The Syk Tyrosine Kinase Is Required for Skin Inflammation in an In Vivo Mouse Model of Epidermolysis Bullosa Acquisita

Abstract

The inflammatory form of epidermolysis bullosa acquisita is caused by autoantibodies against type VII collagen (C7), a component of the dermal-epidermal junction. We have previously shown that myeloid Src family kinases mediate skin inflammation triggered by anti-C7 antibodies. Here we identify the Syk tyrosine kinase as a critical component of autoantibody-induced skin inflammation downstream of Src family kinases. Immobilized C7-anti-C7 immune complexes triggered neutrophil activation and Syk phosphorylation in a Src family kinase-dependent manner. Bone marrow chimeric mice lacking Syk in their hematopoietic compartment were completely protected from skin inflammation triggered by anti-C7 antibodies despite normal circulating anti-C7 levels. Syk deficiency abrogated the accumulation of CXCL2, IL-1β, and leukotriene B₄ at the site of inflammation and resulted in defective in vivo neutrophil recruitment. Syk^-/- neutrophils had a normal intrinsic migratory capacity but failed to release CXCL2 or leukotriene B₄ upon activation by immobilized C7-anti-C7 immune complexes, indicating a role for Syk in the amplification of the inflammation process. These results identify Syk as a critical component of skin inflammation in a mouse model of epidermolysis bullosa acquisita and as a potential therapeutic target in epidermolysis bullosa acquisita and other mechanistically related inflammatory skin diseases such as bullous pemphigoid.

Figure 1

Type VII collagen-containing immune complexes trigger neutrophil activation and Syk phosphorylation through Src family kinases. (a) Respiratory burst of WT mouse neutrophils plated on surfaces treated with His-C7 and/or anti-C7. (b) His-C7–anti-C7 immune complexes trigger spreading of WT neutrophils. (c) Specificity of the His-C7‒anti-C7- and HSA–anti-HSA-induced responses. (d–e) Src family kinases are essential for His-C7‒anti-C7 immune complex-triggered superoxide release and Syk-phosphorylation. Kinetic curves and graphs in a, c, and d show mean and standard error of the mean of 2 to 3 independent experiments. In panels a and d, control data points were subtracted. Panels b and e are representative of 2 to 3 independent experiments. See the text for actual P-values. Scale bar = 20 μm. 3 × SFK KO, Hck^–/–Fgr^–/–Lyn^–/–; C7, collagen type VII; HSA, human serum albumin; IP, immunoprecipitation; KO, knockout; min, minute; WT, wild-type.

Figure 2

Characterization of WT and Syk^–/– bone marrow chimeras and their neutrophils. (a) Flow cytometric analysis of the donor-specific CD45.2 epitope in peripheral blood neutrophils tested 4 weeks after BM transplantation. (b) Immunoblot analysis of Syk expression in various cell types from WT and Syk^‒/‒ BM chimeras. (c) Circulating neutrophil (PMN) counts in WT and Syk^‒/‒ chimeras. (d) Average neutrophil number isolated from the BM of the different chimeras. (e) Cell surface molecule expression of WT and Syk^‒/‒ BM-isolated neutrophils. Panels a, b, and e are representative of 3 to 10 independent experiments. Graphs in panel c and d show mean and standard error of the mean from 9 to 11 mice. See the text for actual P-values. BM, bone marrow; PMN, polymorphonuclear cell; WT, wild-type.

Figure 3

Syk is required for experimental epidermolysis bullosa acquisita. Skin disease was triggered in WT or Syk^–/– bone marrow chimeras by anti-C7 antibodies and was followed by (a) photographing the heads, the trunks, and the ears and clinical assessment of (b) the body surface affected and (c) the overall disease severity. (d) Serum anti-C7 levels were tested by ELISA. (e) Hematoxylin and eosin stained ear histological sections are presented from WT and Syk^–/– chimeras. The right images were magnified from the pictures seen on the left. Original magnification ×10. (a, e) Representative images or (b, c) mean and standard error of the mean from 3 control and 12 to 13 anti-C7–treated mice per genotype from 3 independent experiments are shown. Panel d shows mean and standard error of the mean from 2 to 3 control and 9 to 11 anti-C7–treated mice from 2 independent experiments. See the text for actual P-values. Scale bar = 100 μm. C7, collagen type VII; OD, optical density; WT, wild-type.

Figure 4

Syk is essential for forming the inflammatory environment. (a) The accumulation of neutrophils in the ear upon anti-C7 injection was determined by flow cytometry. (b, c) Mixed bone marrow chimeras with CD45.1-expressing WT and CD45.2-expressing WT or Syk^–/– hematopoietic cells were injected with anti-C7 antibodies and neutrophil accumulation was determined by flow cytometric analysis. In panel b, each dot represents an individual ear, and panel c shows the relative migration of WT and Syk^–/– neutrophils. (d–f) Decreased in vivo CXCL2, IL-1β, and LTB₄ levels in the ears of Syk-deficient chimeras. Graphs represent mean and standard error of the mean from (a) 3 to 5 or (c) 2 to 6 mice and (d–f) 3 to 4 control or 5 to 7 anti-C7–treated mice per group from 2 to 3 independent experiments. See the text for actual P-values. C7, collagen type VII; LTB₄, leukotriene B₄; PMN, polymorphonuclear cell; WT, wild-type.

Figure 5

Syk is required for neutrophil responses on immobilized His-C7‒anti-C7 immune complexes. (a) Respiratory burst, (b) cell spreading, (c) CXCL2, and (d) LTB₄ release of WT and Syk^‒/‒neutrophils. Curves in panel a show mean and standard error of the mean of 3 independent experiments; control data points were subtracted. Panel b is representative of 3 independent experiments. Graphs in c and d show mean and standard error of the mean from 3 independent experiments. See the text for actual P-values. Scale bar = 40 μm. C7, collagen type VII; His-C7, His-tagged fragment of the immunodominant NC1 domain of murine C7; LTB₄, leukotriene B₄; min, minute; WT, wild-type.