ABSTRACT
Toxin-antitoxin (TA) systems are genetic elements composed of a protein toxin and a counteracting antitoxin that is either a noncoding RNA or protein. In type I TA systems, the antitoxin is a noncoding small RNA (sRNA) that base pairs with the cognate toxin mRNA interfering with its translation. Although type I TA systems have been extensively studied in Escherichia coli and a few human or animal bacterial pathogens, they have not been characterized in plant-pathogenic bacteria. In this study, we characterized a chromosomal locus in the plant pathogen Erwinia amylovora Ea1189 that is homologous to the hok-sok type I TA system previously identified in the Enterobacteriaceae-restricted plasmid R1. Phylogenetic analysis indicated that the chromosomal location of the hok-sok locus is, thus far, unique to E. amylovora. We demonstrated that ectopic overexpression of hok is highly toxic to E. amylovora and that the sRNA sok reversed the toxicity of hok through mok, a reading frame presumably translationally coupled with hok. We also identified the region that is essential for maintenance of the main toxicity of Hok. Through a hok-sok deletion mutant (Ea1189Δhok-sok), we determined the contribution of the hok-sok locus to cellular growth, micromorphology, and catalase activity. Combined, our findings indicate that the hok-sok TA system, besides being potentially self-toxic, provides fitness advantages to E. amylovora.
IMPORTANCE Bacterial toxin-antitoxin systems have received great attention because of their potential as targets for antimicrobial development and as tools for biotechnology. Erwinia amylovora, the causal agent of fire blight disease on pome fruit trees, is a major plant-pathogenic bacterium. In this study, we identified and functionally characterized a unique chromosomally encoded hok-sok toxin-antitoxin system in E. amylovora that resembles the plasmid-encoded copies of this system in other Enterobacteriaceae. This study of a type I toxin-antitoxin system in a plant-pathogenic bacterium provides the basis to further understand the involvement of toxin-antitoxin systems during infection by a plant-pathogenic bacterium. The new linkage between the hok-sok toxin-antitoxin system and the catalase-mediated oxidative stress response leads to additional considerations of targeting this system for antimicrobial development.
INTRODUCTION
Toxin-antitoxin (TA) systems are simple genetic loci composed of two adjacent genes: a bacterial toxin and a counteracting antitoxin (1, 2). The toxins in all known cases are proteins, whereas the antitoxins can be either noncoding small RNAs (sRNAs) or proteins (2). Based on the nature of antitoxins and their mode of interactions with the toxins, TA systems have been classified into six distinct classes (3, 4). Although TA systems are abundant in bacterial genomes, the biological functions of many are still unknown, raising the question of whether certain TA modules are redundant (5, 6).
In type I TA systems, the antitoxins are mostly cis-acting sRNAs, which block translation and/or facilitate the degradation of the mRNA encoding the cognate toxin. The hok-sok type I system and its homolog flmA-flmB were first described in plasmids R1 and F, respectively, and they both confer plasmid stability via postsegregational killing (7, 8). The hok gene encodes a 52-amino-acid toxin which causes the cessation of respiration, the loss of membrane potential, and the formation of “ghost cells,” ultimately resulting in irreversible self-poisoning (7, 9). An additional reading frame, mok, was also identified, which overlaps and also encompasses the coding sequence of hok; translation of mok is coupled with that of hok, as evidenced by point mutation analysis (10). An antisense sRNA is transcribed on the opposite strand of mok and sok, sharing 10-bp complementarity to the mok reading frame but not the hok reading frame (10, 11). Sok therefore reverses hok-mediated self-toxicity through inhibiting the translation of mok. As the hok mRNA is more stable than the sRNA Sok, daughter cells that lose the plasmid are selectively killed due to the carryover of hok mRNA (10, 11).
Additional chromosomally encoded Hok homologs sharing 30% to 54% amino acid identity to the plasmid-encoded Hok protein have been identified in Escherichia coli and are sometimes present in multiple copies (12–14). For example, the E. coli K-12 chromosome carries the hok homologs hokA, hokC, and hokD, which, while highly toxic upon ectopic overexpression, may lack functionality in the wild-type (WT) strain due to insertion sequences in the reading frames (9, 12, 14). However, hokC systems without any insertion elements have been found in other strains of E. coli (12). Whether the chromosomally encoded Hok homologs have distinct or redundant biological functions is still unknown.
While some type I TA toxins cleave nucleic acids in the cytoplasm (e.g., SymE and RalR), Hok and the toxin TisB are membrane associated (15). These toxins functionally resemble pore-forming phage holin proteins, and their expression results in the collapse of membrane potential and leakage of intracellular ATP (9, 16, 17). In addition to a role in postsegregational killing, Hok and TisB are thought to contribute to bacteriophage exclusion or to antibiotic persistence, a temporary dormant state conferring tolerance to antibiotics (18–20). Moderate overexpression of hokB in E. coli leads to a significant increase in tolerance to the antibiotics ofloxacin and tobramycin, although no defect of persistence was observed in a knockout mutant of hokB (18). Deletion mutagenesis of the tisB-istR TA system in E. coli resulted in a sharp decrease in persister formation (19). Although the mechanisms regulating the activation of type I TA systems are still largely enigmatic, the GTPase Obg was found to upregulate the transcription of hokB through an unknown mechanism (18). Several type I TA systems, such as tisB-istR, dinQ-agrB, and symE-symR, have been implicated as downstream actors in the cellular SOS response, under the control of the master regulator LexA (21–24).
Although TA systems have garnered much attention in the model organism E. coli and other mammalian pathogens, few type II TA systems have been characterized in plant-pathogenic bacteria (25–30); to our knowledge, no type I TA system has been characterized in a plant pathogen. In this study, we identify a chromosomally encoded hok-sok TA system highly conserved in Erwinia amylovora, a plant pathogen in the family Enterobacteriaceae. Erwinia amylovora is listed as one of the top 10 most significant bacterial plant pathogens (31) and causes the devastating fire blight disease on Rosaceae family hosts, including apple and pear. We explored the additional beneficial roles of the hok-sok locus in E. amylovora. Our results signify a first step toward understanding why the hok-sok locus, a potentially self-destructive system, is maintained in the chromosome of E. amylovora.
RESULTS
Prediction of type I toxin-antitoxin systems in plant-pathogenic bacteria.Protein sequences of known type I toxins (15), listed in Materials and Methods, were used as query sequences in taxon-wide BLAST searches of bacterial genera that are largely comprised of plant pathogens: Acidovorax, Burkholderia, Clavibacter, Dickeya, Erwinia, Pantoea, Pectobacterium, Pseudomonas, Ralstonia, Streptomyces, Xanthomonas, and Xylella. Homologs of hok and symE were identified in a few genera from the families Enterobacteriaceae and Xanthomonadaceae (Fig. 1). No other type I toxin homologs were identified in any lineages. Hok is present in only a few genera of Enterobacteriaceae, i.e., Erwinia and Pantoea, that are phylogenetically most closely related. SymE is present in all of the Enterobacteriaceae and Xanthomonadaceae strains examined, with various copy numbers of the coding genes.
Abundance of toxins of type I toxin-antitoxin systems in major plant-pathogenic bacteria. The numbers of toxin-encoding genes are displayed in the heat map. The phylogenetic tree was constructed by the maximum likelihood method using the 16S rRNA sequences of the bacteria. The bar in the phylogenetic tree represents 1 substitution in 100 bp.
Identification of hok and sok transcripts in E. amylovora.The hok-sok locus identified in E. amylovora was located within the intergenic region between the genes EAM_0497 and EAM_0498 in the chromosomal sequence of E. amylovora strain ATCC 49946 (32). This locus was homologous to loci previously identified and characterized in plasmids R1 and F (denoted flmA-flmB) (7, 8, 10). An alignment between the E. amylovora ATCC 49946 hok-sok locus and that of plasmid R1 showed that the E. amylovora locus contains all of the conserved hok-sok regulatory elements as defined by Gerdes et al. (33) (see Fig. S1 in the supplemental material). As some chromosomally encoded Hok family proteins seem to be inactivated in E. coli by an insertion sequence (IS) within the reading frame of the toxin gene, we searched for IS elements in the 4,000-bp flanking sequence of the hok coding region in all E. amylovora strains with available genome sequences using the built-in BLAST function in ISfinder (34). No significant hits (E value of ≤0.05) were found in this analysis (data not shown), suggesting that the chromosomally encoded Hok protein is likely to be functional in E. amylovora. The predicted E. amylovora Hok protein shares 87% amino acid identity with Hok encoded by plasmid R1. As predicted by TMHMM server v.2.0 (35), both the predicted E. amylovora Hok and plasmid R1 Hok contain a transmembrane domain, spanning the 7th to the 26th deduced amino acids, leaving the N-terminal amino acids and the C-terminal amino acids anchored inside and outside the cell membrane, respectively (Fig. S2). Based on our BLASTn search result, the predicted reading frame of mok, translationally coupled with hok in plasmid R1 (10), overlaps the entire coding sequence of hok, and the translation start site of mok is located 57 bp upstream of the hok coding sequence. We identified the transcriptional start site of the sRNA sok using a 5′ rapid amplification of cDNA ends (RACE) assay (Fig. 2A). The full length of sok was determined to be 61 bp, not 66 bp as described for the plasmid R1. sok is transcribed from the strand opposite the hok gene and shares 9 bp of reverse complementarity with mok but has no complementarity with hok (Fig. 2B). The sRNA sok has a stem-loop structure, and the bases with reverse complementarity to mok were exposed, as predicted by the RNAfold Web server (http://rna.tbi.univie.ac.at/cgi-bin/RNAWebSuite/RNAfold.cgi) (Fig. S3).
(A) Determination of the transcriptional start site of the sRNA sok by 5ʹ RACE-PCR and subsequent TOPO cloning. “+” indicates a positive reaction with the Cap-Clip acid pyrophosphatase enzyme, “−” indicates a negative reaction without the enzyme, and “L” indicates the DNA ladder. (B) Relative chromosomal locations of hok, mok, and sok. (C) Phylogenetic analysis of the Hok family proteins. Sequences were analyzed using the maximum likelihood method with bootstrap analysis (1,000 replicates). Each entry is labeled with the strain name followed by the GenBank accession number of the Hok family protein if annotated. The chromosomally encoded Hok family proteins are in a light blue background, and the plasmid-encoded Hok family proteins are in a light gray background. The bar indicates the number of amino acid changes per site.
Phylogenetic analysis of the Hok family proteins.To date, all Hok family proteins have been exclusively found in bacteria within the Enterobacteriaceae family (13). To better understand the evolution of Hok in E. amylovora, we conducted a comparative analysis of predicted Hok family proteins, including those encoded by all 35 sequenced E. amylovora strains, representative strains of other Erwinia species, and the genome of the E. coli strain O157:H7 EDL933, which contains the greatest number of chromosomally encoded Hok family proteins reported (13, 36). We determined that the hok gene is of a single copy and chromosomally located in all E. amylovora strains examined. Except for one substitution (K36N) in E. amylovora MANB02-1, the deduced amino acid sequences of Hok proteins are identical in all 33 Spiraeoideae-infecting E. amylovora strains (E. amylovora strains are phylogenetically subdivided into strains that infect the Spiraeoideae plant family and those that infect Rubus spp. [37]). One amino acid substitution (V17I) is present in both of the Rubus-infecting E. amylovora strains examined, Ea644 and MR1 (Ea574) (Fig. S4). Phylogenetic analysis indicated that Hok proteins in E. amylovora are within the same clade as other plasmid-encoded Hok proteins and more distant from the chromosomally encoded Hok family proteins in E. coli O157:H7 EDL933 (Fig. 2C). Although chromosomally encoded Hok proteins were also predicted in other Erwinia species, i.e., Erwinia tracheiphila and Erwinia pyrifoliae, these Hok proteins were not members of the plasmid-encoded Hok clade (Fig. 2C). Thus, our phylogenetic analysis suggested that the chromosomal placement of E. amylovora Hok is unique among known members of its clade.
Growth arrest by hok and mok and reversal by sok.To evaluate whether hok in E. amylovora encodes a self-toxic protein, the coding sequence of hok was ectopically overexpressed under the control of the Ptac promoter on pOE-hok in strain Ea1189. The empty vector pEVS143 was used as a control. After induction with 1 mM isopropyl-β-d-1-thiogalactopyranoside (IPTG), cell survival of Ea1189/pOE-hok and Ea1189/pEVS143 was monitored in both Luria-Bertani (LB) medium (rich medium) and Hrp-inducing minimal medium (HrpMM) (low-nutrient medium). Overexpression of hok caused massive killing of the bacterial population in both media (Fig. 3A and Fig. S5). In LB broth, a >100-fold reduction of the bacterial population was observed after 1 h of hok induction. In HrpMM, the killing effect was less pronounced than that in LB broth, with a reduction of approximately 10-fold after 1 h of induction, which was likely due to the slower growth of bacteria in a low-nutrient medium. For optimal induction of Hok-mediated killing, LB medium was used for all subsequent overexpression experiments in this study.
Overexpression of Hok and its truncated derivatives in E. amylovora. (A) Induction of hok by 1 mM IPTG in cell cultures at an OD600 of 0.1 grown in liquid LB medium. CFU were measured hourly by serial dilution plating. (B) Toxicity of N-terminally or C-terminally truncated derivatives of Hok. Results represent the means, and error bars are the standard deviations (SD). Different letters or asterisks indicate significant differences (P < 0.05 using Student’s t test). The assays were done three times, with similar results.
To evaluate whether the full length of Hok in E. amylovora is required to maintain its toxicity, we constructed plasmids that overexpressed truncated hok coding sequences lacking two, four, or six C-terminal residues or lacking one or three N-terminal residues. After 6 h of induction, constructs lacking two or four C-terminal residues (Ea1189/pOE-hok1:50 and Ea1189/pOE-hok1:48) elicited a level of cell death that was comparable or slightly decreased compared to that elicited by Ea1189/pOE-hok (Fig. 3B), respectively, suggesting a minor role of the four C-terminal amino acids in the toxicity of Hok. Truncation of six C-terminal residues abolished toxicity, however (Fig. 3B). Although both Ea1189/pOE-hok2:52 and Ea1189/pOE-hok4:52 elicited significant killing after induction, the effects were not as drastic as those of Ea1189/pOE-hok, suggesting a significant role of the N-terminal amino acids for the full toxicity of Hok.
To determine whether the predicted sRNA sok modulates the Hok-mediated cell-killing phenotype in E. amylovora Ea1189, we generated an overexpression construct of sok, denoted pOE-sok. Induction of sok expression in Ea1189 did not cause any notable growth defect or advantage compared with the empty vector (data not shown). Although the chromosomal location of sok does not directly overlap the coding sequence of hok, it shares 9 bp of reverse complementarity to mok, an upstream open reading frame of hok. Induction of mok was significantly cytotoxic in Ea1189/pOE-mok. By cooverexpressing sok with hok or mok, we found that sok significantly reversed mok-mediated killing but not hok-mediated killing (Fig. 4). The incomplete recovery of mok-mediated killing by sok was likely due to the aberrant balance of mok and sok under the ectopic overexpression conditions. Taken together, our data demonstrated that the chromosomally located hok-sok locus in E. amylovora functions as a TA system.
Cooverexpression of hok or mok with the sok sRNA. Both the genes and the sRNA were induced by 1 mM IPTG for 6 h, and CFU were counted by serial dilution plating. Results represent the means, and error bars are the SD. Different letters above bars suggest significant differences (P < 0.05 using Student’s t test). The assay was done twice, with similar results.
Overexpression of hok causes membrane blebbing during division.To better understand the mechanism of hok-mediated toxicity, hok was overexpressed in E. amylovora Ea1189 for 1 h in LB medium, and the cells were observed by scanning electron microscopy (SEM). Compared with the empty vector control, the pOE-hok cells appeared to be unable to divide successfully, although septa were formed between the daughter cells (Fig. 5). We also observed leakage of cell contents, presumably DNA and protein, accompanied by the formation of one or two spherical “blebs” at the division points. The blebs appear similar to outer membrane vesicles (OMVs), spherical vesicles with periplasmic content with a diameter of approximately 20 to 250 nm commonly found in Gram-negative bacteria (38).
Overexpression of hok causes membrane blebbing during division. hok was induced for 1 h with 1 mM IPTG, and the cells were observed by SEM. (A) Ea1189/pEVS143 cells at the population level; (B) representative cell of Ea1189/pEVS143; (C) Ea1189/pOE-hok cells at the population level; (D) representative cell of Ea1189/pOE-hok. Blue arrows point to membrane blebs, and the yellow arrow points to leakage of cell contents.
The growth of E. amylovora is affected by the hok-sok locus.To further examine if the hok-sok locus participates in additional biological processes in E. amylovora, we generated a hok-sok locus deletion mutant, Ea1189Δhok-sok, and the mutant was complemented with a plasmid copy of the hok-sok locus with its native promoter region, Ea1189Δhok-sok(pJP-hok) (Fig. S6). We first compared the growth rates of the E. amylovora strains and observed that the Ea1189Δhok-sok mutant exhibited slower exponential-phase growth, although it reached a similar bacterial population size at the end of the incubation period compared with the WT parent (Fig. 6A). The doubling time of Ea1189Δhok-sok during exponential growth was 190 min, compared with 99 min for WT Ea1189. When the hok-sok locus was introduced in trans, the growth defect of the Ea1189Δhok-sok strain was largely restored, with a doubling time of 108 min for the complemented strain. These results suggest that the hok-sok locus plays important roles in regulating the growth of E. amylovora.
Effects of the hok-sok locus on bacterial growth (A) and morphology (B) of E. amylovora.
Mutation in the hok-sok locus affects the morphology of a subpopulation of E. amylovora cells.The micromorphology of WT Ea1189 and the Ea1189Δhok-sok mutant was observed by SEM when the bacterial population reached an optical density at 600 nm (OD600) of 0.5 and an OD600 of 1.5, representing cells in exponential phase and stationary phase, respectively. In exponential phase, we observed elongated nondividing cells in cultures of Ea1189Δhok-sok but not in the WT (Fig. 6B). In stationary phase, long bacterial filaments were exclusively formed in populations of Ea1189Δhok-sok (Fig. 6B).
The hok-sok locus affects catalase activity.The deferred bacterial growth and formation of filaments in Ea1189Δhok-sok are reminiscent of the phenotypes observed in a catalase- and peroxidase-defective mutant (denoted Hpx−) of E. coli that grows poorly with aeration and forms bacterial filaments due to the inability to detoxify reactive oxygen species (ROS) produced during growth (39). We wondered whether deletion of the hok-sok locus resulted in defects in detoxifying the ROS that accumulated during growth in E. amylovora. We therefore examined the catalase activity of WT Ea1189 and Ea1189Δhok-sok. Our results suggested that the hok-sok locus contributed to the catalase activity in E. amylovora, as indicated by the reduced height of foam formed after supplementation with hydrogen peroxide in the hok-sok mutant (Fig. 7A and Fig. S7). The heights of foam were converted into catalase activity units, which were calculated based on a standard curve using a purified catalase enzyme of known amounts. Consistent with this result, Ea1189Δhok-sok also exhibited compromised survival in LB plates amended with hydrogen peroxide (Fig. 7B). Catalase activity in E. amylovora has recently been shown to be conferred by two genes, katA and katG, with katA playing the major role (40). To further explore the mechanism of how the hok-sok locus affects catalase activity in E. amylovora, we constructed transcriptional fusions of each of the katA and katG promoters to a green fluorescent protein reporter. We observed decreased transcription from the katA promoter but not the katG promoter in Ea1189Δhok-sok compared to WT Ea1189 (Fig. 7C). The decreased promoter activity of katA in Ea1189Δhok-sok was complemented through the introduction of the hok-sok locus in trans. Taken together, our results suggested that the hok-sok locus contributed to the catalase activity of E. amylovora through a direct or indirect effect on the transcription of the major catalase gene, katA.
The hok-sok locus affects the catalase activity of E. amylovora. (A) Catalase activities were quantified based on the height of the oxygen bubbles formed after hydrogen peroxide was added. Catalase activity units were calculated based on a calibration plot using defined units of catalase enzyme. (B) Hydrogen peroxide sensitivity assay. Serially diluted cultures were plated onto LB plates with or without the addition of 800 μM hydrogen peroxide. (C) Effects of the hok-sok locus on the promoter activities of the catalase genes katA and katG in E. amylovora. Promoter activities were quantified by measuring the relative fluorescence of green fluorescent protein normalized by the OD600 of the corresponding culture. Error bars indicate the standard deviations of the means; the asterisks above the bars suggest significant differences of the means (P < 0.05 using Student’s t test). ns, not significant.
DISCUSSION
In this study, we predicted the type I TA systems in plant-pathogenic bacteria and functionally characterized the biological roles associated with a chromosomally located hok-sok locus in E. amylovora Ea1189, including identifying effects on exponential growth, cell size, and catalase activity. Although the hok-sok locus was first characterized as a postsegregational killing system, the prevalence of this locus on the chromosome of all strains of E. amylovora with genomes available suggests that it is likely to confer important functional roles in this bacterium. The deduced amino acid sequence of Hok in E. amylovora was identical in 32 of 33 Spiraeoideae-infecting strains examined, and a V17I amino acid substitution was identified in the two Rubus-infecting strains. This could indicate possible host adaptation of the Hok protein in E. amylovora, which will be more evident with the availability of more genomes of Rubus-infecting E. amylovora strains. Based on our phylogenetic analysis, the chromosomally encoded Hok protein in E. amylovora is phylogenetically closest to the plasmid-encoded Hok proteins of other Enterobacteriaceae and distinct from the other chromosomally located Hok family proteins. The chromosomal location of an R1-like hok-sok type I TA system is currently unique to E. amylovora. Interestingly, all of the Hok-containing Erwinia species that we detected are either plant pathogens (E. amylovora, E. pyrifoliae, and E. tracheiphila) or insect pathogens (E. iniecta); Hok is missing in other species, such as E. tasmaniensis and E. billingiae, which are epiphytic to pome fruit trees and may be antagonistic to E. amylovora (41–44). This finding suggests that Hok family proteins may play important roles during infection of the pathogenic Erwinia spp. In agreement with this, a previous study indicated that hok-sok family loci were more prevalent in pathogenic E. coli strains than in nonpathogenic strains (36).
We confirmed that the hok-sok locus in E. amylovora Ea1189 encodes a type I TA system. In this system, the hok reading frame encodes a protein that is highly toxic to E. amylovora cells, the mok reading frame overlaps hok and is also highly toxic to E. amylovora cells, and the sok gene encodes an sRNA that counteracts hok-mediated toxicity through interaction with mok. These results agree with previous observations of this system in the plasmid R1 (9, 10). Our results suggested that sok was likely to act as a cis-encoded sRNA that interferes with mok and therefore blocks the translation of the downstream hok coding sequence. As determined by overexpression studies, the last four C-terminal amino acids played a minor role in the full toxicity of Hok, whereas the N-terminal amino acids were more essential for its toxicity. This agrees with our prediction of the transmembrane domain in Hok, where the N-terminal residues are anchored in the cytoplasm and the C-terminal residues are exposed outside the bacterial cell. However, a previous study showed that chemically synthesized truncated derivatives of Hok, Hok(1–28) and Hok(31–52), were toxic to E. coli when introduced into cells by electroporation (45). These phenotypic differences suggest that the core region for the toxicity of Hok in different bacteria may be different and that its toxicity may be affected by how it is delivered into the cells.
After induction of hok, we observed that the division of daughter cells was arrested and accompanied by the leakage of cell contents at the cell division point, presumably due to cell membrane damage. This agrees with the observation that induction of hok in E. coli causes a transparent-center appearance, called “ghost cells,” suggesting leakage of cell contents (7, 36). Interestingly, we also observed membrane blebbing at the division point after hok was induced. Membrane blebbing is a universal phenomenon in bacteria in response to disruption of the plasma membrane (46, 47) and suggests that E. amylovora Hok, like Hok homologs from other species, may function through membrane pore formation (16, 17). The rapid formation of a cell membrane near the division plane might make this location more vulnerable to the damage of Hok, consistent with this location as the site of visible blebbing. The bubbles protruding off the cell outer membrane appeared reminiscent of bacterial outer membrane vesicles (OMVs), round secreted vesicles that can form during normal bacterial growth or as a general envelope stress response (48–50). OMVs may transport nucleic acids, proteins, bacterial endotoxins, and virulence factors during infection (48, 49) or might eliminate harmful materials from the bacterial cell (50). It is not known whether pore-forming toxins such as Hok might help elicit the formation of OMVs.
While the overexpression of many TA system toxins elicits visible stress responses, we demonstrated here that deletion of hok-sok results in a lower rate of exponential growth, formation of filaments in a subpopulation of cells, decreased catalase activity, and reduced expression of the katA catalase-encoding gene. These findings suggest that the entire hok-sok module may be important for stress maintenance in E. amylovora. Filamentation is an anomalous form of bacterial growth caused by the longitudinal growth of cells without septum formation and cell division (51) and can be induced by DNA damage or other stresses that inhibit bacterial replication (52–56). Reactive oxygen species, including hydrogen peroxide, are common sources of DNA damage during bacterial stress (57, 58), and supplementation with hydrogen peroxide or deletion of catalase genes is a trigger of filamentation in some species (59, 60). During infection, E. amylovora must contend with ROS generated by the host defense response (40, 61–63). Regulation of bacterial catalase activity has been previously shown in the type II TA systems MqsR/MqsA and YafQ/DinJ in E. coli, where the antitoxins MqsA and DinJ both affect bacterial catalase by decreasing the level of RpoS, a positive regulator of katG and katE catalase genes (64–66). Catalase gene regulation may be important for a previously observed role of Hok not addressed in this study: ROS are thought to impair the entry of a cell into the dormant persister state, and ROS avoidance mechanisms have been linked to persister formation in some organisms (67). To the best of our knowledge, this study establishes a link between a type I TA system and regulation of catalase activity.
Infection and subsequent systemic spread of E. amylovora through tree hosts are accomplished via the coordinated expression of genes encoding a type III secretion system and biofilm development (68–71). During initial infection, E. amylovora cells trigger a defense response in host cells, during which they are exposed to elevated levels of ROS and to other plant defense compounds such as phytoalexins (63, 72). However, unlike in other bacterial pathogen-host interactions, where triggering plant defense typically results in a disruption of pathogenesis, E. amylovora cells are able to overcome the plant defense response and continue to invade the host (72). It is possible that the hok-sok locus contributes to pathogen survival in response to host defense compounds and also to pathogen growth inside the host, two traits which would distinguish pathogenic hok-sok-carrying Erwinia spp. from the nonpathogenic Erwinia spp. that do not carry this hok-sok locus. Further study will be needed to determine the potential mechanisms by which hok and/or sok may positively regulate catalase or other stress response factors.
MATERIALS AND METHODS
Bacterial strains, plasmids, and growth conditions.The bacterial strains and plasmids used in this study are listed in Table 1. Unless otherwise indicated, bacteria were routinely grown in Luria-Bertani (LB) broth or agar at 28°C. Media were supplemented with ampicillin (Ap) (100 μg/ml), chloramphenicol (Cm) (25 μg/ml), gentamicin (Gm) (10 μg/ml), or kanamycin (Km) (25 μg/ml), as necessary.
Bacterial strains and plasmids used in this study
Prediction of known type I toxin-antitoxin systems and phylogenetic analysis of Hok family proteins.The amino acid sequences of toxins in all known type I toxin-antitoxin systems (listed in reference 15) were downloaded from the NCBI (https://www.ncbi.nlm.nih.gov/) (see Text S1 in the supplemental material), and they were used as the queries in taxon-specific TBLASTN searches against all available genomes of the genera Acidovorax, Burkholderia, Clavibacter, Dickeya, Erwinia, Pantoea, Pectobacterium, Pseudomonas, Ralstonia, Streptomyces, Xanthomonas, and Xylella in the Nucleotide Collection (nr/nt) database (cutoffs of an E value of ≤1e−4, identity of ≥40%, and coverage of ≥50%). The significant hits in representative strains were used for heat map construction using the Interactive Tree of Life Web server (https://itol.embl.de/). For phylogenetic analysis, the amino acid sequences of Hok family proteins were aligned by using MUSCLE (73), and MEGA7 software (74) was used to construct the phylogenetic tree with the maximum likelihood method for 1,000-replicate bootstrap analysis.
5′ RACE assay, gene mutagenesis, and complementation.The reference genome sequence of E. amylovora ATCC 49946 was obtained from the NCBI (accession no. FN666575) (32). Homologs of hok and sok in E. amylovora were identified by BLAST searches. The transcription start site of sok was determined using 5′ rapid amplification of cDNA ends (RACE) as previously described, with some modifications (75). Briefly, 10 μg of bacterial total RNA was extracted and treated with 10 U Cap-Clip acid pyrophosphatase (Epicentre, Madison, WI, USA), an alternative to tobacco acid pyrophosphatase, for 1 h. Control RNA was incubated without Cap-Clip acid pyrophosphatase. After ethanol precipitation, A4 5′ RNA linker and RNA ligase (New England Biolabs, Ipswich, MA, USA) were added to the purified RNA, and this mixture was incubated at 17°C for 12 h. After purification, RNA was reverse transcribed using Applied Biosystems high-capacity cDNA reverse transcription kits (Life Technologies, Waltham, MA, USA). The 5′ transcription start site was determined using the RNA linker-specific primer JVO-0367 and the sok-specific primer sokR_GSP. The resultant PCR product from the Cap-Clip acid pyrophosphatase-treated sample was TOPO cloned into the pCR4-TOPO TA vector according to the manufacturer’s instructions (Thermo Fisher Scientific, Waltham, MA, USA). The resultant recombinant vector was Sanger sequenced to obtain the precise transcriptional start site of sok. The chromosomal hok-sok locus was knocked out routinely via the λ red recombinase system (69, 76). To complement the E. amylovora Δhok Δsok mutant, the mutant was transformed with pJP-hok, ligating the hok reading frame, 150 bp downstream of the reading frame, and 579 bp upstream of the reading frame into pBBR1MCS-5, which contained the hok reading frame, the mok reading frame, the sok small RNA (sRNA) gene, and their corresponding predicted native promoters (77).
DNA manipulations.The native hok coding DNA sequence (CDS) and the hok derivatives were PCR amplified with the primers listed in Table 2. The PCR products were digested with the corresponding restriction enzymes and cloned into pEVS143 to construct isopropyl-β-d-1-thiogalactopyranoside (IPTG)-inducible overexpression plasmids (69). To generate the overexpression construct of the sRNA sok, sok was PCR amplified (Table 2), digested with EcoRI and XbaI, and cloned into pHMB1 (78). For all the assays in this study, induction of a gene or sRNA was obtained by adding a final concentration of 1 mM IPTG to the medium. To construct pPROBE-NT::katA and pPROBE-NT::katG, a ligation-independent PCR cloning method was used (79). Briefly, katA, katG, and pPROBE-NT were PCR amplified with the primers listed in Table 2. The resultant PCR products of the DNA insert and plasmid backbone were mixed in a 1:1 ratio and transformed into E. coli DH5α or E. amylovora Ea1189 using the heat shock method.
Oligonucleotide primers used in this study
Growth curve and growth arrest assay.Cultures grown overnight were washed twice and diluted to an OD600 of 0.1 in fresh LB broth, and 200 μl of the resuspended culture was added to a 96-well plate (Corning Inc., Corning, NY, USA). The OD600 was measured every 30 min with periodic shaking using a Tecan spectrophotometer. The doubling time of bacteria was estimated according to the formula of Poisot et al. (80). To monitor the arrest of bacterial growth mediated by the hok gene and the hok derivatives, equilibrated cultures at an OD600 of 0.1 in LB broth or liquid hrp-inducing minimal medium (HrpMM) (81) were withdrawn every hour after IPTG supplementation; CFU were measured through serial dilution plating on LB plates with selective antibiotics.
Scanning electron microscopy.Cultures grown overnight were washed twice and diluted to an OD600 of 0.05 in 3 ml fresh LB broth in 14-ml polystyrene tubes (Corning Inc.). Cultures were incubated at 28°C with shaking (200 rpm) until the OD600 reached 0.5 or 1.5, representing cells in exponential phase or stationary phase, respectively. Cultures were harvested and washed twice with 0.5× phosphate-buffered saline (PBS). Culture suspensions were fixed with an equal amount of 4% glutaraldehyde in a 0.1 M sodium phosphate buffer (pH 7.4) at 4°C for 1.5 h. Samples were dehydrated using a series of ethanol gradations (25%, 50%, 75%, 95%, and 100%). Samples were coated with osmium in a Neoc-AT osmium chemical vapor deposition coater (Meiwafosis Co. Ltd., Osaka, Japan) and visualized with a JEOL 7500F (field emission emitter) scanning electron microscope (JEOL Ltd., Tokyo, Japan).
Catalase assay and hydrogen peroxide sensitivity assay.The catalase assay was performed according to a previously described method (82), with some modifications. Briefly, cultures grown overnight were washed and resuspended in 0.5× PBS buffer (OD600 of 1.0), after which 200 μl of the cells was thoroughly mixed with 100 μl 1% Triton X-100 (Sigma-Aldrich, St. Louis, MO, USA) and a 100-μl hydrogen peroxide solution (30% [wt/wt] in H2O in 14-ml polystyrene tubes [17-mm diameter by 100-mm height; Corning Inc.]). The mixtures were incubated at 22°C for 5 min. The final heights of the foam formed by O2 were measured, and catalase activity units were calculated by comparison to a standard curve of purified catalase (10,000 U/mg [1 U decomposes 1 μmol of hydrogen peroxide per min at pH 7.0 at 25°C]; Sigma-Aldrich). To measure the sensitivity of bacterial cells to hydrogen peroxide, cultures grown overnight were adjusted to an OD600 of 1.0, and 10-μl samples of appropriate serial dilutions were drop plated onto LB plates with or without the addition of hydrogen peroxide. For hydrogen peroxide-supplemented plates, autoclaved LB agar was cooled to 56°C, and 800 μM hydrogen peroxide was added immediately before medium was poured. Poured plates were dried in a fume hood for an hour before use. In our preliminary experiments, no colonies were formed in the E. amylovora Δhok Δsok mutant when >1 mM hydrogen peroxide was supplemented in the medium. Hydrogen peroxide at 800 μM was therefore used for the best display of the results.
ACKNOWLEDGMENTS
This project was supported by Agriculture and Food Research Initiative Competitive Grants Program grant no. 2015-67013-23068 from the USDA National Institute of Food and Agriculture and by Michigan State University AgBioResearch.
FOOTNOTES
- Received 27 March 2019.
- Accepted 8 May 2019.
- Accepted manuscript posted online 17 May 2019.
Supplemental material for this article may be found at https://doi.org/10.1128/AEM.00724-19.
- Copyright © 2019 American Society for Microbiology.
REFERENCES
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.
- 84.
