Abstract
Acm2, the major autolysin of Lactobacillus plantarum WCFS1, was recently found to be O-glycosylated with N-acetylhexosamine, likely N-acetylglucosamine (GlcNAc). In this study, we set out to identify the glycosylation machinery by employing a comparative genomics approach to identify Gtf1 homologues, which are involved in fimbria-associated protein 1 (Fap1) glycosylation in Streptococcus parasanguinis. This in silico approach resulted in the identification of 6 candidate L. plantarum WCFS1 genes with significant homology to Gtf1, namely, tagE1 to tagE6. These candidate genes were targeted by systematic gene deletion, followed by assessment of the consequences on glycosylation of Acm2. We observed a changed mobility of Acm2 on SDS-PAGE in the tagE5E6 deletion strain, while deletion of other tagE genes resulted in Acm2 mobility comparable to that of the wild type. Subsequent mass spectrometry analysis of excised and in-gel-digested Acm2 confirmed the loss of glycosylation on Acm2 in the tagE5E6 deletion mutant, whereas a lectin blot using GlcNAc-specific succinylated wheat germ agglutinin (sWGA) revealed that besides Acm2, tagE5E6 deletion also abolished all but one other sWGA-reactive, protease-sensitive signal. Only complementation of both tagE5 and tagE6 restored those sWGA lectin signals, establishing that TagE5 and TagE6 are both required for the glycosylation of Acm2 as well as the vast majority of other sWGA-reactive proteins. Finally, sWGA lectin blotting experiments using a panel of 8 other L. plantarum strains revealed that protein glycosylation is a common feature in L. plantarum strains. With the establishment of these enzymes as protein glycosyltransferases, we propose to rename TagE5 and TagE6 as GtfA and GtfB, respectively.