Analysis of the immunoglobulin A protease gene of Streptococcus sanguis

Abstract

The amino acid sequence T-P-P-T-P-S-P-S is tandemly duplicated in the heavy chain of human immunoglobulin A1 (IgA1), the major antibody in secretions. The bacterial pathogen Streptococcus sanguis, a precursor to dental caries and a cause of bacterial endocarditis, yields IgA protease that cleaves only the Pro-Thr peptide bond in the left duplication, while the type 2 IgA proteases of the genital pathogen Neisseria gonorrhoeae and the respiratory pathogen Haemophilus influenzae cleave only the P-T bond in the right half. We have sequenced the entire S. sanguis iga gene cloned into Escherichia coli. A segment consisting of 20 amino acids tandemly repeated 10 times, of unknown function, occurs near the amino-terminal end of the enzyme encoded in E. coli. Identification of a predicted zinc-binding region in the S. sanguis enzyme and the demonstration that mutations in this region result in production of a catalytically inactive protein support the idea that the enzyme is a metalloprotease. The N. gonorrhoeae and H. influenzae enzymes were earlier shown to be serine-type proteases, while the Bacteroides melaninogenicus IgA protease was shown to be a cysteine-type enzyme. The streptococcal IgA protease amino acid sequence has no significant homology with either of the two previously determined IgA protease sequences, that of type 2 N. gonorrhoeae and type 1 H. influenzae. The differences in both structure and mechanism among these functionally analogous enzymes underscore their role in the infectious process and offer some prospect of therapeutic intervention.