Helicobacter pylori 26695 (Tomb et al., 1997) was grown for 2 days on blood agar culture plates containing 10% defibrinated sheep blood at 37°C under microaerophilic conditions, and a bacterial suspension was prepared in Brucella broth and adjusted to an optical density of 0.1 at 600 nm. H. pylori was grown at 37°C for 24 h (exponential phase) and 48 h (stationary phase) in Brucella broth supplemented with 10% fetal bovine serum (FBS; Invitrogen). In addition, Brucella broth adjusted to pH 5.5 or containing either urea [5 μM CO(NH₂)₂], nickel [250 μM NiCl₂] or iron [150 μM (NH₄)₂Fe(SO₄)₂⋅6H₂O], were tested as previously described (Contreras et al., 2003; Wen et al., 2003; Vannini et al., 2014). In all cases, the bacteria were incubated at 37°C with shaking and samples were collected 48 h post-inoculation for RNA extraction. Fold-change transcription was determined by calculating the relative expression of TA genes under different environmental conditions as compared to bacteria growing in Brucella broth.

For cloning of hp0967 and hp0968 coding regions, specific primers (Table 1) containing the NcoI (5′)/HindIII (3′) restriction sites were used to obtain PCR products, which were digested with NcoI and HindIII and then ligated into pBAD-Myc-HisA previously digested with the same restriction enzymes. HP0967-Myc-His₆ and HP0968-Myc-His₆ proteins were purified with Ni-nitrilotriacetic acid resin. Briefly, E. coli BL21(DE3) carrying the pBAD-HP0967-Myc-His₆ and pBAD-HP0968-Myc-His₆ plasmids were grown to mid-logarithmic phase. L(+)-arabinose (Sigma-Aldrich) was added at final concentration of 0.1%, and the bacteria were grown at 6 h at 30°C. The cells were then pelleted by centrifugation, resuspended in PBS buffer and ruptured by sonication. The suspension was centrifuged and the supernatant was filtered through a Ni-nitrilotriacetic acid agarose column (QIAExpress, Qiagen) pre-equilibrated with PBS buffer. After extensive washing with binding buffer containing 50 μM imidazole (100 ml), the bound protein was eluted with 500 μM imidazole. Fractions were analyzed by SDS-PAGE. Protein concentration was determined by the Bradford procedure. Aliquots of the purified protein were stored at -70°C.

Electrophoretic mobility shift assay (EMSA) experiments were performed as previously described (De la Cruz et al., 2007, 2013). Specific primers (Table 1) were used to amplify the hp0968–hp0967 promoter region by PCR. The PCR product (100 ng) was mixed with increasing concentrations of HP0968-Myc-His₆ at room temperature in the presence of the binding buffer 10X (400 μM HEPES, 80 μM MgCl₂, 500 μM KCl, 10 μM DTT, 0.5% NP40, and 1 μg/ml BSA). The reactions were incubated during 30 min at room temperature and then separated in 6% SDS-PAGE gels in Tris-Borate-EDTA buffer. The DNA bands were visualized with ethidium bromide staining.

HP0967-Myc-His₆ protein (0, 0.6, and 1.2 μM, respectively) was incubated with 1 μg of total H. pylori RNA in the presence of 50 μM Tris (pH 8.0) and 10 μM MgCl₂ in a final reaction volume of 20 μl. When necessary, HP0968-Myc-His₆ protein was added to analyze the antagonic effect of the antitoxin on the toxin activity. The reaction was allowed to proceed for 30 min at 25°C, and the samples were then loaded in a bleach agarose gel 2% as previously described (Aranda et al., 2012). RNA was visualized with ethidium bromide staining.

Total RNA was extracted from the different culture conditions using the hot phenol method (Jahn et al., 2008). Purification of RNA, and qRT-PCR were performed as previously reported (Ares et al., 2016). Briefly, DNA was removed with TURBO DNA-free (Ambion, Inc.) and the quality of RNA was assessed using a NanoDrop (ND-1000; Thermo Scientific) and a bleach agarose gel 2% as previously described (Aranda et al., 2012). cDNA was prepared using 1 μg of RNA, random hexamer primers (0.2 μg/μl), and M-MulV-RT (20 U/μl, reverse transcriptase of Moloney Murine leukemia Virus; Thermo Scientific). Specific primers were designed with the Primer3Plus software¹ and are listed in Table 1. The absence of contaminating DNA was controlled by lack of amplification products after 35 qPCR cycles. Control reactions with no template (water) and minus-Reverse Transcriptase were run with all reactions. 16S rRNA (HPrrnA16S) was used as a reference gene for normalization and the relative gene expression was calculated using the 2^-ΔCt method (Livak and Schmittgen, 2001).

AGS gastric epithelial cells were grown to about 75% confluence in RPMI-1640 medium containing 10% FBS, and washed thrice with PBS before adding fresh RPMI media with 10% FBS. H. pylori 26695 was grown in Brucella broth for 24 h, suspended in RPMI, and added to the AGS cell culture at a multiplicity of infection (MOI) of 100. Infected cells were incubated at 37°C under microaerophilic conditions for 0 and 6 h, and bacteria were recovered. RNA was extracted from bacteria to determine gene expression. Fold-change in gene transcription was determined by calculating the relative expression of TA genes with respect to bacteria at time 0 of infection.

Adhesion to abiotic surface (polystyrene) was analyzed using 96-well plates as previously described (Ares et al., 2016). H. pylori 26695 was grown for 2 days on blood agar culture plates containing 10% defibrinated sheep blood at 37°C under microaerophilic conditions, and a bacterial suspension was prepared in Brucella broth supplemented with 10% decomplemented FBS and adjusted to an optical density of 0.1 at 600 nm. Two-hundred μl of the bacterial suspensions were added per well into 96-well plates in octuplicate and incubated at different times at 37°C under microaerophilic conditions. Unbound bacteria were removed by washing the wells three times with PBS, and bound bacteria were stained with 1% Crystal Violet for 20 min. Wells were thoroughly rinsed thrice with PBS and the dye was solubilized in 100 μl of ethanol 70%. Finally, the amount of extracted Crystal Violet was determined by measuring color absorbance at OD₆₀₀ using a spectrophotometer plate reader (Thermo Scientific).

For determination of TA gene expression, the biofilm assay was performed in 6 well plates (3 ml) and supernatant (planktonic) and adhered (sessile) bacteria were recovered for RNA extraction. Fold-change in gene transcription was determined by calculating the relative expression of TA genes within biofilms (sessile bacteria) as compared to the planktonic bacteria.

Helicobacter pylori was grown in Brucella broth supplemented with 10% decomplemented FBS at 37°C for 48 h (stationary phase), with gentle shaking under microaerophilic conditions. The antibiotics ampicillin (Ap, 100 μg/mL), kanamycin (Km, 50 μg/mL), chloramphenicol (Cm, 30 μg/mL), or tetracycline (Tc, 10 μg/mL) were added to the liquid culture and incubated for 1 h as previously described (Christensen-Dalsgaard et al., 2010). Antibiotics concentrations were used as reported for E. coli and Salmonella enterica (Christensen-Dalsgaard et al., 2010; Maisonneuve et al., 2011; Silva-Herzog et al., 2015; Li et al., 2016). Fold-change in gene transcription was determined by calculating the relative expression of TA genes in the presence of each antibiotic as compared to bacteria growing in Brucella broth without antibiotics.

Analyzed sequence data and loci annotations of 260 H. pylori genomes were retrieved from the NCBI database² by a series of custom Perl scripts. In addition, the genomes of 57 Helicobacter non-pylori strains were included in the comparative analysis.

We also selected the sequences from known toxin and antitoxin modules that were classified into super-families as stated previously (Leplae et al., 2011) and compared their homology with previously referred Helicobacter genomes. Briefly, each TA protein pair was queried using PSI-BLAST (Altschul et al., 1997) under the following parameters: matrix = BLOSUM62, word size = 3, PSI-BLAST threshold = 0.005, expect threshold = 10, and without filtering low complexity regions. Hits were carefully examined and selected according to their functional annotation.

Furthermore, we employed the Toxin–Antitoxin Data Base (TADB³) (Shao et al., 2011) to extend our findings.

In order to illustrate the presence of every toxin and antitoxin module in all Helicobacter genomes, a gene content matrix (“heatmap” function) was built using the R program⁴ v3.2.4. Next, these paired loci were hierarchical clustered (“hclust” function, “ward.D” method) according to their loci-content similarity to build a sidelong dendogram.

Three pairs of type II TA proteins have been described in H. pylori: HP0315–HP0316, HP0892–HP0893, and HP0894–HP0895 (Han et al., 2011, 2013; Kwon et al., 2012; Im et al., 2014). Searching for TA-encoding homolog genes of different families, we identified hp0967, a gene that codes for a type II toxin belonging to the VapD protein family. Of note, an intergenic region was identified between hp0967 and hp0969 genes and was designated herein as antitoxin gene hp0968. In contrast, by analyzing the genome sequence of H. pylori J99 strain, a homolog of hp0968 (jhp0902) antitoxin gene was found. The analysis of the H. pylori 26695s genome revealed a stop codon in the coding region of hp0968 gene resulting in a peptide of 19 aminoacids. However, by comparing the nucleotidic sequences of all H. pylori strains deposited in the GenBank, no stop codon was found in this gene, suggesting a problem of annotation in the strain 26695. To demonstrate that hp0968 antitoxin gene is present and transcribed in 26695 strain, we performed RT-PCR experiments using primers that hybridize with TA genes as shown in Figure 1A. We found that both hp0968 and hp0967 genes were transcriptionally expressed and as expected, they presented a single mRNA similar to other type II TA systems (Figure 1B). The analyses of the internal controls for hp0316–hp0315, hp0893–hp0892, and hp0895–hp0894 TA genes also showed that they are genetically organized as bicistronic operons (Figure 1B).

We also amplified the coding region of both hp0967 and hp0968 genes of H. pylori 26695 and cloned them in a pBAD-Myc/HisA vector in order to overexpress and purify the proteins. The sequence analysis of the amplified hp0968 gene revealed that it did not present the stop codon, and that it encodes a functional protein of 93 aminoacids (12.8 kDa). In order to demonstrate the functionality of this antitoxin, EMSA experiments were performed incubating the promoter region of hp0968–hp0967 with the HP0968-Myc-His₆ protein. The DNA-protein complex was observed at 1.2 μM of HP0968-Myc-His₆ antitoxin (Figure 1C). In contrast, the aminoacid sequence of HP0967 presented a PIN domain associated with ribonuclease activity. Total RNA of H. pylori 26695 was incubated with 0.5 and 1.0 μM of HP0967-Myc-His₆ and RNA stability was determined. The assay showed that HP0967-Myc-His₆ (1.0 μM) degraded H. pylori RNA, suggesting that it is indeed a toxin (Figure 1D). In addition, the overexpression of HP0967 toxin affected the growth of E. coli (Figure 1E). To demonstrate the inhibitory effect of HP0968 antitoxin on HP0967 toxin, we added HP0968-Myc-His₆ protein to the ribonuclease activity assay. Ribonuclease activity of HP0967-Myc-His₆ toxin was inhibited in the presence of HP0968-Myc-His₆ antitoxin (Figure 1F). HP0968-Myc-His₆ alone protein did not affect RNA stability. Thus, our data strongly suggests that both HP0967 and HP0968 proteins constitute a new type II TA system of the Vap family in H. pylori.

In order to elucidate the environmental cues that may affect expression of type II TA systems we performed qRT-PCR experiments. Relative expression was determined during both exponential (24 h) and stationary phase (48 h). Interestingly, expression of all the type II TA genes was higher during stationary phase than in exponential phase (Figure 2A); although hp0316–hp0315 and hp0968–hp0967 showed the highest transcriptional levels (Figure 2A), and hp0893–hp0892 and hp0895–hp0894 were poorly expressed in this growth condition. Similarly, expression of cagA and vacA was found increased during stationary growth phase (Figure 2B), in agreement to previously reported data (Thompson et al., 2003; Boonjakuakul et al., 2005).

Of note, the effect of the environmental cues such as pH, urea, nickel, and iron in transcription of type II TA systems has not been reported for H. pylori, although, they are known to affect the expression of other virulence factors (Sachs et al., 2003; Haley and Gaddy, 2015; Keilberg and Ottemann, 2016). We found that during stationary growth phase type II TA genes showed different levels of expression in the presence of the tested cues. While nickel repressed the expression of hp0316–hp0315 (fivefold) and hp0895–hp0894 (threefold) (Figures 2C,E), all conditions tested triggered expression of hp0893–hp0892 TA genes (Figure 2D). Expression of hp0968–hp0967 was repressed in the presence of nickel (20-fold) and iron (10-fold), while urea caused a twofold increase in its expression (Figure 2F). As reaction controls, the transcription of cagA and vacA genes was determined under the same growth conditions. We found that transcription of both virulence genes was differentially regulated: urea slightly boosted transcription (1.5-fold), while acidic pH (2.2-fold), nickel (2.7-fold), and iron (2.7-fold) repressed them (Figure 2G).

Type II TA systems are reported to influence biofilm formation (Wang and Wood, 2011; Wen et al., 2014; Kedzierska and Hayes, 2016). We quantitatively analyzed the biofilm formation of H. pylori on a polystyrene surface at different time points measuring uptake of Crystal Violet. After 72 h of incubation, a mature biofilm was observed (Figure 3A). Thus, in order to evaluate the expression of type II TA genes in a mature biofilm, we determined their expression in both H. pylori planktonic and sessile cells at 72 h. Expression of all type II TA genes was found to be induced in biofilms-associated H. pylori, as compared to planktonic bacteria (Figure 3B). It should be noted that hp0968–hp0967 showed the highest induction in this environment (500-fold). In a mature biofilm, vacA transcription was significantly increased (23-fold), whereas cagA expression was not affected (Figure 3B). On the other hand, the contact with AGS cells significantly increased transcription of type II TA genes. Of note, hp0968–hp0967 genes presented the highest levels of expression over the other TA systems, even higher than expression of cagA and vacA (Figure 3C). These data strongly suggest that hp0968–hp0967 genes might be expressed in the host during stomach colonization.

We analyzed the expression of type II TA genes in the presence of antibiotics, and found that kanamycin and chloramphenicol (Km/Cm) increased the expression of hp0316–hp0315 (6.2/9.9-fold), hp0893–hp0892 (3.1/9.2-fold), and hp0895–hp0894 (4.34/9-fold) systems (Figure 4). In particular, chloramphenicol, kanamycin, and tetracycline stimulated the expression of hp0968–hp0967 genes (50-, 12-, and 2.5-fold, respectively) while ampicillin repressed them (2.8-fold) (Figure 4D). Interestingly, similar transcriptional profiling was found between the hp0893–hp0892 and hp0968–hp0967 genes in presence of chloramphenicol, kanamycin, and tetracycline (Figures 4B,D). Concerning the effect of antibiotics on the expression of cytotoxin genes, only vacA expression was boosted in the presence of kanamycin (10-fold) and chloramphenicol (23-fold) (Figure 4E). These results suggest that type II TA genes may have an important role in the response of H. pylori to the presence of antibiotics by H. pylori, particularly to chloramphenicol and kanamycin.

We performed PCR experiments to identify type II TA genes in a collection of 114 clinical H. pylori isolates. These bacteria were isolated from patients with diverse pathologies, such as non-atrophic gastritis, premalignant lesions, gastric ulcer, and gastric cancer. Whereas hp0968–hp0967 was the most prevalent type II TA system in these isolates (in over 70% of all isolates), hp0895–hp0894 genes were the least prevalent (in less than 50%) (Figure 5A). In addition, a Blast search in sequences deposited in GenBank⁵ using sequences of different TA families (included Vap), revealed the presence of type II TA genes in other H. pylori strains. TA genes of the Vap family were highly prevalent, presenting high identity and constituting a genetic signature inside H. pylori species (Figure 5B). In agreement with our findings in the clinical isolates, the hp0968–hp0967 genes presented the highest prevalence as compared to the other type II TA genes (Figures 5A–C). Interestingly, in both groups of H. pylori clinical and sequenced strains, the hp0968 antitoxin gene was more conserved than hp0967 toxin gene (Figures 5A–C).

Since the number of type II TA systems has been associated to bacterial pathogenicity (Georgiades and Raoult, 2011; De la Cruz et al., 2013), we studied the presence of the type II TA systems of different families in other Helicobacter species, and found that the presence of type II TA genes varied from one species to another. Species phylogenetically related to H. pylori such as H. acinonychis (big cats) and H. cetorum (marin mammals), contain the four type II TA systems (hp0316–hp0315, hp0893–hp0892, hp0895–hp0894, and hp0968–hp0967) and show high identity to the H. pylori sequences (Figures 5B,C). The presence of TA genes marked a separation of H. acinonychis and H. cetorum from the rest of Helicobacter species and they were grouped together with the H. pylori strains.