Identification and characterization of a HEPN-MNT family type II toxin-antitoxin in Shewanella oneidensis

Abstract

Toxin-antitoxin (TA) systems are prevalent in bacteria and archaea. However, related studies in the ecologically and bioelectrochemically important strain Shewanella oneidensis are limited. Here, we show that SO_3166, a member of the higher eukaryotes and prokaryotes nucleotide-binding (HEPN) superfamily, strongly inhibited cell growth in S. oneidensis and Escherichia coli. SO_3165, a putative minimal nucleotidyltransferase (MNT), neutralized the toxicity of SO_3166. Gene SO_3165 lies upstream of SO_3166, and they are co-transcribed. Moreover, the SO_3165 and SO_3166 proteins interact with each other directly in vivo, and antitoxin SO_3165 bound to the promoter of the TA operon and repressed its activity. Finally, the conserved Rx4-6H domain in HEPN family was identified in SO_3166. Mutating either the R or H abolished SO_3166 toxicity, confirming that Rx4-6H domain is critical for SO_3166 activity. Taken together, these results demonstrate that SO_3166 and SO_3165 in S. oneidensis form a typical type II TA pair. This TA pair plays a critical role in regulating bacterial functions because its disruption led to impaired cell motility in S. oneidensis. Thus, we demonstrated for the first time that HEPN-MNT can function as a TA system, thereby providing important insights into the understanding of the function and regulation of HEPNs and MNTs in prokaryotes.

Figure 1

SO_3166 is the toxin and SO_3165 is the antitoxin. E. coli K12 BW25113 hosts containing pCA24N-based constructs were cultured in LB medium supplemented with 30 μg/μl chloramphenicol and 1 mM IPTG (added at OD₆₀₀ of 0.1). Cell growth (A) and viability (CFU/mL) (B) were tested at the time points indicated. The cells induced for 5 h were serially diluted, dropped onto LB plates and incubated at 37°C for 16 hr (C). MR-1 hosts carrying the pHGE-based plasmids were cultured in LB with 50 μg/mL kanamycin, and 1 mM IPTG were added at OD₆₀₀ ∼ 0.1. Cell growth (D) and viability (E) were tested at the time points indicated. (F) MR-1 hosts carrying the pHGE-based plasmids were streaked onto LB plates with 50 μg/mL kanamycin with or without 1 mM IPTG, and were incubated for 16 hr. Data are from two independent cultures, and standard deviations are shown in A, B, D and E.

Figure 2

Co-transcription of SO_3165 and SO_3166. (A) Primers were designed to amplify the whole open reading frame of SO_3165 (I), SO_3166 (II) and the total region that covers the start codon of SO_3165 and stop codon of SO_3166 (III). Approximately 150 ng of cDNA reverse transcribed from MR-1 RNA was used as templates to amplify the three fragments. The same amount of MR-1 genomic DNA and RNA were used as the positive and negative controls, respectively. M indicates the DNA ladder. (B) Primer extension was conducted using the 5′ FAM-labeled reverse transcriptional primer FAM-SO_3166-r and total RNA isolated from E. coli K12 BW25113 carrying the pBS(Kan)-SO_3165-3166 plasmid. The X-axis indicates the length of the cDNA with FAM and the Y-axis indicates intensity of the fluorescence signal.

Figure 3

SO_3165 and SO_3166 form a complex in vivo. (A) Toxin SO_3166 with a C-terminal hexahistidine tag (His-tagged) was constructed together with untagged antitoxin SO_3165 to generate pET28b-SO_3165-3166-CHis. After induction with 1 mM IPTG, the 16.00 kDa SO_3166-CHis and a 15.57 kDa SO_3165 were induced (lane 2). The negative control was included when no IPTG was added (lane 1). During purification, SO_3166-CHis and SO_3165 were co-purified (lane 3). The protein marker (M) was loaded in lane 4. (B) SO_3165 was induced and purified via pET28b-SO_3165 with IPTG induction under the same condition described in (A). The purified SO_3165 cannot bind to the Ni-NTA agarose beads (lane 4). (C). SO_3165-CHis (16.39 kDa) was induced and purified via pET28b-SO_3165-CHis with IPTG induction followed the same condition described in (A). Dimerization of SO_316 was observed (lane 4), and the addition of the reducing agent dithiothreitol (DTT) greatly reduced the dimerization (lane 5).

Figure 4

Antitoxin SO_3165 binds to the promoter of the SO_3165–3166 operon. (A) The sequence of the promoter DNA used for EMSA (296-nt upstream of the translational start of the operon). The double underlines indicated the primers used for PCR amplification for the promoter region. The palindromic sequences are highlighted in blue, while the sequence between the two arms are highlighted in violet. (B) EMSA results demonstrating that purified SO_3165-CHis binds to the 296 bp biotin-labeled promoter DNA of SO_3165. The binding increases with the increasing concentrations of SO_3165-CHis protein. (C) Mid-log-phase cells of the indicated strains carrying the integrated reporter system (300-nt upstream of the translational start of the operon) were collected and tested for β-galactosidase activity. Error bars represent standard deviations for triplicate cultures. Asterisks represent a statistically significant difference between the wild-type and indicated mutants (P < 0.001; n = 3).

Figure 5

Key residues for determining SO_3166 toxicity. (A) Toxicity results of single-site mutagenesis of the R and H in Rx4H region and an adjacent tyrosine of toxin SO_3166 in the pCA24N-SO_3166 plasmid in DH5α. (B) Predicted 3-D structure of SO_3166. The conserved domain Rx4H is situated at the end of one helix (yellow). Other residues obtained by epPCR assay that affected SO_3166 toxicity are shown in red. (C) Toxicity test of seven strains expressing different mutated SO_3166 proteins obtained by epPCR. WT indicates the wild-type SO_3166 protein; the remainders are mutated proteins. The number in the mutated protein indicates the position of the amino acid in SO_3166. Overnight cultures were streaked on 30 μg/mL chloramphenicol LB plates with or without 0.5 mM IPTG. Two independent cultures were evaluated for each; only one representative image is shown here.

Figure 6

SO_3166-SO_3165 confers plasmid stability in E. coli. E. coli K12 BW25113 harboring plasmids pCA24N and pCA24N-SO_3165-3166 were used for the plasmid stability assay. Overnight cultures were 1% diluted in LB medium without any antibiotics, then incubated at 37°C for 12 hr. This process was repeated every 12 hr for 7 days. Three independent cultures were conducted, and the data are shown as means ± standard deviations.

Figure 7

SO_3166-SO_3165 represses motility in S. oneidensis. (A) In-frame deletion mutant of the toxin gene SO_3166 and the SO_3165-SO_3166 operon. Primer LF indicates the long forward primer and SF indicates the short forward primer. Primers SR and LR indicate the short reverse and long reverse primers, respectively. PCR products are indicated with the expected sizes using genomic DNA from the wild-type (WT) and the deletion mutants. (B) Swimming motility test for the SO_3165-3166 deletion mutant versus WT (left plate), and for wild-type MR-1 co-expressing SO_3165 and SO_3166 versus empty vector (right plate). Two independent cultures were evaluated for each; only one representative image is shown here.