The halophilic alkalithermophile Natranaerobius thermophilus adapts to multiple environmental extremes using a large repertoire of Na+(K+)/H+ antiporters

Natranaerobius thermophilus is an unusual extremophile because it is halophilic, alkaliphilic and thermophilic, growing optimally at 3.5 M Na+, pH55°C 9.5 and 53°C. Mechanisms enabling this tripartite lifestyle are essential for understanding how microorganisms grow under inhospitable conditions, but remain unknown, particularly in extremophiles growing under multiple extremes. We report on the response of N. thermophilus to external pH at high salt and elevated temperature and identify mechanisms responsible for this adaptation. N. thermophilus exhibited cytoplasm acidification, maintaining an unanticipated transmembrane pH gradient of 1 unit over the entire extracellular pH range for growth. N. thermophilus uses two distinct mechanisms for cytoplasm acidification. At extracellular pH values at and below the optimum, N. thermophilus utilizes at least eight electrogenic Na+(K+)/H+ antiporters for cytoplasm acidification. Characterization of these antiporters in antiporter-deficient Escherichia coli KNabc showed overlapping pH profiles (pH 7.8–10.0) and Na+ concentrations for activity (K0.5 values 1.0–4.4 mM), properties that correlate with intracellular conditions of N. thermophilus. As the extracellular pH increases beyond the optimum, electrogenic antiport activity ceases, and cytoplasm acidification is achieved by energy-independent physiochemical effects (cytoplasmic buffering) potentially mediated by an acidic proteome. The combination of these strategies allows N. thermophilus to grow over a range of extracellular pH and Na+ concentrations and protect biomolecules under multiple extreme conditions.


Introduction
Extremophilic microorganisms are excellent models for the study and characterization of novel physiologies and biochemical pathways, and are also models for possible extraterrestrial life. To date, the majority of studies have focused on extremophiles growing at one environmental extreme, such as high temperature, high salinity or acid/ alkaline pH. Microorganisms that cope with two extremes (e.g. alkalithermophiles, acidothermophiles) are well documented (Segerer et al., 1988;Schleper et al., 1995;Cook et al., 1996;Prowe et al., 1996;Peddie et al., 1999;Olsson et al., 2003;Huber and Prangishvili, 2006), but to our knowledge, bacteria growing under three environmental extremes have not been studied in detail. Recent work from our laboratory has described the isolation and characterization of novel anaerobic bacteria growing at a unique combination of extreme conditions (Mesbah et al., 2007;Mesbah and Wiegel, 2009). These new bacterial isolates are obligately anaerobic, chemoorganotrophic and grow optimally with a doubling time of a few hours under the combined conditions of elevated temperature (Ն 50°C), alkaline pH (pH 55°C Ն 8.5, equivalent to pH 25°C 9.5), and elevated salt concentration (Ն 3.2 M Na + ) [The superscript denotes the temperature at which the pH was measured and the pH meter calibrated. See Wiegel (1998) for more details]. These microorganisms are thermophilic, alkaliphilic and halophilic, and these growth conditions place them into the unusual group of 'poly extremophiles', termed the halophilic alkalithermophiles. The taxa of halophilic alkalithermophiles isolated thus far represent a novel order within the Firmicutes, the Natranaerobiales (Mesbah et al., 2007;Mesbah and Wiegel, 2009). Their high pH and temperature optima/maxima for growth distinguish them from other anaerobic halophiles within the Firmicutes.
One of these halophilic alkalithermophiles, Natranaerobius thermophilus, was isolated from sediments of the alkaline, hypersaline lakes of the Wadi An Natrun, Egypt (Mesbah et al., 2007). N. thermophilus grows between 35°C and 56°C, with an optimum at 53°C. The pH 55°C range for growth is 8.3-10.6, with an optimum at pH 55°C 9.5 and no growth at pH 55°C 8.2 or below or pH 55°C 10.8 or above. This corresponds to pH 25°C 9.3-11.2, with an optimum at pH 25°C 10.5. At optimum pH 55°C and temperature, N. thermophilus grows in the Na + concentration range of 3.1-4.9 M and optimally between 3.3 and 3.9 M Na + . It is a non-motile and strictly anaerobic chemoorganotroph. As N. thermophilus is a halophile, alkaliphile and thermophile, its physiological and bioenergetic characteristics must be compatible with these complex and extreme growth conditions. As an alkaliphile, N. thermophilus must be able to acidify the cytoplasm in order to preserve the functional and structural integrity of cytoplasmic proteins supporting growth. As a thermophilic and halophilic bacterium, N. thermophilus is confronted with the problem of passive permeation of H + and Na + through the cytoplasmic membrane, processes that increase with high temperature and elevated Na + concentration (Vossenberg et al., 1999;Konings et al., 2002). Na + toxicity occurs due to an increase in intracellular Na + concentrations. While this increase can range from 20 mM in non-halophiles to over 3 M in extremely halophilic archaea, all prokaryotes maintain intracellular Na + concentrations lower than the extracellular concentration (Padan and Krulwich, 2000). Increased proton permeability of the cytoplasmic membrane compromises both cytoplasmic acidification mechanisms that are critical for survival at alkaline pH, and the ability to maintain a membrane potential.
Given these considerations, the bioenergetics of N. thermophilus becomes an intriguing problem. Previous studies have investigated response to alkaline pH in alkaliphiles (Sturr et al., 1994;Aono et al., 1997;Kitada et al., 2000), and alkalithermophiles (Cook et al., 1996;Olsson et al., 2003). These microorganisms exhibited intracellular pH regulation over their extracellular pH ranges for growth, and the transmembrane pH gradient (DpH, pH out-pHin) decreased and, in the case of the anaerobic alkalithermophile Clostridium paradoxum, collapsed under extreme alkaline stress as well as the minimal pH value where growth ceased. The limiting factor for growth at extreme alkaline pH appears to be the reduced capacity for cytoplasm acidification. However, the precise mechanisms allowing intracellular pH homeostasis have not been investigated in detail.
We report on the biochemical and molecular strategies used by N. thermophilus to combat extreme pH at high salinity and temperature. N. thermophilus exhibited cytoplasm acidification, but has the unusual characteris-tic of DpH homeostasis that is different from what has been previously reported for other alkaliphiles and alkalithermophiles (Sturr et al., 1994;Cook et al., 1996;Aono et al., 1997;Olsson et al., 2003). Intracellular pH in N. thermophilus was maintained at a constant 1 unit below the extracellular pH and the DpH did not collapse, even under extreme alkaline stress. In order to determine the mechanism of this unique cytoplasm pH regulation, the genome sequence of N. thermophilus was analysed for putative cation/proton antiporter genes. Of the 12 putative cation/proton antiporters identified, 8 were functionally expressed in the triple antiporterdeficient Escherichia coli KNabc. Various aspects of the biological activity and transcription of these antiporters were characterized, and their roles in physiology of growth at high Na + concentration and pH are discussed.

N. thermophilus exhibits active and passive mechanisms of cytoplasm pH acidification
The obligately anaerobic N. thermophilus is able to grow in the pH 55°C range 8.3-10.6 and no growth is observed below pH 55°C 8.2 and above pH 55°C 10.8 (Fig. 1A). The main organic fermentation products from sucrose at pH 55°C 9.5, 53°C and 3.5 M Na + are acetate and formate (Mesbah et al., 2007). The effect of external pH on intracellular pH was examined in sucrose-energized cell suspensions of N. thermophilus over the pH 55°C range 8.0-11.0 ( Fig. 1B). At the optimal pH 55°C for growth of 9.5, 3.3 M Na + and 53°C, N. thermophilus maintained a DpH of approximately 1 unit (acidin). This DpH of 1 unit was due to active cytoplasm acidification as the gradient (measured at optimal extracellular conditions) was collapsed with inhibitors that dissipate the DpH [e.g. 3,3′,4′,5-tetrachlorosalicylanilide, nigericin ( Fig. S1A)], and the DpH was absent in non-energized suspensions of N. thermophilus (see below). As the external pH 55°C was increased from 8.0 to 11.0, the intracellular pH 55°C increased from 7.2 to 9.9 (Fig. 1B). Strikingly, the DpH across the cell membrane did not collapse, even at the upper and lower pH 55°C boundaries for growth. This suggests that cessation of growth at more alkaline pH values is not due to a decrease in DpH as has been reported for other alkaliphiles and alkalithermophiles (Sturr et al., 1994;Cook et al., 1996;Olsson et al., 2003). To investigate the nature of this DpH homeostasis at high extracellular pH 55°C values, the experiment was repeated with non-energized cells. The DpH of nonenergized cells was close to zero until the external pH 55°C was > 9.5, at which point the same pattern of cytoplasm acidification of energized cells was observed.

Intracellular pH regulation in N. thermophilus 271
This DpH was only partially collapsed by protonophores and ionophores (Fig. S1B), suggesting an energyindependent component to DpH homeostasis (e.g. cytoplasmic buffering capacity).
The electrochemical membrane potential (Dy) of energized cell suspensions varied between -99 and -138 mV (Fig. 1C) and the overall proton motive force (pmf) was between -50 and -89 mV (Fig. 1C) as a result of the oppositely orientated Z DpH. Despite the low overall pmf, the phosphorylation potential was maintained between -477 and -491 mV (data not shown). Intracellular K + concentrations in energized cell suspensions remained constant at approximately 250 mM below extracellular pH 55°C 9.5, but increased sharply at more alkaline pH values reaching 540 mM at pH 55°C 10.6 ( Fig. 1D). The intracellular Na + concentration in both exponentially growing and energized cells was 8 mM at extracellular pH 55°C 9.5, and increased to 33 mM at pH 55°C 10.5 (data not shown).

N. thermophilus harbours a large repertoire of Na + (K + )/H + antiporters in its genome
To determine the mechanism for cytoplasm acidification in N. thermophilus, the genome sequence was analysed for transport systems that could participate in cytoplasm acidification [e.g. Na + (K + )/H + antiporters] using a semiautomated pipeline (Ren et al., 2007). Twelve genes with significant homology to Na + /H + exchangers were identified by these analyses (Table S1). Eight of these genes had homology to members of the NhaC Na + /H + antiporter family (Saier et al., 2006), and were encoded alone and did not appear to be part of an operon. Three genes were 38-48% identical to Na + /H + exchanger proteins from 'Alkaliphilus metalliredigens' and 'Alkaliphilus oremlandii', two obligately anaerobic, non-halophilic and moderately alkaliphilic bacteria (Ye et al., 2004;Fisher et al., 2008). Gene nt-Nha had 35% identity to the shaA (mrPA) gene of Clostridium tetani. The Mrp proteins belong to the monovalent cation/proton antiporter-3 protein family. This family is composed of multi-component Na + /H + and K + /H + antiporters encoded by operons of six or seven genes, and all genes are required for full function in Na + and alkali resistance (Ito et al., 2000). Sequence analysis of the regions surrounding gene nt-Nha, however, did not show that it was part of an operon. This indicates that gene nt-Nha does not encode a subunit of an Mrp system, but rather a mono-subunit antiporter. A. Effect of external pH 55°C on growth of N. thermophilus in batch culture. B. Effect of external pH on intracellular pH in sucrose-energized cell suspensions (ᮀ) and non-energized cell suspensions ( ). The diagonal line represents absence of a DpH. C. Effect of external pH 55°C on Dy (᭜) and proton motive force (ᮀ) in energized cell suspensions of N. thermophilus. D. Effect of external pH 55°C on intracellular K + concentration in energized cell suspensions of N. thermophilus. All values reported are the mean of three independent experiments; the standard error associated with the determinations is shown.

N. thermophilus antiporter genes complement the Na + and alkali sensitivities of E. coli KNabc
To verify the function of the predicted antiporter genes, the individual genes were expressed heterolougsly in antiporter deficient strain E. coli KNabc (DnhaA DnhaB DchaA). E. coli strain KNabc is sensitive to NaCl concentrations of 200 mM and above and pH 37°C values > 8.0 (Goldberg et al., 1987). Of the 12 cloned putative antiporter genes identified, 8 were able to complement the Na + -sensitive phenotype of strain KNabc (Fig. 2). The KNabc-Nt-NhaC1, -Nt-NhaC3 and -Nt-NhaC4 transformants exhibited Na + resistance up to 700 mM at pH 37°C 7.5, while the remaining transformants supported resistance at up to 600 mM NaCl (Fig. 2). None of the transformants supported growth at NaCl concentrations > 750 mM. The same eight N. thermophilus antiporter genes supported modest alkali-resistance allowing growth at pH 37°C values up to 8.5 in the absence of added NaCl. The remaining four genes did not complement the Na + -sensitivity phenotype of E. coli KNabc, but were able to complement the K + -uptake deficiency of E. coli TK2420 (Kdp -, Kup -, Trk -), indicating that they have potential roles in transport of K + into the cell (Fig. S2). Genomic analyses showed that these four genes are encoded alone and do not appear to be part of an operon or any other system. N. thermophilus antiporters catalyse Na + /H + and K + /H + antiport at high pH with a wide range of K0.5 for cations Biochemical assays with inverted membrane vesicles of strain KNabc expressing the individual N. thermophilus antiporters were carried out to determine antiport activity with the fluorescent dye acridine orange (Goldberg et al., 1987) (see Experimental procedures for details). Initial experiments were conducted at pH 37°C 8.5 in order to simulate the intracellular pH measured in energized cell suspensions of N. thermophilus (Fig. 1B). When inverted membrane vesicles of strain KNabc were incubated with Trizma-succinate, strong quenching of acridine orange was observed indicating the formation of a DpH. The signal was reversed by the addition of NaCl indicating Na + /H + antiporter activity. With the exception of Nt-NhaC3, all N. thermophilus antiporter proteins exhibited significant Na + -dependent antiport activity, with a dequenching percentage as high as 77% for Nt-CPA1a (Table 1). K + /H + antiport activity was observed for all eight putative antiporters in the concentration range between 3 and 31 mM K + . Nt-NhaC3 exhibited solely K + /H + antiport activity and did not show any Na + -dependent activities using a range of Na + concentrations (2-100 mM) (Table 1). Antiport activity was examined over a range of NaCl concentrations (Table 1, Fig. 3). In contrast to the Michaelis-Menten kinetics that have been reported for Na + (Li + )/H + antiporters from Gram-positive bacteria (Swartz et al., 2007), non-linear kinetics were observed for all N. thermophilus antiporters (Fig. 3). Antiport activity was inhibited at high concentrations of Na + (> 10 mM, substrate Na + -inhibition) for all proteins. Due to non-Michaelis-Menten kinetics, Km values could not be determined. Instead, half saturation concentrations (K0.5) were determined from v versus [S] plots (Fig. 3). K0.5 values for the antiporter proteins varied significantly between the seven Na + -translocating antiporters, and ranged between 0.8 and 4.4 mM. The K + optimum for the K + /H + specific Nt-NhaC3 was 31 mM (Table 1).
The activity profiles of N. thermophilus antiporters as a function of pH 37°C indicated alkaline pH 37°C optima between 8.5 and 8.8 (Table 2). Activity was greatly reduced at pH 37°C 7.8 and no activity was detected at pH 37°C 7.6 or below. The pH profile for Nt-NhaC2 is distinguished from the remaining antiporters in that it has a pH 37°C optimum of 9.5 and retained activity at pH 37°C 10.0.
Higher pH values could not be tested due to decreased activity of the heterologous E. coli membranes used.
Electrogenic antiport was assessed using the fluorescent probe Oxonol VI (see Experimental procedures). When cation/proton antiporters are used for pH homeostasis, the antiporter needs to be electrogenic rather than electroneutral (Padan et al., 2005). During electrogenic antiport, the inward H + flux is larger than outward Na + flux during a single turnover, such that net positive charge is translocated into the cell. Thus, the antiporter can be energized by Dy (positive outside, relatively negative inside in whole cells). As a result, antiporter activity dissipates the Dy. Electrogenic antiporter activity was observed in membrane vesicles containing each one of the eight N. thermophilus antiporter genes (Fig. 4). Membrane vesicles containing Nt-NhaC1 did not show a complete reversal of quenching upon addition of Na2SO4. This could be due to a different stoichiometry of Na + /H + exchange, thus not completely consuming the Dy. Alternatively, this particular protein may not be able to function optimally in the heterologous system used in this study.

Expression of N. thermophilus antiporters is constitutive
As cation/proton antiporters play a dominant role in adaptation to alkaline stress, expression of all 12 identified antiporters was examined after adaptation of N. thermophilus (~15 doubling times in continuous culture) to either an alkaline (pH 55°C 10.5) or relatively 'acidic' (pH 55°C 8.5) stress using quantitative reverse transcriptase PCR (qRT-PCR). Changes in the expression levels of the 12 proteins were determined in cells in midexponential growth phase after a shift (from optimal pH 55°C 9.5) to either pH 55°C 10.5 or pH 55°C 8.5 for 50 h. Samples were withdrawn from steady-state cultures of N. thermophilus growing continuously at the indicated pH 55°C value, and RNA was extracted and reversetranscribed as described in Supporting information Experimental procedures.
In the presence of both alkaline and 'acid' stress, expression of Nt-CPA1a, Nt-CPA2, Nt-NhaC1, Nt-NhaC2, Nt-NhaC3, Nt-NhaC5, Nt-NhaC6 and Nt-NhaC8 remained constitutive (Fig. 5). Consistent with its relatively low-pH minimum for activity ( Table 2), expression of the Na + (K + )/H + antiporter Nt-CPA1b showed a sixfold increase after adaptation to pH 55°C 8.5. Nt-NhaC4 showed a fourfold increase in expression after acclimation to pH 55°C 10.5 (Fig. 5). Interestingly, Nt-NhaC2 showed a slight decrease in expression at pH 55°C 10.5 even though it had the most alkaline pH optimum when assayed in inverted membrane vesicles. Nt-NhaC7 showed a sevenfold increase in expression at pH 55°C 8.5. This protein did not show any antiport activity in inverted membrane vesicles nor did it complement the Na + -sensitive phenotype of E. coli KNabc. However, it did complement the K + -uptake deficiency of E. coli TK2420 indicating that it has K + -transport capability.

Discussion
The poly extremophile N. thermophilus is a prototype for robust growth at combined extremes of pH, temperature and salt concentration. The mechanism(s) responsible for this extraordinary feat were the focus of this study. N. thermophilus exhibited unusual energy-dependent and -independent DpH homeostasis over the pH 55°C range for growth, a characteristic not observed in the alkalithermophiles studied (Cook et al., 1996;Olsson et al., 2003). At high pH 55°C values (i.e. > 9.5), growth decreased and the intracellular pH increased to values > 8.8. Despite poor growth at these high pH 55°C values, the cells continued to maintain a cytoplasmic pH that was more acidic than the external milieu. The ability of some bacteria to maintain a DpH that is energy-independent has been attributed to the buffering capacity of the cytoplasm to sequester or release protons (Booth, 1985). The buffering capacity of the cytoplasm is due to the presence of titratable groups  (Fig. 3).
(e.g. basic and acidic amino acids, both free and attached to proteins). Analysis of the proteome of N. thermophilus showed that, consistent with being a halophilic microorganism, the isoelectric point of proteins is predominantly acidic, ranging between 4 and 5 (Fig. S3). Thus, at the alkaline intracellular pH of N. thermophilus, a large amount of proteins will be negatively charged. As the internal pH increases, the charge will become more negative and therefore, to reach the demand by the cell for overall neutrality on both sides of the membrane, cations, including protons may enter the cell. This leads to a more acidic interior. It is important to note that when incubated at pH 55°C values greater than 10.3, cells of N. thermophilus grow with extended doubling times (15 versus 3.5 h at optimal pH), do not reach OD600 values greater than 0.1, and quickly lose viability after reaching stationary phase. The intracellular pH under these conditions increases beyond 9.5. Cessation of growth could be due to saturation of the intracellular buffering capacity and alkalinization of the cytoplasm. Intracellular K + , which is tightly regulated at 200-250 mM during growth at optimal conditions, sharply increases to almost 550 mM at extracellular pH 55°C 10.5; and intracellular Na + increases fourfold from 8 to 33 mM. We hypothesize that this sudden increase in K + serves to protect intracellular proteins from the cytotoxic effects of Na + . Alternatively, the increase in intracellular K + (and Na + ) may act to neutralize the net negative charge of the cytoplasm.
Alkaliphiles must maintain a cytoplasm pH that is compatible with the functional and structural integrity of cytoplasm proteins supporting growth. Alkaliphiles generally acidify the cytoplasm, and maintain an intracellular pH that is moderately alkaline (pH 7.5-8.5) (Krulwich et al., 1996;Padan et al., 2005). Among the strategies used for cytoplasm acidification, increased activity of monovalent cation proton antiporters plays a dominant role in cytoplasm pH homeostasis (Padan et al., 2005). We identified 12 putative antiporters in the genome of N. thermophilus. Of these 12 homologues, 8 of them showed strong sequence identity to the Na + /H + antiporter NhaC. The four remaining genes had homology to members of the monovalent cation/proton antiporter-1 and cation/proton antiporter-2 families. All these protein families contain transporters that play roles in cytoplasmic pH regulation, extrusion of intracellular Na + and cell volume regulation (Waser et al., 1992;Southworth et al., 2001;Radchenko et al., 2006;Wei et al., 2007). Eight of the 12 N. thermophilus antiporter homologues displayed antiport activity in membrane vesicles prepared from E. coli KNabc. The antiport capacities of the proteins support their ability to play a major role in cytoplasm acidification in N. thermophilus. Electrogenicity is an important property for Na + /H + antiporters that support alkali resistance (Padan et al., 2005), and all N. thermophilus antiporters showed Na + and K + -dependent consumption of the Dy. Seven antiporter proteins exhibited strong Na + (K + )/H + antiport activity, and one showed only K + /H + antiport activity. Together, the eight antiporters functioned over a range of Na + and K + concentrations, consistent with the ability of this bacterium to grow over a range of salinities. The ability to use K + /H + antiport in addition to Na + /H + antiport for cytoplasm acidification is beneficial for an anaerobic halophilic alkalithermophile. Aerobic microorganisms typically use H + -coupled bioenergetics; there is no competition for the intracellular Na + substrate. On the other hand, the anaerobic alkalithermophiles studied to date have Na + -coupled mechanisms for pH homeostasis and solute transport (Speelmans et al., 1993;Prowe et al., 1996;Ferguson et al., 2006). As a result, the Na + -coupled ATPase will compete with the Na + /H + antiporters for intracellular Na + . In this case, K + /H + antiporters can continue to acidify the cytoplasm after Na + /H + antiporters sufficiently reduce the cytoplasmic Na + content.
The overlapping pH profiles of N. thermophilus antiporters further supports their ability to play a role in mediating concomitant contributions to intracellular pH and/or salt tolerance of N. thermophilus. Optimal antiport activity in inverted membrane vesicles was observed within the pH 37°C range of 8.5-8.8, similar to the intracellular pH in energized cells of N. thermophilus. This indicates that the antiporters are biochemically compatible with the intracellular pH of N. thermophilus. Furthermore, constitutive expression of the majority of antiporter genes under both acidic and alkaline stress implies that their products make synergistic contributions to pH and/or salt tolerance of N. thermophilus and a 'ready' mechanism to combat sudden changes in the external environment. Expression of nt-CPA1b, nt-NhaC4 and nt-NhaC7 were significantly altered under stress conditions indicating pH-dependent regulation of these proteins. It is possible that these proteins in particular play major adaptive roles under the respective stress conditions.
A model summarizing the bioenergetic processes presented and discussed in N. thermophilus is shown in Fig. 6. N.   generates the bulk of its ATP via substrate level phosphorylation producing acetate and formate. The membranebound F-type ATPase plays an important role in establishing the electrochemical gradient of Na + ions, fuelled by ATP that is rapidly produced from substrate level phosphorylation. The result of sodium pumping also leads to the formation of a significant Dy (-124 mV), which contributes to the overall electrochemical gradient of Na + ions. This is vitally important because the cation/ proton antiporters in N. thermophilus are coupled to Dy. N. thermophilus utilizes two distinct mechanisms for cytoplasm acidification under conditions of high salt concentration and elevated temperature. At extracellular pH 55°C values at and below the optimum, acidification of the cytoplasm is achieved via a large cohort of electrogenic cation/proton antiporters that are able to translocate Na + and K + ions in exchange for protons. As the extracellular pH 55°C increases, energy-dependent antiport activity stops, and acidification is probably achieved by physiochemical forces such as cytoplasmic buffering. These strategies allow N. thermophilus to adapt to fluctuations in extracellular pH and Na + ion concentration under conditions where the availability of energy generating substrates varies.

Bacterial strains and culture conditions
Strains used in this study are shown in Table S2. N. thermophilus strain JW/NM-WN-LF T (= DSM 18059 T = ATCC BAA-1301 T ) was grown anaerobically at 53°C in carbonate-buffered medium as described previously (Mesbah et al., 2007). E. coli KNabc (DnhaA DnhaB DchaA) was grown routinely in LBK medium, pH 7.5 (Goldberg et al., 1987). E. coli JM109 was used for routine cloning procedures and was grown in LB medium. When antibiotics were added to the medium for selection or plasmid maintenance, concentrations used were 100 mg ampicillin ml -1 , 50 mg kanamycin ml -1 and 50 mg erythromycin ml -1 .

Determination of bioenergetic parameters
For determination of DpH, DY and pmf, N. thermophilus was grown in batch culture at pH 55°C 9.5, 53°C in the presence of 3.3 M Na + . The superscript preceding the pH value indicates the temperature at which the pH meter was calibrated and pH measured as described (Wiegel, 1998;Mesbah and Wiegel, 2006).
Cells were harvested during mid-exponential phase by centrifugation (6000 g, 30 min, 24°C) and washed three times in sterile anaerobic carbonate-buffered medium, pH 55°C 9.5. All washing steps were carried out anaerobically inside a Coy anaerobic chamber (Coy Laboratory Products, Grass Lake, Michigan). Cells were resuspended to a final OD600 of 1.0 in anaerobic carbonate-buffered medium adjusted to the pH 55°C being studied. These cell suspensions (non-growing) were energized by the addition of sucrose (0.5% wt/vol.) and incubation at 53°C for 20 min. Energization caused an increase in gas pressure at all tested pH values, indicating that the cells remained active during the assay period (subculturing and dilution to extinction further confirmed viability of the cells). The final pH 55°C values were determined, then energized cell suspensions (2.0 ml) were then transferred by syringe to sealed serum vials (10 ml capacity, nitrogen in gas phase) containing one of the following isotopes: [ 14 C]methylamine (5.4 mM) or [ 3 H]tetraphenylphosphonium + (TPP + , 1 mM). After incubation for 5 min at 53°C, 0.9 ml of the culture were rapidly centrifuged through 300 ml of silicone oil (13 000 g, 3 min, 24°C). Twenty microlitres of supernatant were removed. The tubes with remaining contents were frozen at -80°C for at least 2 h. The bottoms of the tubes containing cell pellets were removed with dog-nail clippers; the supernatant and cell pellets were separately dissolved in scintillation fluid, and radioactivity (cpm) was determined with an LKB Wallac 1214 Rack-Beta scintillation counter.
Intracellular volume (5.6 Ϯ 0.34 ml mg -1 protein) was determined from the difference in partitioning of [ 3 H]water (1 mM) and [ 14 C]polyethylene glycol. Polyethylene glycol is not metabolized by N. thermophilus. The membrane potential across the membrane was calculated from the uptake of [ 3 H]TPP + according to the Nernst relationship. Non-specific TPP + binding was estimated from valinomycin-and nigericintreated cells (10 mM each). These inhibitors cause complete growth-arrest when added to cultures of N. thermophilus during the exponential growth phase, thus they are membrane active with this bacterium. The DpH was determined from the distribution of [ 14 C]methylamine with the Henderson-Hasselbach equation, and ZDpH was calculated by 59 mV multiplied by the DpH. The phosphorylation potential, DGp, was calculated with the equation: DGp = DG°+ 2.3RT log[ATP]/[ADP] [Pi]. The value used for DG°was 33.3 kJ mol -1 , or the equivalent of -347 mV (Nicholls and Ferguson, 2002). Intracellular concentrations of ATP, ADP and Pi were determined as described (Olsson et al., 2003).

Measurement of intracellular sodium and potassium ion concentrations
Intracellular Na + and K + concentrations were measured as described (Olsson et al., 2003). Details are explained in Supporting information Experimental procedures.

DNA extraction, cloning and plasmids
PCR was carried out on N. thermophilus genomic DNA using Phusion TM High-Fidelity DNA polymerase (New England BioLabs, Ipswich, MA). PCR primers used are listed in Table S2. All genes were cloned with their native Shine-Dalgarno sequences behind the T7 promoter in pGEM-3Z(f+). Basal levels of expression of the cloned genes, without addition of inducer, were used for all experiments. Complete DNA sequencing was used to ensure that the plasmids ultimately used were free of errors. Details of genome sequencing, identification of putative transporter genes, PCR reactions and plasmid constructions are provided in Supporting information Experimental procedures.

Complementation assays in E. coli Knabc
All complementation assays were done under anaerobic conditions. Recombinant plasmids were transformed into Na + (K + )(Ca 2+ )/H + antiporter-deficient E. coli KNabc. The plasmid pGEM-3Zf(+) was used as a negative control. For studies of complementation of the Na + -and alkali-sensitive phenotypes of E. coli KNabc, the test and control transformants were cultured anaerobically overnight in LBK medium, pH 37°C 7.5. Two hundred microlitres of the overnight grown cultures was transferred into 5 ml of LBK adjusted to different pH values with different NaCl concentrations. Cell growth was monitored by measuring optical density at 600 nm.

Preparation of inverted membrane vesicles
Inverted membrane vesicles were prepared as described previously (Rosen, 1986). Details of buffers used are described in Supporting information Experimental procedures.

Assays of DpH-dependent antiport activity in inverted membrane vesicles
Antiport assays were conducted anaerobically in 10 mM Tris-Cl, 140 mM choline chloride, 15 mM MgCl2, 2.5 mM Trizma nitrate, 0.75 mM acridine orange and 500 mg protein ml -1 . The pH 37°C was adjusted to values between 7.5 and 10.0. Measurements were conducted using a Turner Designs TD-700 laboratory fluorometer with peak excitation 486 nm and peak emission 510 nm. Respiration was initiated by addition of Trizma-succinate to a final concentration of 2.5 mM. The chloride content of the buffer was high enough to ensure that the pmf established upon addition of an electron donor was entirely due to a DpH, acidic inside the inverted membrane vesicles (Rosen and Futai, 1980). Establishment of the DpH was monitored by quenching of the fluorescence of acridine orange. Cation addition results in dequenching of fluorescence that reflects cation-dependent proton movement out of the inverted membrane vesicles. At the end of the assay, 12 mM of ammonium chloride was added to dissipate the remaining pmf and bring fluorescence back to baseline.

Fluorescence-based assays of DY generation and antiporter dependent consumption
DY-dependent fluorescence of oxonol VI was used to measure the generation of a membrane potential, positive inside, with the addition of 0.6 mM of tetra(cyclohexylammonium)-NADPH. Electrogenicity of antiporter genes was evaluated by adding 4 mM Na2SO4 to energized membranes and observing a reversal of the quench. Reversal of quenching represents antiporter-dependent consumption of DY, which reflects activity of an electrogenic antiporter (Padan et al., 2005). Fluorescence was brought back to baseline by adding of 10 mM of carbonyl cyanide mchlorophenylhydrazone, a protonophore that abolishes the DY. The assay mixture contained 10 mM bis-[tris(hydroxymethyl)methylamino]-propane, 5 mM MgSO4, 200 mM K2SO4, 1 mM nigericin and 1 mM oxonol VI (pH 37°C 8.0-9.5). Measurements were made on a Turner Designs TD-700 laboratory fluorometer. The peak excitation and emission wavelengths were 523 nM and 630 nm respectively. Final concentration of vesicle protein was 500 mg ml -1 .

Continuous culture, RNA isolation and qRT-PCR
For continuous culture experiments, N. thermophilus was grown at 53°C with continuous stirring in a custom-made glass fermentor with a working volume of 500 ml. The same carbonate-buffered medium used for the batch cultures was used; the medium was kept anaerobic by a continuous flow of sterile N2. The continuous culture system was operated and maintained at a constant pH 55°C by the adjustment of the medium in the reservoir to the target pH 55°C . Medium pH 55°C was adjusted by addition of either 10 N HCl, 3 M Na2CO3 or 5 M NaOH. The pH 55°C in the fermentor was measured with an autoclavable glass electrode (Femprobe®, Broadley James Corp., CA). The pH 55°C values detected by the immersed electrode were checked for accuracy against an external electrode calibrated at 55°C with appropriately preheated buffers (Wiegel, 1998;Mesbah and Wiegel, 2006). The cultures were maintained at steady-state at OD600 in the range of 0.20-0.22 for at least 50 h (equivalent to approximately 15 doubling times). Continuous culture runs were conducted in duplicate for each culture condition tested.
Samples for RNA extraction were withdrawn from steadystate cultures of N. thermophilus growing at the indicated pH 55°C value. Details of RNA extraction, reverse transcription, and qPCR are provided in Supporting information Experimental procedures.
Differences in the transcript levels of the antiporter genes in response to different extracellular pH 55°C values were calculated according to the model proposed by Pfaffl (2001), using the recA gene as a reference gene.