Cell Surface Glycoside Hydrolases of Streptococcus gordonii Promote Growth in Saliva.

The growth of the oral commensal Streptococcus gordonii in saliva may depend on a number of glycoside hydrolases (GHs), including three cell wall-anchored proteins that are homologs of pneumococcal β-galactosidase (BgaA), β-N-acetylglucosaminidase (StrH), and endo-β-N-acetylglucosaminidase D (EndoD). In the present study, we introduced unmarked in-frame deletions into the corresponding genes of S. gordonii DL1, verified the presence (or absence) of the encoded proteins on the resulting mutant strains, and compared these strains with wild-type strain DL1 for growth and glycan foraging in saliva. The overnight growth of wild-type DL1 was reduced 3- to 10-fold by the deletion of any one or two genes and approximately 20-fold by the deletion of all three genes. The only notable change in the salivary proteome associated with this reduction of growth was a downward shift in the apparent molecular masses of basic proline-rich glycoproteins (PRG), which was accompanied by the loss of lectin binding sites for galactose-specific Erythrina cristagalli agglutinin (ECA) and mannose-specific Galanthus nivalis agglutinin (GNA). The binding of ECA to PRG was also abolished in saliva cultures of mutants that expressed cell surface BgaA alone or together with either StrH or EndoD. However, the subsequent loss of GNA binding was seen only in saliva cocultures of different mutants that together expressed all three cell surface GHs. The findings indicate that the growth of S. gordonii DL1 in saliva depends to a significant extent on the sequential actions of first BgaA and then StrH and EndoD on N-linked glycans of PRG.

The ability of oral bacteria to grow on salivary glycoproteins is critical for dental plaque biofilm development. Little is known, however, about how specific salivary components are attacked and utilized by different members of the biofilm community, such as Streptococcus gordonii. Streptococcus gordonii DL1 has three cell wall-anchored glycoside hydrolases that are predicted to act on host glycans. In the present study, we introduced unmarked in-frame deletions in the corresponding genes, verified the presence (or absence) of encoded proteins on the resulting mutant strains, and compared these strains with wild-type DL1 for growth and glycan foraging in saliva. The results indicate that the growth of S. gordonii DL1 depends to a significant extent on sequential action of these cell surface GHs on N-linked glycans of basic proline-rich salivary glycoproteins, which appears to be an essential first step in salivary glycan foraging.