SGM Journals
Physiology And Biochemistry

Escherichia coli toxin gene hipA affects biofilm formation and DNA release

Junqiao Zhao, Qian Wang, Mingji Li, Björn D. Heijstra, Shengjun Wang, Quanfeng Liang, Qingsheng Qi

  • 1State Key Laboratory of Microbial Technology, National Glycoengineering Research Center, Shandong University, Jinan 250100, PR China
  • 2National Glycoengineering Research Center, Shandong University, Jinan 250100, PR China
  • 3School of Biological Sciences, University of Auckland, Private Bag 92019, Auckland, New Zealand
  • Correspondence
    Quanfeng Liang liangquanfeng{at}sdu.edu.cn

    • Microbiology 2013; 159(Pt 3):633–640 · https://doi.org/10.1099/mic.0.063784-0

    View at publisher PubMed

    Abstract

    Toxin–antitoxin (TA) systems in Escherichia coli may play a role in biofilm formation, but the mechanism involved remains debatable. It is not known whether the TA systems are responsible for extracellular DNA (eDNA) in biofilms. In this study, we investigated the function of the hipBA TA system in biofilm formation by Escherichia coli strain BW25113. First, the deletion of the HipBA TA system in E. coli BW25113 significantly reduced the biofilm biomass without antibiotic stress. Second, treatment of the BW25113 biofilm with DNase I caused a major reduction in biofilm formation, whereas similar treatment of the hipA mutant biofilm had only a minor effect. Third, the inactivation of HipA reduced the level of eDNA present in biofilm formation, and addition of BW25113 genomic DNA stimulated biofilm formation for both the wild-type and hipA mutant. Fourth, the wild-type cells underwent significantly more cell lysis than the hipA mutant. These results suggest that hipA plays a significant role during biofilm development and that eDNA is an important structural component of E. coli BW25113 biofilms. Thus, the TA system may enhance biofilm formation through DNA release.

    • Edited by: D. Demuth

    Introduction

    Most bacteria live in surface-associated communities called biofilms, while free floating micro-organisms are comparatively uncommon (Costerton et al., 1995). Micro-organisms attach via appendages such as fimbriae and flagella, accumulating until microcolonies are formed, further stimulated by the production of the microbial matrix, containing polysaccharides, glycoproteins and DNA (Karunakaran et al., 2011). This multicellular behaviour is crucial for effective host colonization, since 80 % of human bacterial chronic inflammatory and infectious diseases involve biofilms (Barraud et al., 2009).

    Toxin–antitoxin (TA) systems are widely present in bacterial chromosomes, including pathogens (Engelberg-Kulka et al., 2006). In Escherichia coli alone, the number of TA systems has recently increased from five to 37 (Hayes, 2003; Tan et al., 2011). However, the role of TA systems in cell physiology is controversial, with several possible roles identified (Magnuson, 2007): addictive genomic debris, stabilization of genomic parasites, selfish alleles, gene regulation, growth control, persister cell formation, programmed cell arrest, programmed cell death, and anti-phage measures. Recent studies have suggested that bacterial cell death and lysis play a significant role in biofilm formation and development (Rice et al., 2007; Webb et al., 2003). However, reports on the effects of TA systems on biofilm formation are limited.

    Investigation of the hip locus has revealed that it consists of two genes, hipA and hipB (Black et al., 1991; Moyed & Broderick, 1986). The hipA gene encodes a stable toxin, HipA, while hipB encodes the antitoxin HipB. The HipBA system is implicated in biofilm formation, providing multi-drug tolerance (Lewis, 2007, 2008). Biochemical and structural studies have shown that the toxin HipA is a member of the phosphatidylinositol 3/4-kinase superfamily and can phosphorylate the translation factor EF-Tu (Schumacher et al., 2009). This kinase activity is required for the induction of the dormant state of persister cells (Correia et al., 2006). The hipB gene product is a small Cro-like protein with a helix–turn–helix DNA-binding domain. It functions as a repressor for the hipBA operon by binding to four operator sites on the promoter region of hipBA, and inhibits HipA activity through covalent binding to the toxin (Black et al., 1991, 1994; Schumacher et al., 2009). It has been reported that expression of HipA in the wild-type can inhibit protein, RNA and DNA synthesis (Korch & Hill, 2006). Furthermore, low-level expression of HipA in the presence of glucose is sufficiently toxic to inhibit cell growth and affect cell viability (Korch et al., 2003). Overexpression of HipA in the ΔhipBA mutant causes almost all the cells to form filaments, with eventual lysis (Korch et al., 2003).

    Currently, toxin proteins from over 30 chromosomally encoded TA modules have been identified in E. coli. Chromosomal TA modules can be grouped into three main superfamilies based on whether the toxin has an RNase/gyrase-like fold, RNase barnase-like structure, or PilT N terminus (PIN) domain. The corresponding antitoxins contain DNA-binding domains (Buts et al., 2005). However, the TA system hipBA shows no homology to any member of the three TA superfamilies.

    While the mode of action of reported E. coli toxins varies, they also appear to differ in their role regarding bacterial physiology and subsequent biofilm formation (Kim et al., 2009; Kolodkin-Gal et al., 2009). Therefore, we investigated whether hipBA affected biofilm formation, even though the effects of several other TA systems on E. coli biofilm formation have been reported (Kim et al., 2009; Kolodkin-Gal et al., 2009).

    In the study presented here, we further investigated the relationship between TA systems and biofilm formation. Our results demonstrate that the TA system HipBA significantly affects biofilm formation in E. coli BW25113 without antibiotic pressure. Furthermore, treatment of the BW25113 biofilm with DNase I caused significant reduction in biofilm formation, whereas similar treatment of the hipA mutant biofilm had only a minor effect. Additionally, the inactivation of HipA reduced the level of extracellular DNA (eDNA) present during biofilm formation. To our knowledge, this is the first report that HipBA can influence E. coli biofilm formation and that a TA system may affect biofilm formation by producing eDNA in the biofilm.

    Methods

    Bacterial strains, plasmid and growth conditions.

    E. coli strains and the plasmid used in our work are listed in Table 1. We used p1TGL to measure β-galactosidase activity and GFP. p1TGL was constructed from p1TG (Li et al., 2012) by inserting the lacZ gene at the XhoI and SacII sites. The lacZ gene was amplified with primers lacz-up and lacz-down (Table 2). HipBA genes were amplified from wild-type BW25113 with primers hipBA-F and hipBA-R (Table 2) and ligated into plasmid pBAD24, digested by EcoRI and PstI. The recombinant plasmid was named pBAD-hipBA.

    Table 1. Strains and plasmids used in this work
    Table 2. Primers used in this work

    All strains were incubated at 37 °C. Biofilms were grown in the liquid rich medium TSBG (Rice et al., 2007) and minimal medium ABTCAA (Reisner et al., 2006). Crystal violet and bovine pancreatic DNase I with a specific activity of 2000 Kunitz units per milligram were obtained from Sigma. SYTO 9 was obtained from Invitrogen.

    HipBA operon inactivation.

    The hipBA mutant strain of BW25113 was created by using the one-step inactivation method (Datsenko & Wanner, 2000). This method includes the following steps: first, amplification of the resistance gene from pKD4 (KmR) using the primer hipBA-pKDF and hipBA-pKDR (Table 2). The PCR products flanked by FRT (FLP recognition target) sites and homologous sequences to hipBA were purified and electroporated into E. coli cells carrying plasmid pKD46.

    Complementation of the hip mutation by expression of HipBA.

    E. coli strain BW25113ΔhipBA was transformed with pBAD-hipBA or pBAD-24. The transformants were induced with 0.2 % arabinose at time zero, and biofilm formation was quantified with crystal violet stain as described below.

    Quantitative detection of crystal violet-stained biofilm.

    The biofilm attached to the bottom of 96-well plates was detected by spectrophotometric measurements. In brief, cells were grown overnight in ABTCAA medium, and diluted to a final OD600 of 0.05 with fresh medium and incubated in 96-well polystyrene plates (200 µl culture in each well) for 8 or 24 h without shaking. After incubation, cell turbidity at 600 nm was measured. The supernatant was removed, and the wells were washed twice gently with isotonic phosphate buffer (0.15 M, pH 7.2). The adherent biofilm was stained with 0.1 % (w/v) crystal violet (200 µl) and incubated at room temperature for 20 min. The excess crystal violet was removed by three gentle washes with distilled water. The crystal violet was solubilized by adding 95 % (v/v) ethanol, and the extent of biofilm formation was determined by the A540 using a microplate reader (BioTek). Normalized biofilm was calculated by dividing the biofilm value (A540) by the bacterial growth turbidity value (OD600) for each strain.

    Effect on biofilm formation of supplementation with extracellular genomic DNA.

    Overnight cultures were diluted to OD600 0.2 in the same TSBG medium containing 10 µg ml−1 DNA extracted from E. coli BW25113. Then, diluted cultures were incubated in 96-well polystyrene plates for 8 or 24 h without shaking. Crystal violet stain was used for quantification.

    Visualization of biofilm under static conditions.

    E. coli strains BW25113 and BW25113ΔhipA were transformed with p1TGL to express GFP constitutively. Overnight cultures were diluted to a final OD600 of 0.05 and incubated in 24-well polystyrene plates (1 ml culture in each well) without shaking. After 2 days, the polystyrene plates were centrifuged at 3000 g for 10 min. Then, the supernatants were removed and fresh ABTCAA medium was added. After 4 days of culture, the polystyrene plates were washed three times with isotonic phosphate buffer (0.15 M, pH 7.2). Finally, biofilms were visualized by fluorescence microscopy (Nikon Eclipse TE2000 S) with a Plan Fluor ×10/0.30 objective.

    DNase I treatment of biofilms on polystyrene microtitre plates.

    The incubation of biofilms was performed as described above. (1) For the crystal violet method, DNase I was diluted with ABTCAA medium to a final concentration of 100 U ml−1 and 200 µl was added to each well. After 8 h, crystal violet-stained biofilm quantification was performed as described above. Normalized biofilm was calculated by dividing the biofilm value (A540) by the bacterial growth value (OD600) for each strain. (2) For the fluorescence microscopy method, DNase I was diluted with isotonic phosphate buffer (0.15 M, pH 7.2) to a final concentration of 100 U ml−1. Polystyrene microtitre plates with 4 days-incubated attached cells were treated with 1 ml DNase I solution for 1 h. Biofilms were visualized by fluorescence microscopy as described above.

    Extraction and quantification of eDNA.

    BW25113 and its mutants were incubated overnight, after which the cultures were diluted to an OD600 of 0.05 in ABTCAA medium and grown for 96 h in polystyrene Petri dishes at 37 °C. eDNA in biofilms was extracted with a combination of N-glycanase and proteinase K treatment, as described before (Wu & Xi, 2009). eDNA was quantified using a spectrophotometer (NanoDrop ND-1000; Thermo Fisher Scientific) (Wu & Xi, 2009). To account for potential differences in biomass, the average OD600 of each unwashed biofilm was determined and used to calculate the relative OD600 of each biofilm with respect to the OD600 of the untreated BW25113 biofilm (whereby the relative OD600 of the BW25113 biofilm was taken as 1). The mass of eDNA per relative biomass of each biofilm was then calculated by dividing its total eDNA (ng) by its relative OD600 (Rice et al., 2007).

    Measurement of β-Galactosidase activity in culture supernatants.

    β-Galactosidase activity was measured as described elsewhere (Rice et al., 2007; Steinmoen et al., 2002). BW25113 and BW25113ΔhipA, both harbouring p1TGL, were incubated overnight, the cultures were diluted to an OD600 of 0.05 in ABTCAA medium and grown for 48 h. Samples were taken every 12 h and cell turbidity was measured at 600 nm. Cells were then removed by centrifugation, and β-galactosidase activity in the supernatants was determined with ONPG as the substrate.

    Statistical analysis.

    Statistical analyses were performed by using SPSS for Windows (version 13.0). A P value of <0.05 was considered statistically significant.

    Results

    The E. coli TA system HipBA affects biofilm formation

    The hipBA system is implicated in biofilm formation in response to several drugs (Lewis, 2007, 2008). However, whether HipBA affects biofilm formation without antibiotic supplementation has not been reported. Using the E. coli deletion mutants described in Methods, we studied the effect of HipBA on biofilm formation. Biofilm formation in polystyrene wells under static conditions was tested. The normalized biofilm formation by the hipA mutant and the hipBA mutant (0.50±0.09 and 0.27±0.01) was significantly less than the wild-type strain (1.02±0.16 and 0.51±0.02) after 8 and 24 h incubation (Fig. 1). To quantify the specific biofilm-forming ability, normalized biofilm was calculated by dividing the total biofilm by the bacterial growth for each strain. Furthermore, the mutant strain BW25113ΔhipBA was complemented by expression of HipBA and biofilm formation was measured in the 96-well plate crystal violet assay: the normalized biofilm of the hipBA mutant expressing HipBA was 0.98±0.13 and 0.49±0.01 after 8 and 24 h incubation. Hence, the hipBA mutant carrying plasmid pAD-hipBA showed suppression of the mutator phenotype.

    Figure image not available in archive
    Fig. 1.

    Effect of HipBA on E. coli biofilm formation. Normalized biofilm formation (total biofilm/growth) after 8 h (black columns) and 24 h (white columns) incubation at 37 °C for BW25113 and its hipA and hipBA mutants. The data are means from 16 replicate wells. Error bars, sd.

    The TA system may affect biofilm formation by controlling eDNA release

    Recent studies have shown that eDNA can be a structural component of the biofilm matrix (Heijstra et al., 2009; Rice et al., 2007; Spoering & Gilmore, 2006). However, an earlier study indicated that the addition of DNase did not reduce biofilm formation of E. coli MC4100 WT and its Δ5 TA system mutants (Kolodkin-Gal et al., 2009). We investigated whether eDNA was present in biofilm structures of E. coli BW25113 and studied the link between biofilm formation and DNA release mediated by the hipBA system. First, BW25113 and the hipA mutant were able to produce significantly more biofilm mass (all P<0.05) by adding 10 µg ml−1 genomic DNA (Fig. 2). This result suggests that extracellular genomic DNA directly enhances biofilm formation and that eDNA is a structural component of the biofilm matrix in E. coli BW25113. The effect of DNase I on biofilm formation was further investigated by utilizing polystyrene microtitre plates. When biofilm was grown on polystyrene microtitre plates, we observed a significant decrease in BW25113 biofilm formation after DNase I treatment for 8 h (Fig. 3). However, DNase I had only a minor effect on the hipA mutant biofilm (Fig. 3). Further, DNase I treatment for 1 h led to the formation of similar biofilm densities of BW25113 and its hipA mutant under a fluorescence microscope (Fig. 4 c, d). The addition of DNase I to growing cultures had no obvious effect on cell growth for any of the BW25113 strains (data not shown). Last, measurement of the eDNA level during biofilm formation is a possible way to investigate the link between the release of eDNA and the hipA mutant, so the eDNA from 96 h-old biofilms was obtained and quantified. This analysis indicated that the amount of eDNA present in a BW25113 biofilm (9.82±0.544 ng) was almost threefold greater than that present in the both the hipBA and hipA mutant biofilms (3.50±0.239 and 3.335±0.243 ng, respectively) when grown under similar conditions (Fig. 5). These results imply that eDNA is an important structural component in biofilms of E. coli BW25113 and that the inactivation of HipA reduces the level of eDNA present during biofilm formation.

    Figure image not available in archive
    Fig. 2.

    Effect of DNA supplements on biofilm formation. Normalized biofilm formation after incubation at 37 °C for BW25113 and the hipA mutant after 8 h (a) and 24 h (b). Genomic DNA (10 µg ml−1) was supplemented at time zero (black bars); white columns, controls. Each bar represents the mean of eight replicate wells; error bars, sem. A P value of <0.05 was considered statistically significant.

    Figure image not available in archive
    Fig. 3.

    Effect of DNase I on biofilm formation. Normalized biofilm formation (total biofilm/growth) after 8 h incubation at 37 °C for BW25113 and its hipA mutant. White columns, DNase I (100 U ml−1)-treated biofilm; black columns, control. The data are means from 16 replicate wells; error bars, sd.

    Figure image not available in archive
    Fig. 4.

    Biofilms of BW25113/p1TGL and BW25113ΔhipA/p1TGL attached on a 24-well plate after 4 days incubation at 37 °C. Images were acquired by with a Nikon Eclipse TE2000 S with a Plan Fluor ×10/0.30 objective. BW25113/p1TGL (a); BW25113ΔhipA/p1TGL (b); BW25113/p1TGL treated with DNase I (c); BW25113ΔhipA/p1TGL treated with DNase I (d). Scale bar, 40 µm.

    Figure image not available in archive
    Fig. 5.

    Quantification of eDNA present in BW25113, BW25113ΔhipBA and BW25113ΔhipA. eDNA concentration in biofilms after 4 days incubation in ABTCAA medium at 37 °C for BW25113 and the hipBA and hipA mutants. The values are expressed as nanograms eDNA per relative biofilm biomass, as described in Methods. The data were obtained from three independent cultures; error bars, sd.

    It has been suggested that hipA is a lethal toxin, because overproduction of HipA in the absence of HipB leads to decreased viability of E. coli BL21, K-12 and MG1655 (Black et al., 1994; Falla & Chopra, 1998; Korch et al., 2003). However, Korch & Hill (2006) have disputed this claim using E. coli MG1655. To determine whether HipA can mediate cell lysis in E. coli BW25113, we transferred p1TGL into BW25113 and the hipA mutant (p1TGL expresses the lacZ reporter constitutively). Cell lysis was measured by β-galactosidase activity in culture supernatants at 12 h intervals. Low levels of β-galactosidase activity were found in the hip mutant containing p1TGL, whereas higher levels of β-galactosidase activity in BW25113 containing p1TGL were observed at most time points (Fig. 6). No β-galactosidase activity was detected in BW25113 and the hipA mutant without p1TGL. These results indicate that the presence of HipA affects the integrity of E. coli BW25113 cells. Apparently, the lysis of a subpopulation of cells had an influence on eDNA concentrations. Therefore, HipA could trigger cell death and influence the amount of eDNA in the biofilm.

    Figure image not available in archive
    Fig. 6.

    β-Galactosidase activity assay in the culture supernatant. The supernatant β-galactosidase activities (solid lines) of BW25113 (squares) and BW25113ΔhipA (triangles), each harbouring p1TGL, were measured every 12 h. The growth curves of each strain are also shown (dashed lines). Data are means of three independent samples; error bars, sd.

    Discussion

    The first TA system reportedly linked to biofilm formation was the MqsR/MqsA pair of E. coli, where, upon deletion of mqsRA, biofilm formation decreased (Kasari et al., 2010; Ren et al., 2004). Further evidence of the role of TA systems in biofilm formation was obtained with E. coli Δ5 which had five of the most-studied TA systems deleted (Δ5, MazF/MazE, RelE/RelB, ChpB, YoeB/YefM and YafQ/DinJ) from E. coli MG1655 (Kim et al., 2009; Kolodkin-Gal et al., 2009). Kim and co-workers have reported that strain Δ5, compared with the wild-type, shows decreased biofilm formation at 8 h and increased biofilm formation at 24 h of incubation. This result is caused by overexpression of YjgK, a protein that affects the production of fimbriae involved in both biofilm attachment and dispersal (Kim et al., 2009). The five TA systems affect cell death differently during biofilm formation in E. coli strain MC4100 (Kolodkin-Gal et al., 2009). First, the mazEF system is the regulatory module, mediating cell death both in liquid media and during biofilm formation; second, the relBE system is a principal mediator of cell death only in liquid media, but not in biofilm formation; third, the chpBIK system is a back-up death system for the mazEF system in a ppGpp-independent manner; fourth, the yefM-yoeB system mediates cell death only in liquid media under certain conditions, but not in biofilm formation; fifth, the dinJ-yafQ system is a principal mediator of cell death only in biofilm formation. These results, when compared with earlier studies with MG1655 (Kim et al., 2009) and our studies with BW25113, provide different mechanisms for the effects of TA systems on biofilm formation.

    Although HipA is implicated in biofilm formation in the presence of several drugs (Lewis, 2007, 2008), its enhancement of biofilm formation in the absence of antibiotics and the mechanism have not been reported. Our results show that the deletion of HipBA of E. coli BW25113 significantly reduced the biofilm biomass.

    eDNA is an important component of the biofilm matrix (Heijstra et al., 2009; Rice et al., 2007; Spoering & Gilmore, 2006). Recent studies have shown that bacterial programmed cell death has multiple functions, including releasing eDNA that could have a function in structuring biofilms (Rice et al., 2007; Webb et al., 2003; West et al., 2007). In E. coli, some TA systems can mediate programmed cell death to influence cell lysis (Engelberg-Kulka et al., 2005). However, no evidence indicates that the TA system can affect biofilm formation through DNA release. Although Kolodkin-Gal and co-workers attempted to study the link between biofilm and DNA release mediated by TA-mediated cell lysis, they found that the addition of DNase did not reduce biofilm formation of MC4100 wild-type and Δ5. They suggested that the role of cell death in E. coli is not through the release of DNA; thus, MazEF and DinJ-YafQ-mediated cell death has a possible function in biofilm formation through novel, yet unknown, mechanism(s) (Kolodkin-Gal et al., 2009). However, the contribution of cell lysis to biofilm formation appears to vary from species to species and strain to strain (Rice et al., 2009; Wang et al., 2009; Zegans et al., 2009). We suspected hipA-mediated cell lysis and eDNA release to play an important role in biofilm formation in BW25113. To investigate this hypothesis, treatment with DNase I of biofilms of BW25113 and its hipA mutant was investigated. The results showed that DNase I treatment significantly decreased biofilm formation for E. coli strain BW25113, but that the treatment had only a minor effect on the hipA mutant. This hypothesis was also supported by the unequal level of cell lysis in the cultures with or without the presence of hipA in the genome. Additionally, by measuring eDNA levels in BW25113 and mutant biofilms, we directly demonstrated that hipA influences the production of eDNA and that a supplement of extracellular genomic DNA enhances biofilm development in both strain BW25113 and its hipA mutant. These results demonstrate that HipA plays a significant role during biofilm development and that released genomic DNA is an important structural component of the E. coli biofilm.

    Based on this study and earlier reported studies, it is clear that the E. coli TA systems differ in their role in bacterial physiology and biofilm formation (Kim et al., 2009; Kolodkin-Gal et al., 2009). In addition, the roles of TA systems in biofilm formation, cell lysis and DNA release also appear to vary within species and strains (Kim et al., 2009; Kolodkin-Gal et al., 2009; Rice et al., 2009; Wang et al., 2009; Zegans et al., 2009). Therefore, we need careful clarification on a case-by-case basis in future studies of the link between TA systems and their function in biofilm formation.

    Acknowledgements

    This work was supported by National key Basic Research Program of China (973 Program) 2012CB725202, the National Natural Science Foundation of China (31170097), grants from Shandong Science and Technology Development Plan (2010GSF10202, 2011GGE27097) and Independent Innovation Foundation of Shandong University (IIFSDU) 2012ZD029.

    References

      1. Baba T.,
      2. Ara T.,
      3. Hasegawa M.,
      4. Takai Y.,
      5. Okumura Y.,
      6. Baba M.,
      7. Datsenko K. A.,
      8. Tomita M.,
      9. Wanner B. L.,
      10. Mori H.
      (2006). Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection. Mol Syst Biol 2, 0008. doi:10.1038/msb4100050pmid:16738554
      1. Barraud N.,
      2. Storey M. V.,
      3. Moore Z. P.,
      4. Webb J. S.,
      5. Rice S. A.,
      6. Kjelleberg S.
      (2009). Nitric oxide-mediated dispersal in single- and multi-species biofilms of clinically and industrially relevant microorganisms. Microb Biotechnol 2, 370–378. doi:10.1111/j.1751-7915.2009.00098.xpmid:21261931
      1. Black D. S.,
      2. Kelly A. J.,
      3. Mardis M. J.,
      4. Moyed H. S.
      (1991). Structure and organization of hip, an operon that affects lethality due to inhibition of peptidoglycan or DNA synthesis. J Bacteriol 173, 5732–5739.pmid:1715862
      1. Black D. S.,
      2. Irwin B.,
      3. Moyed H. S.
      (1994). Autoregulation of hip, an operon that affects lethality due to inhibition of peptidoglycan or DNA synthesis. J Bacteriol 176, 4081–4091.pmid:8021189
      1. Buts L.,
      2. Lah J.,
      3. Dao-Thi M.-H.,
      4. Wyns L.,
      5. Loris R.
      (2005). Toxin-antitoxin modules as bacterial metabolic stress managers. Trends Biochem Sci 30, 672–679. doi:10.1016/j.tibs.2005.10.004pmid:16257530
      1. Correia F. F.,
      2. D’Onofrio A.,
      3. Rejtar T.,
      4. Li L.,
      5. Karger B. L.,
      6. Makarova K.,
      7. Koonin E. V.,
      8. Lewis K.
      (2006). Kinase activity of overexpressed HipA is required for growth arrest and multidrug tolerance in Escherichia coli. J Bacteriol 188, 8360–8367. doi:10.1128/JB.01237-06pmid:17041039
      1. Costerton J. W.,
      2. Lewandowski Z.,
      3. Caldwell D. E.,
      4. Korber D. R.,
      5. Lappin-Scott H. M.
      (1995). Microbial biofilms. Annu Rev Microbiol 49, 711–745. doi:10.1146/annurev.mi.49.100195.003431pmid:8561477
      1. Datsenko K. A.,
      2. Wanner B. L.
      (2000). One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci U S A 97, 6640–6645. doi:10.1073/pnas.120163297pmid:10829079
      1. Engelberg-Kulka H.,
      2. Hazan R.,
      3. Amitai S.
      (2005). mazEF: a chromosomal toxin–antitoxin module that triggers programmed cell death in bacteria. J Cell Sci 118, 4327–4332. doi:10.1242/jcs.02619pmid:16179604
      1. Engelberg-Kulka H.,
      2. Amitai S.,
      3. Kolodkin-Gal I.,
      4. Hazan R.
      (2006). Bacterial programmed cell death and multicellular behavior in bacteria. PLoS Genet 2, e135. doi:10.1371/journal.pgen.0020135pmid:17069462
      1. Falla T. J.,
      2. Chopra I.
      (1998). Joint tolerance to beta-lactam and fluoroquinolone antibiotics in Escherichia coli results from overexpression of hipA. Antimicrob Agents Chemother 42, 3282–3284.pmid:9835528
      1. Hayes F.
      (2003). Toxins-antitoxins: plasmid maintenance, programmed cell death, and cell cycle arrest. Science 301, 1496–1499. doi:10.1126/science.1088157pmid:12970556
      1. Heijstra B. D.,
      2. Pichler F. B.,
      3. Liang Q.,
      4. Blaza R. G.,
      5. Turner S. J.
      (2009). Extracellular DNA and type IV pili mediate surface attachment by Acidovorax temperans. Antonie van Leeuwenhoek 95, 343–349. doi:10.1007/s10482-009-9320-0pmid:19263234
      1. Karunakaran E.,
      2. Mukherjee J.,
      3. Ramalingam B.,
      4. Biggs C. A.
      (2011). “Biofilmology”: a multidisciplinary review of the study of microbial biofilms. Appl Microbiol Biotechnol 90, 1869–1881. doi:10.1007/s00253-011-3293-4pmid:21538113
      1. Kasari V.,
      2. Kurg K.,
      3. Margus T.,
      4. Tenson T.,
      5. Kaldalu N.
      (2010). The Escherichia coli mqsR and ygiT genes encode a new toxin-antitoxin pair. J Bacteriol 192, 2908–2919. doi:10.1128/JB.01266-09pmid:20233923
      1. Kim Y.,
      2. Wang X.,
      3. Ma Q.,
      4. Zhang X. S.,
      5. Wood T. K.
      (2009). Toxin-antitoxin systems in Escherichia coli influence biofilm formation through YjgK (TabA) and fimbriae. J Bacteriol 191, 1258–1267. doi:10.1128/JB.01465-08pmid:19060153
      1. Kolodkin-Gal I.,
      2. Verdiger R.,
      3. Shlosberg-Fedida A.,
      4. 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, e6785. doi:10.1371/journal.pone.0006785pmid:19707553
      1. Korch S. B.,
      2. Hill T. M.
      (2006). Ectopic overexpression of wild-type and mutant hipA genes in Escherichia coli: effects on macromolecular synthesis and persister formation. J Bacteriol 188, 3826–3836. doi:10.1128/JB.01740-05pmid:16707675
      1. Korch S. B.,
      2. Henderson T. A.,
      3. Hill T. M.
      (2003). Characterization of the hipA7 allele of Escherichia coli and evidence that high persistence is governed by (p)ppGpp synthesis. Mol Microbiol 50, 1199–1213. doi:10.1046/j.1365-2958.2003.03779.xpmid:14622409
      1. Lewis K.
      (2007). Persister cells, dormancy and infectious disease. Nat Rev Microbiol 5, 48–56. doi:10.1038/nrmicro1557pmid:17143318
      1. Lewis K.
      (2008). Multidrug tolerance of biofilms and persister cells. Curr Top Microbiol Immunol 322, 107–131. doi:10.1007/978-3-540-75418-3_6pmid:18453274
      1. Li M.,
      2. Wang J.,
      3. Geng Y.,
      4. Li Y.,
      5. Wang Q.,
      6. Liang Q.,
      7. Qi Q.
      (2012). A strategy of gene overexpression based on tandem repetitive promoters in Escherichia coli. Microb Cell Fact 11, 19. doi:10.1186/1475-2859-11-19pmid:22305426
      1. Magnuson R. D.
      (2007). Hypothetical functions of toxin-antitoxin systems. J Bacteriol 189, 6089–6092. doi:10.1128/JB.00958-07pmid:17616596
      1. Moyed H. S.,
      2. Broderick S. H.
      (1986). Molecular cloning and expression of hipA, a gene of Escherichia coli K-12 that affects frequency of persistence after inhibition of murein synthesis. J Bacteriol 166, 399–403.pmid:3516974
      1. Reisner A.,
      2. Krogfelt K. A.,
      3. Klein B. M.,
      4. Zechner E. L.,
      5. Molin S.
      (2006). In vitro biofilm formation of commensal and pathogenic Escherichia coli strains: impact of environmental and genetic factors. J Bacteriol 188, 3572–3581. doi:10.1128/JB.188.10.3572-3581.2006pmid:16672611
      1. Ren D.,
      2. Bedzyk L. A.,
      3. Thomas S. M.,
      4. Ye R. W.,
      5. Wood T. K.
      (2004). Gene expression in Escherichia coli biofilms. Appl Microbiol Biotechnol 64, 515–524. doi:10.1007/s00253-003-1517-ypmid:14727089
      1. Rice K. C.,
      2. Mann E. E.,
      3. Endres J. L.,
      4. Weiss E. C.,
      5. Cassat J. E.,
      6. Smeltzer M. S.,
      7. Bayles K. W.
      (2007). The cidA murein hydrolase regulator contributes to DNA release and biofilm development in Staphylococcus aureus. Proc Natl Acad Sci U S A 104, 8113–8118. doi:10.1073/pnas.0610226104pmid:17452642
      1. Rice S. A.,
      2. Tan C. H.,
      3. Mikkelsen P. J.,
      4. Kung V.,
      5. Woo J.,
      6. Tay M.,
      7. Hauser A.,
      8. McDougald D.,
      9. Webb J. S.,
      10. Kjelleberg S.
      (2009). The biofilm life cycle and virulence of Pseudomonas aeruginosa are dependent on a filamentous prophage. ISME J 3, 271–282. doi:10.1038/ismej.2008.109pmid:19005496
      1. Schumacher M. A.,
      2. Piro K. M.,
      3. Xu W.,
      4. Hansen S.,
      5. Lewis K.,
      6. Brennan R. G.
      (2009). Molecular mechanisms of HipA-mediated multidrug tolerance and its neutralization by HipB. Science 323, 396–401. doi:10.1126/science.1163806pmid:19150849
      1. Spoering A. L.,
      2. Gilmore M. S.
      (2006). Quorum sensing and DNA release in bacterial biofilms. Curr Opin Microbiol 9, 133–137. doi:10.1016/j.mib.2006.02.004pmid:16529982
      1. Steinmoen H.,
      2. Knutsen E.,
      3. Håvarstein L. S.
      (2002). Induction of natural competence in Streptococcus pneumoniae triggers lysis and DNA release from a subfraction of the cell population. Proc Natl Acad Sci U S A 99, 7681–7686. doi:10.1073/pnas.112464599pmid:12032343
      1. Tan Q. A.,
      2. Awano N.,
      3. Inouye M.
      (2011). YeeV is an Escherichia coli toxin that inhibits cell division by targeting the cytoskeleton proteins, FtsZ and MreB. Mol Microbiol 79, 109–118. doi:10.1111/j.1365-2958.2010.07433.xpmid:21166897
      1. Wang X. X.,
      2. Kim Y.,
      3. Wood T. K.
      (2009). Control and benefits of CP4-57 prophage excision in Escherichia coli biofilms. ISME J 3, 1164–1179. doi:10.1038/ismej.2009.59pmid:19458652
      1. Webb J. S.,
      2. Thompson L. S.,
      3. James S.,
      4. Charlton T.,
      5. Tolker-Nielsen T.,
      6. Koch B.,
      7. Givskov M.,
      8. Kjelleberg S.
      (2003). Cell death in Pseudomonas aeruginosa biofilm development. J Bacteriol 185, 4585–4592. doi:10.1128/JB.185.15.4585-4592.2003pmid:12867469
      1. West S. A.,
      2. Diggle S. P.,
      3. Buckling A.,
      4. Gardner A.,
      5. Griffin A. S.
      (2007). The social lives of microbes. Annu Rev Ecol Evol Syst 38, 53–77. doi:10.1146/annurev.ecolsys.38.091206.095740
      1. Wu J.,
      2. Xi C.
      (2009). Evaluation of different methods for extracting extracellular DNA from the biofilm matrix. Appl Environ Microbiol 75, 5390–5395. doi:10.1128/AEM.00400-09pmid:19561191
      1. Zegans M. E.,
      2. Wagner J. C.,
      3. Cady K. C.,
      4. Murphy D. M.,
      5. Hammond J. H.,
      6. O’Toole G. A.
      (2009). Interaction between bacteriophage DMS3 and host CRISPR region inhibits group behaviors of Pseudomonas aeruginosa. J Bacteriol 191, 210–219. doi:10.1128/JB.00797-08pmid:18952788