Dr. Lawyer and colleagues raise interesting points regarding the possible role of S. aureus in the autoimmunity to proteinase-3 in patients with Wegener granulomatosis. Because proteinase-3 shows significant homology with different members of the human serine proteases family such as T-cell granzymes, leukocyte elastase, cathepsin G and (chymo)trypsin, the possibility of shared epitopes exists [1, 2]. The induction of granzyme B in cytotoxic T lymphocytes by staphylococcal enterotoxin A would therefore be a possible link between antiproteinase-3 Wegener granulomatosis and S. aureus.

Evidence suggests, however, that the antiproteinase-3 antibody associated with Wegener granulomatosis is not directed to other serine proteases, including granzyme B. First, antigen-specific enzyme-linked immunosorbent assay (ELISA), immunoblot techniques, and immunoprecipitation studies with crude neutrophil preparations have failed to show reactivity of sera that is positive for antibodies against proteinase-3 with leukocyte elastase and cathepsin G [1, 3]. Second, on indirect immunofluorescence of ethanol-fixed leukocytes, only granulocytes and monocytes show positive fluorescence with antiproteinase-3-positive sera, with lymphocytes being consequently negative. Negative results were also obtained with cytotoxic T lymphocytes, normal natural killer cells, and large granular lymphocytes from a patient with T lymphocytosis and large granular lymphocyte cell lines [4]. These findings make cross-reactivity with granzymes present in granules of T cells an unlikely hypothesis. We are aware of no reports on antigen-specific tests with highly purified granzymes; these are necessary to answer this question definitively.

The possibility of S. aureus producing proteases with antigenic cross-reactivity with proteinase would be an attractive explanation of the role of this microorganism in the development of disease activity in Wegener granulomatosis, presuming that autoimmunity directed to proteinase-3 plays an important role in the pathophysiology of the disease. We have done some preliminary antigen-specific ELISA studies (unpublished data) on the possible relation between antiproteinase-3 antibodies and antibodies directed to S. aureus V8 proteinase to explain the link between S. aureus and active Wegener granulomatosis. Thirty-seven of 39 serum specimens from patients with the disease reacted with S. aureus V8 proteinase. However, we were unable to show a correlation between antibodies to S. aureus V8 proteinase and antiproteinase-3 antibodies. Furthermore, we could detect no changes in antiproteinase-3 antibodies after absorption of these sera with S. aureus V8 proteinase. Whether other proteins produced by S. aureus could have antigenic cross-reactivity with proteinase 3 is unknown. Sequencing of the genome of S. aureus would certainly be helpful in the search for possible candidates.