Editorial Note
Readers have raised concerns about quantitative results published in bar graphs in Figs 2–3, S3-S7, and S9 of this article [1, 2]. Specifically, it was noted that the bars and axes demarcations are not consistently spaced within the bar graphs; there are some misalignments between error bars and data points, or between axes and their demarcations; some bars have lower edges that appear slightly above or below the X axis line; and some bars are similar to each other.
Also, in Figs 2A, 2B, 2C, S3A, S3B, S3C, S3D, S5B, S5C, and S6B, some data annotations (values) on the graphs do not agree with bars’ alignment with values on the Y axis. The authors stated that the data annotations within the graphs (i.e., above the bars) provide the correct results.
In addition, in Fig S5C, the data values on top of the bars are tenfold higher than the scale on the Y axis. The authors stated that the measured population size was 106 for stained population size and the Y axis for Fig S5C should be labelled, ‘Absolute number of dead cells (x 106)’.
The authors stand by the results and commented that these issues are due to the method(s) used to prepare the graphs. They stated that they used an early illustrator program into which they copied the values of calculated results (means and standard deviations) of three independent experiments for plotting on electronic millimeter grid paper [3].
The authors stated that the original raw data underlying the graphs of concern are not fully available at this time given the time elapsed. In light of this issue, PLOS has been unable to verify the results and the article does not currently comply in full with the applicable version of the PLOS Data Availability policy*. Nevertheless, the editors consider that the explanation provided by the authors sufficiently clarifies how the graphs were generated.
*The PLOS Data Availability policy was updated in 2014. The former policy (which was in place in 2008–09, [4]) still required authors to make data integral to an article’s findings freely available without restriction. The pre-2014 policy advised that data should be provided in Supporting Information files, a public repository, or upon request, but it did not strictly require publication or deposition of all quantitative data.
13 Sep 2023: The PLOS ONE Editors (2023) Editorial Note: A Differential Effect of E. coli Toxin-Antitoxin Systems on Cell Death in Liquid Media and Biofilm Formation. PLOS ONE 18(9): e0291587. https://doi.org/10.1371/journal.pone.0291587 View editorial note
Correction
2 Oct 2015: Kolodkin-Gal I, Verdiger R, Shlosberg-Fedida A, Engelberg-Kulka H (2015) Correction: A Differential Effect of E. coli Toxin-Antitoxin Systems on Cell Death in Liquid Media and Biofilm Formation. PLOS ONE 10(10): e0140184. https://doi.org/10.1371/journal.pone.0140184 View correction
Figures
Abstract
Toxin-antitoxin (TA) modules are gene pairs specifying for a toxin and its antitoxin and are found on the chromosomes of many bacteria including pathogens. Here we report how each of five such TA systems in E. coli affect bacterial cell death differently in liquid media and during biofilm formation. Of all these systems, only the TA system mazEF mediated cell death both in liquid media and during biofilm formation. At the other extreme, as our results have revealed here, the TA system dinJ-YafQ is unique in that it is involved only in the death process during biofilm formation. Cell death governed by mazEF and dinJ-YafQ seems to participate in biofilm formation through a novel mechanism.
Citation: Kolodkin-Gal I, Verdiger R, Shlosberg-Fedida A, Engelberg-Kulka H (2009) A Differential Effect of E. coli Toxin-Antitoxin Systems on Cell Death in Liquid Media and Biofilm Formation. PLoS ONE 4(8): e6785. https://doi.org/10.1371/journal.pone.0006785
Editor: Stefan Bereswill, Charité-Universitätsmedizin Berlin, Germany
Received: June 28, 2009; Accepted: July 24, 2009; Published: August 26, 2009
Copyright: © 2009 Kolodkin-Gal et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This research was supported by grant No. 177/07 from the Israel Science Foundation (ISF) administrated by the Israel Academy of Science and Humanities, and by grant No. 2005029 from United State-Israel Bi-national Science Foundation (BSF), and by Grant Number GM069509 of the National Institute of Health(USA). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Toxin-antitoxin systems consist of a pair of genes that specify two components: a stable toxin and an unstable antitoxin that interferes with the action of the toxin. Several toxin-antitoxin modules have been identified in the chromosome of E. coli. Among them are: mazEF [1]–[3], chpBIK [1], [4], relBE [5]–[7], yefM-yoeB [8]–[10], dinJ-yafQ [11]. The well studied mazEF system was the first to be described as regulatable and responsible for bacterial programmed cell death [12]. mazF encodes for the stable toxin MazF and mazE encodes for the labile antitoxin MazE. MazE is degraded by the ATP-dependent ClpPA serine protease[12]. MazF is an endoribonuclease that cleaves mRNAs at ACA sequences in a ribosome-independent manner [13], [14]. As long as MazE and MazF are co-expressed, MazE counteracts the toxic activity of MazF [12]. Since MazE is a labile protein, preventing MazF-mediated action requires the continuous production of MazE. Thus, any stressful condition that prevents the expression of the chromosomally borne mazEF module will lead to the reduction of MazE in the cell, permitting toxin MazF to act freely. Such conditions include: (i) antibiotics inhibiting transcription and/or translation like rifampicin, chloramphenicol, and spectinomycin [15]; and ii) antibiotics causing DNA damage like trimethiprim or nalidixic acid [16], [17]. These antibiotics, and some other stressful conditions that are well known to cause bacterial cell death [18], [19], have been found to act through the mazEF module [15]–[17].
Another, well studied TA system is relBE, which is induced by stressful conditions that cause growth arrest by allowing the toxin RelE to act freely [20]. RelE causes cleavage of mRNA codons in the ribosomal A site. This cleavage is highly codon specific, occurring between the second and third nucleotides [21]. It has been suggested that RelE might not be an endonuclease itself, but rather may enhance the intrinsic cutting action of the ribosome when it pauses in the translation process [22]. The antitoxin RelB is degraded by the Lon protease [20].
Other than mazEF [23] and relBE [6], [7], [20], most other E. coli TA systems have been discovered only recently. Each of them inhibits translation; however, the various toxins differ in their primary targets and modes of action. chpBIK is partially homologous to mazEF [1]. Like MazF, the toxin ChpK is also an endoribonuclease that cleaves mRNAs in a ribosome-independent manner [14]. Unlike MazF that cleaves mRNAs only at ACA sequences [13], ChpK also cleaves at sequences ACU and ACG [14]. YoeB was also initially thought to function as an endoribonuclease cleaving translated mRNAs [9]. However, recently YoeB was found to be specifically associated with the 50S ribosomal subunit of E. coli and thereby it primarily inhibits translation initiation [24]. YafQ of the dinJ-yafQ system has been recently shown to be a toxin that functions different from other TA toxins. It is an endoribonuclease that associates with the ribosome and blocks translation elongation through sequence–specific and frame dependent mRNA cleavage [25]
Because of the similar organization of the genes in the TA modules, and the similar concept of specifying a stable toxin and labile antitoxin, TA modules are generally viewed as having similar roles in physiological processes. However, based on the fact that the mode of action the studied E. coli toxins varies, they may also differ in their role in bacterial physiology. Therefore, here, using similar experimental conditions, we compared the effects on cell death of the five confirmed E. coli TA systems (mazEF, chpBIK, relBE, yefM-yoeB, and dinJ-yafQ) in liquid media and in bioifilm formation. We found that these TA systems can be divided into four groups: 1) mazEF is involved in the death processes in both liquid and biofilm formation; 2) relBE only participates in cell death in liquid medium; 3) chpBIK and yefM-yoeB only participate in cell death in liquid medium under certain conditions; and 4) dinJ-yafQ only participates in the death process involved in biofilm formation, but not in liquid medium. We have shown, for the first time, that the process of E. coli biofilm formation required cell death which is governed by the TA systems mazEF and dinJ-yafQ. We particularly found that the dinJ-yafQ gene pair is a unique TA system that participates in the cell death process only in biofilm formation.
Results
Overproduction of each of five chromosomal E. coli toxins has a different effect on cell viability in liquid media
We wished to compare the toxicity of each of five overproduced E. coli chromosomal toxins. To this purpose, we used E. coli strain MC4100relA+ transformed with plasmid pBAD33 carrying an insert of each of five E. coli toxins under the regulation of the araBAD promoter. We grew the transformed cells in liquid LB medium to mid-logarithmic phase, and then induced the toxins by adding arabinose. We found that overproducing each of these E. coli toxins led to significantly different results (Figure 1A): The most rapid and dramatic loss of viability was caused by overproducing either MazF or RelE: in only one hour, the number of colony forming units was decreased by 3 orders of magnitude, in 6 hours by more than 4 orders of magnitude. Overproducing YoeB caused a similar reduction in cell viability, but only 2 hours after the toxin was induced. Overproducing ChpK affected cell viability only about half as much. Surprisingly, overproducing YafQ did not affect cell viability at all (Figure 1A); even in defined M9 medium, in which overproducing the other four toxins caused drastic reductions in cell viability, overproducing YafQ had no effect (Figure S1). YafQ is a functional toxin, because, as demonstrated previously and here (Figure 1B, Figures S2A and B), when tested for its ability to induce growth arrest, YafQ prevents cell growth as do other E. coli toxins.
(A) Cells were grown in LB medium at 37°C to mid-logarithmic phase (OD600 = 0.5). At time zero, 0.2% arabinose was added to the cultures to induce the toxin expression (Materials and Methods). (B) Overproduction of five of E. coli toxins in cells grown on solid medium. E. coli strain MC4100 containing BAD33 carrying each of five of the chromosomal E. coli toxins were plated on LB medium plates (left) or LB plates applied with 0.2% arabinose (right) and incubated at 37°C.
Each of five of chromosomal E. coli toxin-antitoxin system has a different effect on cell viability following treatment with various antibiotics
Having found previously that mazEF mediated cell death is triggered by various antibiotics [15]–[17], we asked if cell death triggered by antibiotics might involve additional TA systems. We compared the viability of wild-type E. coli MC4100relA+ to the viability of mutants from which each of the five TA gene pairs had been deleted individually. Deleting each of the five TA systems affected cell survival differently (Figure 2). Mutants deleted for either mazEF or relBE did not die following treatment with various antibiotics (Figure 2). The effect on cell death of deleting yefM-yoeB or chpBIK was dependent on which antibiotics were used to trigger the cell death process: after treatment with antibiotics inhibiting transcription or translation, we observed only about 50% cell death in the ΔyefM-yoeB mutant (Figure 2A,B). In the ΔchpBIK mutant, there was a reduction in cell death only after treatment by DNA damaging agents (Figure 2C,D). Deleting dinJ-yafQ had no significant effect on cell death regardless of the antibiotic used to trigger the death process (Figure 2).
E. coli strains MC4100relA+ (WT) and its ΔmazEF, ΔchpBIK, ΔrelBE, ΔyefM-yoeB, or ΔdinJ-yafQ derivatives were grown in M9 medium to mid-log phase (Materials and Methods). Cells were incubated without shaking at 37°C with: (A) Rifampicin (10 µg/ml) for 10 min; (B) Spectinomycin (1 mg/ml) for 10 min; (C) Trimethoprim (2 µg/ml) for 1 hr; (D) Nalidixic acid (1 mg/ml) for 10 min (Materials and Methods). The results describe the average of three independent experiments that were carried out in triplicate. Error bars indicate standard deviations.
Since our experiments were carried out in bacterial strains that carried an intact relA gene, we also compared the effect on cell death of deletion of each of five TA systems in MC4100relA1 background. Since, as we have described previously, ppGpp has a role in mazEF-mediated cell death [17], using the same experimental conditions as for the relA+ strain we were not able to induce cell death in a relA1 strain (Data not shown). However, using a more drastic treatment with different antibiotics, we did observe mazEF and relBE-mediated cell death (Figure S3). We were surprised that in relA1 background, the ΔchpBIK mutant survived all stressful conditions, suggesting that the presence of ppGpp may disguise the involvement of the chpBIK TA system in cell death (Figure S3).
These results suggest that mazEF and relBE are the principal TA systems participating in cell death induced by antibiotics; that chpBIK and yefM-yoeB are TA systems that only in some cases participate in cell death, and that dinJ-yafQ has no role in this cell death process.
dinJ-yafQ and mazEF are involved in cell death during biofilm formation
We also asked whether the differential effect of TA systems is also reflected in biofilm formation. Using the E. coli deletion mutants described above, we studied the effect of each of the TA systems on biofilm formation. As early as 8 hours, we observed a significant decrease in biofilm formation in both the ΔmazEF and the ΔdinJ-yafQ mutants (Figure S4). Particularly in a defined medium, we observed only a partial decrease in biofilm formation in the ΔchpBIK, ΔrelBE, and ΔyefM-yoeB mutants (Figure S4). We subsequently studied the effects on later biofilm formation, after 24 hours, of deleting mazEF or dinJ-yafQ. Indeed, compared to their parental strain, after 24 hours, biofilm production was significantly reduced in both ΔmazEF and Δdinj-yafQ mutants (Figure 3 and Figures 4A,D). In rich LB medium, deleting dinJ-yafQ or mazEF almost completely abolished biofilm formation (Figure 4A, upper panel). In M9 minimal medium, deleting dinJ-yafQ or mazEF resulted in the formation of very thin, impaired biofilms (Figure 4D, upper panel).
E. coli strains MC4100relA+ (WT) and its ΔmazEF, ΔchpBIK, ΔrelBE, ΔyefM-yoeB or ΔdinJ-yafQ derivatives were grown in 96 well polystyrene plates at 37°C for 24 hr in (A) LB or (B) M9. Quantification of CV-stained attached cells was done as described in Materials and Methods. The results describe the average of three independent experiments that were carried out in triplicate. Error bars indicate standard deviations.
E. coli strains MC4100relA+ (WT), MC4100relA+ΔmazEF, and MC4100relA+ ΔdinJ-yafQ were grown at 37°C for 24 hr in 96 well polystyrene plates in LB (A,B,C) or M9 (D,E,F) media. Attached cells were washed and planktonic cells were removed as described (Materials and methods). Dead cells stained with PI (red) and living cells retained staining with SYTO 9 (green). (A,D) Biofilms were photographed by CLSM with a using ×2.5, ×10 magnifications. The entire well was photographed from above. The image is a representative image from three independent experiments; each experiment was carried out in quadruplicate. Additionally, cells stained with SYTO 9 or PI were quantified as described in “materials and methods”. (B,E) Percentage of dead cells from the total population in the biofilm. (C,F) Absolute number of dead cells. Data were obtained from two independent experiments, performed quadruplicate.
In a bacterial culture, the death of some of the cells may provide extra nutrient, signaling, and extra-cellular matrix molecules for the cells that remain living [27]. It seems likely that biofilm formation would be supported by an increase in the abundance of such molecules. We hypothesized that the affects of deleting dinJ-yafQ or mazEF on biofilm production could be the result of a defect in a cell death pathway. Based on this assumption, we predicted that biofilms produced by the WT strain would contain a larger proportion of dead cells then those produced by the ΔmazEF or the ΔdinJ-yafQ mutants. To test this idea, we used propidium iodide (PI) and SYTO 9, to distinguish between living and dead cells. When used alone, SYTO 9 stains both live and dead bacteria. In contrast, PI penetrates only dead bacteria with damaged membranes. When used together, PI causes a reduction in SYTO 9 fluorescence: live bacteria with intact membranes fluoresce green, while dead bacteria with damaged membranes fluoresce red. We analyzed a 24-hr biofilms produced by each strain, and stained them with both PI and SYTO 9, (Figures 4A,D, lower panels). We calculated the percentage of dead cells in each biofilm by dividing the number of dead cells by the total number of cells (dead+live cells). We found that in both LB and in M9 media, the fraction of dead cells in the biofilm of the WT parental strain contained 100 times more dead cells than the biofilm of the dinJ-yafQ deletion mutant (Figures 4B and 4E). In LB medium, the biofilm of the mazEF deleted mutant contained 10 times fewer dead cells than the biofilm of the parental strain (≈9% versus 45% Figure 4B). Based on these results, we predicted that optimal biofilm formation might require a minimal threshold number of dead cells, so we determined the absolute number of dead cells in each case. Compared to the biofilms of the ΔmazEF or the ΔdinJ-yafQ mutants, the biofilm of the parental WT strain included at least 2 orders of magnitude more dead cells (Figures 4C and 4F). This suggested that mazEF and dinJ-yafQ mediated cell death in E. coli play a pivotal role in biofilm formation. The requirement for the active cell death process mediated by the TA systems mazEF and dinJyafQ for proper biofilm formation is also reflected by the significant difference(s) in the number of total viable cells inside the biofilm. The biofilm of both mazEF and dinJyafQ mutants contain at least 10 times less viable cells than their parental WT strain (data not shown).
Based on the results described above, we reasoned that overexpression of the MazF and YafQ toxins may increase biofilm formation in ΔmazEF and ΔdinJ-yafQ. After overexpressing MazF and YafQ with 0.05% arabinose, we found that both toxins fully complemented the defect in biofilm formation of their cognate TA deleted mutant (Figure S7). This result further support our finding that the defect of the mutants deleted for either dinJ-yafQ or mazEF in biofilm formation is a result of a defect in cell death pathway (Figure 4).
Note that for biofilm formation over a 24 hour period, in both in rich LB medium (Figure S5A lower panel and Figures S5B,C), and in a defined M9 defined medium (Figure S6A lower panel and Figures S6B,C) deleting the TA systems chpBIK, or yefM-yoeB had much less significant effect than deleting either mazEF or dinJ-yafQ. In addition, though deleting the relBE TA gene pair led to a partial defect in biofilm formation (Figures 3, S4, S5A, S6A), the loss of relBE had no effect on the proportion of dying cells during 24 h of biofilm formation comparing with the WT strain (Figure S5 and S6). Thus, we suggest that there may be an additional role for the relBE TA system in biofilm formation that is not connected to cell death.
Discussion
It is well known that bacteria can undergo between two physiological states: a free-swimming plaktonic state and in surface-associated communities called biofilms [26]. Biofilms are viewed as complex communities of bacteria resulting through multi-developmental stages that can be viewed as a multicellular behavior [26]. We have previously described the E. coli TA module mazEF programmed cell death system as one of the facets of multi-cellular behavior in bacterial populations. When challenged by stressful conditions, the bacterial population acts like a single multicellular organism in which a sub-population dies, thereby permitting the continued survival of the bacterial population as a whole [28], [29]. So, here we asked: Is mazEF mediated cell death involved in biofilm formation? Moreover, if mazEF is involved, is it a representative of other E. coli TA systems?
We found that with respect to cell death, the several TA systems behaved differentially (Figure 5). Based on our previous studies [23], [29] and here, mazEF is the regulating module mediating cell death both in liquid media (Figure 2) and in biofilm formation (Figures 3 and 4). relBE seems to be a principal mediator of cell death only in liquid media (Figure 2, [30]), but not in biofilm formation (Figure 3). We found that chpBIK, which is homologous to mazEF [1], [4], seems to be a back-up death system for mazEF. In a relA+ E. coli strain in which ppGpp is produced [31], the principal death mediator was mazEF (Figure 2); however, in the isogenic relA1 strain, in which ppGpp is not produced [31], chpBIK was required to mediate cell death (Figure S3). In addition, under our experimental conditions, the yefM-yoeB module mediates cell death only in liquid media, but only in some cases (Figure 2, and Figure S3), while not at all in bioiflm formation (Figure 3, Figure S5 and S6). The results of this study revealed that that the TA system dinJ-yafQ is unique: even when the toxin YafQ is over-expressed, it may not mediate cell death in liquid media (Figures 1 and 2). However, strikingly, dinJ-yafQ seems to be a principal mediator of cell death in biofilm formation (Figures 3 and 4). This is the first study showing a specific function for dinJ-yafQ.
Principle TA systems for cell death are in red, and secondary in black broken arrow.
It was recently reported that all E. coli TA systems are involved in biofilm formation [32]. However, these experiments were performed in strain MG1655, which is defective in cell death [33]. We report here the following novel findings regarding E. coli TA systems and biofilm formation: (i) Among the five studied E. coli TA systems, only mazEF and dinJ-yafQ have a role; (ii) The role of mazEF and dinJ-yafQ in biofilm formation is related to their role cell death; and (iii) A threshold of dead cells seems to be important for biofilm formation. We are presently studying the specific stage in biofilm formation that may be affected by mazEF and/or dinJ-yafQ mediated cell death and whether these may be related to the planktonic death pathways about which we have reported [28], [29]. It was previously reported that cidA-controlled cell lysis has a significant role in biofilm development in Staphylococus aureus, and that the released genomic DNA is an important structural component of the biofilm of this bacteria [34]. In our work here, we found that the addition of DNAse did not reduce E.coli biofilm formation (Figures S8 and S9). It seems, therefore, that the role of cell death in E.coli biofilm formation is not through the release of DNA. Thus, mazEF and dinJ-yafQ mediated cell death participate in biofilm formation through novel yet unknown mechanism(s).
Materials and Methods
Bacterial strains and plasmids
E. coli strains: MC4100relA+, its ΔmazEF derivative [26], MC4100relA1 and its ΔmazEF derivative [17]. Using Red-mediated homologous recombination [35], we constructed the chloramphenicol resistant derivatives: ΔrelBE, ΔdinJ-yafQ, ΔchpBIK and ΔyefM-yoeB of E. coli strains MC4100relA+ and MC4100relA1. For strain MC4100relA+ ΔdinJ-yafQ, the CamR was eliminated [35]. We used pBAD-mazF [36], and constructed pBAD-yafQ as follows: yafQ gene was PCR amplified from strain MC4100 and cloned using KpnI and HindIII sites into the plasmid pBAD33 [37] bearing an chloramphenicol resistance gene, downstream of the arabinose pBAD promoter.
Materials and media
The bacteria were grown in liquid M9 defined medium [27] with 1% glucose and applied with each amino acid (10 µg/ml) or LB [17] and then plated on rich LB agar plates [17]. The following materials were obtained from Sigma: L-arabinose, nalidixic acid, mitomycin C, trimethoprim, rifampin, serine hydroxamate, chloramphenicol, spectinomycin, crystal violet and DNAse I. PI and SYTO 9 were obtained from Invitrogen (Carlsbad, California).
Determining the effect of each toxin overproduction
E.coli strains MC4100 deleted with each of the five TA modules were transformed with pBAD33 carrying an insert of its cognate toxin and were grown in M9 medium containing 0.5% glycerol as a carbon source or in LB medium with chloramphenicol (50 µg/ml). Then cells were treated as described in Figure 1 and S1 legends. To determine CFU, samples were withdrawn at various time points and spread on LB plates containing 50 µg of chloramphenicol per ml and 0.2% glucose to repress further expression of the toxins. The percentage of the survivors was calculated by comparing the CFU of the induced culture to that of the uninduced culture at time zero. Error bars indicate standard deviations.
Determining the effect of stressful conditions on mazEF-mediated cell death
Cells were grown in M9 medium with shaking (160 rpm) at 37°C for 12 hours. The cells were diluted 1∶100 in 10 ml of M9 medium and were grown with shaking (160 rpm) at 37°C to mid-logarithmic phase (OD600 0.6). Samples of 500 µl were withdrawn into Eppendorf tubes (1.5 ml volume) and were further incubated without shaking at 37°C for 10 min as described for each case. Antibiotics were applied as described in each figure legend. The cells were centrifuged and re-suspended in pre-warmed saline, diluted, plated on pre-warmed LB plates and incubated at 37°C for 12 hours. Cell survival was calculated by comparing the number of the colony-forming units of cells treated by stressful conditions to those of the cells that were not exposed to the treatment.
Quantification of Crystal Violet (CV) -stained attached cells
Cells were grown overnight in LB or M9 medium. Then, cells were diluted 1∶100 in the same medium and grown in standing 96 wells polystyrene (without shaking). Quantification of CV stained attached cells was done as described previously (38) with a few modifications: After the wells were stained with 125 µl of 1.0% CV, rinsed and thoroughly dried, the CV was solubilized by the addition of 200 µl of 95% ethanol. A 125 µl sample of the solubilized CV was removed and added to a fresh polystyrene 96-well dish, and absorbance at OD590 was determined using Microplate Reader from from BMC (Ontario, Canada).
Determination of the Number of Dead cells in E. coli biofilms
E.coli cells were grown in 96 wells polystyrene plates as described in Figures 4, S5 and S6 legends. The culture supernatants were discarded, and the leftover biofilms in each well were fixed in 3% formalin solution for 15 min. The supernatants were discarded, and the leftover biofilms were washed with saline. Biofilms were incubated with a solution of 4 mM PI and 4 mM SYTO 9 for additional 15 min. The wells were washed again in saline and applied with 100 µl of 50% glycerol. Wells were analyzed by confocal laser scanning microscopy (CLSM) using a LSM 410 confocal microscope (Carl Zeiss, Inc., Thornwood, NY) under 1,000 magnification. Image acquisition was performed by using Carl Zeiss LSM software version 3.99 (Carl Zeiss, Inc.). After counting all six fields of view, the number of red cells was divided by the total number of cells (red+green) and multiplied by 100 to calculate the percentage of dead cells in each sample.
Supporting Information
Figure S1.
The effects of overproducing five E. coli toxins in cells grown in liquid M9 minimal medium. The E. coli strains used in the experiments from Fig. 1 were treated as described in the Legend to Fig. 1 except that cells were grown in M9-Glycerol medium (rather than in LB).
https://doi.org/10.1371/journal.pone.0006785.s001
(0.04 MB PDF)
Figure S2.
Growth arrest induced by the overproduction of five of E. coli chromosomally borne toxins. E. coli strain MC4100 was transformed with pBAD33 carrying an insert of one of five E. coli toxins. Cells were grown at 37°C in (A) liquid LB medium with 0.2% glucose (B) glycerol M9 minimal medium to an A600nm = 0.3–0.5. Then, growth medium of was changed into fresh medium with 0.2% arabinose. Bacterial growth was assessed by measuring A600 nm every 30 min.
https://doi.org/10.1371/journal.pone.0006785.s002
(0.05 MB PDF)
Figure S3.
The effect of each of five of chromosomal encoded TA systems on E. coli cell survival following treatment with different antibiotics. E. coli strains: MC4100relA1 (WT) and its DchpBIK, DrelBE, DyefM-yoeB and DdinJ-yafQ derivatives (in gray) were grown to mid-log phase as described in Materials and Methods. Cells were incubated without shaking at 37°C with: (A) Rifampicin (25 µg/ml) for 10 min; (B) Spectinomycin (2 mg/ml) for 10 min; (C) Trimethoprim (2 µg/ml) for 2 hr; (D) Nalidixic acid (2 mg/ml) for 10 min. Cells were plated and CFUs assessed as described in Materials and Methods.
https://doi.org/10.1371/journal.pone.0006785.s003
(0.29 MB PDF)
Figure S4.
The effect of each of five of chromosomal encoded TA systems on early biofilm formation in E. coli. E. coli strains MC4100relA+ (WT) and its DmazEF, DchpBIK, DrelBE, DyefM-yoeB or DdinJ-yafQ derivatives were grown in 96 well polystyrene plates at 37°C for 8 hr in (A) LB or (B) M9. Quantification of CV-stained attached cells was done as described in Materials and Methods.
https://doi.org/10.1371/journal.pone.0006785.s004
(0.18 MB PDF)
Figure S5.
The effects of DchpBIK, DrelBE and DyefM-yoeB on E. coli cell death during biofilm formation in LB medium. E. coli strains: MC4100relA+ (WT) and its derivatives DchpBIK, DrelBE, and yefM-yoeB were grown in 96 wells polystyrene plates at 37°C for 24 hr in LB medium. For the rest of the experiment, see the Legend to Fig. 4.
https://doi.org/10.1371/journal.pone.0006785.s005
(6.81 MB PDF)
Figure S6.
The effect of the deletions DchpBIK, DrelBE and DyefM-yoeB on E. coli cell death during biofilm formation in M9 medium. E. coli strains MC4100relA+ (WT) and its DchpBIK, DrelBE, and DyefM-yoeB derivatives were grown in 96 well polystyrene plates at 37°C for 24 hr in M9 medium. For the rest of the experiment, see the Legend to Fig. 4.
https://doi.org/10.1371/journal.pone.0006785.s006
(5.07 MB PDF)
Figure S7.
Overproduction of YafQ and MazF restores biofilm formation in a DmazEF and DdinJ-yafQ derivatives. E. coli strains MC4100relA+ (WT), MC4100relA+DmazEF/pBAD-mazF and MC4100relA+DdinJ-yafQ/pBAD-yafQ were grown in 96 wells polystyrene plates at 37°C in LB or LB+Arabinose 0.05% for (A) 8 h (B) 24 h. Quantification of CV-stained attached cells was done (Materials and Methods).
https://doi.org/10.1371/journal.pone.0006785.s007
(0.20 MB PDF)
Figure S8.
DNAse treatment does not affect early biofilm formation in E. coli. E. coli strains MC4100relA+ (WT) or its DmazEF and DdinJ-yafQ derivatives were grown in 96 wells polystyrene plates at 37°C for 8 h in (A)LB (B) M9. DNAse (1 or 10 Kunitz) was either applied or not applied to each well. Quantification of CV-stained attached cells was done (Materials and Methods).
https://doi.org/10.1371/journal.pone.0006785.s008
(0.21 MB PDF)
Figure S9.
E. coli strains MC4100relA+ (WT) or its DmazEF and DdinJ-yafQ derivatives were grown in 96 wells polystyrene plates at 37°C for 24 h in (A)LB (B) M9. DNAse (1 or 10 Kunitz) was either applied or not applied to each well. Quantification of CV-stained attached cells was done (Materials and Methods).
https://doi.org/10.1371/journal.pone.0006785.s009
(0.22 MB PDF)
Acknowledgments
We thank F. R. Warshaw-Dadon (Jerusalem, Israel) for her critical reading of the manuscript, and Idan Yelin for his help in preparing the manuscript.
Author Contributions
Conceived and designed the experiments: IKG RV ASF HEK. Performed the experiments: IKG RV ASF. Analyzed the data: IKG RV ASF HEK. Contributed reagents/materials/analysis tools: IKG RV. Wrote the paper: IKG HEK.
References
- 1. Masuda Y, Miyakawa K, Nishimura Y, Ohtsubo E (1993) chpA and chpB, Escherichia coli chromosomal homologs of the pem locus responsible for stable maintenance of plasmid R100. J Bacteriol 175: 6850–6.
- 2. Metzger S, Dror IB, Aizenman E, Schreiber G, Toone M, et al. (1988) The nucleotide sequence and characterization of the relA gene of Escherichia coli. J Biol Chem 263: 15699–704.
- 3. Engelberg-Kulka H, Sat B, Reches M, Amitai S, Hazan R (2004) Bacterial programmed cell death as a target for antibiotics. Trends Microbiol 12: 66–71.
- 4. Masuda Y, Ohtsubo E (1994) Mapping and disruption of the chpB locus in Escherichia coli. J Bacteriol 176: 5861–3.
- 5. Gotfredsen M, Gerdes K (1998) The Escherichia coli relBE genes belong to a new toxin-antitoxin gene family. Mol Microbiol 29: 1065–76.
- 6. Gerdes K, Christensen , SK , Lobner-Olesen A (2005) Prokaryotic toxin-antitoxin stress response loci. Nat Rev Microbiol 3: 371–82.
- 7. Bech FW, Jorgensen ST, Diderichsen B, Karlstrom OH (1985) Sequence of the relB transcription unit from Escherichia coli and identification of relB gene. EMBO J 4: 1059–66.
- 8. Grady R, Hayes F (2003) Axe-Txe, a broad-spectrum proteic toxin-antitoxin system specified by a multidrug-resistant clinical isolate of Enterococcus faecium. Mol Microbiol 47: 1419–32.
- 9. Christensen SK, Maenhaut-Michel G, Mine N, Gottesman S, Gerdes K, et al. (2004) Overproduction of the lon protease triggers inhibition of translation in Escherichia coli: involvement of the yefM-yoeB toxin-antitoxin system. Mol Microbiol 51: 1705–17.
- 10. Cherny I, Gazit E (2004) The YefM antitoxin defines a family of natively unfolded proteins: implications as a novel antibacterial target. J Biol Chem 279: 8252–61.
- 11. Hayes F (2003) Toxins-antitoxins:plasmid maintenance, programmed cell death, and cell cycle arrest. Science 301: 1496–9.
- 12. Aizenman E, Engelberg-Kulka H, Glaser G (1996) An Escherichia coli chromosomal “addiction module” regulated by guanosine 3′,5′-bispyrophosphate: a model for programmed bacterial cell death. Proc Natl Acad Sci U S A 93: 6059–63.
- 13. Zhang Y, Zhang J, Hoeflich KP, Ikura M, Qing G, et al. (2003) MazF cleaves cellular mRNAs specifically at ACA to block protein synthesis in Escherichia coli. Mol Cell 12: 913–23.
- 14. Zhang Y, Zhang J, Hara H, Kato I, Inouye M (2005) Links Characterization of ChpBK, an mRNA interferase from Escherichia coli. J Biol Chem 280: 3143–50.
- 15. Sat B, Hazan R, Fisher T, Khaner H, Glaser G, et al. (2001) Programmed cell death in Escherichia coli: some antibiotics can trigger mazEF lethality. J Bacteriol 183: 2041–5.
- 16. Sat B, Reches M, Engelberg-Kulka H (2003) The Escherichia coli mazEF suicide module mediates thymineless death. J Bacteriol 185: 1803–7.
- 17. Hazan R, Sat B, Engelberg-Kulka H (2004) Escherichia coli mazEF-mediated cell death is triggered by various stressful conditions. J Bacteriol 186: 3663–9.
- 18. Davies W (1998) Antibiotic resistance in bacteria in Emerging infections (Academic press, New York),239–273.
- 19. Ahmad SI, Kirk SH, Eisenstark A (1998) Thymieless death in recombinanation deficient mutants of Escherichia coli. Annu Rev Microbiol 52: 591–625.
- 20. Christensen S K, Mikkelsen M, Pedersen K, Gerdes K (2001) RelE, a global inhibitor of translation, is activated during nutritional stress. Proc Natl Acad Sci U S A 98: 14328–33.
- 21. Pedersen K, Zavialov AV, Pavlov MY, Elf J, Gerdes K, et al. (2003) The bacterial toxin RelE displays codon-specific cleavage of mRNAs in the riboosme A site. Cell 112: 131–40.
- 22. Hayes CS, Sauer RT (2003) Cleavage of the A site mRNA codon during ribosome pausing provides a mechanism for translational quality control. Mol Cell 12: 903–11.
- 23. Engelberg-Kulka H, Hazan R, Amitai S (2005) mazEF: a chromosomal toxin-antitoxin module that triggers programmed cell death in bacteria. J Cell Sci 118: 4327–32.
- 24. Zhang , Y , Inouye M (2009) Inhibitory mechanism of E. coli RelE/RelB toxin/antitoxin module involves a helix displacement near a mRNA interferase active site. J Biol Chem. Papers in Press published on March 18, 2009 as doi:10.1074/jbc.M809656200.
- 25. Prysak MH, Mozdzierz CJ, Cook AM, Zhu L, Zhang Y, et al. (2009) Bacterial toxin YafQ is an endoribonuclease that associates with the ribosome and blocks translation elongation through sequence-specific and frame-dependent mRNA cleavage. Mol Microbiol 71: 1071–87.
- 26. Aguilar C, Vlamakis H, Losick R, Kolter R (2007) Thinking about Bacillus subtilis as a multicellular organism. Curr Opin Microbiol 10: 638–43.
- 27. Kolodkin-Gal I, Hazan R, Gaathon A, Carmeli S, Engelberg-Kulka H (2007) A linear pentapeptide is a quorum-sensing factor required for mazEF-mediated cell death in Escherichia coli. Science 318: 652–5.
- 28. Amitai S, Kolodkin-Gal I, Hananya-Meltabashi M, Sacher A, Engelberg-Kulka H (2009) Escherichia coli MazF leads to simultaneous selective synthesis of both “death proteins “ and “survival proteins”. PLoS Genet 5: e1000390.
- 29. Engelberg-Kulka H, Amitai S, Kolodkin-Gal I, Hazan R (2006) Bacterial programmed cell death and multicellular behavior in bacteria. PLoS Genet 2: e135.
- 30. Godoy VG, Jarosz DF, Walker FL, Simmons LA, Walker GC (2006) Y-family DNA polymerases respond to DNA damage-independent inhibition of replication fork progression. Embo J 25: 868–79.
- 31. Potrykus K, Cashel M (2008) (p)ppGpp: still magical? Annu Rev Microbiol 62: 35–51.
- 32. Kim Y, Wang X, Ma Q, Zhang XS, Wood TK (2009) Toxin-antitoxin systems in Escherichia coli influence biofilm formation through YjgK (TabA) and fimbriae. J Bacteriol 191: 1258–67.
- 33. Kolodkin-Gal I, Engelberg-Kulka H (2008) The extracellular death factor: physiological and genetic factors influencing its production and response in Escherichia coli. J Bacteriol 190: 3169–75.
- 34. Bayles KW (2007) The biological role of death and lysis in biofilm development. Nat Rev Microbiol 5: 721–6.
- 35. Datsenko KA, Wanner BL (2000) One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci U S A 97: 6640–5.
- 36. Amitai S, Yassin Y, Engelberg-Kulka H (2004) MazF-mediated cell death in Escherichia coli: a point of no return J Bacteriol 186: 8295–300.
- 37. Guzman LM, Belin D, Carson MJ, Beckwith J (1995) Tight regulation, modulation, and high-level expression by vectors containing the arabinose pBAD promoter. J Bacteriol 177: 4121–30.
- 38. O'Toole GA, Kolter R (1998) Initiation of biofilm formation in Pseudomonas fluorescens WCS365 proceeds via multiple, convergent signaling pathways: a genetic analysis. Mol Microbiol 28: 449–61.