Identification and functional characterization of type II toxin/antitoxin systems in Aggregatibacter actinomycetemcomitans

Summary Type II toxin/antitoxin (TA) systems contribute to the formation of persister cells and biofilm formation for many organisms. Aggregatibacter actinomycetemcomitans thrives in the complex oral microbial community subjected to continual environmental flux. Little is known regarding the presence and function of type II TA systems in this organism or their contribution to adaptation and persistence in the biofilm. We identified 11 TA systems that are conserved across all seven serotypes of A. actinomycetemcomitans and represent the RelBE, MazEF and HipAB families of type II TA systems. The systems selectively responded to various environmental conditions that exist in the oral cavity. Two putative RelBE‐like TA systems, D11S_1194‐1195 and D11S_1718‐1719 were induced in response to low pH and deletion of D11S_1718‐1719 significantly reduced metabolic activity of stationary phase A. actinomycetemcomitans cells upon prolonged exposure to acidic conditions. The deletion mutant also exhibited reduced biofilm biomass when cultured under acidic conditions. The D11S_1194 and D11S_1718 toxin proteins inhibited in vitro translation of dihydrofolate reductase (DHFR) and degraded ribosome‐associated, but not free, MS2 virus RNA. In contrast, the corresponding antitoxins (D11S_1195 and D11S_1719), or equimolar mixtures of toxin and antitoxin, had no effect on DHFR production or RNA degradation. Together, these results suggest that D11S_1194‐1195 and D11S_1718‐1719 are RelBE‐like type II TA systems that are activated under acidic conditions and may function to cleave ribosome‐associated mRNA to inhibit translation in A. actinomycetemcomitans. In vivo, these systems may facilitate A. actinomycetemcomitans adaptation and persistence in acidic local environments in the dental biofilm.

its ability to disseminate from this niche to other organs of the body, have not been well defined.
The dental biofilm is a complex and dynamic microbial community that is comprised of up to 700 different species of bacteria. [9][10][11][12][13][14] This biofilm is the prime etiological agent of three common oral diseases in humans: dental caries, gingivitis and periodontal disease. 1,15,16 The progression of these diseases is associated with major shifts in microbial populations in the oral biofilm and diseased sites often exhibit increased populations of pathogenic species relative to healthy sites in the oral cavity. [15][16][17] The stimuli that contribute to these population shifts have not been well characterized but the oral cavity is subject to continual environmental flux, including changes in pH, temperature, osmolarity and nutrient supply. Oral bacteria rapidly detect and respond to these environmental fluctuations, allowing them to successfully coexist and thrive in the oral cavity. 16,18,19 A variety of mechanisms contribute to adaptation to environmental flux but there is growing evidence that the activity of toxin/antitoxin (TA) systems play an important role in adapting to and persisting under conditions of environmental stress. 20 The TA systems comprise a variety of genetic elements classified in six different families or types based on the mechanism of action of the antitoxins and are encoded on both plasmids and the bacterial chomosome. [21][22][23][24] Type II TA systems have been most extensively studied and encode protein toxins and antitoxins. Interaction of the antitoxin with the toxin inhibits toxin activity 25 and in many cases, the antitoxin and/ or the toxin-antitoxin complex can also function as a transcriptional repressor to autoregulate their expression. 26,27 Under conditions of environmental stress, cellular proteases (eg, Lon and ClpXP) are activated, which degrade the labile antitoxin, activating the toxin. Type II toxins target important physiological functions such as translation, DNA replication, cell wall synthesis and the assembly of cytoskeletal proteins during cell division, leading to growth arrest. 21 Many type II toxins function as ribonucleases that cleave their target mRNAs in either a ribosome-dependent or independent manner. 22 Very little is known about the presence and function of A. actinomycetemcomitans type II TA systems and how they may contribute to adaptation and persistence of the organism in the dental biofilm. In this study, we identified 11 operons in the A. actinomycetemcomitans D11S genome that encode putative functional type II TA systems and show that these systems respond selectively to various environmental conditions. Two RelBE-like systems were activated under acidic growth conditions and deletion of these systems resulted in reduced metabolic activity of A. actinomycetemcomitans in stationary phase. Finally, we demonstrate that both of these TA systems inhibit translation and function as ribosome-dependent ribonucleases.

| Bacterial strains and plasmids
The bacterial strains and plasmids used in this study are listed in the Supplementary material (Tables S1 and S2), respectively. Luria-Bertani (LB) broth or LB agar (LB broth plus 1.5% agar) was used for the propagation or plating of Escherichia coli. Bacteria were grown at 37°C with shaking for broth cultures or under microaerophilic conditions in an atmosphere of 5% CO 2 for plates. Aggregatibacter actinomycetemcomitans strain 652 was routinely cultured in brain-heart infusion (BHI) broth or BHI agar (BHI plus 1.5% agar) under microaerophilic conditions, either in a candle jar for plates or as standing broth cultures in an atmosphere of 5% CO 2 . For the construction of gene deletion mutants, SOC medium (2% tryptone, 0.5% yeast extract, 10 mm NaCl, 2.5 mm KCl, 10 mm MgCl 2 and 20 mm glucose), TYE (1% tryptone and 0.5% yeast extract) or TYE agar (TYE plus 2% agar) were routinely used for the propagation or plating of A. actinomycetemcomitans. For some experiments, A. actinomycetemcomitans was grown in a chemically defined medium as described by Socransky et al. 28 with some modifications (see Supplementary material, Table S3). When necessary for plasmid selection, medium was supplemented either with 25 μg mL −1 kanamycin, 50 μg mL −1 spectinomycin, 12.5 μg mL −1 tetracycline, 50 μg mL −1 ampicillin, 40% sucrose or 1 mm isopropyl Bd-1-thiogalactopyranoside (IPTG).

| Identification of putative TA systems
Putative type II TA systems in A. actinomycetemcomitans were identified by two methods. First, the genome of A. actinomycetemcomitans strain D11S-1, serotype c 29 was probed for sequence similarity to known E. coli TA systems using the protein basic local alignment search tool (pBLAST, https://www.ncbi.nlm.nih.gov). The sequences for known E. coli toxins that were used in these searches are listed in the Supplementary material (Table S4). In addition, the entire genome of A. actinomycetemcomitans D11S-1 was examined using TAfinder (http://202.120.12.133/TAfinder/TAfinder.php) to identify operons that were not previously identified in the BLAST searches described above. TAfinder detects type II TA loci mainly based on sequence alignments and conserved domain searches against a database of TA families. The A. actinomycetemcomitans peptide sequences identified as putative TA systems in strain D11S-1 were subsequently used to probe 33 other A. actinomycetemcomitans genome sequences that were present in the NCBI database (see Supplemental material, Table   S5) that represent all seven A. actinomycetemcomitans serotypes.

| Expression of putative TA systems under environmental stress
The expression of the putative TA systems was determined under various environmental stress conditions using real-time polymerase chain reaction (PCR). Cells were grown to mid-exponential phase in chemically defined medium supplemented with 20 mm glucose or 20 mm lactate 30 at 37°C and were then exposed to various environmental conditions for 20 min. Environmental conditions examined included: acidic pH (pH 5.0), oxidative stress (0.1% hydrogen peroxide), microaerophilic conditions (5% CO 2 ), elevated temperature (39°C), anaerobic conditions (10% H 2 , 10% CO 2 , 80% N 2 ), iron limitation (250 μm bipyridyl), and reduced temperature (30°C). Bacteria were harvested and RNA was isolated using the cesium chloride step-gradient method as described by Reddy and Gilman. 31 RNA was reverse transcribed to cDNA using random primers provided in the cDNA synthesis kit (Quanta Bio, Beverley, MA), and the resulting cDNA was used in SYBR Green real-time PCR using primers for the TA system as recommended by the manufacturer (Quanta Bio). Data were analyzed using the ΔΔCt method and fold change expression was determined by normalizing the results to the levels of an unstressed control (5S rRNA was used for normalization). The putative TA systems that are coexpressed with a third gene (D11S_0469-470 and D11S_2094-2095) were excluded in these analyses.

| Construction of isogenic deletion mutants
The generation of markerless deletion mutations was carried out as

| Measurement of metabolic activity
To determine the metabolic activity of isogenic mutants, 3-(4,5di methylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) was used as described by Wang et al. 34 with some modifications. Cultures were grown at 37°C to stationary phase, removed from the incubator (t = 0) and incubated at room temperature for 5 days. At various time points, 3 mL of the culture was removed, centrifuged at 5000 g and suspended in fresh BHI broth containing 2.07 mm MTT. Cultures were then incubated for 2 h (approximately one doubling period) before being harvested. Metabolically active cells reduce MTT via NADH + /NADPH + -dependent oxidoreductase enzymes in the electron transport chain to form water-insoluble formazan. 35 The resulting formazan crystals were dissolved in dimethyl sulfoxide and absorbance was measured at 550 nm.

| Growth and analysis of static biofilms
Static biofilms were grown in multi-well tissue culture plates. [36][37][38] The A. actinomycetemcomitans cultures were grown overnight in BHI broth and the optical density at 600 nm (OD 600 ) was measured. To form mature A. actinomycetemcomitans biofilms, cells were diluted into fresh BHI broth to a final OD 600 of 0.005 and this dilution was used to inoculate the plate wells. Cultures were incubated at 37°C for 72 h and the resulting biofilms were then supplied with fresh BHI broth that was adjusted from pH 5.0-pH 8.0 to represent the range of conditions that might typically exist in the gingival pocket. 39,40 The biofilms were subsequently incubated for an additional 24 h, rinsed gently with sterile water and stained with 0.1% crystal violet. After staining, cells were rinsed with sterile water until the water was clear, crystal violet was solubilized with 30% acetic acid and the absorbance at OD 570 was measured.

| Statistical analysis
All assays were carried out in triplicate and data were analyzed using the unpaired t-test with statistical significance defined as P ≤ .05.

| Functional analysis of putative TA systems
Peptides representing the toxin and antitoxin proteins of each TA system (sequences are shown in the Supplementary material, Table   S7) were chemically synthesized (Biosynthesis, Inc., Lewisville, TX) at > 90% purity. Synthetic peptides were dissolved in 0.05 m phosphate buffer, pH 7.8, containing 300 mm NaCl and 0.01% trifluoroacetic acid and peptide purity and size was confirmed using a NuPAGE Bis-Tris SDS-PAGE gel (Thermo Fisher Scientific, Waltham, MA). Samples were prepared in NuPAGE SDS sample buffer according to the recommendation of the manufacturer and gels were electrophoresed using NuPAGE MES Buffer in an XCell SureLock MiniCell. To test the ability of the synthetic toxins and antitoxins to interfere with protein translation, a cell-free protein synthesis system was used. Toxin

| Identification of type II TA systems in A. actinomycetemcomitans
Type II TA systems encode two small proteins (each between ~40 and 200 amino acids) that can regulate bacterial cell function including the formation of biofilms and the cellular response to various environmental stresses (eg, antibiotic treatment, starvation). To determine if the A. actinomycetemcomitans genome encodes type II TA systems, peptide sequences of known TA systems that are present in E. coli were used to perform BLAST searches of the A. actinomycetemcomitans D11S-1 genome. As shown in Table 1, this search identified nine putative type II TA systems. Two of these TA operons exhibit sequence similarity to the MazEF-family of TA systems, which function as site-specific endoribonucleases that cleave mRNA independent of ribosome binding, resulting in inhibition of translation. One A. actinomycetemcomitans TA system was related to the HipAB-family, which acts as a serine/threonine kinase targeting tRNA to inhibit translation. The remaining six putative type II TA systems of A. actinomycetemcomitans were related to the RelBE-family, which function as ribosome-dependent endoribonucleases to cleave mRNA.
As approximately 32% of the genes in the A. actinomycetemcomitans genome have not been assigned a function, it is possible that A. actinomycetemcomitans encodes additional type II TA systems that may not be related to known E. coli TA system sequences. Hence, the entire A. actinomycetemcomitans genome was further examined using TAfinder, which detects type II TA loci using both sequence alignment and conserved domain searches against a database of diverse TA families. As shown in Table 1, TAfinder identified five additional putative type II TA systems. None of these exhibited significant sequence similarity to E. coli TA systems. In addition, D11S_0469-470 and D11S_2094-2095 were unique in that each operon contained a third open reading frame encoding a gene unrelated to toxins or antitoxins, diadenosine tetraphosphatase and O-succinylbenzoate synthase, respectively. Finally, sequence analysis suggested that three of the putative TA systems may contain pseudogenes (indicated by an asterisk in Table 1). Hence, the A. actinomycetemcomitans D11S-1 genome encodes at least 14 putative type II TA loci, three of which may contain pseudogenes and may be non-functional.
To determine the extent to which the type II TA systems identified Sequence analysis indicated that this operon contained a toxin and/or antitoxin pseudogene. b Identified by TAfinder but present in a three-gene operon with a non-toxin/antitoxin (TA) -related gene (see text

| Putative TA systems in A. actinomycetemcomitans respond to environmental stress
For many of the known type II TA systems, the complex of the toxin and antitoxin proteins functions as a repressor that autoregulates its own operon expression. Under environmental stress, proteases such as Lon are activated and degrade the labile antitoxin, activating the toxin and de-repressing operon expression. To assess the effect of environmental stress on A. actinomycetemcomitans TA expression, cells were exposed to various stress conditions and TA expression was determined by real-time PCR. As shown in Figure 1A which suggests that some TA systems are functionally more specific.
Only one TA system, D11S_2133-2134, did not respond to any of the environmental stress conditions that were tested.

| Functional characterization of D11S_1194-1195 and D11S_1718-1719
During growth of A. actinomycetemcomitans in broth culture, the pH of the medium drops from an initial pH 7.5 to pH 5. A. actinomycetemcomitans was cultured at pH 5.0 ( Figure 1D), suggesting that these systems may contribute to overall fitness of the organism at reduced pH. These two TA systems were selected for further study. A gene deletion mutant of each TA system was constructed and metabolic activity of stationary phase cultures of these strains at pH 5.0 were compared with the wild-type strain.
As shown in Figure 2A, when stationary phase cultures were incubated at room temperature, the metabolic activity of wild-type cultures decreased gradually over 120 h. In contrast, the Δ1718/1719 mutant strain was more sensitive to sustained exposure to acidic pH and lost metabolic activity to a greater extent than the wild-type, especially between 72 and 120 h. Complementation of the deletion mutant with a functional copy of D11S_1718-1719 TA system restored metabolic activity to wild-type levels. As shown in Figure 2B, the Δ1194/1195 mutant phenotype was similar to the wild-type strain through 72 h but exhibited a significant decrease in metabolic activity relative to the wild-type at the 96 h and 120 h time points. Complementation of the mutant restored metabolic activity to the wild-type level. Interestingly, the OD 600 of the cultures did not significantly decrease over the incubation period for any of the strains (data not shown), suggesting that cell lysis may not be occurring as metabolic activity decreases.

| Δ1718-1719 and Δ1194-1195 mutants exhibit reduced biofilm growth
To determine if the D11S_1718-1719 and D11S_1194-1195 TA systems influence A. actinomycetemcomitans biofilm growth, static biofilms were grown for 72 h and then provided with fresh medium at pH 5.0 to pH 8.0. After incubation for an additional 24 h, biofilms were quantified by crystal violet staining. As shown in Figure 3A, biofilm biomass of the wild-type strain increased after the addition of fresh medium at pH 6.0, 7.0 or 8.0, consistent with the previous findings of Bhattacharjee et al 41 However, a significant decrease in biomass occurred when the wild-type biofilm was incubated in fresh medium at pH 5.0. Biomass of the D11S_1718-1719 mutant was significantly less than that of the wild-type at pH 6.0 and pH 5.0, but was restored to wild-type levels when the deletion strain was complemented with a functional copy of the TA system. Biomass of the D11S_1194-1195 deletion mutant did not differ significantly from wild-type; however, complementation of this strain resulted in a significant increase (P ≤ .01) in biofilm biomass relative to wild-type at pH 5.0 and 6.0, possibly arising from the presence of the TA system in multiple copy in the complemented strain. Together, these results suggest that the TA systems may contribute to the persistence of A. actinomycetemcomitans in mature biofilms exposed to acidic conditions.

| D11S_1194-1195 and D11S_1718-1719 encode a ribonuclease and inhibit translation
To functionally characterize the D11S_1194-1195 and D11S_1718-1719 TA systems, each of the full-length putative toxin proteins and their corresponding antitoxins was obtained as a synthetic peptide. These peptides were tested for inhibition of translation F I G U R E 2 Metabolic activity of stationary phase Aggregatibacter actinomycetemcomitans wild-type, toxin-antitoxin (TA) system deletion mutants, and complemented strain cultures. Metabolic activity was measured using 3-(4,5-dimethylthiazol-2-yl)-2,5diphenyl tetrazolium bromide (MTT) as described in the Materials and methods. Asterisks indicate the time points where metabolic activity of the mutant strain was significantly reduced relative to the wildtype (P ≤ .05) and restored to wild-type levels when complemented with a functional copy of the TA system

| DISCUSSION
Type II TA systems are widely distributed in prokaryotes and increasing evidence indicates that they contribute to adaptation and persistence under conditions of stress. [21][22][23][24] The human oral cavity is an environmentally diverse niche that is populated by a complex microbial community comprised of over several hundred bacterial species, 10 53 The physiologic outcomes that arise from activation of the D11S_1194-1195 and D11S_1718-1719 TA systems in vivo are not clear. TA system activation is associated with the formation of persister cells and is also important for biofilm formation. [20][21][22][23][24] Aggregatibacter actinomycetemcomitans thrives in a complex multispecies biofilm in the oral cavity and in this context, A. actinomycetemcomitans has been shown to closely associate with commensal oral streptococci and benefits from a cross feeding relationship in which streptococci metabolize sugars to produce lactate, which serves as a primary energy source for A. actinomycetemcomitans. 54,55 However, A. actinomycetemcomitans is also acid sensitive and grows poorly in an acidic environment. 41