Purification and characterization of the major antigen WI-1 from Blastomyces dermatitidis yeasts and immunological...

The lack of well-defined antigens from Blastomyces dermatitidis has hampered the ability to reliably diagnose human infection and study the immunobiology of blastomycosis. We recently discovered a novel surface protein on B. dermatitidis yeasts, designated WI-1, and demonstrated it to be a key antigenic target of humoral and cellular responses during infection. In the present article, we purified and characterized WI-1 and compared it immunologically with the only Blastomyces antigen commercially available, A antigen. WI-1 was purified by high-performance liquid chromatography over a DEAE-cellulose column. It eluted from the column at a point on the salt gradient corresponding to 460 to 490 mM NaCl, reflecting its acidic pI of approximately equal to 5.2. Purified WI-1 had a molecular mass of 120 kDa and contained a large amount of cysteine (85 residues) and aromatic amino acids but undetectable carbohydrate. In contrast, A antigen had a molecular mass of 135 kDa and contained 37% carbohydrate. Immunological comparison of the two antigens showed that, when radiolabeled, WI-1 was more reactive with anti-Blastomyces antisera than A antigen but did not cross-react with anti-Histoplasma antisera. Proteinase digestion of WI-1 eliminated its recognition by anti-WI-1 and anti-Blastomyces antisera. Proteinase treatment of A antigen had no effect on its recognition by anti-Blastomyces or anti-Histoplasma antisera, but periodate treatment abolished recognition by anti-Histoplasma antisera, indicating that the cross-reactive determinant(s) of A antigen is displayed on the accompanying carbohydrate. In further studies, anti-WI-1 antiserum reacted with A antigen and, conversely, anti-A antiserum and monoclonal antibodies (MAbs) reacted with WI-1, indicating a shared determinant on the two antigens. A recombinant 25-amino-acid repeat, recently cloned from WI-1 and found to be the major target of antibody recognition of WI-1, reacted strongly with anti-A antiserum and MAbs. In MAb competition tests, MAbs specific for the 25-residue repeat abolished binding of anti-A antiserum to A antigen. In antigen inhibition tests, the recombinant repeat abolished binding of anti-A antiserum to A antigen. These results demonstrate that the repeat is the major site of antibody recognition of both WI-1 and A antigen and that the recombinant, nonglycosylated peptide could replace either native antigen in formatting better diagnostic tests for blastomycosis. Moreover, they suggest that producing fungal protein antigens as nonglycosylated peptides in a procaryotic expression system may circumvent problems of antigen cross-reactivity that are due to posttranslational modification.