Zonula occludens toxins and their prophages in Campylobacter species

The presence of Zot proteins in Campylobacter species was examined by searching for proteins in Campylobacter species that have the Zot family domain in multiple protein databases including NCBI protein database, InterPro and Pfam [22‐24]. The Zot proteins in these databases were annotated based on the presence of Zot family domain.

It is shown in the InterPro database that the Zot family proteins (InterPro entry identity: IPR008900) belong to p-loop containing nucleoside triphosphate hydrolase (p-loop NTPase) superfamily. The proteins of p-loop NTPase superfamily have Walker A and Walker B motifs [25]. In this study, we examined the presence of Walker A and Walker B motifs in Campylobacter Zot proteins by protein alignment using Clustal Omega software [26]. The Walker A motif has a sequence of GxxxxGK[S/T], where x is any residue and the Walker B motif has a sequence of hhhh[D/E], where h is a hydrophobic residue [25].

To examine the genetic relationship between the zot genes in different Campylobacter species, the nucleotide sequences of the zot genes in Campylobacter species identified above were obtained from NCBI genome database. These sequences were used to generate a phylogenetic tree using the maximum likelihood method implemented in molecular evolutionary genetics analysis software version 6.0 [27]. In order to differentiate the zot genes in different Campylobacter species, the zot genes found in different Campylobacter species were indicated by the last four digits of their locus tag numbers.

To examine whether the zot genes in different Campylobacter species are carried by prophages similar to that in C. concisus, the zot-containing prophages in these Campylobacter species were identified by examination of the genes adjacent to their zot genes. The prophages were defined based on the presence of integrase, hypothetical proteins and attachment sites [28]. The attachment sites were identified by the presence of repetitive sequences located at both ends of the prophages [28].

The proteins in the identified Campylobacter prophages in this study were compared with the corresponding proteins in C. concisus prophages CON_phi2 and CON_phi3. The comparison was performed using Clustal Omega [26]. Proteins sharing more than 40 % sequence identity were considered to have high similarities and were recorded [29].

To examine whether the zot containing prophages identified in Campylobacter species are similar to previously reported prophages, proteins of Campylobacter prophages were compared with viral proteins in the NCBI non-redundant protein sequence database (taxonomy identity for viruses: 10,239) using protein BLAST with default settings [22]. The identified viral proteins with E-values lower than 10 were noted. For prophage proteins which shared sequence similarities with multiple viral proteins, the viral proteins with the lowest E-values were noted. The full length sequences of the identified viral proteins were then aligned with the Campylobacter prophage proteins using Clustal Omega [26]. The protein identities were calculated by dividing the number of identical amino acids by the total number of amino acids in proteins from Campylobacter prophages. The Campylobacter prophage proteins were also compared with viral proteins in the virus pathogen database and analysis resource (ViPR) using BLAST with a cut-off E-value of 10. In addition to C. concisus, the zot gene was also found in other bacterial species such as Vibrio cholerae and Neisseria meningitis [30, 31]. The identities between Campylobacter Zot, V. cholerae Zot and N. meningitidis Zot proteins were also compared in this study.

Secreted proteins in zot-containing prophages in different Campylobacter species were predicted using SignalP 4.1 and SecretomeP 2.0. The software SignalP 4.1 predicts secreted proteins based on the presence of the signal peptide at the N-terminus [32]. The software SecretomeP 2.0 predicts non-classical (not signal peptide triggered) protein secretion based on analysis of post-translational and localizational aspects of the proteins [33]. Transmembrane proteins were predicted using the software Phobius [34].

Twelve Zot proteins were found in nine Campylobacter species/subspecies. Three Campylobacter species including C. concisus, C. ureolyticus and C. corcagiensis had two Zot proteins and the remaining six Campylobacter species/subspecies including C. gracilis, Campylobacter jejuni subsp. doylei, Campylobacter jejuni subsp. jejuni, Campylobacter hyointestinalis subsp. hyointestinalis, C. hyointestinalis subsp. lawsonii and C. iguaniorum had one Zot protein (Table 1).

The Zot family domains were localized at the N-terminal side of the identified Campylobacter Zot proteins, prior to the transmembrane domains. The Zot family proteins had p-loop NTPase domains and the entry identity for p-loop NTPase domain was IPR027417 in InterPro database. We identified the Walker A and Walker B motifs in Campylobacter Zot proteins, which were at the N-terminal side of the Zot proteins (Fig. 1).

The Campylobacter zot genes formed two clusters. Cluster 1 contained three zot genes, including C. concisus zot2276, C. ureolyticus zot3935 and C. corcagiensis zot6485. Cluster 2 contained nine Campylobacter zot genes (Fig. 2).

The Zot proteins encoded by Cluster 1 and Cluster 2 Campylobacter zot genes were referred to as Zot_{CampyType_1} and Zot_{CampyType_2} proteins respectively. The three Zot_{CampyType_1} proteins had high similarities; they shared 171 identical amino acids and 77 conservative mutations (Fig. 3). The Zot_{CampyType_2} proteins were less similar to each other as compared to Zot_{CampyType_1} proteins; they had 71 identical amino acids and 65 conservative mutations (Fig. 4).

We identified the zot-containing prophages in different Campylobacter species. Each of these prophages began with an integrase and had a number of hypothetical proteins (Fig. 5; Additional file 1). The attachment sites were found (Table 2). These prophages were inserted within either tRNA-Met or tRNA-Ser genes, except for URE_phiZB, which was inserted into tRNA-Leu gene (Table 2). For the prophage in C. iguaniorum, two tRNA genes were found after the integrase, suggesting that multiple insertions have occurred.

Prophages containing Zot_{CampyType_1} proteins had high similarities to CON_phi2. Eight proteins in URE_phiZA and nine proteins in COR_phiZA were found to have more than 40 % identities (41-73 %) with proteins in CON_phi2 (Fig. 5a; Additional file 1). Proteins in prophages containing Zot_{CampyType_2} proteins were more diverse. However, three to five proteins in these prophages had more than 40 % identities with that in CON_phi3 (40–72 %) (Fig. 5b; Additional file 1).

The zot-containing Campylobacter were compared with known viral proteins in NCBI non-redundant protein sequence database. The proteins within each individual prophage showed low similarities to multiple phage proteins, except for CCC13826_0188 in CON_phi3 that shared 43 % identity with a phage transferase from an uncultured phage (Additional file 2). The Zot proteins in Campylobacter prophages had low similarities with the Zot proteins in V. cholerae and N. meningitidis (15–21 %) (Additional files 3, 4). None of the Campylobacter prophage proteins shared significant similarities with viral proteins in ViPR database. These data suggest that the zot-containing prophages in Campylobacter species are new prophages that have not been previously characterized.

Proteins secreted via both classical secretory pathway (with signal peptides) and non-classical secretory pathway (without signal peptides), as well as transmembrane proteins were found in all Campylobacter prophages (Fig. 5).