SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

For a sustained infection, enteric bacterial pathogens must evade, resist or tolerate a variety of antimicrobial host defence peptides and proteins. We report here that specific organic acids protect stationary-phase Escherichia coli and Salmonella cells from killing by a potent antimicrobial peptide derived from the human bactericidal/permeability-increasing protein (BPI). BPI-derived peptide P2 rapidly halted oxygen consumption by stationary-phase cells preincubated with glucose, pyruvate or malate and caused a 109-fold drop in cell viability within 90 min of addition. In marked contrast, O2 consumption and viability were not significantly affected in stationary-phase cells preincubated with formate or succinate. Experiments with fdhH, fdoG, fdnG, selC and sdhO mutants indicate that protection by formate and succinate requires their oxidation by the Fdh-N formate dehydrogenase and succinate dehydrogenase respectively. Protection was also dependent on the BipA GTPase but did not require the RpoS sigma factor. We conclude that the primary lesion caused by this cationic peptide is not gross permeabilization of the bacterial cytoplasmic membrane but may involve specific disruption of the respiratory chain. Because P2 shares sequence similarity with a range of other antimicrobial peptides, its cytotoxic mechanism has broader significance. Additionally, protective quantities of formate are secreted by E. coli and Salmonella during growth suggesting that such compounds are important determinants of bacterial survival in the host.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Antibacterial peptides and proteins are ubiquitous in nature and serve as a critical first line of host defence against infection (Martin et al., 1995). It follows that many commensal and pathogenic bacteria must avoid, resist or tolerate such components if they are to survive within a host. This is particularly the case for enteric pathogens, such as Escherichia coli and Salmonella, which are exposed to a wide array of antibacterial peptides and proteins in diverse host environments. These antimicrobial compounds include the cryptidins, which are produced by the Paneth cells of the small intestine, the β-defensins, which are secreted by a wide variety of epithelial cells, and proteins such as azurocidin/CAP37 and bactericidal/permeability-increasing protein (BPI, also known as CAP57) of granulocytes (Elsbach et al., 1979; 1993; Shafer et al., 1984; Hovde and Gray, 1986; Gabay et al., 1989; Joiner et al., 1989; Ouellette and Luardi, 1990; Stolzenberg et al., 1997).

The cationic protein BPI is of particular interest because of its extreme potency and specificity towards Gram-negative bacteria. The specificity of the protein is due to its high affinity for the lipid A region of lipopolysaccharide (LPS), to which it binds with Kd values in the low nanomolar range (Gazzano-Santoro et al., 1992; Capodici et al., 1994). Following binding to LPS, BPI disrupts the bacterial outer membrane, thereby rendering it permeable to hydrophobic compounds. The addition of trypsin or high concentrations of Mg2+ ions to BPI-treated bacteria removes most of the bound protein and reverses the damage to the outer membrane (Weiss et al., 1976, 1984). However, bacterial growth is not restored, indicating that the primary cytotoxic target for BPI resides elsewhere in the cell. Several lines of evidence suggest that the principal target is located at the cytoplasmic membrane. Not only are oxygen uptake and energy-dependent transport of amino acids inhibited before bacterial death (Hovde and Gray, 1986; In’t Veld et al., 1988), but the kinetics of bacterial killing by BPI closely coincide with damage to the cytoplasmic membrane (In’t Veld et al., 1988; Mannion et al., 1990). Despite these results, however, the molecular basis for killing remains to be elucidated.

Functional mapping of BPI has established that peptides containing residues 85–99 of BPI retain significant antibacterial activity (Gray and Haseman, 1994; Little et al., 1994). This region, which forms an amphipathic β-turn in the crystal structure of BPI (Beamer et al., 1997), also binds LPS and shares sequence similarity with several other antibacterial and LPS-binding proteins, including the anti-LPS factor of Limulus (Hoess et al., 1993; Little et al., 1994). The cytotoxic mechanism used by such BPI-derived peptides is currently unknown but may be similar to that of the native protein (Gray and Haseman, 1994; Little et al., 1994). Peptides bearing residues 85–99 also retain the ability to render the outer membrane permeable to hydrophobic compounds (H.S.B. and C.D.O'C, unpublished results).

In view of the variety of antibacterial peptides produced by host organisms, successful enteric pathogens are likely to have evolved multiple strategies for evading, tolerating or resisting these defences. Known genetic loci in Salmonella implicated in such resistance to antimicrobial peptides include eight sap (sensitivity to antimicrobial peptides) loci as well as the pmrAB, phoPQ and bipA loci. The sap loci were identified on the basis of hypersensitivity to the cationic peptide protamine and some have been functionally characterized (Groisman et al., 1992). The sapABCDF operon, for example, specifies an ABC transporter system that has been previously implicated in transport of K+ ions across the cytoplasmic membrane. Similarly, the sapG and sapJ loci also specify components of a low-affinity K+-transport system (Parra-Lopez et al., 1994).

The pmrAB loci specify a two-component regulatory system that controls Salmonella resistance to several antimicrobial peptides, including polymyxin, protamine and BPI (Roland et al., 1993). The system is activated by mild acid but also by another two-component system, PhoP–PhoQ, which is activated by both low pH and by decreases in the extracellular concentration of Mg2+ ions. Recent studies indicate that both systems regulate LPS alterations, including the attachment of aminoarabinose and palmitate to lipid A and the modification of phosphate groups in the core region (Guo et al., 1997). Such alterations will reduce the charge density associated with LPS and also modulate the fluidity of the outer membrane, thereby decreasing the affinity of cationic peptides for the bacterial surface.

Proteomic studies on Salmonella have uncovered a novel member of the GTPase superfamily that is strongly induced in response to BPI (Qi et al., 1995). This protein, termed BipA (BPI-inducible protein A), participates in several virulence-associated processes in enteropathogenic E. coli and in Salmonella, including resistance to BPI and BPI-derived antibacterial peptides (Farris et al., 1998; N. Kinsella, A. Grant, M. Farris, A. White, P. Adams and C. D. O′Connor, unpublished results). The pleiotropic effects of bipA null mutants suggest that the protein functions as a global regulator. Moreover, its striking similarity to ribosome-binding GTPases such as elongation factor G indicates that it may operate by an unusual mechanism. However, little is presently known about its mode of action or how it is involved in resistance to antimicrobial peptides.

In this paper, we report that formate and certain other carboxylic acids protect stationary-phase E. coli and Salmonella cells from a potent antimicrobial peptide derived from BPI. We also present evidence suggesting that the peptide acts at the level of the respiratory chain and show that protection mediated by formate involves the BipA GTPase and is dependent on its oxidation by a formate dehydrogenase. As millimolar amounts of such organic acids are secreted by E. coli and Salmonella during fermentative growth, we suggest that they play a similar protective role during infection.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Cell lysis due to the BPI-derived peptide P2 is delayed in stationary-phase E. coli

Previous studies showing that stationary-phase cells are frequently more resistant to environmental stresses (Foster and Spector, 1995; Hengge-Aronis, 1996; 1999) raise the possibility that they might also respond differently to antimicrobial peptides. We therefore examined the ability of P2, a 22-residue cationic peptide derived from the host defence protein BPI, to lyse E. coli ML35 cells at different stages of growth. Lysis of the cytoplasmic membrane of ML35 was monitored by measuring the rate of o-nitrophenol β-d-galactopyranoside (ONPG) hydrolysis in the growth medium. Because this strain constitutively produces β-galactosidase but is deficient in Lac permease, ONPG is only hydrolysed at a significant rate if the cytoplasmic membrane is breached. Following P2 addition, ONPG hydrolysis began after 10 min with exponential phase cells but only commenced after 40 min with stationary-phase cells (Fig. 1). Once lysis had commenced, the rate of ONPG hydrolysis was similar for both types of cell. We conclude that lysis of stationary-phase cells treated with peptide P2 occurs later than with cells in the exponential phase of growth.

image

Figure 1. Lysis of the cytoplasmic membrane mediated by peptide P2 in exponential (1) and stationary-phase (2) ML35 cells. ONPG hydrolysis (A400) was monitored at 60 s intervals following addition of P2 (50 µg ml−1) to cells. The C source used was 1% TSB. β- Galactosidase assays indicated that similar levels of the enzyme were present in exponential- and stationary-phase cells (data not shown). No significant hydrolysis of ONPG was observed in untreated ML35 cells over the time course of the experiment.

Download figure to PowerPoint

P2-mediated inhibition of DNA and protein biosynthesis is slower in stationary-phase cells than in exponential phase cells

We next compared the onset of inhibition of DNA and protein synthesis after addition of peptide P2. In exponential phase ML35 cells, both processes were completely inhibited 15 min after P2 addition (Fig. 2A and B). In contrast, with stationary-phase cells relatively little inhibition of DNA synthesis occurred at 15 min and complete inhibition only occurred 30 min after the addition of peptide P2 (Fig. 2C). Similarly, P2-mediated inhibition of protein synthesis was also slower in stationary-phase cells (Fig. 2D). These results indicate that, like lysis of the cytoplasmic membrane, inhibition of DNA and protein synthesis by P2 is also delayed in stationary-phase cells.

image

Figure 2. Effect of peptide P2 on DNA and protein synthesis in exponential- (A and B) and stationary-phase ML35 cells (C and D). [3H]-thymidine (A and C) and 14C-labelled-amino acid (B and D) incorporation into TCA-precipitable material was measured at 15 min intervals in the presence (○) or absence (●) of P2 (50 µg ml−1). Results are an average of three independent experiments.

Download figure to PowerPoint

Inhibition of bacterial O2 consumption is the earliest detectable event caused by peptide P2

Studies with native BPI have shown that it rapidly inhibits bacterial oxygen consumption (Hovde and Gray, 1986; In’t Veld et al., 1988). Because peptide P2 mimics many of the properties of the holoprotein, we tested its ability to block bacterial O2 consumption in stationary-phase cells. Addition of P2 to ML35 cells blocked O2 uptake in a dose-dependent manner, with 50 µg of the peptide causing complete inhibition of O2 consumption within 15 min (Fig. 3). Because permeabilization of the cytoplasmic membrane and complete inhibition of macromolecular synthesis occur much later under the same experimental conditions (see above), cessation of O2 consumption appears to be a very early effect of P2 on cells.

image

Figure 3. Effect of peptide P2 on oxygen consumption in stationary-phase ML35 cells. O2 consumption was monitored in a Clark-type oxygen electrode using the standard assay mix with 1% TSB. The arrow indicates the point of P2 addition: The curve corresponds to addition of the following amounts of P2: 0 µg (1); 10 µg (2); 25 µg (3); and 50 µg (4).

Download figure to PowerPoint

Stationary-phase E. coli and Salmonella cells survive exposure to P2 in the presence of specific organic acids

The rapid inhibition of respiration on addition of P2 suggested that the primary target for this peptide might be a component of the bacterial respiratory chain. We also postulated that alterations to the respiratory chain that occurred on entry of cells into the stationary phase might account for the delay in P2-mediated cell lysis and inhibition of protein and DNA synthesis. To test these ideas, we studied stationary-phase cells respiring single-carbon sources. P2 halted O2 consumption in cells incubated with glucose, pyruvate and malate (Fig. 4) and also caused a 109-fold decrease in cell viability over 90 min (Fig. 5). In marked contrast, when formate was used as a C source, bacterial O2 consumption decreased less than twofold and the viability of cells decreased by only one log order over 90 min (Figs 4 and 5). A protective effect was also observed when succinate or d/l-lactate was included with cells (Fig. 5).

image

Figure 4. Effect of peptide P2 on the rate of oxygen consumption of stationary-phase E. coli ML35 cells in the presence of different carbon sources. The shaded and lightly shaded bars indicate cells in the presence and absence of P2 (50 µg ml−1) respectively. Asterisks indicate cases where the rate of O2 consumption was undetectable. All carbon sources were present at 2.5 mM ,except TSB, which was added to 1%, final concentration.

Download figure to PowerPoint

image

Figure 5. Effect of different C sources on the viability of stationary-phase ML35 cells following addition of peptide P2. cfu ml−1 were determined by plating serial dilutions of samples taken at the indicated time points after P2 addition. (○), control (no peptide); (▪_),formate; (◆_),d + l-lactate; (●_),glucose; (▾_), succinate; (▴_), l-lactate. Other conditions were as described in the legend to Fig. 4.

Download figure to PowerPoint

To rule out the possibility that the protective effect was strain specific, similar viability assays were performed with Salmonella cells. Although Salmonella strain SL1344 was not as sensitive to P2 as E. coli ML35, showing only a ≈ 104-fold decrease in cell viability over 180 min, addition of formate once again had a protective effect. Inclusion of formate alleviated P2-mediated inhibition of oxygen consumption (Fig. 6A) and increased cell viability (Fig. 6B). Protection was obtained with formate at concentrations between 0.1 and 5 mM; however, higher concentrations of the acid were deleterious to cells. As before, formate did not protect exponential phase cells exposed to P2 (data not shown). Succinate was also found to protect SL1344 against peptide P2 (Fig. 6A and B). Taken together, these results indicate that certain organic acids, specifically formate and succinate, protect stationary-phase cells from the antibacterial effects of BPI-derived peptide P2.

image

Figure 6. A. Effect of formate or succinate on O2 consumption in stationary-phase cells of Salmonella SL1344 in the presence or absence of peptide P2. Formate or succinate were included at 5 mM, final concentration.

B. Effect of formate or succinate on viability of stationary-phase cells of Salmonella SL1344 treated with peptide P2. Succinate (▾_) was included at 5 mM whereas formate was included at the following concentrations: 0.1 mM (◆_); 1 mM (▴_); and 5 mM (▪_). Untreated cells and cells treated with P2 in the absence of formate are shown by (○_) and (●_) respectively. Cells were prepared as described in Experimental procedures and cfu ml−1 were determined by plating serial dilutions of samples taken at the indicated time points after addition of P2 (50 µg).

Download figure to PowerPoint

Formate protects but does not rescue bacterial cells treated with peptide P2.

We next tested if formate inhibits killing by P2 when it is added to cells after addition of the peptide. Figure 7 shows that the addition of formate 5 min prior to addition of peptide P2 was sufficient to protect stationary-phase ML35 cells. In contrast, formate was unable to protect ML35 cells if it was added 5 min after peptide P2. Indeed, a slight reduction in viability relative to cells treated with peptide P2 alone was consistently found under these conditions. We conclude that formate has a protective role but cannot rescue cells previously treated with the peptide.

image

Figure 7. Formate protects but does not rescue E. coli ML35 cells. Formate (5 mM, final concentration) was added to ≈ 107 stationary-phase ML35 cells 5 min before (▪_) or 5 min after (▴_) the addition of peptide P2 (25 µg ml−1). For comparison untreated cells (○) and cells treated with just peptide P2 (●) were also tested. In each case samples were taken at 30 min intervals and cell viability determined as the number of colonies formed after incubation at 37°C for 18–24 h.

Download figure to PowerPoint

Oxidation of formate by the Fdh-N formate dehydrogenase is required for protection of stationary-phase cells.

To determine if protection by formate required its metabolism, the cell viability experiments were repeated using E. coli mutants lacking one or more formate dehydrogenase. In addition to fdhH, fdnG and fdoG mutants, coding for subunits of formate dehydrogenases-H, -N and –O respectively, we also tested a selC mutant. This gene specifies a selenocysteine-specific tRNA and is essential for insertion of this amino acid into each of the three known formate dehydrogenases of E. coli (Sawers et al., 1991; Sawers, 1994). Thus, selC null mutants lack detectable formate dehydrogenase activity. Figure 8A shows that formate protected parental E. coli cells and fdhHand fdoG derivatives from peptide P2 (2.5 h exposure), with a ³ 100-fold increase in viability relative to control cells. In contrast, both the selC and fdnG mutants were not protected under the same conditions. These results indicate that a formate dehydrogenase, specifically Fdh-N, is required for protection of stationary-phase cells by formate.

image

Figure 8. A. Susceptibility of formate dehydrogenase mutants to the antimicrobial action of peptide P2 in the presence or absence of formate. E. coli MC4100 and RK4353 or their selC, fdhH, fdoG and fdnG null mutant derivatives (≈ 5 × 107 cells) were incubated with peptide P2 (0.1 mg ml−1, final concentration) in the presence or absence of 2.5 mM formate. Cell viability was then determined as the number of colonies formed after incubation at 37°C for 18–24 h. Asterisks indicate < 10 cfu ml−1.

B. Susceptibility of a succinate dehydrogenase mutant to the antimicrobial action of peptide P2 in the presence or absence of succinate. E. coli PL2024 and its isogenic sdhO-and frdA11-mutants (≈ 5 × 107 cells) were incubated with peptide P2 (0.1 mg ml−1, final concentration) in the presence or absence of 2.5 mM formate. Cell viability was then determined as above. Asterisks indicate < 10 cfu ml−1.

Download figure to PowerPoint

Similar experiments were carried out with mutants deficient in succinate dehydrogenase to determine if protection by succinate also required its metabolism. The presence of succinate increased the viability of the parental E. coli strain ³ 1000-fold relative to control samples, whereas isogenic sdhO mutants were not protected (Fig. 8B). Collectively, these experiments indicate that the oxidation of protective organic acids by their cognate dehydrogenases is required for protection of stationary-phase cells from peptide P2.

The BipA GTPase but not RpoS is needed for protection of stationary-phase Salmonella cells by formate.

The RpoS sigma factor is essential for increased resistance to many stresses and plays an important role in cell survival during the stationary phase (Foster and Spector, 1995; Hengge-Aronis, 1996, 1999). We therefore compared the sensitivities of Salmonella SL1344 and an isogenic SL1344 rpoS::kan mutant towards peptide P2 in the presence and absence of formate. Cells that had been incubated with formate were less sensitive to peptide P2 than the control (Fig. 9A). However, in both cases, the kinetics of cell death were altered relative to the parent strain (see Fig. 6). The decreased sensitivity of the formate-treated SL1344 cells to peptide P2 suggests that RpoS does not play an important role in the protection mechanism.

image

Figure 9. A. Viability of peptide P2-treated SL1344 rpoS::kan in the presence (▪) or absence (●) of 5 mM formate. Cells were treated with 50 µg of peptide P2 or left untreated (○). In each case samples were taken at 30 min intervals and cell viability determined as the number of colonies formed after incubation at 37°C for 18–24 h.

B. Comparison of peptide P2 effects on O2 consumption in wild-type, BipA null mutant and transcomplemented Salmonella cells. Conditions were as described above.

C. Viability of peptide P2-treated SL1344 bipA::cat cells in the presence or absence of 5 mM formate. Samples were taken at 30 min intervals and cell viability determined as the number of colonies formed after incubation at 37°C for 18–24 h.

Download figure to PowerPoint

Previous studies have uncovered a novel GTPase termed BipA that is induced when Salmonella is exposed to sublethal amounts of the antimicrobial protein BPI (Qi et al., 1995). Because peptide P2 is derived from BPI and retains many of its properties, we also examined the role of BipA in the protective effect. Figure 9B shows that peptide P2 inhibited respiration in SL1344 bipA::cat cells even when formate was present. Indeed, the degree of P2 inhibition of O2 consumption was greater in the bipA::cat strain relative to the parent cells as the mutant consistently showed higher rates of formate-driven respiration. Introduction of pBipA, a plasmid expressing a wild-type bipA gene from an EPEC strain (Farris et al., 1998), increased the ability of formate to block inhibition of O2 consumption. In keeping with these results, formate failed to increase the survival of a BipA null mutant (Fig. 9C). We therefore conclude that the BipA GTPase, but not RpoS, plays an important role in the formate-mediated protection of stationary-phase Salmonella cells against peptide P2.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Both pathogenic and commensal bacteria encounter antimicrobial peptides in the host environment and have presumably evolved appropriate adaptive strategies. In this paper, we have shown that E. coli and Salmonella responses to a potent antimicrobial peptide, P2, depend critically on the stage of cell growth and that the cytotoxic effects of the peptide are delayed in stationary-phase cells. Further study of these responses has revealed that specific organic acids, notably formate and succinate, protect stationary-phase cells from peptide P2. The formate effect was investigated further and shown to be abolished in selCand fdnG mutants. Similarly, the protective effect of succinate was not observed in an sdhO mutant. These results suggest that oxidation of formate and succinate, by the Fdh-N and succinate dehydrogenases respectively, is required for protection. Finally, experiments with Salmonella mutants indicated that the BipA GTPase but not RpoS participates in the protective mechanism.

Given these findings, how might peptide P2 operate and what is the molecular basis for cellular protection by specific organic acids? Many antimicrobial peptides form pores or channels in membranes (Martin et al., 1995). Thus, one possibility is that peptide P2 forms pores in the cytoplasmic membrane of E. coli and Salmonella cells. If pore formation was sufficiently slow in stationary-phase cells metabolizing formate (or succinate) it might allow the cell time to induce additional, as yet unknown, defences. Although this model explains some of the experimental data, it is difficult to reconcile it with the observation that cell survival does not correlate with the relative rate of O2 consumption with different carbon sources (Fig. 5). Moreover, if pore formation was dependent on the growth rate, starved cells should be protected from the cytotoxic effects of P2. This does not appear to occur, as stationary-phase cells stored in phosphate-buffered saline are efficiently killed by the peptide (data not shown).

An alternative model that is more consistent with the experimental data is that peptide P2 targets a component in the respiratory chain. The resulting inhibition of electron flow through the chain would rapidly halt bacterial O2 consumption and lead to a decrease in the proton gradient. This, in turn, would reduce ATP production, thereby arresting DNA and protein synthesis and ultimately causing the disintegration of the cytoplasmic membrane. The model predicts that O2 consumption should be restored if the lesion in the respiratory chain is bypassed, i.e. electron flow is restored further along the chain. In agreement with this prediction compounds that feed electrons into the chain on oxidation by specific membrane-bound dehydrogenases restored bacterial O2 consumption and protected cells from killing by peptide P2 (Figs 4, 5 and 8).

The observation that fdnG and selC mutants were no longer protected by formate provides further evidence that P2 affects the respiratory chain and that oxidation of this organic acid is required to bypass the lesion. There are three known formate dehydrogenases in E. coli, and each functions in different conditions. Synthesis of Fdh-H is optimal under fermentative conditions, whereas Fdh-O is synthesized in the presence of oxygen and nitrate and Fdh-N is induced by anaerobiosis and nitrate (Lester and DeMoss, 1971; Sawers, 1994). The involvement of Fdh-H and Fdh-O can be ruled out as strains lacking these activities were protected as effectively as the parent strain (Fig. 8A). In contrast, results with an fdnG mutant strongly suggest that the Fdh-N formate dehydrogenase is required for formate-mediated protection of stationary-phase cells. Fdh-N forms part of a formate–nitrate respiratory chain that is made if nitrate is available to anaerobic cells. It therefore appears that, even though the stationary-phase cultures were grown aerobically, there is sufficient oxidation of formate by Fdh-N to allow protection of cells from peptide P2. It remains to be seen, however, if the high level induction of Fdh-N leads to a further increase in protection of cells by formate. Similar results with succinate, and the observation that cells lacking succinate dehydrogenase could not be protected (Fig. 8B), are also consistent with the hypothesis that peptide P2 affects the respiratory chain.

The positions of the formate, succinate and d-lactate dehydrogenases in the respiratory chain suggest that the target for peptide P2 may be at or close to complex I, the energy-generating NADH:ubiquinone oxidoreductase (Fig. 10). It is noteworthy that complex I of E. coli mainly operates in anaerobic electron transport and is essential for fumarate respiration. In contrast, aerobic and nitrate respiration mainly operate through the alternative NADH dehydrogenase (Tran et al., 1997; Tran and Unden, 1998). These results correlate well with the observation that the (anaerobic) Fdh-N is involved in protection by formate. In E. coli, complex I consists of 13 subunits which have been reconstituted in vitro (Leif et al., 1995; Friedrich, 1998). To date, however, attempts to obtain inhibition of purified complex I using P2 have been unsuccessful (T. Friedrich, data not shown). While this may reflect the difficulty of accurately reproducing in vitro the complex architecture of this part of the respiratory chain (for example, soybean rather than E. coli phospholipids are used for reconstitution), it is also possible that peptide P2 acts further along the respiratory pathway. It is noteworthy, for example, that E. coli mutants defective in complex I (nuo-) or in the non-energy-generating alternative NADH dehydrogenase (ndh-) have levels of sensitivity to peptide P2 that are similar to the parent strain (A. Cottle and C. D. O′Connor, unpublished results). Thus, further biochemical studies are required to pinpoint the target for P2 more precisely.

image

Figure 10. Scheme to illustrate the positions of protective and non-protective compounds used in this study in relation to the respiratory chain of E. coli. The proposed position of the inhibitory lesion mediated by BPI-derived peptide P2 is also shown. ‘NADH’ and ‘Q’ indicate the positions of the NADH:ubiquinone oxidoreductase complex and the common quinone pool, respectively, whereas ‘Cyt. bd and bo’ indicate the terminal quinol oxidases.

Download figure to PowerPoint

In view of the different responses of exponential- and stationary-phase cells to peptide P2, we tested the involvement of the RpoS sigma factor, which regulates resistance to a wide range of other stresses (Foster and Spector, 1995; Hengge-Aronis, 1996, 1999). Our results indicate that RpoS does not play a significant role in resistance to peptide P2, as the formate protection effect was still found in a Salmonella rpoS::kan mutant. It therefore appears that Salmonella cells in the stationary phase have additional pathways for the regulation of survival responses to antimicrobial peptides.

Salmonella induces the synthesis of the BipA protein on exposure to the host defence protein BPI (Qi et al., 1995). This GTPase is involved in a number of virulence-related processes in enteropathogenic E. coli and Salmonella, including resistance to BPI and its peptide derivative P2 (Farris et al., 1998; N. Kinsella, M. Farris, A. Grant, A. White, M. Woodward and C. D. O′Connor, in preparation). The results obtained in the present work extend these results by demonstrating that, for Salmonella cells at least, the formate-mediated protection mechanism also involves BipA. As BipA null mutants are hypersensitive to both BPI and peptide P2, these antimicrobial molecules may use the same mechanism for killing bacteria. Further support for this proposal comes from the finding that peptide P2 and BPI have similar effects on bacterial physiology. For example, both render the bacterial outer membrane permeable to hydrophobic compounds and rapidly inhibit bacterial O2 consumption (Hovde and Gray, 1986; In’t Veld et al., 1988; this study; H. Barker, A. Jaspe and C. D. O’Connor, in preparation). It is possible therefore that the BPI residues contained in peptide P2 constitute a ‘warhead’ and that the main function of the rest of the holoprotein is to deliver this region to its target in the cytoplasmic membrane. On reaching the membrane, we propose that the region of BPI corresponding to peptide P2 interferes with the respiratory chain, thereby ultimately causing an imbalance between endogenous catabolism and macromolecular synthesis. In this respect, it is interesting to note that entry of cells into the stationary phase also creates such an imbalance between respiration and the synthesis of key macromolecules (Nyström, 1998; Dukan and Nyström, 1999). Thus, there are intriguing parallels between the two phenomena. Because similar sequences to P2 are found in several other antibacterial peptides (Hoess et al., 1993; Little et al., 1994), the cytotoxic mechanism characterized here, and bacterial responses to it, may be of wider significance.

In summary, the studies reported here indicate that specific organic acids protect E. coli and Salmonella from killing by a cationic antimicrobial peptide. Because bacteria secrete millimolar quantities of d-lactate and formate during fermentative growth (Böck and Sawers, 1996), similar protective conditions are likely to exist in the gastrointestinal tract of the host. It is noteworthy that adherent strains of E. coli cells secrete more d-lactate than non-adherent strains when incubated with host cells (McCabe et al., 1998). Moreover, succinic acid has been reported to inhibit the bactericidal activity of neutrophils (Abdul-Majid et al., 1997). The results presented here provide a biological rationale for such observations and suggest that a primary function of such compounds is to protect bacteria in the host.

Experimental procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Bacterial strains and culture conditions

The E. coli strains used in this work were ML35 (Lehrer et al., 1989), MC4100 and its ΔselC::kan,ΔfdhF and fdoGderivatives (Birkmann et al., 1987; Abaibou et al., 1997;), RK4353 and RK4353 fdnG (Li and Stewart, 1992), and PL2024 and its sdhO and frdA11-derivatives (Creaghan and Guest, 1978). The S. typhimurium strains used were SL1344 (Wray and Sojka, 1978), SL1344K (rpoS::kan;Coynault et al., 1996), SL1344 bipA::cat and SL1344 ΔbipA. Construction of the bipA null mutants will be described elsewhere (N. Kinsella, M. Farris, A. Grant, A. White, M. Woodward and C. D. O′Connor, in preparation). The bipA mutants were confirmed by Southern blotting, PCR and DNA sequencing of the region.

Stationary-phase cells were prepared by inoculating tryptic soy broth (TSB) medium (10 ml) with a single colony of the relevant strain prior to 18 h incubation at 37°C with moderate shaking. For exponential phase cells, 0.5 ml of the culture was used to inoculate a further 10 ml of TSB and incubation continued to OD600 = 0.5. Cultures were washed twice with an equal volume of 0.9% NaCl and resuspended in 20 mM sodium phosphate buffer, pH 6.0, to give an OD600 = 0.5.

Peptide synthesis

Peptide P2, consisting of residues 86–104 of BPI flanked by serine and cysteine residues (sequence SKISGKWKAQKRFLKMSGNFGC), was synthesized using standard F-moc chemistry on an ABI model 430A synthesizer. Peptides were cleaved from solid supports with trifluoroacetic acid and purified by gel filtration and reverse-phase high-performance liquid chromatography (HPLC) using a C18 column and an elution gradient of 0–60% acetonitrile with 0.1% trifluoroacetic acid. The purity and size of the peptide was confirmed by analytical HPLC and by mass spectrometry using an electrospray orthogonal acceleration time-of-flight instrument (Micromass LCT). Treatment of peptide P2 with trypsin abolished its cytotoxic effects on bacterial cells, indicating that the bactericidal activity of P2 preparations used in these studies was due to the peptide itself and not to a contaminant.

Measurement of ONPG hydrolysis

Permeabilization of the bacterial cytoplasmic membrane was measured using the method of Lehrer et al. (1989). Two hundred microlitres of cell suspension, prepared as described above, was added to 800 µl of an assay mix consisting of 1% TSB in phosphate buffer (20 mM, pH 6.0), 1.67 mM ONPG. Following the addition of peptide P2, the hydrolysis of ONPG was recorded at 400 nm using a programmable Uvikon spectrophotometer (Kontron) maintained at 37°C. Readings were taken at 1 min intervals for up to 2.5 h and the reference cell containing the assay mix and cells without peptide.

Measurement of DNA and protein synthesis

To measure inhibition of DNA and protein synthesis concurrently with ONPG hydrolysis 14C-labelled amino acids and [3H]-thymidine were added to a final concentration of 1 µCi ml−1 and 10 µCi ml−1 respectively. After addition of peptide P2, 25 µl samples were removed at time points from both the test and reference cuvettes. Each sample was treated with ice-cold 10% trichloroacetic acid (2.5 ml), collected onto Whatman G/FC filters and washed three times with 5% trichloroacetic acid (1 ml) and then with methanol (1 ml). The dried filters were placed in ‘Optiphase HiSafe 3′ scintillant and radioactivity counted using a Beckman LS6500 scintillation counter.

Assay for O2 consumption

The assay mix used to measure ONPG hydrolysis was also used to monitor oxygen consumption, with the exceptions that ONPG was omitted and TSB was replaced by alternative C sources in some experiments; the final assay volume was 2 ml. Oxygen uptake was measured at 37°C with a Clark-type oxygen electrode (Rank Brothers). Cells, suspended at 5 × 107 cfu ml−1, were stirred throughout the assays. O2 uptake before and after the addition of a carbon source was first measured prior to recording the rate following addition of the peptide. The rates of O2 consumption were then calculated from the slopes of the lines recorded.

Preparation and assay of E. coli complex I

E. coli complex I was purified and the NADH:ubiquinone oxidoreductase activity assayed as previously described (Leif et al., 1995). The NADH oxidase activity of purified cytoplasmic membranes was measured using a Clark-type oxygen electrode as described previously (Friedrich et al., 1994).

Bactericidal assay

Bacterial cells, prepared as described above, were diluted to the desired number of cfu ml−1 and 100 µl added to 400 µl of 20 mM sodium phosphate, pH 6.0, containing 1–10 mM of the relevant carbon source and 0–50 µg of peptide P2. Cells were incubated at 37°C and samples taken at timed intervals for plating on Luria–Bertani (LB) agar. Cell viability was determined as the number of colonies formed after incubation at 37°C for 18–24 h. All assays were performed at least in duplicate and usually in triplicate.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

H.B. and N.K. made equal contributions to this work. We thank A. Böck, J. Guest, M. Mandrand-Berthelot, R. Lehrer, F. Norel and V. Stewart for strains, A. Cottle for technical assistance and G. Sawers, B. Poole and M. Akhtar for helpful discussions. These studies were supported by project and equipment grants from the BBSRC and the Wellcome Trust (to C.D.O′C.).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  • Abaibou, H., Giordano, G., Mandrand-Berthelot, M.A. (1997) Suppression of Escherichia coli formate hydrogenlyase activity by trimethylamine N-oxide is due to drainage of the inducer formate. Microbiology 143: 26572664.
  • Abdul-Majid, K.B., Kenny, P.A., Finlay-Jones, J.J. (1997) The effect of the bacterial product, succinic acid, on neutrophil bactericidal activity. FEMS Immunol Med Microbiol 17: 7986.
  • Beamer, L.J., Carroll, S.F., Eisenberg, D. (1997) Crystal structure of human BPI and two bound phospholipids at 2.4 angstrom resolution. Science 276: 18611864.
  • Birkmann, A., Zinoni, F., Sawers, G., Böck, A. (1987) Factors affecting transcriptional regulation of the formate-hydrogen-lyase pathway of Escherichia coli. Arch Microbiol 148: 4451.
  • Böck, A. & Sawers, G. (1996) Fermentation. In Escherichia coli and Salmonella: Cellular and Molecular Biology, 2nd edn. Neidhardt, F.C., et al. (eds) . Washington, DC: American Society for Microbiology Press, pp. 262282.
  • Capodici, C., Chen, S., Sidorczyk, Z., Elsbach, P., Weiss, J. (1994) Effect of lipopolysaccharide (LPS) chain length on interactions of bactericidal/permeability-increasing prtoein and its bioactive 23-kilodalton NH2-terminal fragment with isolated LPS and intact Proteus mirabilis and Escherichia coli. Infect Immun 62: 259265.
  • Coynault, C., Robbe-Saule, V., Norel, F. (1996) Virulence and vaccine potential of Salmonella typhimurium mutants deficient in the expression of the RpoS (sigma S) regulon. Mol Microbiol 22: 149160.
  • Creaghan, I.T. & Guest, J.R. (1978) Succinate dehydrogenase-dependent nutritional requirement for succinate in mutants of Escherichia coli K-12. J Gen Microbiol 107: 113.
  • Dukan, S. & Nyström, T. (1999) Oxidative stress defense and deterioration of growth-arrested Escherichia coli cells. J Biol Chem 274: 2602726032.
  • Elsbach, P. & Weiss, J. (1993) Bactericidal/permeability increasing protein and host defence against Gram-negative bacteria and endotoxin. Curr Opin Immunol 5: 103107.
  • Elsbach, P., Weiss, J., Franson, R.C., Becker-Quagliate, S., Schneider, A., Harris, L. (1979) Separation and purification of a potent bactericidal/permeability-increasing protein and a closely-associated phospholipase A2 from rabbit polymorphonuclear leukocytes. Observations on their relationship. J Biol Chem 254: 1100011009.
  • Farris, M., Grant, A., Richardson, T.B., O’Connor, C.D. (1998) BipA: a tyrosine-phosphorylated GTPase that mediates interactions between enteropathogenic Escherichia coli (EPEC) and epithelial cells. Mol Microbiol 28: 265279.
  • Foster, J.W. & Spector, M.P. (1995) How Salmonella survive against the odds. Annu Rev Microbiol 49: 145174.
  • Friedrich, T. (1998) The NADH: ubiquinone oxidoreductase (complex I) from Escherichia coli. Biochim Biophys Acta 1364: 134146.
  • Friedrich, T., Van Heek, P., Leif, H., Ohnishi, T., Forche, E., Kunze, B., et al. (1994) Two binding sites of inhibitors in NADH: ubiquinone oxidoreductase (complex I). Relationship of one site with the ubiquinone-binding site of bacterial glucose: ubiquinone oxidoreductase. Eur J Biochem 219: 691698.
  • Gabay, J.E., Scott, R.W., Campanelli, D., Griffith, J., Wilde, C., Marra, M.N., et al. (1989) Antibiotic proteins of human polymorphonuclear leukocytes. Proc Natl Acad Sci USA 86: 56105614.
  • Gazzano-Santoro, H., Parent, J.B., Grinna, L., Horwitz, A., Parsons, T., Theofan, G., et al. (1992) High-affinity binding of the bactericidal permeability-increasing protein and a recombinant amino-terminal fragment to the lipid-A region of lipopolysaccharide. Infect Immun 60: 47544761.
  • Gray, B.H. & Haseman, J.R. (1994) Bactericidal activity of synthetic peptides based on the structure of the 55-kilodalton bactericidal protein from human neutrophils. Infect Immun 62: 27322739.
  • Groisman, E.A., Parra-Lopez, C., Salcedo, M., Lipps, C.J., Heffron, F. (1992) Resistance to host antimicrobial peptides is necessary for Salmonella virulence. Proc Natl Acad Sci USA 1992 89: 1193911943.
  • Guo, L., Lim, K.B., Gunn, J.S., Bainbridge, B., Darveau, R.P., Hackett, M., et al. (1997) Regulation of lipid A modifications by Salmonella typhimurium virulence genes phoP-phoQ. Science 276: 250253.
  • Hengge-Aronis, R. (1996) Regulation of gene expression during entry into stationary phase. In Escherichia Coli and Salmonella: Cellular and Molecular Biology, 2nd edn. Neidhardt, F.C., Curtis, III, R., Ingraham, J.L., Lin, E.C.C., Low, K.B., Magasanik, B., et al. (eds). Washington, DC: American Society for Microbiology Press, pp. 14971512 .
  • Hengge-Aronis, R. (1999) Interplay of global regulators and cell physiology in the general stress response of Escherichia coli. Curr Opin Microbiol 2: 148152.
  • Hoess, A., Watson, S., Siber, G.R., Liddington, R. (1993) Crystal structure of an endotoxin-neutralizing protein from the horseshoe crab, Limulus anti-LPS factor, at 1.5 X resolution. EMBO J 12: 33513356.
  • Hovde, C.J. & Gray, B.H. (1986) Physiological effects of a bactericidal protein from human polymorphonuclear leukocytes on Pseudomonas aeruginosa. Infect Immun 52: 9095.
  • In’T Veld, G., Mannion, B., Weiss, J., Elsbach, P. (1988) The effects of the bactericidal/permeability increasing protein of polymorphonuclear leukocytes on isolated bacterial cytoplamic membrane vesicles. Infect Immun 56: 12031208.
  • Joiner, K.A., Ganz, T., Albert, T., Rothosen, D. (1989) The opsonizing ligand on Salmonella typhimurium influences the incorporation of specific, but not azurophil, granule constituent into neutrophil phagosomes. J Cell Biol 109: 27712782.
  • Lehrer, R.I., Barton, A., Daher, K.A., Harwig, S.S., Ganz, T., Selsted, M.E. (1989) Interaction of human defensins with Escherichia coli. Mechanism of bactericidal activity. J Clin Invest 84: 553561.
  • Leif, H., Sled, V.D., Ohnishi, T., Weiss, H., Friedrich, T. (1995) Isolation and characterization of the proton-translocating NADH: ubiquinone oxidoreductase from Escherichia coli. Eur J Biochem 230: 538548.
    Direct Link:
  • Lester, R.L. & DeMoss, J.A. (1971) Effects of molybdate and selenite on formate and nitrate metabolism in Escherichia coli. J Bacteriol 105: 10061014.
  • Li, J. & Stewart, V. (1992) Localisation of upstream sequence elements required for nitrate and anaerobic induction of fdn (formate dehydrogenase-N) operon expression in Escherichia coli K-12. J Bacteriol 174: 49354942.
  • Little, R.G., Kelner, D.N., Lim, E., Burke, D.J., Conlon, P.J. (1994) Functional domains of recombinant bactericidal/permeability-increasing protein (rBPI23). J Biol Chem 269: 18651872.
  • McCabe, K., Mann, M.D., Bowie, M.D. (1998) d-Lactate production and [14C]succinic acid uptake by adherent and nonadherent Escherichia coli. Infect Immun 66: 907911.
  • Mannion, B.A., Weiss, J., Elsbach, P. (1990) Separation of the sublethal and lethal effects of the bactericidal/permeability increasing protein on Escherichia coli. J Clin Invest 85: 853860.
  • Martin, E., Ganz, T., Lehrer, R.I. (1995) Defensins and other endogenous peptide antibiotics of vertebrates. J Leukocyte Biol 58: 128136.
  • Nyström, T. (1998) To be or not to be: the ultimate decision of the growth-arrested bacterial cell. FEMS Microbiol Rev 21: 283290.
  • Ouellette, A.J. & Luardi, J.C. (1990) A novel mouse gene family coding for cationic, cysteine-rich peptide regulation in small intestine and cells of myeloid origin. J Bacteriol 265: 98319837.
  • Parra-Lopez, C., Lin, R., Aspedon, A., Groisman, E.A. (1994) A Salmonella protein that is required for resistance to antimicrobial peptides and transport of potassium. EMBO J 13: 39643972.
  • Qi, S.-Y., Li, Y., Szyrolki, A., Giles, I.G., Moir, A., O’Connor, C.D. (1995) Salmonella typhimurium responses to a bactericidal protein from human neutrophils. Mol Microbiol 17: 523531.
  • Roland, K.L., Martin, L.E., Esther, C.R., Spitznagel, J.K. (1993) Spontaneous pmrA mutants of Salmonella typhimurium LT2 define a new two-component regulatory system with a possible role in virulence. J Bacteriol 175: 41544164.
  • Sawers, G. (1994) The hydrogenases and formate dehydrogenases of Escherichia coli. Antonie Van Leeuwenhoek 66: 5788.
  • Sawers, G., Heider, J., Zehelein, E., Böck, A. (1991) Expression and operon structure of the sel genes of Escherichia coli and identification of a third selenium-containing formate dehydrogenase isoenzyme. J Bacteriol 173: 49834993.
  • Shafer, W.M., Martin, L.E., Spitznagel, J.K. (1984) Cationic antimicrobial proteins isolated from human neutrophil #granulocytes in the presence of diisopropyl fluorophosphate. Infect Immun 45: 2935.
  • Stolzenberg, E.D., Anderson, G.M., Ackermann, M.R., Whitlock, R.H., Zasloff, M. (1997) Epithelial antibiotic induced in states of disease. Proc Natl Acad Sci USA 94: 86868690.
  • Tran, Q.H., Bongaerts, J., Vlad, D., Unden, G. (1997) Requirement for the proton-pumping NADH dehydrogenase I of Escherichia coli in respiration of NADH to fumarate and its bioenergetic implications. Eur J Biochem 244: 155160.
  • Tran, Q.H. & Unden, G. (1998) Changes in the proton potential and the cellular energetics of Escherichia coli during growth by aerobic and anaerobic respiration or by fermentation. Eur J Biochem 251: 538543.
  • Weiss, J., Muello, K., Victor, M., Elsbach, P. (1984) The role of lipopolysaccharides in the action of the bactericidal/permeability-increasing protein on the bacterial cell envelope. J Immunol 132: 31093115.
  • Weiss, J., Schmeidler, K., Beckerdite-Quagliata, S., Franson, R.C., Elsbach, P. (1976) Reversible envelope effects during and after killing of Escherichia coli by a highly purified rabbit polymorphonuclear leukocyte fraction. Biochim Biophys Acta 436: 154169.
  • Wray, C. & Sojka, W.J. (1978) Experimental Salmonella typhimurium in calves. Res Vet Sci 25: 139146.
Footnotes
  1. †Present address: Department of de Nutricón y Bromatología III, Universidad Complutense, Madrid 28040, Spain.