Hsp90 forms a complex with MMP-12 as well as with MMP-2, but not MMP-9, in sera of EBA patients, implicating that these two basement-degrading factors are client proteins of Hsp90 and that their activity is dependent on its chaperone function. A close interaction between MMPs and Hsp90 secreted by cancer cells has been previously described in the context of tumor cell invasion and metastasis formation, and there is evidence that disruption of such Hsp90-MMP complexes by pharmacological Hsp90 blockade can inhibit the protease-dependent mechanism of tissue destruction by malignant cells that shares some similarity with the processes of blister formation in EBA (7,8). These findings might therefore have direct translational implications for the management of patients with EBA and related diseases.It has been previously shown that leukocyte-driven dermalepidermal separation in the ex vivo EBA model can be induced by serum from EBA patients with either the inflammatory or mechanobullous variant and that protease inhibitors can abrogate this effect regardless of the phenotype (5,6). While only sera of inflammatory EBA were used in the cryosection assay of this study, our co-immunoprecipitation experiments revealed that Hsp90-MMP complex formation can be found in both types of EBA. This suggests that Hsp90 inhibition may also be effective in mechanobullous EBA, although it remains speculative until a specific model for this disease subtype will be developed in the future.In summary, our findings add to the knowledge of the multimodal anti-inflammatory effects of Hsp90 blockade and implicate that Hsp90 is closely involved in the effector mechanisms of neutrophil-driven autoantibody-induced tissue damage, thus being a relevant therapeutic target in patients with neutrophilmediated autoimmune diseases such as inflammatory types of EBA.
AcknowledgementsThis work was supported by Deutsche Forschungsgemeinschaft (DFG) Excellence Cluster 'Inflammation at Interfaces' (EXC 306/2) and DFG KA 3438/1-1 as well as grants from the Medical Faculty of the University of L€ ubeck (E16-2012, E22-2013, and Focus Program 'Autoimmunity').
Author contributionMK, ST, TL, and RJL conceived and designed the experiments. ST, LH, CU, and MH performed the experiments. MK, ST, TL, RJL, DZ, and US analysed the data and provided reagents/materials/analysis tools. MK and ST wrote the paper.
Conflict of interestsThe authors have declared no conflicting interests.
Supporting InformationAdditional supporting data may be found in the supplementary information of this article. Figure S1. Cell viability. Data S1. Supplementary methods.