Fig. 1.

TraA contains a variable domain that determines recognition specificity. (A) Domain architecture of TraA: Type I signal sequence (SS); Cys-rich region and MYXO-CTERM are a putative stalk and a sorting tag, respectively (11). (B) Sequence alignment of the TraA variable domain from six recognition groups. Conserved residues are shaded red. Amino acid differences between the TraA^A96 and TraA^{M. fulvus} are in bold. Predicted secondary structures by I-TASSER (22) are indicated (loops, α-helices; arrows, β-strands). (C) Modeled 3D structure of the TraA^{M. fulvus} variable domain by I-TASSER where the highest C-score was selected. (D) Domain architecture of TraB; Thrombospondin type 3 (TSP-3, Pfam02412). β-Barrel prediction was made with BOCTOPUS2 (32).