Effects of chromosomal deletion of the operon encoding the multiple resistance and pH-related antiporter in Vibrio cholerae

Abstract

To examine the possible physiological significance of Mrp, a multi-subunit cation/proton antiporter from Vibrio cholerae, a chromosomal deletion Δmrp of V. cholerae was constructed and characterized. The resulting mutant showed a consistent early growth defect in LB broth that became more evident at elevated pH of the growth medium and increasing Na⁺ or K⁺ loads. After 24 h incubation, these differences disappeared likely due to the concerted effort of other cation pumps in the mrp mutant. Phenotype MicroArray analyses revealed an unexpected systematic defect in nitrogen utilization in the Δmrp mutant that was complemented by using the mrpA′-F operon on an arabinose-inducible expression vector. Deletion of the mrp operon also led to hypermotility, observable on LB and M9 semi-solid agar. Surprisingly, Δmrp mutation resulted in wild-type biofilm formation in M9 despite a growth defect but the reverse was true in LB. Furthermore, the Δmrp strain exhibited higher susceptibility to amphiphilic anions. These pleiotropic phenotypes of the Δmrp mutant demonstrate how the chemiosmotic activity of Mrp contributes to the survival potential of V. cholerae despite the presence of an extended battery of cation/proton antiporters of varying ion selectivity and pH profile operating in the same membrane.

Introduction

Vibrio cholerae, the causative agent of cholera, causes severe watery diarrhoea in humans. The pathogen is highly adaptable to changing ion concentrations, pH and osmolarities by a variety of molecular mechanisms that enable V. cholerae to adapt to fluctuations in its natural environment, brackish and estuarian waters (Miller et al., 1984; Singleton et al., 1982a, b), as well as to colonize the gastroenteric tract of the human host (Häse & Mekalanos, 1999). The sodium cycle in energetic exchange seems to be one of the mechanisms boosting the survival potential of V. cholerae. It includes a number of primary (e.g. NQR) and secondary (Na⁺/H⁺ antiporters) sodium pumps that remove Na⁺ ions from the cytoplasm, creating the sodium motive force (SMF), which energizes a plethora of SMF consumers performing osmotic and mechanical work (Habibian et al., 2005; Häse & Mekalanos, 1999; Häse et al., 2001; Dibrov et al., 2005; Resch et al., 2010; Minato et al., 2014).

Mrp (multiple resistance and pH-related antiporter) is a multi-subunit cation/proton antiporter (Krulwich et al., 2009) residing in the membrane of V. cholerae cells, exchanging cytoplasmic Na⁺, Li⁺ or K⁺ ions for extracellular H⁺ electrogenically with an extremely alkaline maximum of activity at pH 9.0–9.5. Kinetic analyses suggested that the Na⁺(Li⁺)/H⁺ antiport rather than the K⁺/H⁺ antiport should be a physiological mode of Mrp activity (Dzioba-Winogrodzki et al., 2009). When the cloned Mrp was expressed in trans in an antiport-deficient Escherichia coli strain, EP432, it protected against sodium and lithium ions; this protection increased considerably as the pH of the media was progressively raised from 6.5 to 8.5 (Dzioba-Winogrodzki et al., 2009). In addition, at neutral pH, Mrp protected EP432 transformants against natural bile salts such as sodium cholate and taurocholate (Dzioba-Winogrodzki et al., 2009). However, the potential physiological role(s) of Mrp in its native host, V. cholerae, remained obscure. Here we characterize the unique antiporter, Mrp, from V. cholerae. We discovered that the role of Mrp in V. cholerae physiology is more complicated than was previously displayed in an antiport-deficient E. coli strain. Loss of Mrp altered V. cholerae growth and motility in rich and minimal media, caused a systematic defect in nitrogen utilization and resulted in an increased sensitivity to weak acids. This study is one step in understanding the complex antiporter systems in V. cholerae.

Methods

Bacterial strains, plasmids and routine culture conditions used in this study.

Bacterial strains and plasmids used in this study are referenced in Table S1 (available in the online Supplementary Material). All strains were maintained at −80 °C until use. Single colonies were pre-cultured in LB Lennox broth (Difco) overnight at 37 °C in a roller drum (New Brunswick Scientific set to 8) prior to the assays. Antibiotics were used at the following concentrations: streptomycin, 100 µg ml⁻¹; ampicillin, 100 µg ml⁻¹; kanamycin, 50 µg ml⁻¹.

Mutant construction.

Deletion of the Mrp operon mrpA′-CDEFG (VCA0157-VCA0152) in V. cholerae O395N1 was constructed using overlap extension PCR and double homologous recombination as described previously (Donnenberg & Kaper, 1991; Ho et al., 1989). Briefly, genomic DNA from V. cholerae O395N1 was isolated using DNeasy Blood and Tissue kit (Qiagen). Primers used to construct this mutant are listed in Table S1. Regions 1 kb upstream (primers ‘MRP del 1’ and ‘MRP del 2’) and downstream (primers ‘MRP del 3’ and ‘MRP del 4’) of the targeted locus were PCR amplified. Primers ‘MRP del 2’ and ‘MRP del 3’ have complementary BglII sites engineered into the primers to allow combination of the two fragments. The mrp upstream and downstream PCR fragments were templates in a second round of PCR with primers ‘MRP del 1’ and ‘MRP del 4’. This resulted in a PCR fusion product of the mrpA′-CDEFG operon flanking regions, creating a six-gene-deletion fragment, Δmrp. To create the deletion construct, the Δmrp fragment was cloned into pWM91, an allelic exchange suicide vector containing an ampicillin resistance marker, the R6Kγ origin of replication that requires the product of the pir gene for replication and the sacB gene for sucrose counter-selection (Metcalf et al., 1996). The deletion construct, pWM91-Δmrp, was then transformed into E. coli DH5αλpir for propagation. The pWM91-Δmrp construct was isolated and transformed into SM10λpir for conjugation with wild-type V. cholerae O395N1. V. cholerae O395N1 ex-conjugates were selected on LB ampicillin agar for plasmid integration and counter-selected on LB 5 % sucrose agar lacking NaCl [10 g l⁻¹ tryptone (Difco), 5 g l⁻¹ yeast extract (EMD) and 50 g l⁻¹ sucrose (Sigma)] for excision of the plasmid. PCR using primers ‘MRP del 1’ and ‘MRP del 4’ was used to screen colonies for loss of the mrpA′-CDEFG operon.

Construction of complementation strains.

The V. cholerae complementation strains and their vectors are listed in Table S1. V. cholerae wild-type and ∆mrp strains were cultured in LB overnight at 37 °C with aeration. The overnight cultures were diluted in fresh LB and grown to mid-log phase. Cells were washed three times with ice-cold 1 mM CaCl₂. Electroporation was used to introduce the plasmids into V. cholerae wild-type and Δmrp strains using 0.2 cm Gene Pulser cuvettes (BioRad) in the MicroPulser Electroporator (BioRad) on the EC2 setting (2.5 kV). Transformants were recovered on LB ampicillin agar, and the presence of the mrpA′-G operon in the Δmrp complemented strain was confirmed using PCR (Δmrp pVcMrp).

LB growth assays.

Overnight pre-cultures of the V. cholerae strains were diluted to an OD₅₉₅ of 0.1 in either LB− [10 g l⁻¹ tryptone (Difco) and 5 g l⁻¹ yeast extract (EMD), lacking NaCl] or LBB− [LB− buffered with 60 mM BIS-TRIS propane (Sigma)] at pH 8.5, 7.5, 6.5 and 5.5 and in the presence of NaCl, KCl and LiCl at final concentrations of 0.1 (only NaCl), 1, 2 or 3 % (w/v). A total of 200 µl of each strain were transferred in triplicate to a 96-well tissue culture plate (Nunc) and incubated at 37 °C on an orbital shaker (Thermo Scientific MaxQ 4000) set to 250 r.p.m. Growth measurements (OD₅₉₅) were recorded with the iMark plate reader (BioRad). The data represent at least three independent experiments.

Calculating doubling times.

The growth measurements (OD₅₉₅) for each V. cholerae strain were plotted individually and fit to an exponential curve to determine the maximal growth rate. The doubling time for each strain under each condition was calculated from the maximal growth rate by applying a least square fitting exponential method (http://mathworld.wolfram.com/LeastSquaresFittingExponential.html). The results presented are the mean±standard error of three independent cultures.

Biolog Phenotype MicroArray analyses.

Biolog Phenotype MicroArray (PM) analyses were conducted with V. cholerae O395N1 wild-type and Δmrp strains using PM1–PM20 as previously described (Bochner et al., 2001; Zhou et al., 2003). The strains were sent to Biolog (Hayward, CA) for PM testing of panels 1–20. PM panels 1–10 are metabolic tests: PM1 and PM2A, carbon; PM3B, nitrogen; PM4A, phosphorus and sulfur; PM5, nutrient supplements; PM6–PM8, peptide nitrogen; PM9, osmolytes and PM10, pH. PM panels 11–20 test chemical sensitivities. Gained or lost phenotypes are scored based on the difference in cellular respiration between the V. cholerae wild-type and Δmrp strains. A negative score indicates a lost phenotype and a positive score indicates a gained phenotype.

M9 growth assays.

Overnight LB pre-cultures of the V. cholerae strains were pelleted in a microcentrifuge (Beckman Coulter Microfuge 18 centrifuge) at 5000 gfor 5 min. Pellets were washed once in M9− (12.8 g l⁻¹ Na₂HPO₄·7H₂O, 3 g l⁻¹ KH₂HPO₄ and 0.5 g l⁻¹ NaCl, with no carbon or nitrogen source), re-pelleted by centrifugation and then re-suspended in M9−. To compare different nitrogen sources, strains were diluted to an OD₅₉₅ of 0.1 in M9− supplemented with 20 mM sodium succinate (Sigma), 0.1 mM CaCl₂, 2 mM MgSO₄ and either 1 g l⁻¹ (~19 mM) NH₄Cl, 5 mM glycine (Sigma), 5 mM asparagine (Sigma), 5 mM thiamine (Sigma), 5 mM glutamine (Sigma), 5 mM methionine (Sigma), 5 mM l-glutamic acid (Sigma), 5 mM serine (Sigma), 5 mM N-acetyl-d-glucosamine (Sigma) or 5 mM l-arginine (Sigma) as the sole nitrogen source or in M9− supplemented with 20 mM sodium succinate (Sigma), 0.1 mM CaCl₂, 2 mM MgSO₄, 1 g l⁻¹ NH₄Cl and either 5 mM l-glutamic acid (Sigma), 5 mM asparagine (Sigma), 5 mM glycine (Sigma) or 5 mM thiamine (Sigma) as additional nitrogen sources. For the carbon source assays, strains were diluted to an OD₅₉₅ of 0.2 in M9 (containing 1 g l⁻¹ NH₄Cl as the nitrogen source) supplemented with 0.1 mM CaCl₂, 2 mM MgSO₄ and either 20 mM sodium succinate (Sigma), 20 mM sodium fumarate (Alfa Aesar), 20 mM sodium pyruvate (Alfa Aesar) or 20 mM d-glucose (Sigma) with 20 mM NaCl to account for the additional sodium conjugated with the other carbon sources. For the pH assays, strains were diluted to OD₅₉₅ of 0.1 in 50 mM phosphate-buffered M9 supplemented with 20 mM sodium succinate, 0.1 mM CaCl₂ and 2 mM MgSO₄ at pH 7.5, 6.5, 5.5 or 4.5 and either 1 g l⁻¹ (~19 mM) NH₄Cl or 2.52 g l⁻¹ (~19 mM) l-glutamine (Sigma) as the sole nitrogen source. Each condition had an additional 1, 2 and 3 % (w/v) NaCl, KCl or LiCl added. For the initial experiments to confirm the nitrogen source growth defect, the strains were grown at 37 °C in test tubes containing 5 ml of the media on a roller drum (New Brunswick Scientific set to 8) and an OD₆₀₀ measurement was taken after 24 h incubation. Once we had determined that the cultures would sufficiently grow in a 200 µl volume on a 96-well plate format, all subsequent experiments were conducted in 96-well tissue culture plates (Nunc). The carbon source plates were incubated at room temperature (~23–24 °C) on the iMark plate reader (BioRad) set to mix on medium speed and measurements (OD₅₉₅) were taken every 15 min for 24 h. The nitrogen and pH assays were incubated for 24 h at 37 °C on an orbital shaker (Thermo Scientific MaxQ 4000) set to 250 r.p.m. and growth was measured at OD₅₉₅ on the iMark plate reader (BioRad). The data represent the mean±standard error of at least three independent experiments.

Motility.

The LB− and LBB− soft agar motility plates were made with 10 g l⁻¹ tryptone, 5 g l⁻¹ yeast extract and 0.3 % (w/v) agar at pH 5.5, 6.5, 7.5 and 8.5. The LBB− plates were buffered with 60 mM BIS-TRIS propane (Sigma). The M9 soft agar motility plates were made with 12.8 g l⁻¹ Na₂HPO₄·7H₂O, 3 g l⁻¹ KH₂PO₄, 0.5 g l⁻¹ NaCl, 1 g l⁻¹ NH₄Cl, 0.1 mM CaCl₂, 2 mM MgSO₄, 20 mM sodium succinate (Sigma) and 0.3 % (w/v) agar at pH 5.5, 6.5 or 7.5. For complementation, 0.2 % arabinose (w/v) (Sigma) and 100 µg ml⁻¹ ampicillin (Sigma) were added to the soft agar for induction of the mrpA′-G operon on the pBAD vector. Single colonies were picked from overnight cultures grown on 1.5 % agar plates and stabbed into the soft agar motility plates using a 10 µl pipette tip. Plates were photographed after 24 h incubation at 37 °C and analysed with ImageJ (http://imagej.nih.gov/ij). Area of growth was measured in pixels thrice. Results are the mean±standard error of at least four independent replicates.

Biofilms.

Overnight cultures of the V. cholerae strains were pelleted by centrifugation at 5000 g for 5 min, washed once in M9− (12.8 g l⁻¹ Na₂HPO₄·7H₂O, 3 g l⁻¹ KH₂PO₄ and 0.5 g l⁻¹ NaCl, no carbon or nitrogen source) and re-pelleted. The washed pellets were re-suspended to an OD₅₉₅ of 0.1 in 50 mM phosphate-buffered M9− containing 20 mM sodium succinate, 0.1 mM CaCl₂ and 2 mM MgSO₄ at pH 4.5, 5.5, 6.5 or 7.5 with either 1 g l⁻¹ NH₄Cl or 2.52 g l⁻¹ l-glutamine (Sigma) as the sole nitrogen source. Then, 200 µl of each strain was transferred in triplicate into a 96-well tissue culture plate (Nunc) and incubated stationary at 37 °C for 24 h. Growth measurements (OD₅₉₅) were taken using the iMark plate reader (BioRad). Measurement of biofilm formation was adapted from previous publications (Watnick et al., 1999; Watnick & Kolter, 1999). Plates were washed three times with DI water and left to dry; 250 µl of 0.1 % (w/v) crystal violet stain was added to each well and incubated on an orbital shaker (Thermo Scientific MaxQ 4000) set to 250 r.p.m. for 30 min at room temperature. Excess crystal violet stain was washed off using DI water and wells were allowed to dry; 300 µl of DMSO was added to each well and incubated on an orbital shaker (Thermo Scientific MaxQ 4000) set to 250 r.p.m. for 30 min at room temperature prior to reading. Biofilm formation measurements were taken on the iMark plate reader (BioRad) at OD₅₇₀. Results represent the mean±standard error of three independent experiments.

Weak acid resistance.

Overnight cultures of V. cholerae strains were diluted to an OD₅₉₅ of 0.1 in LB alone and LB containing twofold serial dilutions of either dioctyl sodium sulphosuccinate (Sigma), n-lauroylsarcosine (Sigma), probenecid (Sigma) or sodium cholate (Sigma). A total of 200 µl of each diluted strain were added in triplicate to a 96-well tissue culture plate (Nunc) and incubated on an orbital shaker (Thermo Scientific MaxQ 4000) set to 250 r.p.m. at 37 °C for 24 h. End point growth measurements (OD₅₉₅) are obtained using the iMark plate reader (BioRad). Results presented are the mean±standard error of at least three independent experiments.

Statistical analyses.

GraphPad Prism 7.0 was used for all statistical analyses. A two-tailed, unpaired Student’s t-test was used for pairwise comparisons of V. cholerae wild-type and Δmrp strains. Two-way ANOVA with Tukey’s multiple comparisons test was used to analyse V. cholerae wild-type pBAD, Δmrp pBAD and Δmrp pVcMrp complementation results. P values of <0.05 were considered statistically significant.

Results

Effects of pH and increasing NaCl, KCl and LiCl concentrations on growth of V. cholerae Δmrp mutant in LB

Previously we demonstrated that the V. cholerae Mrp complex transported Na⁺, Li⁺ and K⁺ at an optimal pH of 9.0–9.5 (Dzioba-Winogrodzki et al., 2009). Based on these findings, we constructed a V. cholerae Δmrp chromosomal deletion and tested this strain’s sensitivity to various concentrations of Na⁺, Li⁺ and K⁺ in LB at pH ranging from acidic to alkaline. In LBB− (LB− buffered with 60 mM BIS-TRIS propane with no added NaCl, KCl or LiCl), the Δmrp mutant had a modest growth defect at pH 8.5 (Table 1, Fig. S1a). Increasing amounts of NaCl caused the Δmrp strain to have an observable growth defect at pH 8.5, 7.5, 6.5 and 5.5. When KCl was added in increasing amounts, a similar trend was observed where the Δmrp strain had a slight growth defect at pH 8.5, 7.5 and 6.5, but not at pH 5.5 even with 3 % KCl (Table 1, Fig. S1b). When 1 % LiCl was added, the Δmrp strain had a dramatic growth defect at pH 8.5 and 5.5 (Table 1, Fig. S1c). The presence of 2 % LiCl in the medium severely inhibited growth of both V. cholerae wild-type and Δmrp strains at pH 8.5. At pH 7.5, 6.5 and 5.5, both strains were able to grow in the 24 h incubation but the Δmrp strain had a growth defect at pH 5.5 and a significant growth defect at pH 6.5 (Table 1). Both strains were unable to grow at pH 8.5, 7.5 and 5.5 with 3 % LiCl in the growth medium; however, the wild-type was able to grow at pH 6.5 but not the Δmrp mutant.

Table 1. Doubling time of wild-type and Δmrp strains in LBB− supplemented with different concentrations of NaCl, KCl and LiCl.

Growth measurements (OD₅₉₅) of each strain were plotted and doubling times were calculated from exponential rate of growth for each strain. Asterisks indicate statistical significance with wild-type (*P<0.05, **P<0.01 and ***P<0.001). A ‘–’ indicates no growth during the course of the experiment.

LBB− media	Strain	Doubling time (h)†
		pH 8.5	pH 7.5	pH 6.5	pH 5.5
0 %	Wild-type	1.59±0.16	1.88±0.01	1.78±0.17	3.23±1.19
	Δmrp	2.04±0.20*	1.85±0.03	1.99±0.11	3.47±0.56
0.1 % NaCl	Wild-type	1.63±0.14	1.45±0.08	1.33±0.04	2.05±0.10
	Δmrp	2.12±0.25*	1.61±0.08	1.64±0.17*	2.46±0.19*
1 % NaCl	Wild-type	1.45±0.12	1.21±0.07	1.31±0.06	1.56±0.08
	Δmrp	1.83±0.18*	1.64±0.21*	1.67±0.17*	1.99±0.20*
2 % NaCl	Wild-type	1.46±0.03	1.48±0.04	1.52±0.12	1.88±0.06
	Δmrp	1.73±0.03***	1.82±0.08**	1.79±0.08*	2.24±0.14*
3 % NaCl	Wild-type	1.66±0.04	1.53±0.05	1.86±0.11	2.05±0.10
	Δmrp	2.01±0.19*	1.82±0.17*	1.92±0.10	2.45±0.29*
1 % KCl	Wild-type	1.30±0.21	1.47±0.08	1.70±0.04	2.22±0.16
	Δmrp	1.86±0.19*	1.95±0.17*	1.92±0.10*	2.26±0.29
2 % KCl	Wild-type	1.37±0.11	1.44±0.09	1.39±0.14	2.18±0.16
	Δmrp	1.94±0.11**	1.84±0.11**	1.87±0.11**	2.07±0.12
3 % KCl	Wild-type	1.55±0.11	1.40±0.08	1.37±0.12	2.49±0.18
	Δmrp	1.92±0.19*	1.72±0.11*	1.94±0.01**	2.64±0.18
1 % LiCl	Wild-type	5.30±0.53	2.29±0.33	1.67±0.15	2.07±0.23
	Δmrp	31.66±4.59***	2.66±0.49	1.88±0.24	3.00±0.51*
2 % LiCl	Wild-type	12.92±1.39	4.41±1.69	1.79±0.15	6.51±4.04
	Δmrp	25.54±15.70	3.45±0.59	2.29±0.12*	18.84±21.28
3 % LiCl	Wild-type	–	–	6.33±2.71	–
	Δmrp	–	–	–	–

†Results presented are the mean±standard error of three independent experiments.

Since the Δmrp mutant exhibited a growth defect in LBB− at pH ranging from alkaline to acidic with increasing amounts of cations, we explored whether unbuffering the pH of the medium would abolish this growth phenotype. Therefore, we repeated the growth data using unbuffered LB at pH 8.5, 7.5, 6.5 and 5.5 with increasing amounts of NaCl, KCl and LiCl (Table 2, Fig. S2). No growth defect was observed by the Δmrp strain in LB− (unbuffered containing no cations). Interestingly, when 0.1 % NaCl was added, the Δmrp strain only had a growth defect at pH 5.5 (Table 2, Fig. S2a). The Δmrp strain also had a growth defect only at pH 6.5 with 1 % NaCl and at pH 8.5 with 3 % NaCl. On the contrary, addition of 1, 2 and 3 % KCl resulted in a significant growth defect at pH 8.5, 7.5 and 6.5 for the Δmrp strain (Table 2, Fig. S2b). Unsurprisingly, increasing amounts of LiCl had a negative effect on growth for both wild-type and Δmrp strains; however, growth overall was slightly better in LB− compared to LBB− with LiCl (Figs S1c and S2c). Again the Δmrp strain had a growth defect at pH 8.5 with 1 % LiCl but not at pH 5.5 as was seen in LBB− (Tables 1 and 2). With 2 % LiCl, the Δmrp strain only had a growth defect at pH 8.5. For 3 % LiCl, both strains grew slightly at pH 8.5 and 7.5 but not enough to calculate a doubling time (Table 2). The growth data suggest that the Δmrp strain has a growth defect at pH 6.5 with 3 % LiCl; however, it was not statistically significant. Lastly, only the wild-type strain grew sufficiently to calculate a doubling time in 3 % LiCl at pH 5.5. Taken together, these data suggest that the Δmrp strain has a transient growth defect under these conditions that is presumably overcome by the numerous cation antiporters present in V. cholerae.

Table 2. Doubling time of wild-type and Δmrp strains in LB− supplemented with different concentrations of NaCl, KCl and LiCl.

Growth measurements (OD₅₉₅) of each strain were plotted and doubling times were calculated from exponential rate of growth for each strain. Asterisks indicate statistical significance with wild-type (*P<0.05 and **P<0.01). A ‘–’ indicates no growth during the course of the experiment.

LB− media	Strain	Doubling time (h)†
		pH 8.5	pH 7.5	pH 6.5	pH 5.5
0 %	Wild-type	1.96±0.12	2.40±0.21	3.38±1.19	8.10±1.56
	Δmrp	1.99±0.16	2.23±0.11	2.99±0.81	12.49±0.80
0.1 % NaCl	Wild-type	1.75±0.06	1.97±0.17	2.60±0.27	4.26±1.80
	Δmrp	1.82±0.23	1.91±0.24	2.48±0.33	10.03±1.59*
1 % NaCl	Wild-type	1.47±0.16	1.29±0.02	1.35±0.02	1.99±0.08
	Δmrp	1.51±0.02	1.45±0.12	1.60±0.16*	2.00±0.18
2 % NaCl	Wild-type	1.58±0.05	1.60±0.08	1.49±0.12	1.85±0.08
	Δmrp	1.57±0.03	1.50±0.07	1.62±0.14	1.94±0.12
3 % NaCl	Wild-type	1.75±0.13	1.74±0.12	2.18±0.07	2.26±0.21
	Δmrp	1.98±0.06*	1.87±0.02	2.03±0.16	2.20±0.14
1 % KCl	Wild-type	1.47±0.17	1.62±0.19	1.58±0.18	2.08±0.02
	Δmrp	1.78±0.08*	2.13±0.02**	1.96±0.15*	2.19±0.13
2 % KCl	Wild-type	1.42±0.11	1.38±0.04	1.38±0.08	1.89±0.11
	Δmrp	1.67±0.11*	1.73±0.11**	1.86±0.11**	1.96±0.13
3 % KCl	Wild-type	1.43±0.07	1.28±0.14	1.39±0.16	2.01±0.08
	Δmrp	1.65±0.10*	1.75±0.05**	1.82±0.06*	2.03±0.13
1 % LiCl	Wild-type	3.85±1.03	2.58±0.10	1.84±0.15	2.03±0.07
	Δmrp	9.04±0.32**	2.97±0.61	1.72±0.06	2.49±0.32
2 % LiCl	Wild-type	6.14±1.18	2.72±0.54	1.98±0.12	2.82±0.67
	Δmrp	16.12±4.87*	3.02±0.24	1.96±0.31	3.10±0.36
3 % LiCl	Wild-type	–	–	3.07±0.58	10.51±0.31
	Δmrp	–	–	25.41±30.04	–

†Results presented are the mean±standard error of three independent experiments.

Biolog PM

Since the Δmrp strain had a modest growth phenotype in LB, we chose to further characterize the physiological roles of Mrp by testing the strain in growth-independent PM. The V. cholerae wild-type and Δmrp strains were sent to Biolog for comparative PM analyses using PM panels 1–20, which included metabolic and chemical sensitivity tests. Surprisingly, the Δmrp mutant displayed a systematic defect in nitrogen utilization (Table S2). To investigate this nitrogen utilization defect and confirm the PM results, we performed growth assays using minimal M9 medium supplemented with 20 mM sodium succinate and various nitrogen sources. The Δmrp cells grown with glycine, asparagine, thiamine, serine or ammonium chloride as the sole nitrogen source, or the combination of ammonium chloride with glycine, asparagine and thiamine, had a significant growth defect (Figs 1a and S3). We did observe that not all nitrogen sources indicated as lost phenotypes from the PM analyses, which are growth independent, were confirmed in growth-dependent assays (Fig. S3). When the mrpA′-G operon was reintroduced into the Δmrp strain on the pBAD complementation vector, growth was restored to wild-type level (Fig. 1b), thus confirming that the deletion of the mrp operon has an effect on the efficiency of utilization of nitrogen in V. cholerae. To exclude the possibility that the observed growth defect was the result of growth on succinate and not nitrogen utilization, we examined the growth of the Δmrp strain in M9 media with ammonium chloride as the nitrogen source, and with glucose, fumarate, pyruvate or succinate as the carbon source (Table 3, Fig. S4). The Δmrp strain had significant growth defects in M9 supplemented with fumarate, pyruvate and glucose similar to what was observed for succinate. This excludes the possibility that the growth phenotype observed by the Δmrp mutant is a result of a succinate defect, and further supports the Biolog PM analyses of a nitrogen utilization defect.

Fig. 1.

Loss of mrp results in a V. cholerae nitrogen utilization defect in minimal media. V. cholerae strains were grown overnight at 37 °C in LB broth and then diluted to OD₆₀₀=0.1 in M9 minimal media. Cultures were incubated with aeration at 37 °C for 24 h. Results presented are the mean±standard error of two independent experiments. Asterisks indicate statistical significance (*P<0.05, **P<0.01, ***P<0.001 and ****P<0.0001). (a) M9 minimal media containing either a single nitrogen source (19 mM NH₄Cl, 5 mM glycine, 5 mM asparagine or 5 mM thiamine) or 19 mM NH₄Cl with an additional nitrogen source (5 mM glycine, 5 mM asparagine or 5 mM thiamine); (b) complementation of the nitrogen utilization defect.

Table 3. Doubling time of wild-type and Δmrp strains in M9 minimal media supplemented with different carbon sources.

Growth measurements (OD₅₉₅) of each strain were plotted and doubling times were calculated from exponential rate of growth for each strain. Asterisks indicate statistical significance with wild-type (**P<0.01).

Strain	Doubling time (h)†
	Succinate	Fumarate	Pyruvate	Glucose
Wild-type	14.38±0.26	11.97±0.47	9.78±0.15	7.55±0.26
Δmrp	23.83±1.57**	16.29±0.28**	22.14±3.9**	10.17±0.59**

†Results presented are the mean±standard error of three independent experiments.

Effects of pH and nitrogen source

To further understand the Δmrp-dependent nitrogen utilization defect, we compared growth of the V. cholerae wild-type and Δmrp strains in M9 with two different nitrogen sources, ammonium chloride and glutamine, at pH 7.5, 6.5, 5.5 and 4.5 after 24 and 48 h growth (Fig. 2). pH 8.5 was not tested due to precipitation of the in media at this pH (unpublished observation). For both nitrogen sources at pH 7.5 and 6.5, the Δmrp strain had a significant growth defect at 24 h incubation. However, after 48 h incubation, the Δmrp strain achieved wild-type growth (Fig. 2a, b, e, f). Surprisingly, the Δmrp mutant growth achieved or even surpassed the wild-type strain at pH 4.5 and 5.5, respectively, so no growth difference was observed after 48 h incubation (Fig. 2c, d, g, h). These data suggest that the M9 nitrogen utilization defect is pH dependent.

Fig. 2.

The Δmrp nitrogen source utilization defect is dependent on pH. V. cholerae strains were grown overnight at 37 °C in LB broth, washed once with M9− and then diluted to OD₅₉₅=0.1 in 200 µl of M9− minimal media (12.8 g l⁻¹ Na₂HPO₄·7H₂O, 3 g l⁻¹ KH₂HPO₄, 0.5 g l⁻¹ NaCl, 0.1 mM CaCl₂, 2 mM MgSO₄ and 20 mM succinate) containing either (a–d) 1 g l⁻¹ (~19 mM) NH₄Cl or (e–h) 19 mM glutamine as the sole nitrogen source at pH 7.6, 6.5, 5.5 and 4.5 on a 96-well plate. Cultures were incubated with aeration at 37 °C for 48 h and measurements (OD₅₉₅) were taken after 24 and 48 h incubation. Results presented are the mean±standard error of three independent experiments. Asterisks indicate statistical significance compared to wild-type (*P<0.05, **P<0.01, and ****P<0.0001).

Next we determined whether increasing amounts of NaCl, KCl and LiCl at pH 7.5, 6.5, 5.5 and 4.5 would result in the Δmrp strain no longer achieving wild-type growth by (1) 48 h at pH 7.5 and 6.5 and/or (2) at pH 5.5 and 4.5 at 24 and 48 h growth. Increasing amounts of NaCl (1–3 %) caused a growth defect for the Δmrp strain at 24 and 48 h at pH 4.5 and 5.5, for both nitrogen sources (Table S3). In general, when 2 and 3 % NaCl was added to M9, the Δmrp strain was unable to achieve wild-type growth even after 96 h incubation when grown with glutamine or ammonium chloride as the nitrogen source. The addition of 1 % NaCl was not enough to prevent the Δmrp strain from achieving wild-type growth by 48 h at pH 7.5 and 6.5, but adding 2 and 3 % NaCl inhibited the mutant for up to 96 h for both nitrogen sources. A similar trend was observed when adding 1, 2 and 3 % KCl to the M9 media at pH 7.5, 6.5, 5.5 and 4.5 (Table S4). For LiCl, only 1 % was tested due to toxicity (Table S5). Growth was severely inhibited for both strains at acidic pH with 1 % LiCl, so it is difficult to conclude anything from these data. However, at pH 7.5 and 6.5 with ammonium chloride as the nitrogen source, the presence of 1 % LiCl did prevent the Δmrp strain from reaching wild-type growth by 48 h, but the strain was able to overcome the LiCl toxicity and grow by 72 h incubation (Table S5). Surprisingly, 1 % LiCl completely inhibited growth of the Δmrp strain at pH 7.5 and 6.5 when grown with glutamine as a nitrogen source for 96 h of the experiment. In general, increasing amounts of cations NaCl, KCl and LiCl inhibited Δmrp growth in M9 media, even under conditions when the growth defect was absent.

Effects on motility

The flagellar motor of V. cholerae is powered by the SMF (Häse & Mekalanos, 1999; Kojima et al., 1999) and mediates a massive Na⁺ influx into the cytoplasm (Magariyama et al., 1994). Thus, the rate of flagellar rotation is probably highly sensitive to the magnitude of the SMF on the membrane that is created by a well-balanced operation of primary and secondary sodium pumps, such as Na⁺/H⁺ antiporters. Therefore, we examined the motility of V. cholerae wild-type and Δmrp strains to determine whether loss of the Mrp antiporter affects bacterial motility. Since the Δmrp strain had the most dramatic growth phenotypes in M9 media, we chose to examine motility in M9 media with two different nitrogen sources and at pH 7.5, 6.5 and 5.5. Unexpectedly, the Δmrp strain displayed a general increase in swarm circle diameter compared with the wild-type strain for both nitrogen sources, ammonium chloride (Fig. S5a–c) and glutamine (Fig. S5d–f) at pH 7.5, 6.5 and 5.5. Similar to the growth data, after 48 h incubation both strains had the same swarm circle diameters. Addition of the pVcMrp vector did reduce the swarm circle diameter to either at or below the wild-type level at 24 and 48 h incubation (Fig. 3c, d). Since we unexpectedly observed hypermotility by the Δmrp strain in M9 media conditions that resulted in a growth defect, we examined motility of the Δmrp strain in LBB− and LB− at pH 8.5, 7.5, 6.5 and 5.5. After 24 h incubation, the Δmrp strain had swarm circle diameters that were at or above wild-type in LBB− (Fig. 3a) and LB− (Fig. 3b). Addition of the pVcMrp vector complemented this phenotype in the Δmrp strain. These data suggest that loss of Mrp causes a slight hypermotility phenotype.

Fig. 3.

Loss of mrp causes increased swimming motility in soft agar. V. cholerae strains were grown overnight at 37 °C on LB agar. Single colonies were stabbed using a 10 µl pipette tip into (a) LBB− soft agar (10 g l⁻¹ tryptone, 5 g l⁻¹ yeast extract, 0.3 % agar, 60 mM BIS-TRIS propane, 0.2 % arabinose and 200 µg ml⁻¹ ampicillin) and (b) LB− soft agar (10 g l⁻¹ tryptone, 5 g l⁻¹ yeast extract, 0.3 % agar, 0.2 % arabinose and 200 µg ml⁻¹ ampicillin) at pH 8.5, 7.5, 6.5 and 5.5 and incubated for 24 h. Single colonies were stabbed using a 10 µl pipette tip into M9 minimal media soft agar (12.8 g l⁻¹ Na₂HPO₄·7H₂O, 3 g l⁻¹ KH₂HPO₄, 0.5 g l⁻¹ NaCl, 1 g l⁻¹ NH₄Cl, 0.1 mM CaCl₂, 2 mM MgSO₄, 20 mM succinate, 0.3 % agar, 0.2 % arabinose and 200 µg ml⁻¹ ampicillin) and incubated for (c) 24 or (d) 48 h. Soft agar plates were incubated at 37 °C for the indicated times and pictures were taken. Results presented are the mean±standard error of at least three independent experiments. Asterisks indicate statistical significance (*P<0.05, **P<0.01 and ***P<0.001).

Biofilm formation

Both motility and growth are important factors in biofilm formation. Based on our motility results, we examined biofilm formation in M9 at pH 7.5, 6.5, 5.5 and 4.5 with either l-glutamine or ammonium chloride as the nitrogen source. Since the Δmrp mutant has a growth defect under these conditions (Fig. 2), growth (as OD₅₉₅) as well as biofilm formation (as OD₅₇₀) were measured. In accord with our previous findings (Fig. 2), the mutant had a significant growth defect under all conditions after 24 h standing culture (Fig. 4a). Nevertheless, it had wild-type levels of biofilm formation for all conditions tested (Fig. 4b). It is important to note that static cultures of the Δmrp mutant at pH 5.5 and 4.5 have a growth defect that is not present when incubated with aeration (Fig. 2). It appears that loss of the Mrp antiporter results in wild-type level biofilm formation despite the defect of growth, possibly due to the hypermotility phenotype observed.

Fig. 4.

The Δmrp mutant has a wild-type biofilm phenotype in M9 minimal media but not in LB media. For M9 biofilm formation, V. cholerae strains were grown overnight at 37 °C in LB broth, washed once with M9− and then diluted to OD₅₉₅ of 0.1 in 200 µl of M9− minimal media (12.8 g l⁻¹ Na₂HPO₄·7H₂O, 3 g l⁻¹ KH₂HPO₄, 0.5 g l⁻¹ NaCl, 0.1 mM CaCl₂, 2 mM MgSO₄ and 20 mM succinate) containing either 1 g l⁻¹ (~19 mM) NH₄Cl or 19 mM glutamine as the sole nitrogen source at pH 7.6, 6.5, 5.5 and 4.5 on a 96-well plate. For LB biofilm formation, V. cholerae strains were grown overnight at 37 °C in LB broth, washed once with LB− and then diluted to OD₅₉₅0.1 in 200 µl of LBB− or LB− at pH 8.5, 7.6, 6.5 and 5.5 on a 96-well plate. Cultures were statically incubated at 37 °C for 24 h. End point growth measurements (OD₅₉₅) were taken. Each well was washed with DI water and stained with crystal violet (0.1 % w/v). DMSO was added to solubilize the crystal violet stain for measurement (OD₅₇₀) to quantify the biofilm. Results presented are the mean±standard error of three independent experiments. Asterisks indicate statistical significance compared to wild-type. (*P<0.05, **P<0.01, ***P<0.001 and ****P<0.0001). End point growth measurements (OD₅₉₅) for (a) M9, (c) LBB− and (e) LB−. Biofilm quantification (OD₅₇₀) for (b) M9, (d) LBB− and (f) LB−.

Since we observed a hypermotility phenotype in LB media, we also examined biofilm formation for wild-type and Δmrp strains in LBB− and LB− at pH 8.5, 7.5, 6.5 and 5.5 (Fig. 4c–f). The Δmrp mutant only had a growth defect at pH 5.5 for LBB− and LB− (Fig. 4c, e). Surprisingly, the Δmrp strain had a general biofilm defect at all pH values for both LBB− and LB−, despite the hypermotility phenotype and wild-type growth for these conditions (Fig. 4d, f). Unlike the M9 biofilm data, loss of Mrp resulted in decreased biofilm formation in LB media at all pH values tested. Therefore, the contribution of Mrp to motility and biofilm formation is more complex than anticipated.

Weak acid resistance

One of the most unusual biochemical properties of Mrp-type antiporters is their ability to protect bacterial growth against certain physiologically relevant weak acids, such as cholate (Ito et al., 1999, 2000) and other natural bile salts (Dzioba-Winogrodzki et al., 2009). Working with V. cholerae Mrp, we noticed that the heterologously expressed Mrp provides protection for E. coli against a surprisingly broad spectrum of weak acids added to the growth medium (unpublished observations).

Here, we compared susceptibility of the V. cholerae wild-type and Δmrp strains to various weak acids, including dioctyl sodium sulphosuccinate, n-lauroylsarcosine, probenecid, furosemide, nalidixic acid, ampicillin, carbenicillin and penicillin G. The Δmrp mutant of V. cholerae exhibited higher susceptibility than wild-type to n-lauroylsarcosine, dioctyl sodium sulphosuccinate and probenecid (Fig. 5). Curiously, no statistical difference was observed between wild-type and Δmrp strains for furosemide, nalidixic acid, ampicillin, carbenicillin and penicillin G (data not shown). Since we previously observed Mrp protecting E. coli from bile salts (Dzioba-Winogrodzki et al., 2009), we examined growth of V. cholerae wild-type and Δmrp strains in the presence of various concentrations of sodium cholate (Fig. S6). Surprisingly, loss of Mrp in V. cholerae did not increase sensitivity to sodium cholate. Overall, Mrp appears to contribute some weak acid resistance in V. cholerae but not to bile salts.

Fig. 5.

Mrp is involved in weak acid resistance. V. cholerae wild-type and mrp strains were grown overnight in LB broth, and diluted to an OD₅₉₅=0.1 in 200 µl LB or LB containing 312.5 µg ml⁻¹ n-lauroylsarcosine, 156.2 µg ml⁻¹ dioctyl sodium sulphosuccinate or 6.25 mg ml⁻¹ probenecid. Cultures were incubated with aeration at 37 °C for 24 h. End point growth measurements (OD₅₉₅) were taken. Results presented are the average±standard error of three independent experiments. Asterisks indicate statistical significance compared to wild-type (*P<0.05 and **P<0.01).

Discussion

Given the wide variations in ion specificity among Mrp systems from different micro-organisms (Swartz et al., 2005), it does not seem surprising that the monovalent cation/proton antiport mediated by Mrp antiporters was reported to contribute to the alkali, Na⁺ and K⁺ resistance of the hosting species (Hamamoto et al., 1994; Ito et al., 1999; Kosono et al., 1999; Putnoky et al., 1998). When assayed in sub-bacterial vesicles, Mrp predominantly transports Na⁺ or Li⁺, with K⁺ being a much less (kinetically) favoured substrate (Dzioba-Winogrodzki et al., 2009). It exchanges alkali cations for protons with maximal activity at pH 9.0–9.5 (Dzioba-Winogrodzki et al., 2009). Expression of V. cholerae Mrp protects growth of the antiporterless mutant of E. coli against external sodium and lithium (Dzioba-Winogrodzki et al., 2009). This detailed examination of the growth phenotype of the Δmrp mutant of V. cholerae allows us to gain perspective into the actual role of Mrp on its natural genetic background with respect to Na⁺, Li⁺ and K⁺ homeostasis under different physiological conditions. Generally, the extensive growth data collected here bring us to the conclusion that a functional Mrp is needed for maximal protection of V. cholerae cells against alkali cations at neutral and alkaline growth media.

However, our results pose the intriguing question of how an antiporter with an optimal Na⁺/H⁺ exchange activity at pH 9.0–9.5 could affect the salt resistance of cells at a much lower pH of the external medium, especially if it operates against the background of six more Na⁺/H⁺ antiporters? At this moment, no clear answer emerges and an extended (preferably comprehensive) collection of antiport-deficient V. cholerae mutants must be assembled and examined as a first step towards a solution of this conundrum. Two points should be stressed here, however: (i) with its broad cation specificity, V. cholerae Mrp could contribute to the stabilization of cellular ion homeostasis by a heterologous cation exchange (Na⁺/K⁺) rather than ‘normal’ Na⁺/H⁺ exchange, and the pH optimum of this activity might be shifted toward more acidic pH. (ii) Given the relatively small osmotically active volume of the bacterial cytoplasm and the rather huge transmembrane fluxes of H⁺ and Na⁺, even a small amount of V. cholerae Mrp activity that might be ongoing at neutral and slightly alkaline pH in the cytoplasm could make for the difference reflected in growth rates.

On a more general note, the very fact of Mrp influencing the growth of V. cholerae, however subtle the effect, suggests that the extravagant number of alkali cation/proton antiporters in this organism is not at all excessive [also see (Krulwich et al., 2009; Resch et al., 2010) for extended general discussion]. Rather, all these systems are working in concert, ensuring robust growth in challenging and changing environments. Our data show that the mrp mutant is able to catch up to the growth levels of the wild-type strain, possibly due to the concerted effort of other antiporters in V. cholerae. Clearly, these intriguing observations merit further research to fully understand the relative contributions of the many antiporters present in V. cholerae on bacterial growth.

Another curious feature of the Δmrp phenotype is a systematic defect in nitrogen utilization. This phenomenon clearly deserves further investigation. The simplest explanation might be that elimination of Mrp disturbs a finely tuned pH homeostasis in the cytoplasm of V. cholerae, and this general shift in internal pH affects nitrogen utilization at the level of transport and/or internal metabolism. To check such a possibility, we plan to directly measure the internal pH in cells of the wild-type and Δmrp strains under different conditions, as was recently done for another V. cholerae antiporter mutant (Quinn et al., 2012).

The striking difference in growth observed between the wild-type and Δmrp strains in M9 supplemented with alkali cations at all tested pH from 4.5 to 7.5 could not be attributed to the defect in nitrogen utilization alone. Most probably, a modest Δmrp growth defect observable in the LB data sets becomes more severe when cells are grown in a minimal medium. Such an effect would to be expected: when external resources are scarce, ion/pH homeostasis in bacterial cytoplasm is more fragile, and thus contributions of individual ion-exchanging systems are becoming more pronounced.

In various bacterial species, Mrp-type antiporters have been shown to influence many metabolic traits that seemingly have no obvious connection to cation/proton exchange per se. Mrp has been shown to modulate sporulation in Bacillus subtilis (Kosono et al., 2000) and symbiotic nitrogen fixation in Rhizobium meliloti (Putnoky et al., 1998). Analysis of a deletion mutant suggests an important role of Mrp in the pathogenesis of Pseudomonas aeruginosa (Kosono et al., 2005). Interestingly, these multi-subunit antiporters also render host cells resistant to natural bile salts such as sodium cholate and taurocholate (Dzioba-Winogrodzki et al., 2009; Ito et al., 1999, 2000). Moreover, when expressed in antiport-deficient strains of E. coli, V. cholerae Mrp protected their cells against such structurally different compounds as SDS, n-lauroylsarcosine, dioctyl sodium sulphosuccinate (docusate) and nalidixic acid (J. L. Winogrodzki and others, unpublished observations). Loss of Mrp increased the sensitivity of V. cholerae cells to weak acids dioctyl sodium sulphosuccinate, n-lauroylsarcosine and probenecid. However, Mrp does not protect against sodium cholate, a natural conjugated bile salt. This could result from V. cholerae possessing a potent Mrp-independent system for protection against bile (Chatterjee et al., 2004) that would probably mask possible effects of Mrp. Mrp does protect cells of its natural host, V. cholerae, against amphiphilic anions, and this protection might have an impact on survival in the small intestine where bile is one of the major constituents.

Rotation of the polar flagellum of V. cholerae is energized by the transmembrane electrochemical gradient of Na⁺ (Asai et al., 2003; Kojima et al., 1999). Assuming that the Na⁺ motor of V. cholerae rotates as rapidly as homologous molecular assemblies in other Vibrio species, and requires ~1000 ions per revolution, as happens for bacterial H⁺-motors, total sodium influx mediated by the polar flagellum in V. cholerae could amount to 5 % of the entire cytoplasmic Na⁺ per second (Magariyama et al., 1994). One could therefore expect that even a minor lowering in overall Na⁺ efflux mediated by primary and secondary sodium pumps, including Na⁺/H⁺ antiport, should affect flagellar rotation and thus motility. However, influence of the mrp deletion on motility might be of a more complex nature. In general, the deletion mutant demonstrated hypermotility under all conditions. The influence of Mrp on motility is probably not limited/determined solely by the contribution of the antiporter to overall SMF on the membrane. It should be noted here that the diameter of bacterial population in experiments of this kind is a complex function of multiple parameters, including (i) speed of flagellar rotation, (ii) chemotactic behaviour (frequency of clockwise to counterclockwise switches in flagellar rotation) and (iii) growth rate. In addition, one should bear in mind that the motility in V. cholerae is controlled by the hierarchical regulatory cascades on many levels, including transcription (Prouty et al., 2001).

Moreover, motility in V. cholerae is tightly coupled to exopolysaccharide production and biofilm formation (Heithoff & Mahan, 2004; Lauriano et al., 2004). It has been suggested that, in situ, the Na⁺ motive flagellar motor of V. cholerae acts as a mechanosensor, guiding micro-colony formation and the eventual three-dimensional cell stratification in ascendant biofilm (see Heithoff & Mahan, 2004, and references therein). Since the deletion of the mrp operon was found to affect motility, one could expect it also to modulate the ability of V. cholerae to form biofilms. Indeed, for example in minimal M9 medium, where the Δmrp mutation resulted in a hypermotile phenotype, cells of the deletion mutant demonstrated wild-type biofilm formation despite an overall reduction in growth. This effect was observed at all tested pH values from 4.5 to 7.5. These results seem somewhat of a paradox because they do not support the simple and intuitive idea that a deceleration of flagellar rotation promotes biofilm formation governed by the VpsR/VpsT signalling cascade (Heithoff & Mahan, 2004). Possible conciliation could be reached, though, if it turns out that the increased swarm circle diameter of the Δmrp mutant on M9 motility agar is due not to a faster flagellar rotation but to other factors such as altered chemotaxis. Alternatively, it could be that the chemotactic cascade itself has a key impact on biofilm formation, at least under certain conditions, or that the increased biofilm formation observed for the Δmrp strain is the result of a stress response when experiencing a growth defect. Resolution of this matter would require future detailed analysis of motility and chemotaxis as factors in biofilm formation by the V. cholerae wild-type and the Δmrp mutant.

Perhaps the most important result of this work is that it demonstrates how careful one should be when interpreting the results concerning ion transporters obtained in model systems such as heterologous membrane vesicles (Dzioba et al., 2002; Wiens et al., 2014 and many other similar works). As useful as the previous studies of V. cholerae Mrp expressed in E. coli are (Dzioba-Winogrodzki et al., 2009), based on those data one could not predict diverse effects of Mrp on the overall physiology of its natural host, V. cholerae. In particular, the unforeseen effects of the Δmrp deletion on nitrogen metabolism, cell motility and biofilm formation reported in this communication could only be obtained from examination of a chromosomal deletion mutant of V. cholerae. Another lesson that could be extracted from the above analyses is that each individual component of the seemingly excessive secondary ion exchange machinery in this pathogen is probably strongly embedded in the general physiology of V. cholerae.

Abbreviations:

PM

Phenotype MicroArray

SMF

sodium motive force