This content is PDF only. Please click on the PDF icon to access.
Abstract
PURPOSE: A unique cornea-associated antigen (CO-Ag) has been purified previously from stromal extracts. The protein is the target for autoantibodies in patients with Mooren's ulcer. In this study, the amino acid sequence of CO-Ag was analyzed and the structure-function properties of CO-Ag was determined. METHODS: Purified CO-Ag was subjected to N-terminal sequencing by automated Edman degradation. Binding of calcium (Ca2+) to CO-Ag was measured by a direct (45)Ca2+ -binding assay. RESULTS: The complete amino acid sequence of CO-Ag has been determined. The protein contains 70 amino acids in a single chain and lacks cysteine, tryptophan, and methionine residues. A computerized data base search of protein and nucleic acid sequences revealed strong homology to the Ca2+ -binding proteins of the S-100 family. The sequence of CO-Ag shows a high homology with calgranulin C (CaG-C) previously purified from pig granulocytes. The functional Ca2+ -binding sites of CO-Ag and CaG-C were different based on homology with known Ca2+ -binding domains and their Ca2+ -binding properties. There are three amino acid substitutions in the N-terminal Ca2+ -binding domain. Differences were functionally conserved and compatible, with minimum single-base changes in the codon structures. The greatest difference was located in the C-terminal Ca2+ -binding domain. Five consecutive amino acid changes from D63-K-K-G-A67 in CO-Ag to M63-Q-D-E-Q67 occurred in CaG-C. These differences alter the structure of CO-Ag, which no longer can bind Ca2+ ions. The existence of this nonfunctional Ca2+ -binding site was corroborated by its Ca2+ -binding properties. The number of Ca2+ -binding sites for the CO-Ag sub-unit is approximately half that of the CaG-C monomer, although these two proteins have a similar low binding constant of approximately 2 x 10(-4) M. CONCLUSIONS: These results suggest that CO-Ag is a new member of the Ca2+ -binding protein of the S-100 family heretofore undescribed in the cornea. Sequence data provide an important framework to search for sequence similarity with microbial proteins as possible substrates for molecular mimicry and for the identification of possible pathogenic epitopes in CO-Ag.