Structural Determinants for Antitoxin Identity and Insulation of Cross Talk between Homologous Toxin-Antitoxin Systems

ABSTRACT

Toxin-antitoxin (TA) systems are ubiquitous in bacteria and archaea, where they play a pivotal role in the establishment and maintenance of dormancy. Under normal growth conditions, the antitoxin neutralizes the toxin. However, under conditions of stress, such as nutrient starvation or antibiotic treatment, cellular proteases degrade the antitoxin, and the toxin functions to arrest bacterial growth. We characterized the specificity determinants of the interactions between VapB antitoxins and VapC toxins from nontypeable Haemophilus influenzae (NTHi) in an effort to gain a better understanding of how antitoxins control toxin activity and bacterial persistence. We studied truncated and full-length antitoxins with single amino acid mutations in the toxin-binding domain. Coexpressing the toxin and antitoxin in Escherichia coli and measuring bacterial growth by dilution plating assayed the ability of the mutant antitoxins to neutralize the toxin. Our results identified two single amino acid residues (W48 and F52) in the C-terminal region of the VapB2 antitoxin necessary for its ability to neutralize its cognate VapC2 toxin. Additionally, we attempted to alter the specificity of VapB1 by making a mutation that would allow it to neutralize its noncognate toxin. A mutation in VapB1 to contain the tryptophan residue identified herein as important in the VapB2-VapC2 interaction resulted in a VapB1 mutant (the T47W mutant) that binds to and neutralizes both its cognate VapC1 and noncognate VapC2 toxins. This represents the first example of a single mutation causing relaxed specificity in a type II antitoxin.

IMPORTANCE Toxin-antitoxin systems are of particular concern in pathogenic organisms, such as nontypeable Haemophilus influenzae, as they can elicit dormancy and persistence, leading to chronic infections and failure of antibiotic treatment. Despite the importance of the TA interaction, the specificity determinants for VapB-VapC complex formation remain uncharacterized. Thus, our understanding of how antitoxins control toxin-induced dormancy and bacterial persistence requires thorough investigation of antitoxin specificity for its cognate toxin. This study characterizes the crucial residues of the VapB2 antitoxin from NTHi necessary for its interaction with VapC2 and provides the first example of a single amino acid change altering the toxin specificity of an antitoxin.

INTRODUCTION

Toxin-antitoxin (TA) systems are ubiquitous in bacteria and archaea, where they play a pivotal role in the establishment and maintenance of dormancy. Originally discovered as addictive elements involved in postsegregational killing during cell division, they have since been identified as playing a critical role in responding to stress (1–3). TA systems are typically encoded by a single operon that produces a stable protein toxin and its cognate labile antitoxin, with the gene encoding the antitoxin located upstream of the toxin in most cases (4). Under normal growth conditions, the antitoxin, which may be a protein (types II, IV, and V) or an RNA (types I and III), neutralizes the protein toxin or its expression, thus preventing its toxic activity. The toxin-antitoxin complex also autoregulates the TA operons by repressing transcription (5, 6). However, under conditions of stress, such as nutrient starvation or treatment with antibiotics, the antitoxin is degraded by proteases, such as Lon or Clp, which release the toxin from the complex, allowing the toxin to arrest cellular growth (4, 6, 7). This growth arrest leads to a dormant state, which could allow for antibiotic tolerance and the establishment of latent infections (8–10). Previous studies have found that gradually deleting 10 of the 12 type II TA systems in Escherichia coli led to significantly decreased persistence during antibiotic treatment, suggesting that there is a link between bacterial persistence and multidrug tolerance and TA systems (11).

TA systems are classified into five types depending on an antitoxin's mechanism of action (12). Of the five identified types, the type II systems are the most prevalent. Type II TA systems are classified into families based on sequence homology (4, 5, 13). The family exemplified by HipA toxin inhibits translation by the phosphorylation of glutamyl-tRNA synthetase (9). The RelE family arrests translation by cleaving mRNA at the ribosomal A site (14). The MazF family cleaves mRNA, 16S rRNA, 23S rRNA, and some tRNAs in a sequence-specific manner (15–19). The Doc toxin family phosphorylates elongation factor Tu to halt translation elongation (20, 21). HicA and Kis family toxins cleave mRNA independent of the ribosome (22, 23). CcdB and ParE family toxins inhibit DNA replication by inactivating DNA gyrase (24, 25). The targets of a number of VapC toxins have been identified in different organisms. VapC toxins from the enteric bacteria Salmonella enterica and Shigella flexneri cleave initiator tRNA^fMet (26). VapC20 from Mycobacterium tuberculosis cleaves 23S rRNA at the sarcin-ricin loop (27). VapC1 and VapC29 from M. tuberculosis cleave single-stranded RNAs in GC-rich sequences (28, 29), and VapC4 cleaves specific tRNA isoacceptors (30).

Type II TA systems are composed of a protein antitoxin, which directly interacts with its cognate toxin. These antitoxins are typically modular proteins, usually composed of an N-terminal DNA-binding motif, which allows for autoregulation of the TA operon, and a C-terminal toxin-binding region (31). The DNA-binding motifs are classified into four classes: helix-turn-helix (HTH), ribbon-helix-helix (RHH), looped-hinge-helix (AbrB), and Phd/YefM (4). The modular nature has been verified experimentally with the antitoxins MazE and Phd, where mutations in the N-terminal region disrupted DNA-binding ability, and mutations in the C-terminal region prevented toxin neutralization (32, 33).

VapBC systems are the most abundant family of type II TA systems, consisting of a VapB antitoxin that neutralizes its cognate VapC endoribonuclease toxin. Studies have demonstrated that VapBC1 and VapBC2 are significantly upregulated in NTHi during infection, where they function to regulate growth and enhance survival (34, 35). The VapC toxin contains a PIN domain, a roughly 100-amino-acid domain that has an active site containing four conserved acidic amino acids (36). These conserved residues coordinate a Mg⁺ ion in the active site, which facilitates hydrolytic cleavage of their target molecule. VapBC pairs are highly specific, and a VapB antitoxin will not neutralize its noncognate VapC toxin. Despite the importance of this interaction in bacterial dormancy, very little is understood about what confers antitoxin specificity for its cognate toxin or what amino acid residues are important for this interaction. Characterizing TA interactions allows for a better understanding of how the antitoxin controls toxin activity and bacterial persistence. Crystal structures of VapBC complexes from S. flexneri (61% identity, 76% similarity to NTHi VapC2), M. tuberculosis (23% identity, 51% similarity), Neisseria gonorrhoeae (27% identity, 47% similarity), and Rickettsia felis (43% identity, 65% similarity) suggest the toxin-binding domain of VapB antitoxins bind in a cleft containing the VapC active site, including the conserved amino acid residues of the PIN domain, through multiple amino acid contacts (7, 37–41). Similarly, previous work in M. tuberculosis demonstrated that two amino acid mutations in the VapB4 antitoxin were required to disrupt the toxin-antitoxin interaction (42). However, in this study, we identified single amino acid mutations in the VapB2 antitoxin from nontypeable Haemophilus influenzae (NTHi) that prevent it from neutralizing its cognate VapC2 toxin. Additionally, our experiments demonstrated that the mutation of a single amino acid residue in the VapB1 antitoxin from NTHi allowed it to neutralize both its cognate VapC1 toxin and its noncognate VapC2 toxin. Previous work had shown that altering at least 4 residues in ParD1 caused a specificity swap that allowed for interaction with ParE2 in Mesorhizobium opportunistum (43). Similarly, four amino acid substitutions allow MqsA to neutralize a synthetic, noncognate toxin (50). However, this study demonstrates that a single amino acid mutation can alter antitoxin specificity.

MATERIALS AND METHODS

Bacterial strains and growth media.

Escherichia coli LMG194 [F⁻ ΔlacX74 galE thi rpsL ΔphoA (PvuII) Δara714 leu::Tn10] and E. coli TOP10 [F⁻ mcrA Δ(mrr-hsdRMS-mcrBC) ϕ80lacZΔM15 ΔlacX74 deoR recA1 araD139 Δ(araA-leu)7697 galU galK rpsL endA1 nupG] carrying the indicated plasmids were grown at 37°C in M9 medium supplemented with 0.2% glucose, 0.2% Casamino Acids, 1 mM thiamine, and appropriate antibiotics or LB medium supplemented with appropriate antibiotics.

Plasmid construction. (i) pJSB31-sfGFP.

The superfolder green fluorescent protein (sfGFP)-encoding DNA was amplified by PCR using OSB1233 and OSB1234 from the SuperFolder GFP expression plasmid (Sandia Biotech). The PCR product was digested with BglII and KpnI and cloned into the same sites of pJSB31 plasmid. The pJSB31-sfGFP plasmid was constructed with a 6×-glycine linker at the N termini of the superfolder GFP.

(ii) pJSB31-VapB1-sfGFP and pJSB31-VapB2-sfGFP.

The VapB1-encoding (HI0321) and VapB2-encoding (HI0948) DNA was PCR amplified from NTHi 86-028NP using OSB1385 and OSB1386 (VapB1) or OSB1289 and OSB1257 (VapB2). The PCR product was digested with BglII and NcoI and ligated into the corresponding sites of pJSB31-sfGFP. The resulting plasmid expresses C-terminally sfGFP-tagged VapB in the presence of isopropyl-β-d-thiogalactopyranoside (IPTG).

(iii) pBAD-VapC1 and pBAD-VapC2.

VapC1-encoding (HI0322) and VapC2-encoding (HI0947) DNA was PCR amplified from NTHi 86-028NP using OSB829 and OSB830 (VapC1) or OSB834 and OSB835 (VapC2). The PCR product was digested with NcoI and XbaI and ligated into the corresponding sites of pBAD/Myc-His B (Invitrogen). The resulting plasmid expresses C-terminally Myc-His-tagged VapC in the presence of l-arabinose. DNA sequencing analysis verified all plasmids.

Construction of deletion mutants.

Mutants containing deletions in the N-terminal domain of VapB2 were created by PCR amplification of portions of the DNA encoding VapB2 using oligonucleotides OSB1353 and OSB1356 (for the VapB2 Δ20 mutant, with a deletion of 20 amino acids at the N terminus), OSB1354 and OSB1356 (for the VapB2 Δ35 mutant, with a deletion of 35 amino acids at the N terminus), and OSB1355 and OSB1356 (for the VapB2 Δ45 mutant, with a deletion of 45 amino acids at the N terminus). The PCR product was digested by NcoI and BglII and ligated into the corresponding sites in pJSB31-sfGFP. DNA sequencing analysis verified all plasmids.

Site-directed mutagenesis.

Oligonucleotide-directed site-specific mutagenesis was carried out by a modification of the method of Fisher and Pei (44). The template plasmid was amplified in 50 μl of reaction mixture containing 10 ng of DNA template, 0.2 mM dinucleoside triphosphates (dNTPs), 0.2 μM each primer with the appropriate base changes, 9% dimethyl sulfoxide (DMSO), and 1 U of iProof polymerase (Bio-Rad) in the supplied reaction buffer. PCR conditions consisted of one cycle of 98°C for 3 min, and then 30 cycles of 98°C for 30 s, 65°C for 30 s, and 72°C for 2 min, followed by one cycle of a final extension step at 72°C for 5 min. The PCR product was digested with 5 U of DpnI at 37°C for 1 h. E. coli TOP10 cells were transformed with 10 μl of the DpnI-treated PCR product. DNA sequencing analysis verified all constructs.

Cell growth spotting assay.

E. coli strains were grown at 37°C overnight in M9 glucose (0.2%) medium supplemented with 100 μg/ml ampicillin and 30 μg/ml chloramphenicol. The saturated cultures were diluted in sterile water to an A₆₀₀ of 0.02 and then diluted 10-, 100-, and 1,000-fold. An aliquot of 5 μl of each dilution was pipetted onto M9 glucose (0.2%) medium supplemented with 100 μg/ml ampicillin and 30 μg/ml chloramphenicol in the presence or absence of 0.2% l-arabinose and/or 500 μM IPTG, unless otherwise indicated, and grown for 16 h at 37°C.

Western blot analysis.

E. coli cells carrying the indicated plasmids were grown at 37°C to an A₆₀₀ of 0.3 to 0.5 and induced by the addition of appropriate inducers for 30 min. One milliliter of each cell culture was removed and the cells collected by centrifugation at 16,000 × g for 2 min at room temperature. The cell pellet was resuspended in 10 μl of 1× SDS loading buffer (10% SDS, 10 mM 2-mercaptoethanol, 20% glycerol, 0.05% bromophenol blue, 0.2 M Tris-HCl [pH 6.8]) and boiled for 5 min. Ten microliters of proteins in the sample was separated by electrophoresis on a 10% polyacrylamide–SDS gel. Proteins were transferred onto a nitrocellulose membrane (Bio-Rad) by electroblotting at 90 mA overnight at 4°C. The proteins were detected using anti-GFP monoclonal antibody (1:1,000; Roche) for sfGFP-tagged protein. The primary antibody was detected with anti-mouse IgG (goat) antibody (1:5,000; PerkinElmer) and developed using a Western Lightning Plus-ECL system (PerkinElmer). GroEL served as a loading control for E. coli and was detected by anti-GroEL antibody (1:10,000; Sigma-Aldrich) and anti-rabbit IgG (goat) secondary antibody (1:5,000; PerkinElmer) and developed using the Western Lightning Plus-ECL system (PerkinElmer).

Fluorescence assays.

TOP10 strains carrying the indicated plasmids were grown at 37°C to an A₆₀₀ of 0.1 to 0.2 and induced by the addition of appropriate inducers for 2 h. Two hundred microliters of each culture was added to wells in a fluorescence transparent microtiter plate (catalog no. 3370; Costar). The absorbance in each well at 600 nm was measured using a PowerWave XS microplate reader (BioTek). The GFP fluorescence of each culture was analyzed on a Typhoon 9410 imager (GE Biosciences) using the 488 nM blue laser for excitation and the 520BP40 emission filter. Images were analyzed using the program ImageQuant.

Protein copurification assay.

TOP10 cells carrying the indicated plasmids were grown at 37°C to an A₆₀₀ of 0.3 to 0.5 and induced by the addition of appropriate inducers for 30 min. Fifteen units of cells at A₆₀₀ were removed and collected by centrifugation at 12,000 × g for 10 min at 4°C. The cell pellet was resuspended in 5 ml of lysis buffer (50 mM NaH₂PO₄ [pH 8.0], 300 mM NaCl, 20 mM β-mercaptoethanol, 20 μl of bacterial protease inhibitor cocktail [Sigma]). Cells were broken by two passages through a French pressure cell (Spectronic; Thermo Scientific) at 14,000 lb/in². The soluble fraction of 1 ml of cell lysate was clarified by centrifugation at 16,000 × g for 10 min at room temperature. The supernatant was transferred into a 1.5-ml tube containing 100 μl of nickel nitrilotriacetic acid (Ni-NTA)–agarose resin (Qiagen) preequilibrated in wash buffer (50 mM NaH₂PO₄ [pH 8.0], 300 mM NaCl, 20 mM β-mercaptoethanol). The suspension was nutated at room temperature for 1 h. The resin was collected by centrifugation at 700 × g for 3 min at room temperature and washed 3 times with wash buffer (50 mM NaH₂PO₄ [pH 8.0], 700 mM NaCl, 35 mM β-mercaptoethanol) The purified proteins were eluted by the addition of 200 μl of elution buffer (50 mM NaH₂PO₄ [pH 8.0], 300 mM NaCl, 250 mM imidazole, 20 mM β-mercaptoethanol). Twenty-five microliters of cell lysate or purified protein from each strain was separated by electrophoresis on a 10% polyacrylamide–SDS gel and the proteins visualized by Western blotting.

RESULTS

Identification of the minimal antitoxin domain of NTHi VapB2.

While the interaction between VapB antitoxins and VapC toxins is crucial for controlling bacterial growth and dormancy, very little is known about the molecular details of this interaction. Our study aimed to characterize this interaction through mutagenesis, followed by functional assays of antitoxin activity. In these experiments, compatible plasmids expressed VapB2 and VapC2 from the tac and ara promoters, respectively (Fig. 1A). Liquid cultures grown in M9 glucose medium were diluted 10-, 100-, and 1,000-fold and plated on M9 glucose plates containing either 0.2% arabinose (to induce VapC2), 0.5 mM IPTG (to induce VapB2), or both inducers. The induction of both toxin and antitoxin allows cells to grow, indicating neutralization of the toxin by the antitoxin. Lack of growth indicates failure of the antitoxin to neutralize the toxin.

FIG 1.

Expression of VapB2 or VapB2-sfGFP in trans inhibits toxicity of VapC2. (A) Diagram illustrating the expression of the open reading frame of VapB2 from H. influenzae from the tac promoter of plasmid pJSB31 in trans to VapC2 expressed from the ara promoter on pBAD/Myc-His B. (B) The growth of E. coli strains compared using the cell growth spotting assay, as described in Materials and Methods. Strains carried VapB2 or VapB2-sfGFP, VapB2 or VapB2-sfGFP induced with IPTG (0.5 mM), and VapC2 induced with arabinose (Ara) (0.2%).

In order to study toxin-antitoxin interaction, the experiments took advantage of a fusion of superfolder GFP (sfGFP) to the C terminus of VapB. The fusion to the wild-type and mutant antitoxins allowed live cell quantification of the proteins by measurement of GFP fluorescence. We tested whether the sfGFP tag would not interfere with VapB antitoxin activity. As in the case of the M. tuberculosis VapB4 and VapB5 antitoxins, the GFP fusion had no substantial effect on the ability of the wild-type antitoxin to inhibit the growth defect caused by the cognate toxin VapC2 (42) (Fig. 1B).

Type II antitoxins typically feature a C-terminal domain that binds specifically to their cognate toxins (33, 42, 45). Accordingly, we created N-terminal deletion mutations to identify the portion of the 77 amino acids NTHi VapB2 required for its function as a VapC2 antitoxin (Fig. 2A). Spot plating tests revealed that VapC2 inhibits cell growth upon the induction of expression from the arabinose promoter with 0.01% arabinose (Fig. 2B). The induction of antitoxin expression with 0.5 mM IPTG caused growth inhibition by the VapB2 Δ20 mutant and VapB2 Δ45 mutant, but not the wild type or VapB2 Δ35 mutant (Fig. 2B). In contrast, each of the deletion mutants grew well upon induction with 0.063 mM IPTG and reversed the growth defect caused by the expression of VapC2. The wild-type VapB2 did not reverse the toxicity of VapC2 at this lower level of induction. Quantitative measurement of the fluorescence levels of each of these GFP fusion proteins revealed higher levels of VapB2 Δ20 and VapB2 Δ45 mutants than those of the wild type and VapB2 Δ35 mutant in the presence of 0.5 mM IPTG, suggesting that the elevated levels of these mutant proteins may impair growth of the cells and allow function at lower levels of IPTG than with the full-length antitoxin (Fig. 2C). Consistent with this inference, Western blot analysis indicated that all of the fusion proteins remain intact in the cell (Fig. 2D). The ability of each of the deletion mutants to suppress VapC2 toxicity at a lower level of induction suggests that the C-terminal 33 amino acids of VapB2 confer antitoxin activity to a chimeric fusion protein.

FIG 2.

The C-terminal region of VapB2 (positions 45 to 77) is sufficient for antitoxin activity. (A) The open reading frames of VapB2 and its deletion mutants were cloned into pJSB31-sfGFP plasmid, creating sfGFP fusions. (B) Spotting assay for VapB2 deletion mutants carried out as described in the legend to Fig. 1B. Strains were grown on M9 glycerol (0.2%) medium supplemented with 50 μg/ml ampicillin, 30 μg/ml chloramphenicol, and indicated inducers. Strains carried VapB2-sfGFP or VapB2 deletion mutants, as well as either VapC2 or the empty pBAD vector. (C) E. coli strain LMG194 carrying the indicated plasmids was grown in M9 glycerol (0.2%) medium supplemented with 50 μg/ml ampicillin and 30 μg/ml chloramphenicol at 37°C to an A₆₀₀ of 0.1 to 0.2 and induced with 0.02% l-arabinose and 500 μM IPTG for 2 h. GFP fluorescence was measured in a Typhoon 9410 imager and normalized to the A₆₀₀ of the culture. Error bars illustrate the standard deviations from the results from three biological replicates. (D) Western blot analysis of cells carrying pJSB31 carrying the indicated VapB2 alleles and pBAD-VapC2 were grown in M9 glycerol (0.2%) at 37°C and induced with 0.02% l-arabinose and 500 μM IPTG for 30 min. Blots were probed with anti-GFP antibody for VapB2 or its mutant proteins, and anti-GroEL served as a loading control. Lanes 1 to 4 show a 2-fold dilution series of VapB4-sfGFP as a control for signal linearity.

Identification of mutations that disrupt NTHi VapB2 antitoxin activity.

To identify the structural determinants of VapB2 antitoxin activity, we used its C-terminal 33 amino acids to search the NCBI database for proteins with related sequence. After removing sequences from Haemophilus influenzae isolates, we identified 29 bacterial proteins with significant homology to VapB2 (Fig. 3A). These proteins share five identical amino acids with VapB2, suggesting conservation of side chain function. These residues are also conserved in the C terminus of the VapB from Shigella flexneri, whose crystal structure model in complex with its VapC partner implicated W48, F51, F52, D61, F62, and R66 (Fig. 3A; NTHi VapB2 numbering) in key contacts with the VapC toxin (7). We employed site-directed mutagenesis to test whether these amino acids play an important role in the function of VapB2 in the context of the full-length antitoxin-sfGFP fusion. Spot plating tests revealed that the W48G, F51V, and F62V mutations abolished VapB2 antitoxin activity, while the F52V, D61A, and R66G mutations had no discernible effect (Fig. 3B). Measurement of fusion protein-GFP fluorescence and Western blot analysis indicated comparable expression of all of the mutants except that with F62V (Fig. 3C and D). These results suggest that W48 and F51 each play an essential role in the ability of VapB2 to suppress the toxicity of VapC2. We note that we cannot fully discount the possibility that these changes disrupt the folding of the antitoxin. Physical studies indicate that type II antitoxins exist as unfolded polypeptides prior to interacting with their cognate antitoxins, so the defects identified here may inhibit this transition (46, 47).

FIG 3.

Effects of single mutations on the antitoxin activity of VapB2. (A) Boxshade illustration of a Clustal alignment of VapB2 paralogues identified by a BLAST search using VapB2. Only sequences similar to the C-terminal 33 amino acids of NTHi VapB2 are shown. str., strain. (B) Spotting assay for VapB2 antitoxin activity in E. coli strain TOP10 carrying the indicated plasmids, as described in the legend to Fig. 1B. (C) GFP fluorescence of VapB2-sfGFP proteins was measured in a Typhoon 9410 imager and normalized to the A₆₀₀ of the culture. Error bars illustrate the standard deviations from the results from three biological replicates. (D) Western blot analysis of VapB2 mutant proteins as described in the legend to Fig. 2D. Strains were grown in M9 glucose (0.2%) medium and induced with 0.2% arabinose and 0.5 mM IPTG for 30 min.

Mutation of conserved amino acids in NTHi VapB1 does not abolish antitoxin activity.

NTHi contains a second vapBC operon, vapBC1. The two antitoxins VapB1 and VapB2 belong to the VagC (AbrB) superfamily of DNA-binding proteins and demonstrate 30% identity and 50% similarity. The toxins VapC1 and VapC2 share 39% identity and 60% similarity. These facts suggest that the operons may have arisen by duplication of an ancestral vapBC operon, which raises questions about the relative specificities of the antitoxins. To test conservation of function, we asked if the antitoxin function of VapB1 requires amino acids conserved between itself and VapB2 (Fig. 4A). Wild-type and mutant VapB1-sfGFP fusion proteins all functioned to reverse the growth defect caused by the expression of VapC1 when tested by spot plating (Fig. 4B). Fluorescence and Western blotting assays revealed stable expression of each of the fusion proteins (Fig. 4C and D). Although the VapB1-D58A mutant seemed to exhibit weak inhibition of VapC1 activity in the plate assay, multiple repeats do not support a clear antitoxin defect for this mutant. These findings suggest that the single mutations in VapB1, similar to those causing defects in VapB2, do not have important effects on VapB1 antitoxin function.

FIG 4.

Effects of single mutations on the antitoxin activity of VapB1. (A) Comparison of the amino acid sequences of VapB1 and VapB2. Inferred functional homologues of important VapB2 residues are highlighted in bold. (B) Spotting assay for VapB1 antitoxin activity in E. coli strain TOP10 carrying the indicated plasmids, as described in the legend to Fig. 1B. (C) GFP fluorescence of VapB1-sfGFP proteins was measured in a Typhoon 9410 imager and normalized to the A₆₀₀ of the culture. Error bars illustrate the standard deviations from the results from three biological replicates. (D) Western blot analysis of VapB1 mutant proteins as described in the legend to Fig. 2D. Strains were grown in M9 glucose (0.2%) medium and induced with 0.2% arabinose and 0.5 mM IPTG for 30 min.

A single amino acid change converts VapB1 to a VapC2 antitoxin.

Next, we asked if mutations in either of the antitoxins would alter their specificity for their cognate toxins. The expression of VapB1 or VapB2 specifically neutralized the growth defect caused by their cognate, but not noncognate, toxins (Fig. 5A). Several observations suggested that W48 might play a key role in the specificity of VapB2 for VapC2. First, VapB2 requires this aromatic side chain for its antitoxin activity (Fig. 3). Second, crystal structure models of S. flexneri, Rickettsia felis, and Mycobacterium tuberculosis (VapBC5) toxin-antitoxin complexes show the paralogous tryptophan nestled in a hydrophobic pocket in the cognate toxin (7, 38). Third, Mycobacterium tuberculosis VapB4 requires the paralogous tryptophan for full antitoxin activity (42). Accordingly, we created a T47W mutation in full-length VapB1 and tested its specificity for neutralizing VapC1 and VapC2. The results reveal that this single amino acid change allows the VapB1-T47W mutant to function as an antitoxin for VapC1 and VapC2 (Fig. 5B). We tested the ability of these proteins to interact by (i) purifying the VapC polypeptides by virtue of their C-terminal 6×His tag and (ii) assaying for copurification of the VapB antitoxins by Western blotting. The results revealed that the VapB1-T47W mutant copurifies with both VapC1 and VapC2 (Fig. 5D, lanes 6 and 8). Interestingly, the assay reveals slight binding of VapB1 to VapC2, suggesting that it may interact weakly with its noncognate toxin (Fig. 5D, lane 4). Additionally, while this weak interaction is not sufficient to neutralize VapC2 toxicity, the increase in interaction when the T47W mutation is introduced is now enough to allow neutralization of VapC2 (Fig. 5D, lanes 4 and 8). These findings indicate that a single amino acid change converts VapB1 into a functional VapC2 antitoxin.

FIG 5.

Insulation of VapBC1 and VapBC2 systems and relaxed specificity of VapB1. (A) Spotting assay for specificity of VapB1 and VapB2 antitoxin activities in E. coli strain TOP10 carrying the indicated plasmids, as described in the legend to Fig. 1B. (B) Spotting assay for specificity of VapB1 antitoxin mutants in E. coli strain TOP10 carrying the indicated plasmids, as described in the legend to Fig. 1B. (C) GFP fluorescence of VapB1-sfGFP proteins was measured in a Typhoon 9410 imager and normalized to the A₆₀₀ of the culture. Error bars illustrate the standard deviations from the results from three biological replicates. (D) Western blot analysis of copurified proteins from E. coli TOP10 cells carrying plasmids expressing the indicated VapB1 alleles and VapC1-Myc-6×His, VapC2-Myc-6×His, or an empty vector control. Cells were induced with 0.2% l-arabinose and 500 μM IPTG for 30 min at 37°C. Western blots were probed with anti-GFP antibody for VapB1 or its mutant proteins and anti-GroEL antibody for GroEL proteins.

Finally, we tested whether other mutations in VapB1 or VapB2 result in relaxed specificity for toxin neutralization but found no evidence for such changes (see Fig. S1 in the supplemental material). We also carried out a comprehensive mutagenesis of the VapB2 C-terminal module and selected for mutants that would neutralize VapB1, but we found none that altered its specificity. Thus, VapB2 may have acquired multiple amino acid contacts after its split from the proposed progenitor VapBC, which determines its specificity for VapC2.

DISCUSSION

Type II antitoxins play a critical role in controlling the ability of their cognate toxins to inhibit gene expression and elicit bacterial dormancy (8, 48). The susceptibility of the antitoxin to cellular proteases, such as Lon, provides a mechanism by which stress response systems, such as the stringent response, activate type II toxins. In cases where related toxins, such as VapC proteins, exist within a single bacterium, antitoxins typically inhibit only the activity of their cognate toxin, thereby insulating the pairs from one another. This raises several interesting issues regarding TA specificity. First, it remains unclear whether the insulation of multiple homologous TA pairs reflects specific regulatory requirements of the cell. Second, the evolution of insulation between homologous TA pairs that arose from gene duplication requires successful passage through intermediate variants with relaxed interaction specificity. Third, small-molecule inhibitors of TA interaction might demonstrate specificity that reflects structural differences between the insulated TA pairs. To understand the basis for VapB toxin specificity, we sought to identify whether VapBC interaction requires single conserved amino acid side chains in the antitoxin and whether any of these amino acids govern specificity for the cognate antitoxin.

Our findings support the conclusion that the C-terminal 33 amino acids of the NTHi VapB2 antitoxin are sufficient for its ability to suppress the toxicity of its cognate VapC2 antitoxin. This result aligns with previous conclusions from studies that revealed distinct DNA- and toxin-binding domains in type II TA systems. The Phd antitoxin employs distinct N-terminal DNA-binding and C-terminal toxin-binding domains to control the expression and activity of the Doc toxin (45). Indeed, single amino acid substitutions in the Phd C-terminal domain disrupt its ability to suppress the toxicity of Doc (32). Studies in the MazEF TA system also revealed the modular C-terminal toxin-binding domain in MazE but found that only multiple mutations disrupted its ability to neutralize MazE (33). Similarly, recent structure-function analysis of the M. tuberculosis VapB4 antitoxin identified its C-terminal VapC4 toxin-binding module and revealed that only double mutations abrogate its interaction with VapC4 (42). Our present findings support the modular nature of type II antitoxin function but also indicate that single amino acid changes can disrupt the ability of VapB2 to inhibit its cognate toxin VapC2.

Our finding that the W48G mutation in full-length VapB2 disrupts its antitoxin activity for VapC2 supports the conclusion that tryptophan side chains in this region of VapB antitoxins can play a significant role in the protein's interaction with its cognate toxin. Indeed, our previous work on the M. tuberculosis VapB4 antitoxin revealed that tryptophan in a similar position (W48) provides a key interaction site for binding to VapC4 (42). Structural similarities with M. tuberculosis VapB5 antitoxin, as well with VapB antitoxins from Shigella flexneri and Rickettsia felis, support a model in which tryptophan in this region of the antitoxins fits into hydrophobic pockets in the toxins, thereby providing a critical interaction that stabilizes binding of the antitoxin to its cognate toxin (7, 38). In the case of NTHi VapB2, the loss of toxin neutralization by this single amino acid change suggests that small molecules that bind in the toxin pocket might interfere with antitoxin binding.

Although the equivalent of VapB2 W48 is not conserved among all antitoxins of the VapB class, it exists within the close relatives of VapB2 (Fig. 3). Interestingly, VapB1 does not contain a W48 equivalent, despite the homology between NTHi VapB1 and VapB2 and the possibility that they arose from gene duplication (Fig. 4). Consistent with the key role of W48 in the function of VapB2, the T47W mutation in VapB1 converts it to a VapC2 antitoxin (Fig. 5). Interestingly, the VapB1-T47W mutant retains its ability to bind to and neutralize its cognate toxin VapC1. Thus, the VapB1-T47W mutant may exemplify an intermediate relaxed specificity state in the evolution of insulation between duplicated TA operons. Recently, Aakre and colleagues carried out a comprehensive analysis of the interaction specificity of toxins and antitoxins among ParDE homologs, which provided evidence that antitoxins in TA pairs arising from gene duplication likely pass through promiscuous intermediates on their way to fully insulated pairs in which the antitoxins neutralize only their cognate toxins (43). Our findings support their model and indicate that a single amino acid change can give rise to the relaxed specificity intermediate.

Type II TA systems have attracted interest as targets for the development of inhibitors that would prevent binding, thereby freeing the toxin to halt cell growth (36, 49). Since models based on crystal structures show that VapB C-terminal domains make extensive contacts with their cognate toxin, the possibility that small molecules could inhibit TA interactions seems limited (7, 37, 38, 40, 41). Indeed, an exhaustive search for single amino acid substitutions that abrogate M. tuberculosis VapBC4 interaction revealed only double mutations with this property (42). In the present case, and in that of the Phd-Doc system, single mutations do inhibit binding (32). Thus, some type II TA pairs may be potential targets for small-molecule inhibitors that would find uses as probes to study the function of TA systems and in the treatment of bacterial infections.

In summary, the findings reported here identify the toxin-binding module of the VapB2 antitoxin from NTHi. Single amino acid substitutions disrupt the ability of the antitoxin to neutralize the toxic effect of its cognate toxin, VapC2, on cell growth. The two homologous VapBC pairs from NTHi are functionally insulated from one another in that the antitoxins do not neutralize their noncognate toxins. Nevertheless, a single amino acid substitution can convert VapB1 into a relaxed specificity inhibitor that binds both resident toxins. This represents the first example of a single mutation causing relaxed specificity in a type II antitoxin.