Dermatitis herpetiformis is a papulovesicular skin disease with an associated gluten sensitive enteropathy. Both the skin and bowel disease improve with gluten restriction. The skin lesions are characterized by the accumulation of neutrophils in dermal papillae while immunopathology reveals IgA and complement in dermal papillae both involved and "uninvolved" skin. It has been suggested that cutaneous IgA is directed against gluten or possibly a cross-reacting antigen in the skin with complement fixation and subsequent chemotaxis of neutrophils. This paradigm of local pathogenesis has been obscured by three basic problems. 1) IgA and complement have been said to be equally distributed throughout the entire skin, implying that an additional qualitative or quantitative factor is needed to trigger the inflammatory cascade; 2) Gluten antigen has not been identified in the skin; 3) The origin, antigenic specificity and complement binding characteristics of the cutaneous IgA have not been established. The last problem is a result of the failure to isolate the IgA from either the circulation or the skin. We have hypothesized that: 1) The immunoreactants responsible for chemotaxis (possibly including gluten, specific complement components, IgA and other immunoglobulins) are qualitatively or quantitatively unique in or near active lesions. 2) Clearance of functionally inactive immunoreactants is slow, accounting for their persistence in previously involved skin (referred to as "uninvolved" skin by others). 3) Significant amounts of immunoreactants are absent or markedly decreased in never-involved skin. 4) Isolation of IgA from preparations of papillary dermis will allow its characterization. Preliminary data suggest that: 1) IgA is not equally distributed through the entire skin. 2) A relative quantification of IgA in skin can be accomplished using computerized analysis of fluorescence. 3) Gluten antigen may be present uniquely in areas of inflammation. We plan to test our hypotheses by: 1) obtaining biopsies from erythematous, previously involved and never-involved skin, 2) obtaining suction blister fluid from erythematous and previously involved skin, 3) analyzing skin and blister fluid for immunoglobulin, complement components, and chemotactic activity, 4) correlating chemotactic activity with immunoreactant deposition, 5) Lastly, we will isolate IgA from preparations of papillary dermis using polyacrylamide gel electrophoresis.