting; every other day over 10 weeks starting one day before immunization) and concurrently on any newly developed skin lesion of the body (therapeutic setting; every other day until week 10) compared with vehicle. (b) Representative clinical presentations of vehicle- and 17AAGtreated mice at the end of the treatment (week 10). (c) Semiquantitative histological 24

ACCEPTED MANUSCRIPT evaluation of dermal neutrophil infiltration revealed lower scores (0-3: no, mild, moderate, and severe infiltration, respectively) in 17AAG-treated mice at the end of the treatment. (d) Representative images of hematoxylin-eosin-stained ear skin sections of vehicle- and 17AAG-treated mice at the end of the treatment; scale bars represent 100 µm. (e) COL7-

RI PT

specific IgG autoantibody serum levels were similar in vehicle- and 17AAG-treated mice at the end of the treatment, as measured by ELISA. The dashed line represents the background level of the assay. Significances were calculated with t-test and expressed as mean ± SEM (8-

SC

9 mice per group). *P<0.05, **P<0.01.

M AN U

Figure 3. Topical 17AAG inhibits NFκB activation and affects MMP2, MMP9, MMP12, Flii, and Hsp70 expression in the skin of mice with experimental EBA. In comparison with vehicle controls, 17AAG-treated skin of mice showed reduced levels of (a, b) p-IκBα, (c-e) MMP2, (f-h) MMP9, (i-k) MMP12, and (l-n) Flii as well as (o-q) increased levels of

TE D

Hsp70 at the end of the antibody transfer- and immunization-induced EBA experiments, as analyzed by fluorescent immunohistochemistry and/or qRT-PCR. Immunofluorescent pictures with corresponding higher magnifications show representative staining of the respective

EP

markers (in red) and DAPI nuclear counterstaining (in blue) in skin samples from vehicle- and 17AAG-treated mice at the end of the treatment; scale bars represent 50 µm. Significances

AC C

were calculated with t-test and expressed as relative mean ± SEM (5-9 mice per group). *P<0.05, **P<0.01, ***P<0.001.

Figure 4. Inhibitory effects of topical 17AAG on cellular and molecular pathology are confirmed in a local model of experimental EBA. Short-term topical treatment with 17AAG (500 nmol) applied to both ears (daily for three days starting 1 day before local antiCOL7 antibody application causing locally restricted inflammation of the ear) led to (a) lower degree of dermal neutrophil infiltration based on semiquantitative histological evaluation (025

ACCEPTED MANUSCRIPT 3: no, mild, moderate, and severe infiltration, respectively), reduced cutaneous levels of (b) pIκBα, (c) MMP2, (d) MMP9, (e) MMP12, and (f) Flii as well as (g) increased skin expression of Hsp70 at the end of the 3-day treatment period compared with vehicle, as analyzed by fluorescent immunohistochemistry and/or qRT-PCR. Significances were calculated with t-test

AC C

EP

TE D

M AN U

SC

RI PT

and expressed as (relative) mean ± SEM (6 mice per group). *P<0.05, **P<0.01, ***P<0.001.

26

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT