Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Jan 25.
Published in final edited form as: Curr Biol. 2016 Jan 14;26(2):195–206. doi: 10.1016/j.cub.2015.11.056

Auto Poisoning of the Respiratory Chain by a Quorum Sensing Regulated Molecule Favors Biofilm Formation and Antibiotic Tolerance

Ronen Hazan a,b,c,d,#, Yok Ai Que a,b,c,, Damien Maura a,b,c, Benjamin Strobel a,b, Paul Anthony Majcherczyk e, Laura Rose Hopper a,b, David J Wilbur f, Teri N Hreha g, Blanca Barquera g, Laurence G Rahme a,b,c
PMCID: PMC4729643  NIHMSID: NIHMS744309  PMID: 26776731

Summary

Bacterial programmed cell death and quorum sensing are direct examples of prokaryote group behaviors, wherein cells coordinate their actions to function cooperatively like one organism for the benefit of the whole culture. We demonstrate here that 2-n-heptyl-4-hydroxyquinoline-N-oxide (HQNO), a Pseudomonas aeruginosa quorum sensing -regulated low-molecular-weight excreted molecule, and triggers autolysis by self-perturbing the electron transfer reactions of the cytochrome bc1 complex. HQNO induces specific self-poisoning by disrupting the flow of electrons through the respiratory chain at the cytochrome bc1 complex, causing a leak of reducing equivalents to O2 whereby electrons that would normally be passed to cytochrome c are donated directly to O2. The subsequent mass production of reactive oxygen species (ROS) reduces membrane potential and disrupts membrane integrity, causing bacterial cell autolysis and DNA release. DNA subsequently promotes biofilm formation and increases antibiotic tolerance to beta-lactams, suggesting that HQNO-dependent cell autolysis is advantageous to the bacterial populations. These data both identify a new programmed cell death system, and a novel role for HQNO as a critical-inducer of biofilm formation and antibiotic tolerance. This newly identified pathway suggests intriguing mechanistic similarities with the initial mitochondrial-mediated steps of eukaryotic apoptosis.

Introduction

Programmed-cell-death is a common hallmark of multicellular organisms and encompasses intracellular pathways that control self-death. While it was first identified in multicellular eukaryotes, programmed-cell-death systems have subsequently been also observed in unicellular eukaryotes such as yeast [1] and bacteria [2]. Noteworthy cases of bacterial programmed-cell-death pathways that operate during both planktonic or sessile growth are: (i) toxin-antitoxin modules [3], which can respond to environmental cues [4], and to cell-to-cell communication signals [5]; (ii) cell wall synthesis genes that mediate autolytic activity in certain Gram-positive bacteria in response to environmental signals [6]; and (iii) fratricidal killing that occurs during biofilm formation, induces eDNA release, microcolony development and cell dispersal [7].

Cell lysis is a common phenomenon in bacterial cultures with debatable biological relevance. While such lysis was shown to be advantageous for some processes [7, 8], including the prevention of phage spread [9, 10] it is not commonly thought to impart any selective advantage to bacterial populations [11]. This is because in many cases cell lysis is mediated via selfish elements, such as bacteriophages or secreted lytic components including bacteriocins [12], and bacteriolytic antibiotics [13], or being a byproduct of the assay conditions.

Numerous bacterial paradigmatic group or population level phenotypes, referred also as “multicellular”-like behaviors, have been documented in Pseudomonas aeruginosa (P. aeruginosa) [14], a clinically significant Gram-negative pathogenic bacterium that in many infections defies eradication by antibiotics and forms biofilms [15]. One such well-studied bacterial group behavior is quorum sensing, a process of cell-to-cell communication, mediated by small excreted molecules that may act as infochemicals, enabling cells to coordinate their behavior and act together in a group manner.

P. aeruginosa possesses three major cell-to-cell communication systems, which control almost every aspect of its life cycle. These systems differ in their secreted signals. One of these systems, MvfR (PqsR) [16] is induced by hydroxyl-alkyl-quinolones [17]. The MvfR transcriptional regulator controls the expression of a plethora of P. aeruginosa virulence factors and directs the synthesis of approximately 60 excreted low molecular weight molecules [17, 18] most of which are synthesized through the PqsA-D enzymes [17, 19]. Among these secreted molecules are the hydroxyl-alkyl-quinolones (HAQs); 4-hydroxy-2-heptylquinoline (HHQ) [17] and its derivative 3,4-dihydroxy-2-heptylquinoline (PQS) [20], which are the specific inducers of MvfR [19, 21]. Another HAQ, 2-n-heptyl-4-hydroxyquinoline-N-Oxide (HQNO), has been thought to lack quorum sensing signaling activity and instead to function as a P. aeruginosa virulence factor against host cells and other competitive microorganisms in polymicrobial settings [22]. The MvfR-regulated genes pqsABCD [17] and the LasR-regulated gene pqsL [23] are required for HQNO synthesis [18]. HQNO levels are in tight correlation with the other MvfR regulated QS molecules, whose synthesis is also catalyzed by the PqsABCD enzymes [18]. Yet, unlike other quorum sensing molecules, HQNO it not known to exert its activity via a transcriptional QS regulator. HQNO is a well-known inhibitor of cytochrome bc1 (also known as cytochrome c reductase or complex III) in the respiratory chain of bacteria and mitochondria of eukaryotes [24, 25]. It binds to the Qi site of this complex, interfering with the branch of the Q-cycle in which reducing equivalents are cycled back to the Q-pool, which results in occupancy of a semiquinone intermediate at the Qo, electron accepting site. With the Qi site blocked, the enzyme is only able to accept electrons one at a time. The semiquinone radical at the Qo site can react with O2, producing superoxide, a reactive oxygen species (ROS). In mitochondria, the production of ROS brings about a membrane potential shift and opening of the mitochondrial permeability transition (MPT) pores [26], which then leads to mitochondrial dysfunction and secretion of their contents, actions that ultimately result in cellular apoptosis [26, 27].

P. aeruginosa cells naturally undergo autolysis during their planktonic growth. That this behavior occurs in most of its isolates since its discovery [28–30], suggests that it could provide a selective advantage to P. aeruginosa. Nevertheless, the exact mechanism and its potential biological significance remain almost unexplored. It has generally been assumed that induction of lysis in P. aeruginosa involves the activation of autonomous selfish elements, such as prophages or pyocins [29, 30], and may also be related to cell-to-cell communication [30, 31], despite the fact that no such elements or cell autolysis signaling molecules have been identified. Here we show that HQNO functions as a self-poisoning molecule, in P. aeruginosa, mediating a programmed cell autolysis that benefits biofilm formation and antibiotic tolerance. The P. aeruginosa cell autolysis reported here is similar in many ways to the programmed-cell-death occurring in higher organisms.

Results

Pseudomonas aeruginosa autolysis is a programmed cell death process induced by the quorum sensing regulated molecule HQNO

After growing for 48 hours, planktonic cultures of P. aeruginosa began to exhibit cell lysis, reflected by a decrease in OD600nm (Figures 1A and S1A, B–C) and also in the number of colony forming units (CFUs) (Figures 2A and S1D). This observation is not limited to a specific strain (Figure S1A–B) or a specific growth condition as it is observed both in 96 wells plates (Figure 1A–B) and in tubes, and in both PA14 and PA01 strains (Figure S1C–D). In contrast, in the mvfR− mutant and in several mutants, pqsA−, pqsBC−, pqsD− and pqsL−, deficient in quorum sensing and synthesis of HQNO [18], cells did not lyse, and the OD600nm of the cultures remained essentially constant during stationary phase. Moreover, PA14 cell autolysis was abolished by the anthranilic acid analog, 6-CABA (2-amino-6-chlorobenzoic acid) (Figure 1C), an MvfR regulon inhibitor [32] and the MvfR inhibitor M64 [33], indicating that the MvfR regulon controls P. aeruginosa autolysis. In contrast, cultures of PA14 isogenic mutants pqsE− and pqsH− lysed similarly to parental PA14 (Figure 1D). pqsE− produces wild-type (WT) hydroxyl-alkyl-quinolones levels and pqsH− produces all hydroxyl-alkyl-quinolones except PQS [18]. This correlation between autolysis and hydroxyl-alkyl-quinolones production indicates P. aeruginosa autolysis is mediated via an MvfR-regulated molecule synthesized by the PqsABCD enzymatic pathway. Furthermore, that an MvfR regulated small molecule controls this autolysis suggests that this behavior is coordinated by cell density; quorum sensing. We next asked if the autolysis-deficient phenotype of hydroxyl-alkyl-quinolones -deficient pqsA− cells can be rescued by adding the most abundant MvfR-regulated, PqsA synthesized quorum sensing products; HHQ, PQS, or HQNO [18] at the onset of stationary growth. Figure 1E shows that only HQNO restored autolysis in a dose-dependent manner (Figure 1F) and at a level corresponding to its physiological concentration in stationary PA14 strain cultures (Figure 1G). Additionally, the pqsL− mutant, which produces HHQ and high levels of PQS [30], but not HQNO [18], only lysed upon HQNO addition (Figure 1H).

Figure 1. P. aeruginosa autolysis in liquid cultures is an MvfR-dependent process controlled by the quorum sensing regulated molecule HQNO.

Figure 1

Open in a new tab

(A–F, H) Culture growth curves (OD600nm). Absorbance was recorded every 15 min for 48 h. Curves are color-coded according to either strains or compound concentrations, and match the related legends. Results are representative of at least triplicate experiments. (A) In contrast to PA14 (positive control), mvfR−, pqs operon mutants pqsA−, pqsBC− and pqsD−, as well a pqsL− cells do not lyse. (B) Image of 48 h P. aeruginosa cultures. In contrast to PA14 WT cultures, pqsA− cultures do not lyse unless HQNO is added. (C) Pharmacologic inhibition of MvfR and pqs operon activity. The anthranilic acid analog 6-CABA [32] and the recently developed MvfR inhibitor, M64 [33], abolishes PA14 autolysis. (D) The pqsE− and pqsH− mutants lyse similar to PA14, which together with (B–C), suggest the involvement of an mvfR− regulated small molecule in the autolysis process. (E) Unlike HHQ or PQS, the exogenous addition of HQNO to pqsA− cultures restores autolysis. HQNO restores autolysis in pqsA− (F) and pqsL− (H) cultures in a dose-dependent manner, with maximum autolysis occurring at a concentration similar to that typically observed in PA14 cultures. (G) Typical concentrations of the major MvfR-regulated secreted small molecules measured in the supernatants of PA14 cultures at specified time points. Autolysis is conserved. It occurs in many P. aeruginosa strains, including isolates (Figure S1A–B) and in various growth conditions (Figure S1C–D).

Figure 2. The HQNO-mediated decrease in OD600nm absorbance is due to cell death.

Figure 2

Open in a new tab

(A) The colony forming units of 24 h and 48 h cultures reveal massive loss of cell viability in PA14 and HQNO-treated pqsA− cultures, compared to pqsA− cultures (p<0.05, Unpaired T test with Bonferroni correction). (B) The relative levels of eDNA were measured by qPCR of the P. aeruginosa genes rpoD and ldhD in cell-free supernatants at 17 h, 24 h, and 31 h of PA14 (black), pqsA− (red), and HQNO-treated pqsA− (green) cultures. The results were calibrated for each condition to levels measured at 17 h.

To verify that the decreased OD600nm corresponds to autolysis, we compared the number of viable cells in 24 and 48 h parental and pqsA− cultures, with and without HQNO. Approximately 90% and 10% of the cells died between 24 and 48 h for PA14 and pqsA− cultures, respectively, unless HQNO was exogenously added to pqsA− cultures (Figure 2A). Extracellular DNA (eDNA), which is released into the culture medium during autolysis [34], is another measure of cell lysis. As such, we used quantitative PCR of the rpoD and ldhD housekeeping genes to measure PA eDNA in cell-free supernatants of WT versus pqsA− cultures, with and without HQNO. The pqsA− culture exhibited significantly lower eDNA levels at 48 h (~ 6 log2 fold reduction; p < 0.01) versus the parental, and pqsA− plus HQNO, cultures (Figure 2B).

HQNO-induced programmed cell death is mediated by specific interaction with the Qi site of Complex III and depends on an intact Qo site

HQNO is well known as an inhibitor of the mitochondrium cytochrome bc1 complex by binding to the Qi, quinol binding site (Figure 3A), leading to swelling and rupture of the organelle and eventually cellular apoptosis [25]. Such effects prompt us to examine whether HQNO also causes programmed-cell-death in P. aeruginosa via the respiratory chain when molecular oxygen is the final electron acceptor in the presence of nitrate (Figure 3B). Figure 3C shows that pqsA− cultures supplemented with antimycin A, which also binds at the Qi, it induced pqsA− cell autolysis, indicating that P. aeruginosa programmed-cell-death can occur via interference with the electron transfer functions of cytochrome bc1 complex. To further confirm this, we assessed autolysis in mutants of P. aeruginosa in which the following components of the cytochrome bc1 complex were genetically deleted or disabled: the cytochrome b subunit (cytb1, PA14_57560) and the Rieske subunit (PA14_57570) (Figure 3A). Figure 3D demonstrates that both mutants grew less well overall, did not autolyse, and were insensitive to added HQNO, as compared to WT cells. Moreover, in contrast to the Qi site inhibitors HQNO (Figure 1), and antimycin A (Figure 3C), the known Qo site inhibitor myxothiazol, prevented WT cell autolysis (Figure 3E) and lysis of pqsA− cells induced by externally added HQNO (Figure 3F). These results reveal that HQNO-mediated programmed-cell-death in P. aeruginosa depends on both specific inhibition of the Qi site, and an intact functional and uninhibited Qo site.

Figure 3. HQNO-dependent autolysis is due to inhibition of the Qi site of cytochrome bc1 complex.

Figure 3

Open in a new tab

(A) A rendering of the structure of the bc1 complex, (respiratory chain complex III) together with its major active sites, relevant inhibitors (upper panel, adapted with permission from http://www.life.illinois.edu/crofts/bc-complex_site/), and structure of its operon in the PA14 strain (bottom panel). (B) Schematic representation of the aerobic respiratory chain of Pseudomonas aeruginosa. The role of the bc1 complex in ROS production is highlighted and its activity in various growth stages is shown in Table S1. (C, D, E) Growth curves are plotted as in Figure 1. (C) pqsA− cells grown in LM media were treated with various concentrations of the bc1 complex Qi site inhibitor antimycin A (blue), using PA14 (black), untreated pqsA cells (red), and HQNO-treated pqsA− cells (green) as controls. (D) Mutants in the cytochrome b and the Rieske genes of bc1 complex operon did not lyse and were insensitive to HQNO treatment. (E–F) The cytochrome bc1 complex, cytochrome Qo site inhibitor, myxothiazol, prevented autolysis in both PA14 (E), and in HQNO-treated pqsA− (F), cells. PA14, pqsA− cultures with and without HQNO served as positive and negative controls, respectively.

To confirm the role of cytochrome bc1 complex in programmed-cell-death we verified that this enzyme was present and active in the cells at the appropriate times. To this end, we directly assessed the activity of the cytochrome bc1 complex by measuring the rate of electron transfer from quinone to cytochrome c, catalyzed by cell membranes. As depicted in Figure 3B, the entire electron flow from quinone to oxygen must pass through both the cytochrome bc1 complex and cytochrome c. In the presence of potassium cyanide, which blocks the cytochrome c oxidases (cytochromes cbb3 and aa3) and since there is no pathway from cytochrome c to oxygen and cytochrome c will remain reduced [35]. Thus, the activity of the cytochrome bc1 complex in membranes of P. aeruginosa can be measured by adding ubiquinol and ferri- cytochrome c, in the presence of cyanide, and measuring the reduction of the cytochrome c spectrophotometrically at 550 nm [36]. PA14, pqsA−, and pqsA− cells that received HQNO exogenously were harvested at the late exponential phase (16 hours), mid-stationary phase (24 hours) and cell death phase (30 hours) and the cytochrome bc1 complex activity in the cell membranes was measured as a function of total membrane protein. Table S1, shows significant cytochrome bc1 complex activity at each time point, confirming the presence of a sufficient active complex throughout the growth cycle to account for the formation of ROS during respiratory activity in these cells. That pqsA− mutant cells also exhibit the presence of an active cytochrome bc1 complex throughout the growth cycle, suggests the expression of this enzyme is not linked to the synthesis of HQNO (Table S1).

The ROS burst mediated by HQNO inhibition of bc1 complex Qi site promotes cell autolysis

Inhibition of the mitochondrial cytochrome bc1 complex Qi site induces a burst of reactive oxygen species (ROS), in particular, superoxides [37], which are potent cell death inducers [38]. We quantified ROS production in P. aeruginosa cells grown pre- (24 h), during- (31 h), and post-autolysis (48 h and 60 h) (Figure 4A), using electron paramagnetic resonance (EPR) spin trapping to quantify the decay of the EPR signal of the stable radical 2,2,6,6-tetramethyl piperidine-N-oxyl (TEMPO) (Figure S2A). In contrast to pqsA− cells, both WT and HQNO-treated pqsA− cultures exhibited a large burst of ROS during autolysis (Figure 4A). These results were validated by measuring emission changes of the ROS sensitive dye H2DCFDA (Figure S2B). To further confirm that increased superoxides during stationary phase stimulated autolysis, we added the superoxide-inducing agent N,N′-dimethyl-4,4′-bipyridinium dichloride (paraquat) to a pqsA− culture. Paraquat-induced ROS restored autolysis in a dose-dependent manner (Figure 4B). Furthermore, ROS scavenging agents prevented autolysis: addition of either glutathione (Figures 4C, 4E), or L-cysteine (Figures 4D, 4F), to PA14 (Figures 4C–D), or to HQNO-plus pqsA− cultures (Figures 4E–F), decreased the autolysis rate in a dose-dependent manner. Thus, as in mitochondria [37], HQNO inhibition of P. aeruginosa cytochrome bc1 complex Qi site generates a ROS burst, which is prevented by the addition of exogenous ROS scavenging agents. Interestingly, natural intracellular glutathione levels were sharply reduced at stationary phase (Figure 4G). As such, the timing of autolysis at the onset of stationary growth is likely the outcome of a fine-tuned and time-sensitive interplay between the accumulation of antioxidants versus ROS-inducing agents, like HQNO, in concert with the antagonism of positive and negative signals regulating MvfR quorum sensing [17].

Figure 4. HQNO-induced autolysis requires the generation of Reactive Oxygen Species (ROS).

Figure 4

Open in a new tab

(A) Relative amounts of ROS detected from the decay of the EPR signal of the stable radical TEMPO normalized to bacterial density (Figure S1A) in pre-lysis (24 h), during (31 h), and post-lysis (48 h and 60 h) from cultures of PA14 (black), pqsA− (red), or HQNO-treated pqsA− (green). The differences between pqsA− and the other two samples in 24 h, 31h, 48 h and 60 h are statistically significant (p<0.05, Unpaired T test with Bonferroni correction). These results were validated by H2DCFDA fluorescence dye (Figure S2B). (B–F) Growth of PA14 (C, D), pqsA−. (B) or HQNO-treated pqsA− (E, F) cultures followed a similar pattern, as shown in Figure 1. Prior to the start of autolysis, either oxidative agent paraquat, (B) the antioxidant glutathione (C, E), or L-cysteine (D, F) was added (arrow), and the cultures were further incubated for up to 48 h. The colors of the curves match the font of the concentrations (mM) of the added compounds. (G) The levels of cellular glutathione are reduced at the onset of the stationary phase in PA14 (black), pqsA− (red), and HQNO-treated pqsA− (green) cultures. The connected lines show growth and the dashed lines represent relative glutathione levels.

Pyocins and PF5 phage are not involved in HQNO-induced autolysis

P. aeruginosa autolysis has previously been proposed to result from prophage induction [30], as seen for autolysis in other bacteria. The PA14 genome harbors phage-like elements such as the PF5 prophage, and the R and F type pyocins clusters that encode for P2 and lambda phage-like particles which are produced upon oxidative stress [39]. However, figure S3 shows that these putative selfish genetic elements are not required for the HQNO-induced autolysis since mutant with a knockout of all 38 genes of the R-F pyocin cluster (Figure S3A), or insertions mutants in single R-F and S-type pyocins (Figure S3 A–B) or PF1 lysogen genes (Figure S3C), did not exhibit impaired lysis.

A high-throughput genetic screen identifies autolysis related genes

We performed a high-throughput genetic screen to identify downstream elements leading to bacterial cell death. Using the PA14 non-redundant mutant library [40], we recorded the OD600nm of 5658 mutants daily over a 7 day period. Mutants with impaired autolysis were collected, compared to pqsA− and PA14 cultures, and tested for HQNO sensitivity to autolysis. A total of 41 mutants defective in autolysis were identified (Table S2 and Figure S4). The autolysis phenotype was restored by HQNO in 21 of these mutants. Among these strains, we isolated pqsA-D− and mvfR−, as well as lasR− mutations. Isolation of the lasR− mutant is in agreement with previously published results that a mutation in lasR− allowed escape from lysis [31], likely because LasR positively controls pqsL [23]. The pqsL− mutant was not identified in the screen because the relevant mutant is absent from the mutant library [40]. Using a Search Tool for the Retrieval of Interacting Genes/Protein (STRING, http://string-db.org/), we discovered that three of the isolated mutants were linked to tryptophan metabolism which is a precursor of anthranilic acid, the precursor of HQNO [17, 18], including PA14_09630 (a putative acyl-CoA dehydrogenase), PA14_10320 (a putative transcriptional regulator), and PA14_30750 (a putative tryptophan oxygenase). In addition, a mutation in the gene ogt (PA14_51440) attenuated autolysis, which was restored by HQNO. ogt is located adjacent to pqsA− and annotated as a methylated-DNA-protein-cysteine-methyltransferase, which participates in DNA repair. This gene had not been linked to HQNO previously. Five more genes (PA14_60780, PA14_65420, PA14_44440, PA14_50200, and PA14_10160) are annotated as components of putative pumps or transporters, although their respective involvements in the excretion/uptake of HQNO are unknown. Finally, five additional autolysis impaired strains carried gene mutations with no obvious link to HQNO production or action.

HQNO failed to restore the autolysis phenotype to 20 of the isolated mutants, suggesting that their products are involved in autolysis either downstream of HQNO production, or function in a parallel pathway. First, as expected, were the respiratory chain components described above, including cytb1 (PA14_57560) and PA14_57570 (the Rieske subunit) (Figure 3A). We identified five additional genes related to the respiratory chain, including roxR (PA14_58300) and roxS (PA14_58320) (Figure S4), which encode a two component regulatory system controlling the expression of a cyanide insensitive oxidase known to play a role in P. aeruginosa interactions with airway epithelial cells [41]. We also identified a gene encoding polyferredoxin (PA14_44420), which is under the control of RoxRS [42], and a cytochrome c biogenesis factor (cycH, PA14_45280). These screen results also suggest that rpoN has a role in the autolysis of P. aeruginosa (Table S2). The rpoN gene (PA14_57940) encodes the alternative sigma factor σ54, a global regulator that plays a role in adaptation and survival of P. aeruginosa and other bacteria [43].

No obvious link to the autolysis process for the remaining 14 genes identified in the screen was found. In conclusion, this screen provides supportive evidence that P. aeruginosa spontaneous autolysis is an MvfR-dependent process that involves HQNO-mediated inhibition of the cytochrome bc1 complex.

HQNO mediated bacterial membrane disruption leads to cell autolysis

Our screen did not identify any lytic genes (e.g. bacteriophages or holin homologs) or cell wall synthesis-associated genes (e.g. cidA or lrgA) known to be involved in bacterial lysis [6] that could serve as effectors to mediate autolysis. Since inhibition of the Qi site of the cytochrome bc1 complex in mitochondria induces the opening of MPT to effect the disintegration of the membrane and secretion of mitochondrial contents [26], we asked if P. aeruginosa autolysis is due to a similar mechanism. As such, we assessed bacterial membrane potential and permeability during P. aeruginosa programmed-cell-death. Figure 5 provides evidence that HQNO similarly mediates membrane disruption to cause autolysis. For this experiment we used the fluorescent membrane potential sensitive cyanine indicator, DiOC2(3), which is typically used to determine mitochondrial and bacterial [44] membrane potential. Using flow cytometry, we showed that the DiOC2(3)red/green fluorescence ratio was similar in PA14 cultures and pqsA− cultures, with or without HQNO, at 24 h pre-autolysis (Figure 5A). At 72h post-autolysis the red/green ratio remained high in non-lysed pqsA− cells, indicating the majority maintained high membrane potential. In contrast, this ratio in the PA14 and HQNO-plus pqsA− autolysing cells was greatly decreased, demonstrating membrane potential disruption (Figure 5A). Altered potentials were also clearly observed via the red/green profiles over time (Figures 5B–C). We further assessed membrane permeability with propidium iodide, a red dye that only penetrates cells with disrupted membranes, and is commonly used for bacterial live/dead staining (Figure 5D). Figure 5E shows that the quantification of intact PA14 and HQNO-treated pqsA− cells were both significantly lower versus that of untreated pqsA− cells (P < 0.01).

Figure 5. HQNO-induced autolysis is associated with membrane damage.

Figure 5

Open in a new tab

(A–B) Membrane potential is altered during autolysis. Cells were stained with DiOC2(3) at various time points. (A) The ratio between red and green fluorescence was calculated using population mean intensities. (B) Analysis using red fluorescence parameters reveals a shift of PA14 and HQNO-treated pqsA− cells, but not of untreated pqsA− cells from the non-lysing (NL) to the lysing (L) gates, with relative fluorescence intensities of 38–107 and 7–28, respectively. The populations at each time point are overlayed. Flow cytometric data were collected with log amplification. (C) The ratio between cells counted in the NL compared to the L. (D) Live/dead staining using propidium iodide dye at 24 h, 31 h, and 48 h, for PA14, pqsA, and HQNO-treated pqsA cultures. Bacteria with intact membranes stain green, while bacteria with damaged membranes stain red. (E) The percentage of green cells per microscopic field. In all sub figures, black denotes WT, red is pqsA mutant and green is pqsA mutant treated with HQNO. (C and E) The differences between pqsA− and the other two samples in 48 h and 72 h time points are statistically significant (p<0.05, Unpaired T test with Bonferroni correction).

HQNO mediated autolysis promotes biofilm formation via eDNA release and increases antibiotic tolerance

There is no clear advantage for a unicellular organism to commit suicide, unless this benefits the long-term success of the cell population within the context of the group behavior of bacteria [45]. As such, what might be the selective advantage for P. aeruginosa to specifically induce autolysis? We hypothesized that one such advantage could be the promotion of biofilm formation because: i. Biofilm formation in P. aeruginosa is regulated by quorum sensing [46]; ii. eDNA is secreted during cell lysis (Figure 2B) and was shown to contribute to biofilm growth [46]; iii. biofilm formation of P. aeruginosa is accompanied by eDNA secretion [8]; and most significantly, iv. pqsA− cells support less biofilm compared to WT cells [47] and Figure 6A]. Thus, we tested if HQNO-mediated lysis promotes biofilm formation. Cells were grown in polystyrene 96-well plates and biofilm was allowed to form for 48 hours in the presence or absence of HQNO and stained using crystal violet. Indeed, HQNO addition restored pqsA− deficient biofilm formation (Figure 6A). However, it could be argued that this biofilm enhancement was due to increased amounts of cellular debris resulting from lysis rather than increased CFU, thus we enumerated the viable cells within the biofilm. Figure 6B demonstrates that the number of viable cells within the biofilm is higher in both parental and pqsA− mutant plus HQNO compared to pqsA−. To determine whether the release of eDNA during cell autolysis could have a structural role to the biofilm build-up, we added DNAse to the biofilm medium. As expected, DNAse addition lead to reduced biofilm in both PA14 and pqsA− mutant cells with added HQNO (Figure 6B), indicating that DNA release during HQNO-promoted autolysis contributes to the HQNO-mediated biofilm formation. One of the key characteristics of biofilms is their inherent tolerance to antibiotics. We therefore asked whether the HQNO-mediated biofilm increase is associated with a higher tolerance to antibiotics. Figure 6C shows that pqsA− biofilm becomes more tolerant to the beta-lactam antibiotic meropenem in the presence of HQNO and that DNAse addition reverses this tolerance (Fig. 6D). These data strongly suggest that HQNO-mediated enhancement of biofilm formation, and its associated antibiotic tolerance, provides a potential selective advantage.

Figure 6. P. aeruginosa autolysis can be beneficial to the cell population by promoting biofilm formation and increasing antibiotic tolerance via eDNA release.

Figure 6

Open in a new tab

(A, B) HQNO promotes biofilm formation via the release of eDNA. (A) Biofilm biomass quantified by crystal violet staining. pqsA cells make less biofilm than PA14 (p<0.001) and HQNO addition to pqsA rescues biofilm (p<0.001, Unpaired T test with Bonferroni correction). (B) Effect of DNAse I on the biofilm viable cell concentration assessed by CFU measurements. pqsA cells make less biofilm than PA14 (p<0.05) and HQNO addition to pqsA rescues biofilm (p<0.05, Unpaired T test with Bonferroni correction). The addition of DNAse in PA14 or pqsA + HQNO reduces biofilm (p<0.015 and p<0.01 respectively) whereas addition to pqsA does not (p<0.01, Unpaired T test with Bonferroni correction). (C, D) HQNO is promoting antibiotic tolerance in biofilm via the release of eDNA. (C) Survival fraction of biofilm cells exposed to various concentrations of the antibiotic meropenem. The survival fraction represents the ratio between biofilm viable cells after and before antibiotic treatment. pqsA biofilm is more sensitive than PA14 to 1, 10 or 100 μg/mL Meropenem (p<0.015, p<0.001 and p<0.005 respectively) and HQNO addition reduces pqsA sensitivity (p<0.05, p<0.01 or P<0.5 respectively, Unpaired T test with Bonferroni correction). (D) Effect of DNAse I on the survival fraction of biofilm cells exposed to meropenem (100 μg/mL). The addition of DNAse in PA14 or pqsA + HQNO increases biofilm sensitivity to meropenem (p<0.01 and p<0.6 respectively) whereas addition to pqsA does not (p<0.005, Unpaired T test with Bonferroni correction). Error bars show mean +/− SEM of at least 3 replicates. No differences in MIC (0.25 mg/L) were found between parental and isogenic pqsA− mutant.

Discussion

This study describes a novel quorum sensing-regulated bacterial mechanism that controls a type of programmed-cell-death autolysis in P. aeruginosa that provides a fitness benefit to the collective. We provide evidence that this “programmed-cell-death” pathway is induced by specific self-poisoning of the respiratory chain via the MvfR- quorum sensing -regulated small molecule, HQNO. HQNO disrupts the flow of electrons through the respiratory chain at the cytochrome bc1 complex, leading to the formation of ROS, membrane damage, and ultimately cell death by autolysis (Figure 7A). We show that this autolysis is beneficial, as it contributes to biofilm formation and increases antibiotic tolerance (Figure 6). These data both identify a new programmed-cell-death system, and a novel role of HQNO as its inducer and provide direct evidence that this process provides a fitness benefit to the collective.

Figure 7. Similarities between HQNO-mediated cell autolysis and mitochondrial MPT pore formation.

Figure 7

Open in a new tab

(A) Model for HQNO-mediated P. aeruginosa autolysis. HQNO production depends on the PqsABCD enzymes controlled by the hydroxyquinolone quorum sensing transcription regulator MvfR and on PqsL controlled by the homoserine-lactone quorum sensing regulator LasR (1). Secreted HQNO diffuses into the environment and signals neighboring cells to inhibit the Qi site of cytochrome bc1 complex (2). This inhibition induces a ROS burst, which is not counteracted by glutathione, since reduction of glutathione gene expression occurs at the onset of stationary growth (3). ROS-induced membrane damage results in cell autolysis and death (4). As a consequence, nutrients and eDNA are released into the medium to promote planktonic growth/survival of remaining cells (5), as well as biofilm formation and antibiotic tolerance (6). (B–C) The MPT inhibitor cyclosporine A [50] on P. aeruginosa autolysis. PA14 (Figure 7B) and HQNO-treated pqsA− cells (Figure 7C) were grown to early stationary phase, and treated with cyclosporine A. cyclosporine A reduced P. aeruginosa autolysis in a dose-dependent manner in both PA14 and pqsA−-HQNO treated cells.

There are several reasons why the P. aeruginosa HQNO-dependent autolysis we describe here could be a notable example of an altruistic and advantageous “programmed-cell-death” mechanism that provides a fitness benefit to the collective: (i) it is induced by bacterial components per se and not by selfish elements (Figure S3), as previously assumed that P. aeruginosa lysis is a result of lysogenic phage induction [30]; (ii) it is well conserved among P. aeruginosa isolates (Figure S1) [28, 29], suggesting that it provides an evolutionary advantage to bacterial populations; (iii) unlike bacterial metabolic waste that may accumulate and kill the bacterial cell, (i.e acetic acid), HQNO belongs to the well-regulated hydroxyquinolone family of signaling molecules known to promote virulence [18], suggesting that it is unlikely that this molecule is produced inadvertently to poison P. aeruginosa; (iv) it is produced in bacterial cultures and in vivo in infected animal and human tissues [48], indicating that it is not a byproduct of the assay conditions; and (iv) it promotes biofilm and antibiotic tolerant/persister cells formation, which allows cells to evade antibiotic-induced killing (Figure 6) and potentially a wide variety of other environmental stresses or microbial anti-Pseudomonas compounds in polymicrobial settings. These findings strongly suggest that the autolytic “programmed-cell-death” of single cells is advantageous to the surviving of the group. Moreover, both biofilm and antibiotic tolerance contribute to bacterial survival and both may have critical clinical implications in chronic, persistent and relapsing infections caused by this pathogen.

HQNO effects on biofilm formation and antibiotic tolerance appear to be due to extracellular DNA release as the presence of DNAse inhibits those two phenotypes (Figure 6). eDNA is known to facilitate biofilm formation and maturation as it reduces repulsive forces between cells and surfaces, promotes acid-base interactions, provides structural stability and guides cell motility inside the biofilm for maturation [34]. Consequently, eDNA may significantly change the biofilm architecture and reduce antibiotic penetration inside biofilms, therefore promoting survival of the biofilm population. As such, our findings provide another example for bacterial coordinated group behavior that during their life cycle, sacrifices subpopulations of cells for the benefit/fitness of the entire population [7]. That autolysis is advantageous for long-term bacterial survival since it favors the generation of biofilm and persister cells raises the question of whether antibiotic treatment using lysis-inducing antibiotics, such as β-lactams, or the suggestive phage therapy [49], could be a double-edged sword, in that they could actually promote chronic infections.

How do the bacterial cells time their autolysis? We have found that at least two processes are involved in this determination: the accumulation of the quorum sensing -regulated HQNO, and the levels of glutathione that drop at the onset of the stationary phase (Figure 4G), (note that this occurs even in the pqsA− mutant, which does not make HQNO, showing that this is not simply the consumption of glutathione by ROS as part of the autolysis pathway itself). Thus, it seems that P. aeruginosa autolysis is tightly regulated and not simply a stochastic event.

Notably, P. aeruginosa seems to lack the conventional programmed-cell-death systems previously observed in other bacteria, and its cell wall degrading enzymes, which share homology to those responsible for programmed-cell-death in other bacteria, have little effect on cell death [8]. However, the mode of action of HQNO reported here reveals a novel programmed-cell-death system that also uses ROS to induce cell death (Figure 4) via membrane depolarization, similar to toxin-antitoxin systems [3].

Although P. aeruginosa lacks the toxin-antitoxin system, differentially-induced autolysis may occur in this pathogen as well, because WT cells naturally autolyse in the planktonic phase (Figure 1), and in biofilms [47], but not on plates [30]. Nevertheless, it has not been determined if the lysis observed on plates is selfish element dependent.

There are intriguing similarities between the P. aeruginosa programmed-cell-death mechanism described here and the first steps of eukaryotic mitochondrial apoptosis [26]. Both processes can be induced through the inhibition of the Qi site of the cytochrome bc1 complex by either HQNO or antimycin A (Figures 1F and 3B), and both are inhibited by the Qo site inhibitor myxothiazol (Figures 3D–E), and cyclosporine A, which is a well-known inhibitor of eukaryotic mPTP pores [50], also inhibited P. aeruginosa autolysis (Figures 7B–C). The cyclosporine A findings strongly suggest that the eukaryotic mPTP exists in P. aeruginosa and that it has structural homology with the mPTP sufficient for the inhibitor to work. However, the genetic screen performed did not reveal the existence of a pore complex. A possible explanation may be that the mPTP-related genes are essential.

HQNO disrupts the cytochrome c Q-cycle mechanism by inhibiting the recycling of reducing equivalents back to the Q-pool. As such, Qi site inhibition leads to the accumulation of ROS and disruption of the enclosing membrane (Figure 4A) [37]. In the case of P. aeruginosa, ROS-induced loss of membrane integrity results in the release of intracellular contents into the medium, while in mitochondria, the contents of the organelle are released into the cytoplasm. [26]. Among the contents released when mitochondria are disrupted is cytochrome c, which initiates the eukaryotic cell apoptotic programmed-cell-death cascade [27]. The bacterial content release as a result of the autolysis promoted by HQNO increases biofilm formation and antibiotic tolerance (Figure 6). It is noteworthy that cytochrome c plays a critical role in generation of ATP, and also that lower levels of proton motive force (pmf) and lower ATP concentrations promote bacterial dormancy and persister cells, corroborating our findings on HQNO relevance on antibiotic tolerance.

Several studies have shown similarities between bacterial programmed-cell-death and the cytoplasmic steps of eukaryotic apoptosis [6]. In contrast, the current study identifies a bacterial “programmed-cell-death” system with similarities to the initial apoptotic events in the mitochondria and provides initial evidence of the potential benefits of the bacterial programmed autolysis observed. Considering the endosymbiotic hypothesis for the bacterial origin of mitochondria, this study raises the question whether the initial steps of the eukaryotic apoptosis pathway originated from bacterial autolysis as it is mechanistically related to the HQNO-mediated pathway of the paradigmatic bacterium P. aeruginosa.

Experimental Procedures

A detailed description of all experimental and statistical methods used is included in the supplemental experimental procedures section.

Bacterial strains growth conditions and compounds

The P. aeruginosa strains used in this manuscript are from our strain collection. The mutant screen was performed using the non-redundant library of PA14 mutants [40].

Cytochrome c reductase activity measurements

Cells were prepared as detailed in Supplemental Experimental Procedures. Cytochrome bc1 complex activity in cell membranes was measured spectrophotometrically by following the reduction of cytochrome c, coupled to the oxidation of reduced ubiquinone. The quinol was prepared as reported previously.

ROS detection by Spin Trapping Measurements Coupled with Electron Paramagnetic Resonance (EPR) Spectroscopy

For EPR in bacterial cultures, samples were treated with the spin trapping agent 2,2,6,6-tetramethyl piperidine-N-oxyl (TEMPO). The decay of TEMPO signal as a result of ROS was calculated as the slope of the linear regression of differences between the high and low signal peaks (Figure S2A).

Microscopy, live/dead staining, and ROS staining

Membrane integrity staining was performed using the LIVE/DEAD BacLight Kit. For ROS detection, H2DCFDA was used.

Membrane potential detection

Membrane potential was assessed using the Baclight Membrane potential kit. Cells were stained with DiOC2 and analyzed by flow cytometry with an excitation wavelength of 488 nm and emission wavelengths of 530 nm (Green) and 675 nm (Red). The gate “non-lysing” (NL, Figure 5B) was set on the pqsA− cells at the early time point and the “lysing” gate (L; Figure 5B) was set on the pqsA− + HQNO cells at the late time point.

Biofilm quantification

To assess biofilm formation, cells were grown for 48 h in M63 minimal media supplemented with 0.2% glucose, 0.5% casaminoacids and 1mM magnesium sulfate. Biofilm biomass was measured by crystal violet staining and cells number was assessed by CFU count.

Hydroxyl-alkyl-quinolones detection

hydroxyl-alkyl-quinolones in bacterial culture supernatants were quantified by LC/MS as described [18].

Glutathione detection

Relative glutathione levels were determined using the GSH/GSSG-Glo Assay Kit (Promega, Madison, WI), with the bacteria lysing step modified as detailed in Supplemental Experimental Procedures.

High-throughput screen of the non-redundant PA14 mutant library is also detailed in Supplemental Experimental Procedures section.

Supplementary Material

supplement

Acknowledgments

We thank Drs. Scott Stachel and Alicia Ballok for the critical reading of this manuscript. This work is dedicated to the memory of Scott Stachel.

Footnotes

Author Contributions

R.H, Y.A.Q, D.M, B.S, P.A, L.R, D.W and T.N.H, conducted the experiments. R.H, Y.A.Q, D.M, B.B and L.G.R designed the experiments and wrote the paper. B.B and L.G.R supplied equipment and material.

Financial disclosure

This work was supported by the research grants to LGR, Shriners #8770, Cystic Fibrosis Foundation, #11P0 and NIAID R33AI105902. RH was supported by a Shriners Hospitals Research Fellowship #8494. YAQ was supported by a Swiss National Science Foundation/Swiss Medical Association (FMH) grant #PASMP3-123226 and a grant from the SICPA Foundation. BB and TH were supported by the research grant NSF- MCB1052234 to BB. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • 1.Buttner S, Eisenberg T, Herker E, Carmona-Gutierrez D, Kroemer G, Madeo F. Why yeast cells can undergo apoptosis: death in times of peace, love, and war. The Journal of cell biology. 2006;175:521–525. doi: 10.1083/jcb.200608098. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Tanouchi Y, Lee AJ, Meredith H, You L. Programmed cell death in bacteria and implications for antibiotic therapy. Trends Microbiol. 2013;21:265–270. doi: 10.1016/j.tim.2013.04.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Hu MX, Zhang X, Li EL, Feng YJ. Recent advancements in toxin and antitoxin systems involved in bacterial programmed cell death. Int J Microbiol. 2010;2010:781430. doi: 10.1155/2010/781430. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Hazan R, Sat B, Engelberg-Kulka H. Escherichia coli mazEF-mediated cell death is triggered by various stressful conditions. J Bacteriol. 2004;186:3663–3669. doi: 10.1128/JB.186.11.3663-3669.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Kolodkin-Gal I, Hazan R, Gaathon A, Carmeli S, Engelberg-Kulka H. A linear pentapeptide is a quorum-sensing factor required for mazEF-mediated cell death in Escherichia coli. Science. 2007;318:652–655. doi: 10.1126/science.1147248. [DOI] [PubMed] [Google Scholar]
  • 6.Rice KC, Bayles KW. Molecular control of bacterial death and lysis. Microbiology and molecular biology reviews: MMBR. 2008;72:85–109. doi: 10.1128/MMBR.00030-07. table of contents. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Thomas VC, Hancock LE. Suicide and fratricide in bacterial biofilms. Int J Artif Organs. 2009;32:537–544. doi: 10.1177/039139880903200902. [DOI] [PubMed] [Google Scholar]
  • 8.Bayles KW. The biological role of death and lysis in biofilm development. Nat Rev Microbiol. 2007;5:721–726. doi: 10.1038/nrmicro1743. [DOI] [PubMed] [Google Scholar]
  • 9.Refardt D, Bergmiller T, Kummerli R. Altruism can evolve when relatedness is low: evidence from bacteria committing suicide upon phage infection. Proceedings Biological sciences/The Royal Society. 2013;280:20123035. doi: 10.1098/rspb.2012.3035. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Hazan R, Engelberg-Kulka H. Escherichia coli mazEF-mediated cell death as a defense mechanism that inhibits the spread of phage P1. Molecular genetics and genomics: MGG. 2004;272:227–234. doi: 10.1007/s00438-004-1048-y. [DOI] [PubMed] [Google Scholar]
  • 11.Wells JE, Russell JB. Why do many ruminal bacteria die and lyse so quickly? J Dairy Sci. 1996;79:1487–1495. doi: 10.3168/jds.S0022-0302(96)76508-6. [DOI] [PubMed] [Google Scholar]
  • 12.Cascales E, Buchanan SK, Duche D, Kleanthous C, Lloubes R, Postle K, Riley M, Slatin S, Cavard D. Colicin biology. Microbiology and molecular biology reviews: MMBR. 2007;71:158–229. doi: 10.1128/MMBR.00036-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Dwyer DJ, Camacho DM, Kohanski MA, Callura JM, Collins JJ. Antibiotic-induced bacterial cell death exhibits physiological and biochemical hallmarks of apoptosis. Mol Cell. 2012;46:561–572. doi: 10.1016/j.molcel.2012.04.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Haussler S. Multicellular signalling and growth of Pseudomonas aeruginosa. Int J Med Microbiol. 2010;300:544–548. doi: 10.1016/j.ijmm.2010.08.006. [DOI] [PubMed] [Google Scholar]
  • 15.Harmsen M, Yang L, Pamp SJ, Tolker-Nielsen T. An update on Pseudomonas aeruginosa biofilm formation, tolerance, and dispersal. FEMS Immunol Med Microbiol. 2010;59:253–268. doi: 10.1111/j.1574-695X.2010.00690.x. [DOI] [PubMed] [Google Scholar]
  • 16.Cao H, Krishnan G, Goumnerov B, Tsongalis J, Tompkins R, Rahme LG. A quorum sensing-associated virulence gene of Pseudomonas aeruginosa encodes a LysR-like transcription regulator with a unique self-regulatory mechanism. Proceedings of the National Academy of Sciences of the United States of America. 2001;98:14613–14618. doi: 10.1073/pnas.251465298. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Deziel E, Lepine F, Milot S, He J, Mindrinos MN, Tompkins RG, Rahme LG. Analysis of Pseudomonas aeruginosa 4-hydroxy-2-alkylquinolines (HAQs) reveals a role for 4-hydroxy-2-heptylquinoline in cell-to-cell communication. Proceedings of the National Academy of Sciences of the United States of America. 2004;101:1339–1344. doi: 10.1073/pnas.0307694100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Lepine F, Milot S, Deziel E, He J, Rahme LG. Electrospray/mass spectrometric identification and analysis of 4-hydroxy-2-alkylquinolines (HAQs) produced by Pseudomonas aeruginosa. J Am Soc Mass Spectrom. 2004;15:862–869. doi: 10.1016/j.jasms.2004.02.012. [DOI] [PubMed] [Google Scholar]
  • 19.Xiao G, Deziel E, He J, Lepine F, Lesic B, Castonguay MH, Milot S, Tampakaki AP, Stachel SE, Rahme LG. MvfR, a key Pseudomonas aeruginosa pathogenicity LTTR-class regulatory protein, has dual ligands. Mol Microbiol. 2006;62:1689–1699. doi: 10.1111/j.1365-2958.2006.05462.x. [DOI] [PubMed] [Google Scholar]
  • 20.McKnight SL, Iglewski BH, Pesci EC. The Pseudomonas quinolone signal regulates rhl quorum sensing in Pseudomonas aeruginosa. J Bacteriol. 2000;182:2702–2708. doi: 10.1128/jb.182.10.2702-2708.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Gallagher LA, McKnight SL, Kuznetsova MS, Pesci EC, Manoil C. Functions required for extracellular quinolone signaling by Pseudomonas aeruginosa. J Bacteriol. 2002;184:6472–6480. doi: 10.1128/JB.184.23.6472-6480.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Mitchell G, Seguin DL, Asselin AE, Deziel E, Cantin AM, Frost EH, Michaud S, Malouin F. Staphylococcus aureus sigma B-dependent emergence of small-colony variants and biofilm production following exposure to Pseudomonas aeruginosa 4-hydroxy-2-heptylquinoline-N-oxide. BMC Microbiol. 2010;10:33. doi: 10.1186/1471-2180-10-33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Schuster M, Lostroh CP, Ogi T, Greenberg EP. Identification, timing, and signal specificity of Pseudomonas aeruginosa quorum-controlled genes: a transcriptome analysis. J Bacteriol. 2003;185:2066–2079. doi: 10.1128/JB.185.7.2066-2079.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Hacker B, Barquera B, Crofts AR, Gennis RB. Characterization of mutations in the cytochrome b subunit of the bc1 complex of Rhodobacter sphaeroides that affect the quinone reductase site (Qc) Biochemistry. 1993;32:4403–4410. doi: 10.1021/bi00067a033. [DOI] [PubMed] [Google Scholar]
  • 25.Van Ark G, Berden JA. Binding of HQNO to beef-heart sub-mitochondrial particles. Biochimica et biophysica acta. 1977;459:119–127. doi: 10.1016/0005-2728(77)90014-7. [DOI] [PubMed] [Google Scholar]
  • 26.Armstrong JS. The role of the mitochondrial permeability transition in cell death. Mitochondrion. 2006;6:225–234. doi: 10.1016/j.mito.2006.07.006. [DOI] [PubMed] [Google Scholar]
  • 27.Tsujimoto Y, Shimizu S. Role of the mitochondrial membrane permeability transition in cell death. Apoptosis. 2007;12:835–840. doi: 10.1007/s10495-006-0525-7. [DOI] [PubMed] [Google Scholar]
  • 28.Hadley P. The Variation in Size of Lytic Areas and Its Significance. J Bacteriol. 1924;9:397–403. doi: 10.1128/jb.9.4.397-403.1924. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Zierdt CH. Autolytic nature of iridescent lysis in Pseudomonas aeruginosa. Antonie Van Leeuwenhoek. 1971;37:319–337. doi: 10.1007/BF02218503. [DOI] [PubMed] [Google Scholar]
  • 30.D’Argenio DA, Calfee MW, Rainey PB, Pesci EC. Autolysis and autoaggregation in Pseudomonas aeruginosa colony morphology mutants. J Bacteriol. 2002;184:6481–6489. doi: 10.1128/JB.184.23.6481-6489.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Heurlier K, Denervaud V, Haenni M, Guy L, Krishnapillai V, Haas D. Quorum-sensing-negative (lasR) mutants of Pseudomonas aeruginosa avoid cell lysis and death. J Bacteriol. 2005;187:4875–4883. doi: 10.1128/JB.187.14.4875-4883.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Lesic B, Lepine F, Deziel E, Zhang J, Zhang Q, Padfield K, Castonguay MH, Milot S, Stachel S, Tzika AA, et al. Inhibitors of pathogen intercellular signals as selective anti-infective compounds. PLoS Pathog. 2007;3:1229–1239. doi: 10.1371/journal.ppat.0030126. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Starkey M, Lepine F, Maura D, Bandyopadhaya A, Lesic B, He J, Kitao T, Righi V, Milot S, Tzika A, et al. Identification of anti-virulence compounds that disrupt quorum-sensing regulated acute and persistent pathogenicity. PLoS Pathog. 2014;10:e1004321. doi: 10.1371/journal.ppat.1004321. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Okshevsky M, Meyer RL. The role of extracellular DNA in the establishment, maintenance and perpetuation of bacterial biofilms. Critical reviews in microbiology. 2015;41:341–352. doi: 10.3109/1040841X.2013.841639. [DOI] [PubMed] [Google Scholar]
  • 35.Garcia-Horsman JA, Barquera B, Rumbley J, Ma J, Gennis RB. The superfamily of heme-copper respiratory oxidases. J Bacteriol. 1994;176:5587–5600. doi: 10.1128/jb.176.18.5587-5600.1994. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Margoliash E, Walasek OF. In: Methods in Enzymology. Estabrook RW, Pullman ME, editors. X. Academia Press; New York: 1967. pp. 339–348. [Google Scholar]
  • 37.Drose S, Brandt U. The mechanism of mitochondrial superoxide production by the cytochrome bc1 complex. J Biol Chem. 2008;283:21649–21654. doi: 10.1074/jbc.M803236200. [DOI] [PubMed] [Google Scholar]
  • 38.Chandra J, Samali A, Orrenius S. Triggering and modulation of apoptosis by oxidative stress. Free Radic Biol Med. 2000;29:323–333. doi: 10.1016/s0891-5849(00)00302-6. [DOI] [PubMed] [Google Scholar]
  • 39.Michel-Briand Y, Baysse C. The pyocins of Pseudomonas aeruginosa. Biochimie. 2002;84:499–510. doi: 10.1016/s0300-9084(02)01422-0. [DOI] [PubMed] [Google Scholar]
  • 40.Liberati NT, Urbach JM, Miyata S, Lee DG, Drenkard E, Wu G, Villanueva J, Wei T, Ausubel FM. An ordered, nonredundant library of Pseudomonas aeruginosa strain PA14 transposon insertion mutants. Proceedings of the National Academy of Sciences of the United States of America. 2006;103:2833–2838. doi: 10.1073/pnas.0511100103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Hurley BP, Goodman AL, Mumy KL, Murphy P, Lory S, McCormick BA. The two-component sensor response regulator RoxS/RoxR plays a role in Pseudomonas aeruginosa interactions with airway epithelial cells. Microbes Infect. 2010;12:190–198. doi: 10.1016/j.micinf.2009.11.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Comolli JC, Donohue TJ. Pseudomonas aeruginosa RoxR, a response regulator related to Rhodobacter sphaeroides PrrA, activates expression of the cyanide-insensitive terminal oxidase. Mol Microbiol. 2002;45:755–768. doi: 10.1046/j.1365-2958.2002.03046.x. [DOI] [PubMed] [Google Scholar]
  • 43.Damron FH, Owings JP, Okkotsu Y, Varga JJ, Schurr JR, Goldberg JB, Schurr MJ, Yu HD. Analysis of the Pseudomonas aeruginosa regulon controlled by the sensor kinase KinB and sigma factor RpoN. J Bacteriol. 2012;194:1317–1330. doi: 10.1128/JB.06105-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Leys N, Baatout S, Rosier C, Dams A, s’Heeren C, Wattiez R, Mergeay M. The response of Cupriavidus metallidurans CH34 to spaceflight in the international space station. Antonie Van Leeuwenhoek. 2009;96:227–245. doi: 10.1007/s10482-009-9360-5. [DOI] [PubMed] [Google Scholar]
  • 45.Shapiro JA. Thinking about bacterial populations as multicellular organisms. Annu Rev Microbiol. 1998;52:81–104. doi: 10.1146/annurev.micro.52.1.81. [DOI] [PubMed] [Google Scholar]
  • 46.de Kievit TR. Quorum sensing in Pseudomonas aeruginosa biofilms. Environ Microbiol. 2009;11:279–288. doi: 10.1111/j.1462-2920.2008.01792.x. [DOI] [PubMed] [Google Scholar]
  • 47.Allesen-Holm M, Barken KB, Yang L, Klausen M, Webb JS, Kjelleberg S, Molin S, Givskov M, Tolker-Nielsen T. A characterization of DNA release in Pseudomonas aeruginosa cultures and biofilms. Mol Microbiol. 2006;59:1114–1128. doi: 10.1111/j.1365-2958.2005.05008.x. [DOI] [PubMed] [Google Scholar]
  • 48.Que YA, Hazan R, Ryan CM, Milot S, Lepine F, Lydon M, Rahme LG. Production of Pseudomonas aeruginosa Intercellular Small Signaling Molecules in Human Burn Wounds. J Pathog. 2011;2011:549302. doi: 10.4061/2011/549302. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Nobrega FL, Costa AR, Kluskens LD, Azeredo J. Revisiting phage therapy: new applications for old resources. Trends Microbiol. 2015;23:185–191. doi: 10.1016/j.tim.2015.01.006. [DOI] [PubMed] [Google Scholar]
  • 50.Halestrap AP, Connern CP, Griffiths EJ, Kerr PM. Cyclosporin A binding to mitochondrial cyclophilin inhibits the permeability transition pore and protects hearts from ischaemia/reperfusion injury. Mol Cell Biochem. 1997;174:167–172. [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

supplement
NIHMS744309-supplement.pdf (1.2MB, pdf)

RESOURCES