Ectodomain Shedding Generates Neoepitopes on Collagen XVII, the Major Autoantigen for Bullous Pemphigoid

Abstract

As a type II transmembrane protein in basal keratinocytes, collagen XVII provides stable adhesion between epidermis and dermis in the skin. Its ectodomain can be shed from the cell surface, and autoantibodies in certain blistering diseases preferentially recognize the shed form. Major epitopes of collagen XVII are clustered within the juxtamembranous noncollagenous 16th A domain, and ectodomain shedding occurs within this region, suggesting that cleavage generates neoepitopes. However, the candidate cleavage sites have been controversial, and the mechanism of neoepitope generation is unclear. In this study, we investigated cleavage sites in the noncollagenous 16th A domain to understand the generation of neoepitopes and their pathological role. Polyclonal Abs recognizing the stretch Leu524-Gly532 preferentially reacted with the shed ectodomain, but not with the full-length form, indicating that a neoepitope was localized at this site. The neoepitope-specific Ab fixed complement and induced granulocyte-dependent dermal-epidermal separation in cryosections of normal human skin. The physiological cleavage sites were identified using mass spectrometry. N termini were found at Asp514, Leu524, Glu525, and Gly526, among which Asp514 and Glu525 were blocked by acetylation and pyroglutaminate. In silico prediction of B cell epitopes indicated that the antigenicity of the Leu524-Gly532 region increased substantially after shedding, regardless of the cleavage sites. Correspondingly, neoepitopes were found in the skin and blister fluids of patients with bullous pemphigoid, and bullous pemphigoid sera reacted with the peptide Leu524-Gly532. Taken together, these data demonstrate that physiological shedding of collagen XVII generates neoepitopes, which may serve as a target of blister-inducing autoantibodies.

Dynamic Interactions of Epidermal Collagen XVII with the Extracellular Matrix

“…16 We recently disclosed the major sheddase cleavage sites in the NC16A domain and the physiologic N-termini of the shed ectodomain. 17 All these are located C-terminally from the putative ␣6 integrin binding site. Thus, the integrin binding site is not part of the shed Ecto-ColXVII, and ␣6 integrin-collagen XVII interactions may rather regulate cellular events than be relevant for ECM binding.…”

“…The cleavage-site-specific rabbit polyclonal Ab HK139 was prepared as previously described 17 and was used at a concentration of 10 g/mL for IIF and 1 g/mL for immunoblotting. For detection of laminin 332 by IIF, the monoclonal Ab 6F12 to the ␤3 chain 25 was used in a 1:5000 dilution.…”

“…Stably expressing cells were selected under 200 g/mL of hygromycine B (Invitrogen) as previously described. 17 …”

“…16 This would be juxtamembranous and upstream of the recently disclosed major sheddase cleavage sites. 17 Thus, the interaction of the shed ectodomain of collagen XVII (Ecto-ColXVII) with ␣6␤4 integrin still remains elusive. A study of HaCaT keratinocytes proposed a p38MAPK-dependent functional relationship between collagen XVII and collagen IV and/or other components in Matrigel (BD Biosciences, Franklin Lakes, NJ) that regulates cell attachment and migration.…”

“…This problem has been overcome by a recently generated novel antibody (Ab), HK139, which specifically detects cleaved Ecto-ColXVII. 17 Herein, we used the Ab HK139 to follow shedding of the Ecto-ColXVII and its deposition in ECM in normal human keratinocytes (NHKs). We show that these processes are dynamic and that distinct ECM proteins play essential roles in interacting with collagen XVII in keratinocytes in vitro and in human skin in vivo.…”