Fig. 6.

Phylogenetic analysis of known and predicted glycan pyruvyltransferases and structures of Pvg1P and WbbZ. The phylogenetic tree was established by a multiple sequence alignment (SI Appendix, Fig. S8) of WbbZ orthologs encoded in polysaccharide-biosynthesis loci from bacteria with known polysaccharide structures (SI Appendix, Table S3) MAFFT was used to generate the alignment and build the tree, and bootstrap values greater than 20 (from 100 iterations) are shown. The pyruvate linkage is shown (omitting absolute configuration). Species names are abbreviated as follows: Ab; Acinetobacter baumannii, Bf; Bacteroides fragilis, Cv; Caudovirales sp., Ep; Erwinia pyrifoliae, Ec; Escherichia coli, Kp; Klebsiella pneumoniae, Pa; Paenibacillus alvei, Pm; Proteus mirabilis, Pv; Proteus vulgaris, Rm; Rhizobium meliloti, Sp; Schizosaccharomyces pombe, Sd; Shigella dysenteriae, Sp; Streptococcus pneumoniae, Xo; Xanthomonas oryzae. The entries in red in (A) have a solved structure or were modeled using AlphaFold and are shown in (B). Crystal structures for WbbZ (PDB: 6x1l) and Pvg1p (PDB: 5ax7) are shown, and WcfO and CsaB are AlphaFold models. Structures are colored in rainbow from blue (N terminus) to red (C terminus) to facilitate comparison. The predicted active site is expanded, and the residues shown as sticks are some of those conserved in the multiple sequence alignment in SI Appendix, Fig. S6.