crescentus NA1000 were used. The figure was generated using the WebLogo server , and the height of the residue symbol indicates the degree of conservation within the ortologous groups. The
sequence numbering shown below the alignment corresponds to the respective C. crescentus NA1000 proteins. The complete representation of the motifs for the CzrA and NczA orthologous groups are shown in Additional file 2: Figure S1. (C) Cartoon representation of the CzrA structure model in which the conserved motifs MI-MV and the Loop are colored in yellow. The sub-domains DC, DN, PC1, PC2, PN1 and PN2 are Anlotinib colored in yellow, blue, dark green, red, violet and orange, respectively. The CzrA structure model was obtained using the Phyre2 program with CusA structure as a model (PDB: 3 k07, ). The structure was generated using PyMOL . The secondary structure elements indicated were predicted using the PHYRE2
program ; red ovals and amino acid sequences indicate α-helix; orange arrows and amino acid sequences indicate β-strands. In order to localize the identified signatures in the CzrA protein structure, we performed a homology MLN2238 order modeling analysis utilizing the structure of E. coli CusA as model (PDB: 3 k07), since it is the only metal-transporting RND protein structure so far available in the data bases. All of the motifs described above, with the exception of MV, are located in the periplasmic domain of CzrA structural model (Figure 6C). MV is located in TM8 in CzrA (Figure 6C), which in E. coli CusA suffers a significant conformational change when it binds Cu+ or Ag+, and was proposed to be involved in transmembrane signaling and in initiation of proton translocation across the membrane
. MI and MII are located in two close loops in the sub-domain PN1, MIII is located in the sub-domain DN and MIV is located in the sub-domain-PC2 (Figure 6C). The Etofibrate PC2 sub-domain in E. coli CusA was proposed to move, creating a cleft between PC1 and PC2 when CusA binds to Cu+ or Ag+. The most conspicuous difference between the CzrA and NczA groups is the length of the loop located in PN2, called here Large Loop for CzrA and Small loop for NczA. The periplasmic PN2 region is involved in the interaction between E. coli CusA and one molecule of the CusB dimer [25, 45]. When we superimpose the CzrA model on the CusAB2 complex structure (PDBID: 3NE5), the results suggest that the Large Loop could affect the interaction between CzrA and the adaptor protein (not shown). The predicted adaptors for the C. crescentus HME-RND systems, CzrB and NczB, share no significant amino acid sequence identity with CusB . Nevertheless, most of the interface residues at the sub-domain DC in CusA involved in the interaction with one molecule of the CusB dimer are conserved in the CzrA and NczA orthologs, although the two residues located in PN2, D155 and R147, are not conserved in members of either group.