Na 1 -dependent pH Regulation by the Amitochondriate Protozoan Parasite Giardia intestinalis *

Giardia intestinalis is a pathogenic fermentative parasite, which inhabits the gastrointestinal tract of animals and humans. G. intestinalis trophozoites are exposed to acidic fluctuations in vivo and must also cope with acidic metabolic endproducts. In this study, a combination of independent techniques ( 31 P NMR spectroscopy, distribution of the weak acid pH marker 5,5-dimethyl-2,4-oxazolidinedi-one (DMO) and the fluorescent pH indicator 2 * ,7 * -bis (car-boxyethyl)-5,6-carboxyfluorescein (BCECF)) were used to show that G. intestinalis trophozoites exposed to an extracellular pH range of 6.0–7.5 maintain their cytosolic pH (pH i ) within the range 6.7–7.1. Maintenance of the resting pH i was Na 1 -dependent but unaffected by amiloride (or analogs thereof). Recovery of pH i from an intracellular ac- idosis was also Na 1 -dependent, with the rate of recovery varying with the extracellular Na 1 concentration in a saturable manner ( K m 5 18 m M ; V max 5 10 m M H 1 min 2 1 ). The recovery of pH i from an acid load was inhibited by amilo- ride but unaffected by a number of its analogs. The postulated involvement of one or more Na 1 /H 1 exchanger(s) in the regulation of pH i in G. intestinalis

Maintenance of the cytosolic pH (pH i ) 1 within a physiological range is critical for survival of any organism, because virtually all biological processes are affected by pH. Eukaryote cells use a range of membrane transport mechanisms to import/export H ϩ and HCO 3 Ϫ from the cytosol. These include the electroneutral Na ϩ /H ϩ exchangers (NHEs) and the HCO 3 Ϫ -dependent exchangers that mediate HCO 3 Ϫ /Cl Ϫ exchange or Na ϩ HCO 3 Ϫ /Cl Ϫ exchange. H ϩ efflux can also be mediated by ATP-dependent proton pumps. These include P-type H ϩ -ATPases (as found in yeast, e.g. Ref. 7), P-type H ϩ ,K ϩ -ATPases (as found in the apical membrane of the gastric mucosa, e.g. Ref. 8), and V-type H ϩ -ATPases (found in the intracellular vacuolar membranes of cells of higher eukaryotes (e.g. Ref. 9) as well as in the plasma membrane of some cell types, including a number of parasitic protozoa (10 -12)).
There are a number of techniques available for the estimation of pH i . These include NMR spectroscopy, measurement of the distribution of weak acids, and the use of intracellular pH-sensitive fluorophores.
The most widely used form of the NMR method takes advantage of the pH dependence of the chemical shift (i.e. spectral position) of the peak arising from intracellular inorganic phosphate (P i ) in a 31 P NMR spectrum of intact cells (13). The low inherent sensitivity of the 31 P NMR method necessitates the use of high cell densities (which can lead to the metabolic viability of the cells being compromised) as well as limiting the time resolution of the technique.
Calculation of pH i from the distribution of radiolabeled weak acids such as 5,5-dimethyl-2,4-oxazolidinedione (DMO) assumes that the uncharged form of the molecule equilibrates across the membrane, whereas the ionized species is membrane impermeable. The relative concentrations of the charged and uncharged species is pH-dependent and hence if the intraand extracellular DMO concentrations ([DMO] i and [DMO] o , respectively) and the extracellular pH (pH o ) are known, the pH i can be determined (14). This method is more straightforward than the 31 P NMR method and requires lower cell numbers. However the calculation of pH i from the measured DMO distribution relies on the assumption that the marker is neither actively transported nor bound nor accumulated in compartments within the cell.
In recent years pH-sensitive fluorophores, such as the ratiometric pH indicator 2Ј,7Ј-bis (carboxyethyl)-5,6-carboxyfluorescein (BCECF) have come to represent the method of choice for the determination of pH i . This technique permits on-line meas-urements of pH i , allowing the effect of additions of ions and inhibitors to be determined in real time (15).
The aim of the present study was to use a combination of the techniques outlined above to estimate the resting pH i of G. intestinalis trophozoites (over a range of extracellular pHs), and to investigate the mechanism(s) involved in extruding H ϩ from the cell, thereby protecting it from intracellular acidosis.

EXPERIMENTAL PROCEDURES
Organism and Culture of G. intestinalis-Portland 1 strain was grown as described previously (16). For each experiment, cells were grown to late exponential phase, harvested by centrifugation at 650 ϫ g for 5 min and resuspended in the appropriate buffers. Cells were counted using a hemocytometer.
Determination of pH i by 31 P NMR-The method adopted for 31 P NMR analysis of G. intestinalis trophozoites was based on that described previously (17). Approximately 10 9 trophozoites were harvested from 900 ml of culture medium (60 ϫ 15 ml tubes) washed in PBS, and then resuspended in the appropriate MES-and HEPES-buffered saline (1.5 ml). An aliquot of this suspension, containing ϳ5 ϫ 10 8 cells, was transferred to an insert tube (5-mm outer diameter), which was placed inside a 10-mm outer diameter NMR tube for acquisition of spectra. 31 P NMR spectra were obtained (at 37°C) on a Bruker AM300 spectrometer operating in the Fourier transform mode. Spectra (averaged over 512 scans) were collected using a pulse angle of 60°and a repetition rate of 0.85 s. Chemical shifts are expressed relative to 85% phosphoric acid but were measured against a primary standard of 0.6 M TEP (triethyl phosphate, Ref. 18), contained in a 1-mm internal diameter capillary. A standard titration curve, relating the 31 P NMR chemical shift of P i to the pH was obtained from MES-and HEPES-buffered saline supplemented with 100 mM P i and adjusted to a range of pH values. Peaks were assigned by comparison with published spectra (17) and/or by addition of authentic compounds to the sample. The extracellular P i peak evident in the spectrum was derived from residual P i carried over from the initial PBS wash.
Determination of pH i from [2-14 C]DMO Distribution-G. intestinalis trophozoites were harvested by centrifugation and resuspended in MES-and HEPES-buffered saline of the required pH (pH 6.0 -7.5) at a density of ϳ6 ϫ 10 6 cells ml Ϫ1 . Aliquots (100 l) of the equilibrated cell suspension were transferred, in triplicate, to microcentrifuge tubes containing 100 l of MES-and HEPES-buffered saline at the same pH containing [2-14 C] DMO (0.05 Ci) together with unlabeled DMO (to give a final DMO concentration of 1-100 M) and layered over 100 l of oil (a mixture of dibutyl phthalate and di-iso-octyl phthalate, 4:1, 1.03 g ml Ϫ1 ). The sample was incubated at 37°C for 5 min before being centrifuged (10,000 ϫ g, 1 min) to sediment the cells beneath the oil layer. A 100-l aliquot of the extracellular solution (from above the oil layer) was collected for scintillation counting (thereby allowing an estimate of [DMO] o ). The oil layer was then aspirated, and the cell pellets were lysed with 0.1% (v/v) Triton X-100 (400 l), deproteinized with 6% (v/v) trichloroacetic acid (750 l), and then centrifuged at 10,000 ϫ g for 1 min. The supernatant solutions were transferred to vials for scintillation counting. The intra-and extracellular solution volumes in the cell pellets were determined using [ 3  ) ϫ (10 p K a ϩ 10 pHo ) Ϫ 10 p K a ) where the pK a (for DMO) is 6.3.
Determination of pH i Using the pH-sensitive Fluorescent Indicator BCECF-G. intestinalis trophozoites were harvested and resuspended in PBS (ϳ5 ϫ 10 6 cells ml Ϫ1 ) containing BCECF-AM (1 g ml Ϫ1 ), and then incubated for 15 min at 37°C. The neutral, membrane-permeant AM-ester readily enters the parasite cytosol, where the ester groups are removed by endogenous esterases, rendering the molecule charged and impermeant. Excess dye was removed by centrifugation and resuspension of the cells in PBS.
The suspension was transferred to a cuvette that was placed in the temperature-controlled chamber of a Perkin Elmer LS-50B luminescence spectrophotometer, maintained at 37°C. Using a dual excitation Fast Filter accessory, the sample was excited at 495 nm and 440 nm successively, and the fluorescence emission was measured at 535 nm. The ratio of the fluorescence intensity measured using the two excitation wavelengths (495 nm/440 nm) provides a quantitative measure of pH. The luminescence spectrophotometer was linked to a computer, allowing real-time monitoring of pH i . Data was imported into graphics software for analysis. Calibration of the fluorescence signal was performed using nigericin, as described previously for Leishmania major and mammalian tumor cells (19,20). Linear regression of 3-5 point calibration curves (pH versus the fluorescence intensity ratio) consistently yielded regression coefficients of Ͼ0.99. The pH i traces shown in the results are, in each case, representative of at least three separate experiments.
Cells were viewed by confocal laser scanning microscopy. The fluorescence was observed to be originating from the cell cytosol without interference from intracellular structures.
Acidification of the Cell Cytosol-In experiments designed to investigate the ability of G. intestinalis trophozoites to respond to an intracellular acid load, pH i was reduced using the NH 4 ϩ prepulse technique (21). Cells suspended were exposed briefly (6 -7 min) to 40 mM NH 4 Cl, after which they were centrifuged (10,000 ϫ g, 15 s), and resuspended in an appropriate NH 4 ϩ -free solution. This procedure consistently resulted in a decrease in pH i of 0.3-0.5 pH units.

Intracellular Buffering Power and H ϩ Efflux Determination-
The intracellular buffering power of G. intestinalis was calculated from the pH i changes resulting from the addition of a range of concentrations of NH 4 Cl (16). The buffering power (␤, mM/pH unit) is given by ϩ concentration following the addition of NH 4 Cl, calculated using a pK of 9.21 (in the Henderson-Hasselbach equation) and assuming that [NH 3 ] i is equal to [NH 3 ] o . ⌬pH i is given for changes in the intracellular pH ranging from 7.0 to 6.2. It was assumed that the buffering capacity of the cell remained approximately constant through the range of 6.2-6.6, as shown for mammalian cells (16).
Initial rates of pH i recovery (⌬pH i /⌬t) for cells suspended in medium of varying Na ϩ concentration (NMG-Cl solution with added NaCl 5-50 mM) and acidified using the NH 4 Cl prepulse technique were calculated by fitting the initial phase of the pH i recovery trace to a straight line, using linear regression. The slope of the line, ⌬pH i /⌬t, was converted to H ϩ equivalent efflux rate (J H , mM min Ϫ1 ) using the equation J H ϭ ␤ ϫ ⌬pH/⌬t. The graph of J H versus [Na ϩ ] was fitted to the Michaelis-Menten equation using the Enzyme Kinetics 1.11 computer program (Trinity Software).
Inhibitors-A number of transport inhibitors were tested for their effect on both the resting pH i and the recovery of pH i following an imposed intracellular acid load. These included amiloride (1 mM The bafilomycin A 1 and vanadate stocks used in this study were shown to be active in control experiments in which they were tested for their effect on the intracellular pH of the malaria parasite Plasmodium falciparum (12) and the yeast Saccharomyces cerevisiae (7), respectively (not shown). 31 P NMR spectrum of G. intestinalis trophozoites suspended in MES-and HEPES-buffered saline is shown in Fig. 1. At an extracellular pH of 6.5 (the reduction of pH o from an initial value of 7.0 to 6.5 was because of the generation of acidic metabolic products), there were two P i resonance peaks, indicating a significant difference between the intracellular and extracellular pH. Peaks attributable to sugar phosphates (2 to 6 ppm) and nucleotides (Ϫ2 to Ϫ8 ppm) were also observed. The largest peak corresponds to the internal standard, TEP. The pH i was estimated from the chemical shift of the intracellular P i resonance (2.28 ppm) to be 6.7 Ϯ 0.05 (n ϭ 3).

P NMR Studies-The
Following the acquisition of the initial spectrum, the suspension was supplemented with either 25 mM arginine, 25 mM glucose, or no metabolic substrate, and the extracellular and intracellular pH were monitored. In all cases, pH o fell. The total changes in pH o (⌬pH o ) were 0.52, 0.79, and 0.78 over 60 min in the presence of 25 mM arginine, 25 mM glucose, or no metabolic substrate, respectively (single measurements). As pH o fell, pH i also decreased, though not to the same extent, reaching a minimum value of 6.4 at an extracellular pH of 5.4 (Fig. 2, open circles).
The generation of acidic metabolic products by the cells confirms their biochemical viability. The ability of the parasites to limit the fall in pH i and maintain a substantial transmembrane pH gradient in an acidic medium illustrates their ability to regulate pH i .
Determination of pH i from [2-14 C]DMO Distribution-On addition of [2-14 C]DMO to G. intestinalis trophozoites suspended in MES-and HEPES-buffered saline, the radiolabel distributed between the intra-and extracellular compartments at a distribution ratio ([DMO] i /[DMO] o ) that was independent of the total DMO concentration over the range 1-100 M (not shown).
The pH i of G. intestinalis trophozoites suspended in medium having a pH range of 6.0 -7.5 was calculated from the DMO distribution ratio measured at a DMO concentration of 10 M (Fig. 2, closed squares). The cells maintained a pH i between 6.7 and 7.0 over a pH o range of 6.0 -7.5. The pH i estimated using DMO for the pH o range 6.0 -6.5 was in close agreement with that estimated using 31 P NMR (Fig. 2). Decreasing the temperature from 37 to 25°C had no significant effect on pH i , which was maintained at 6.7-6.8 over a pH o range of 6.0 -7.5 (data not shown).
Determination of pH i from BCECF Fluorescence-The pH i of G. intestinalis trophozoites in PBS (pH o ϭ 7.2) estimated using the pH-sensitive fluorescent indicator BCECF was in the range 6.7-7.1 with a mean value of 6.95 Ϯ 0.10 (n ϭ 24). As pH o was varied over the range 5.4 -7.7, the pH i estimated using BCECF varied from 6.3-7.2, correlating well with the estimates made using the 31 P NMR and DMO methods (Fig. 2). The fluorescence method proved to be the most straightforward of the three methods tested, allowing on-line monitoring of pH i , and it was therefore adopted as the method of choice for the further characterization of the pH i regulatory mechanisms in these cells.
Na ϩ -dependence of pH i Regulation-When G. intestinalis trophozoites in PBS (pH o ϭ 7.2) were transferred to a Na ϩ -and K ϩ -free medium (NMG-Cl solution) there was a rapid intracellular acidification (Fig. 3A). On addition of NaCl (30 mM) to the medium, the pH i recovered.
When cells subjected to an intracellular acid load using the NH 4 Cl prepulse technique were resuspended in NMG-Cl solution there was no recovery of pH i (Fig. 3B) . Under these conditions, addition of KCl (30 mM) had no effect. By contrast, addition of NaCl (30 mM) resulted in a recovery of pH i to its normal resting value.
To quantify the rate of extrusion of H ϩ from G. intestinalis trophozoites following the acid-loading procedure it was necessary to measure the intracellular buffering power (␤). This was determined to be 32.4 Ϯ 0.8 mM pH Ϫ1 (n ϭ 3, at pH i 6.2-6.3). From this value the rate of pH recovery from an acid load (⌬pH/⌬t) could be converted to the rate of extrusion of H ϩ equivalents (see "Experimental Procedures"). Fig. 4A shows the recovery of pH i in G. intestinalis trophozoites subjected to an intracellular acidification (via an NH 4 Cl prepulse) then resuspended in medium having a range of Na ϩ concentrations. The rate of pH i recovery increased with increasing extracellular Na ϩ . The initial phase of the pH i traces shown in Fig. 4A were fitted to a straight line, the slope of which (⌬pH/⌬t) was converted to a H ϩ equivalent efflux rate (J H ϭ ␤ ϫ ⌬pH/⌬t). J H was a saturable function of [Na ϩ ] o , well described by the Michaelis-Menten expression with a K m and V max of 18 Ϯ 7 mM Na ϩ and 10 Ϯ 2 mM H ϩ min Ϫ1 , respectively (Fig. 4B).
Effect of Inhibitors on pH i Regulation-The Na ϩ /H ϩ exchanger inhibitor amiloride (up to 2 mM) and its analogues EIPA and HMA (up to 200 M) had no effect on the resting pH i of G. intestinalis trophozoites suspended in PBS. As shown in Fig. 5A, the addition of 1 mM amiloride did cause a small, instantaneous decrease in the BCECF fluorescence ratio. However an equivalent decrease was observed when the experiment was repeated for cells killed by heat-fixation (incubation at 60°C for 15 min) and can therefore be attributed to an optical effect of amiloride rather than an effect on pH i . The fluoresecence ratio was unaffected by both EIPA and HMA (each at 200 M).
In contrast to the lack of effect of amiloride on the resting pH Regulation in G. intestinalis pH i , 1 mM amiloride inhibited the pH i recovery following an imposed intracellular acidification (Fig. 5B). EIPA (50 M) and HMA (50 M) were both without effect (not shown).
Addition of the carboxyl-blocking agent DCCD (100 M), to G. intestinalis trophozoites, resulted in a slight decrease in the resting pH i (of between 0.1-0.2 pH units; Fig. 6A). DCCD also inhibited pH i recovery from imposed intracellular acidification (Fig. 6B). The P-type H ϩ -ATPase inhibitor vanadate (up to 200 M), the V-type ATPase inhibitor bafilomycin A 1 (up to 200 nM), and the general ATPase inhibitor NEM (200 M) were all without effect on either the resting pH i or the recovery of pH i from an imposed intracellular acidosis (not shown). DISCUSSION Central to the parasite pathogenicity is its ability to survive the dynamic environment of its host. In the case of G. intestinalis trophozoites, they must withstand exposure to a wide range of pH ranges as found in the gastrointestinal tract. In this study G. intestinalis trophozoites exposed to pH o values in the range 6.0 -7.5 (similar to those experienced in vivo) were shown by three different techniques to maintain a resting pH i of 6.7-7. In suspensions of G. intestinalis trophozoites incubated for up to an hour, the pH o decreased significantly and to a greater extent than the pH i . The decrease in pH o was larger for cells to which either glucose (⌬pH o ϭ 0.79) or no substrate (⌬pH o ϭ 0.78) were added than for cells to which arginine was added (⌬pH o ϭ 0.52). In the absence of exogenous substrate, it is likely that the trophozoites were reliant on endogenous glucose stores (e.g. glycogen, Ref. 25) and that the metabolism was therefore essentially the same as that occurring in cells supplemented with glucose. Glucose metabolism in G. intestinalis results in the production of the acidic end-product acetic acid (26,27), whereas arginine catabolism generates ammonia and ornithine, alkaline and neutral products, respectively (6). The pronounced extracellular acidification observed in suspensions in which the cells were utilizing glucose is likely to be due primarily to the production of acetic acid whereas the lesser acidification observed in cells supplemented with arginine probably represents the combined effects of arginine and glucose (glycogen) metabolism.
The buffering capacity of G. intestinalis was estimated to be 32 mM pH Ϫ1 (at pH 6.2-6.3). This value is in the range reported for other cell types (21) and very close to the value estimated for the protozoan P. falciparum (35 mM pH Ϫ1 , (24)).
Maintenance of a steady-state pH i in G. intestinalis was shown to be Na ϩ -dependent with pH i declining sharply in cells suspended in Na ϩ -free medium (Fig. 3A). The recovery of pH i in cells subjected to an intracellular acid load was also Na ϩdependent (Fig. 3B). Maintenance of the resting pH i and recovery from an acid load were both inhibited by the carboxylblocking agent DCCD and unaffected by the two amiloride analogs, EIPA and HMA. However, amiloride itself was found FIG. 3. Extracellular ion-dependence of the pH i of G. intestinalis trophozoites (A) under resting conditions and (B) following an intracellular acidification. A, BCECF-loaded cells suspended initially in PBS were washed and then resuspended in Na ϩ -free NMG-Cl solution to which was subsequently added 30 mM NaCl (as indicated by the arrow). pH i was monitored fluorometrically. B, BCECF-loaded cells suspended initially in PBS were subjected to an intracellular acid load using the NH 4 ϩ prepulse technique then resuspended (at the point of removal of the NH 4 ϩ in Na ϩ -free NMG-Cl solution, to which was subsequently added (as indicated by the arrows) KCl (30 mM) then NaCl (30 mM). The bar indicates the interval during which the cells were centrifuged (10,000 ϫ g, 15 s) and resuspended in NH 4 Cl-free NMG-Cl solution. pH Regulation in G. intestinalis to inhibit the recovery of pH i from an imposed intracellular acidification whereas having no significant effect on the ability of the parasite to maintain its normal resting pH i . Amiloride, an inhibitor of Na ϩ /H ϩ exchangers in a wide range of cell types is not highly selective and is known to affect a range of biological processes. Nevertheless, the finding that the recovery of pH i from an intracellular acidification, observed in the nominal absence of HCO 3 Ϫ from the medium, is both Na ϩ -dependent and amiloride-sensitive is at least consistent with the pathway involved being a type of Na ϩ /H ϩ exchanger. The flux of H ϩ via this pathway was a saturable function of the extracellular Na ϩ concentration, with an apparent K m for Na ϩ of 18 mM. This value is comparable with that reported for Na ϩ /H ϩ exchangermediated H ϩ extrusion from vertebrate cells: e.g. thymic lymphocytes (59 mM, 28), apical membrane cells of amphibian gall bladder (11 mM,Ref. 29) and dog kidney cells (42 mM, Ref. 30).
The finding that the resting, steady-state pH i is unaffected by amiloride would suggest that the maintenance of a normal resting pH i is achieved via a H ϩ extrusion mechanism that differs from that which plays the major role in the efflux of H ϩ following an intracellular acidification. Although amilorideinsensitive, this pathway is Na ϩ -dependent and DCCD-sensitive. The available data are consistent with it being an amiloride-insensitive Na ϩ /H ϩ exchanger. Higher eukaryotes are known to have number of Na ϩ /H ϩ exchanger isoforms (the NHE family; Ref. 31) and it is not uncommon for vertebrate tissues to express more than one Na ϩ /H ϩ isoform, having different Na ϩ affinities, pH dependences and amiloride sensitivities (e.g. NHE1 is known to be sensitive to amiloride whereas NHE3 is resistant to amiloride and amiloride analogues, Ref. 31). Lower eukaryotes, such as S. cerevisiae, also contain more than one putative Na ϩ /H ϩ exchanger (e.g. Ref. 32) and although the pharmacology of these different transporters has not been fully elucidated, differences between isoforms might be expected. The possibility of there being more than one Na ϩ /H ϩ exchanger serving somewhat different functions in G. intestinalis is therefore not unprecedented.
In some protozoa, the major H ϩ efflux mechanisms are thought to be electrogenic H ϩ pumps (e.g. Refs. 10 -12). However, the pronounced Na ϩ -dependence of H ϩ extrusion from G. intestinalis trophozoites, together with the lack of significant inhibitory effect of the H ϩ pump inhibitors NEM (a general H ϩ pump inhibitor), bafilomycin A 1 (a specific V-type H ϩ pump inhibitor), or vanadate (a specific P-type H ϩ pump inhibitor) on either the resting pH i or the extrusion of H ϩ following an intracellular acidification would suggest that such pumps do not play a significant role in pH i regulation in G. intestinalis trophozoites.
DCCD is perhaps best characterized as an inhibitor of electrogenic H ϩ pumps and an effect of DCCD on pH i is often taken as being indicative of the involvement of this class of proteins in the pH regulation. However, DCCD is a general carboxyl blocking agent, which has been shown to inhibit Na ϩ /H ϩ exchangers of higher organisms (e.g. Refs. 33,34). The effect of DCCD on both the maintenance of the normal resting pH i and the recovery from an intracellular acid load is therefore consistent with the postulated involvement of Na ϩ /H ϩ exchangers in both processes.
There are, to our knowledge, two previous reports of the presence of a Na ϩ /H ϩ exchanger in parasitic protozoa. It has been proposed that in P. falciparum trophozoites a Na ϩ /H ϩ exchanger serves as the major pathway for the extrusion of H ϩ . The apparent K m for Na ϩ in this system was estimated to be ϳ6 mM; however, this analysis was complicated by the presence of both a saturable and a linear transport component (24) and, in any case, the role of a Na ϩ /H ϩ exchanger in pH i regulation in this organism has recently been questioned (12). Leishmania promastigotes have also been proposed to possess an Na ϩ /H ϩ   FIG. 5. Effect of amiloride on pH i of G. intestinalis trophozoites. A, the resting pH i of BCECF-loaded cells in PBS was monitored before and after the addition of amiloride (1 mM). The slight, instantaneous decrease in pH i on addition of amiloride is the result of an optical effect of amiloride on the BCECF fluorescence ratio rather than an effect on pH i . B, cells in PBS buffer were acidified by the NH 4 ϩ prepulse technique. Recovery of pH i was monitored in PBS in the presence and absence of amiloride (1 mM). The bar indicates the interval during which the cells were centrifuged (10,000 ϫ g, 15 s) and resuspended in NH 4 Cl-free solution. exchanger involved in pH control (35); however, this too has been disputed (36).
In summary, the results of the present study demonstrate the presence in G. intestinalis trophozoites of active H ϩ extrusion mechanisms that act to maintain the pH i within an appropriate range. Available evidence indicates the presence of at least two Na ϩ -dependent H ϩ extrusion mechanisms. One is insensitive to amiloride and serves a housekeeping role in maintaining the normal resting pH i (in the face of a constant metabolic acid-load). The other is amiloride-sensitive and is activated in response to a cytosolic acidification. The two pathways are postulated to be members of the Na ϩ /H ϩ exchanger family.
Na ϩ /H ϩ exchangers are ubiquitous from higher eukaryotes to bacteria (31). In higher eukaryotic cells the stoichometry of the Na ϩ /H ϩ exchange is 1:1 (37) whereas in lower eukaryotes and bacteria Na ϩ /H ϩ antiporters operate with a Na ϩ /H ϩ ratio of 1:2 (38). The biochemical characteristics of Giardia are often regarded as being a hybrid of those of eukaryotes and prokaryotes (6), with the antiquity of this organism being the focus of much interest (39). Further characterization of the Na ϩ -dependent H ϩ extrusion mechanisms shown in the present study to operate in Giardia trophozoites, and the identification of the proteins involved may therefore be of significant interest in terms of the evolution of Na ϩ /H ϩ exchangers.