Na+/H+ exchangers (NHE) are ubiquitous transporters participating in regulation of cell volume and pH. Cell shrinkage, acidification, and growth factors activate NHE by increasing its sensitivity to intracellular H+ concentration. In this study, the kinetics were studied in dog red blood cells of Na+ influx through NHE as a function of external Na+ concentration ([Na+]o). In cells in isotonic media, [Na+]o inhibited Na+ influx >40 mM. Osmotic shrinkage activated NHE by reducing this inhibition. In cells in isotonic media + 120 mM sucrose, there was no inhibition, and influx was a hyperbolic function of [Na+]o. The kinetics of Na+-inhibited Na+ influx were analyzed at various extents of osmotic shrinkage. The curves for inhibited Na+ fluxes were sigmoid, indicating more than one Na+ inhibitory site associated with each transporter. Shrinkage significantly increased the Na+ concentration at half-maximal velocity of Na+-inhibited Na+ influx, the mechanism by which shrinkage activates NHE.
- cell volume regulation
- kinetics of sodium ion influx
na+/h+ exchangers (nhe) are ubiquitous transporters. They exchange an external Na ion for an intracellular proton in an obligatory, electroneutral, and reversible fashion. Physiologically, the inwardly directed Na+ gradient provides the driving force for outward transport of H+ (22, 34). Eight mammalian isoforms of NHE have been described to date, NHE-1 to NHE-8 (5, 7, 13, 20, 23, 29, 31, 35). NHE-1 is the most widely distributed isoform (29), responsible for regulation of pH in most cells (34). NHE-1, in parallel with the Cl−/HCO3− exchanger, can promote the influx of NaCl and osmotically obliged water, thereby contributing to the regulation of cell volume. NHE-1 has a membrane topology similar to the other NHE isoforms, with 12 membrane-spanning domains and a large cytoplasmic COOH-terminal domain responsible for regulation of transport (22). NHE-1 is the isoform in dog red blood cells (1), the object of this study.
NHE-1, like other isoforms, is stimulated by shrinkage and by acidification of the cells and also by growth factors. Intracellular acidification activates NHE-1 through an allosteric site, where H+ binding causes an alkaline shift in the set point for activation of NHE by H+ (3). Growth factors and osmotic shrinkage also shift the set point for activation of the exchanger by intracellular H+ (9, 19, 24). The molecular identity of this allosteric H+ binding site is not known, but structure and function studies have provided clues about its location in the NHE molecule and also of the location of the shrinkage-sensitive domains (4, 33). Activation of NHE-1 by growth factors involves phosphorylation at multiple sites on the transporter (28). However, increased phosphorylation of NHE-1 does not accompany activation by cell shrinkage (10).
We report here studies on the kinetics of activation of NHE-1 in dog red blood cells by external Na+. Mature dog red blood cells are unusual in lacking Na+-K+ pumps and having intracellular Na+ and K+ near their plasma concentrations (25). This property is shared with cat red blood cells and may be characteristic of red blood cells of carnivores in general. At physiological external Na+ concentration ([Na+]o), there is little Na+ gradient, and at low [Na+]o, there is a large, outwardly directed Na+ gradient.
At low [Na+]o in isotonic media, there was activation of NHE by increasing [Na+]o. At [Na+]o >40 mM in isotonic media, there was a progressive decline in NHE as though Na+ inhibited it. In osmotically shrunken cells (isotonic media + 120 mM sucrose), NHE was greatly stimulated, >10-fold at 150 mM [Na+]o. In these shrunken cells, there was no inhibition of NHE at high [Na+]o, and Na+ influx through NHE was a hyperbolic function of [Na+]o, typical of activated NHE in other cells (16, 18, 21, 34) and consistent with interaction of Na+ with a single external substrate site per transporter. K1/2, the Na+ concentration at half-maximal activation of NHE, was 58.5 mM in shrunken dog red blood cells. The results suggest that there are external sites on dog red cells where Na+ inhibits NHE, and shrinkage activates NHE in part by reducing this inhibition by [Na+]o. The K1/2 for Na+ binding to these inhibitory sites in isotonic media was 82.3 mM.
MATERIALS AND METHODS
Red blood cells.
Blood was drawn by venipuncture into heparinized containers from beagles maintained at Marshall Farms USA, North Rose, NY. The red cells were washed free of plasma and white cells by three brief, successive centrifugations and resuspensions in an isotonic, physiological medium (290 mosmol/kgH2O, measured by using a vapor pressure osmometer; Wescor, Logan, UT) containing (in mM) 145 NaCl, 5 KCl, 10 Tris·HCl, 5 glucose, pH 7.4. The cells were used the first or second day after the blood was drawn.
Unidirectional influxes of Na+ were measured by using 22Na as a tracer (PerkinElmer, Boston, MA). Cells were washed in media appropriate for the experiments; the washing required 5 min. The fluxes were measured over 10 min; all measurements were made in triplicate. NHE was taken as the amiloride-inhibitable Na+ influx (1 mM amiloride). Two amiloride derivatives, 5-(N,N-hexamethylene)-amiloride (HMA) and 5-(N-ethyl-N-isopropyl)amiloride (EIPA), specific inhibitors of NHE (17), were used to confirm that the amiloride-inhibitable Na+ influx is NHE. N-methyl-d-glucamine (NMDG+) was used as the substitute cation for Na+. The methods for calculating the influxes were slight modifications of earlier methods (27). Fluxes are expressed as millimoles per liter cells per hour when performed in isotonic media. Fluxes measured in shrunken cells were corrected to the original, physiological cell volume by using the hemoglobin concentrations of the flux samples. These fluxes are expressed as millimoles per original liter cells per hour.
It is important that the influx of Na+ be a linear function of time when the measurements of fluxes are made. Figure 1 shows the time course of amiloride-inhibitable Na+ influxes during 8 min in cells in 150 mM Na+, both isotonic and hypertonic. The data were well fit by straight lines. One other experiment of similar design gave the same results.
Na+ influxes vs. [Na+]o were fitted to the Hill equation using a nonlinear least-squares procedure (SigmaPlot, Jandel Scientific, San Rafael, CA). This yielded estimates of the kinetic constants ± asymptotic SE. The Hill equation is (1) where J and Jmax are Na+ influx and maximal Na+ influx, respectively, and nH is the Hill coefficient, a function of the number of interacting sites. The nH equals the number of sites only if cooperativity between sites is strong. K′ is a constant comprising 1) the interaction factors between the sites, 2) an apparent dissociation constant, and 3) nH. K1/2 equals K′nH (30).
The method for measuring intracellular pH was a modification of earlier methods (1, 6). After the desired treatments, 0.5-ml cell samples were packed by centrifugation. The cells were suspended in 15 volumes of unbuffered isotonic medium and packed again. The packed cells were lysed by three cycles of freezing in dry ice/methanol and thawing in a room temperature water bath. The lysates were diluted with 1 ml of deionized water. The pH of each lysate was determined in triplicate by using a combination electrode.
Percent cell water.
Cells, ∼0.4 ml, were packed in 1.5-ml tared microfuge tubes by centrifugation for 5 min in a Fisher Scientific microcentrifuge. The supernatant solutions were carefully aspirated from the tubes, along with the top ∼1 mm of cells. The tubes were weighed, giving wet weights. The tubes were dried to constant weight at 80°C and weighed again, giving dry weights. A correction was made for loss of ∼1.5 mg weight of the tubes during drying, determined by using two empty tubes in each experiment. All measurements were made in triplicate. Dry weights were corrected to dry volume by using a density of 1.17 g/ml for dry cell contents, mostly hemoglobin (15). This permitted expression of percent cell water (volume/volume).
Intracellular Na+ concentrations.
Cell samples of ∼0.04 ml were washed three times in isotonic NMDG Cl and lysed in 5 ml of deionized water. The samples were analyzed by using a PerkinElmer AAnalyst 100 atomic absorption spectrometer in the emission mode. Standards of 20–80 μM Na+ were used.
Differences among more than two means were analyzed by using a one-factor ANOVA and the Fisher protected least significant difference test between pairs of means.
Na+ influx vs. [Na+]o in cells in isotonic media.
Figure 2 shows unidirectional Na+ influxes in dog red blood cells in isotonic media over a range of [Na+]o values, total influxes, and influxes plus amiloride. Total Na+ influx increased with increasing [Na+]o to ∼40 mM [Na+]o. At higher [Na+]o values, influx declined with increasing [Na+]o, a surprising result. The first question that these results raise is if the Na+ influx activated by [Na+]o up to 40 mM and inhibited by amiloride is, in fact, mediated by NHE.
Inhibition of Na+ influx by amiloride analogs.
The Na+ influxes at low [Na+]o in isotonic medium may be mediated by an amiloride-sensitive Na+ channel, which functions only at low [Na+]o; amiloride is not fully specific for NHE. Effects of amiloride were compared with those of two amiloride derivatives, HMA and EIPA, specific inhibitors of NHE (17), each at 0.1 mM, concentrations fully inhibitory of rat NHE-1 (21). Figure 3 shows that, from 2 to 40 mM [Na+]o, inhibition by amiloride, HMA, and EIPA was all about the same. These results show that the amiloride-inhibitable flux is NHE.
Amiloride-sensitive Na+ flux at low [Na+]o is NHE, but there are two possible explanations for the apparent activation of NHE by Na o+, which would mean that the decline of NHE at [Na+]o >40 mM is not inhibition of NHE. These possibilities are 1) activation of NHE at low [Na+]o due to acidification of the cells, owing to NHE running in reverse, and 2) activation of NHE at low [Na+]o due to cell shrinkage caused by net Na+ efflux. These causes of activation of NHE would be reduced as [Na+]o increased and the outward Na+ gradient decreased. This decrease would appear as inhibition of NHE at higher [Na+]o.
Activation of NHE by acidification?
At low [Na+]o, NHE runs in reverse, owing to the high cell Na+ concentration ([Na+]c) and the outwardly directed Na+ gradient. This may cause acidification of the cells that would activate NHE. To test this, intracellular pHs were measured in cells in isotonic media over a range of [Na+]o values after incubation for 5–30 min with or without amiloride. There was no suggestion of a change in pH with either time or [Na+]o (Table 1); there was no hint of acidification as time proceeded. Protons taken up by reversed NHE at low [Na+]o are buffered, and the activation of NHE at low [Na+]o is not due to acidification of the cells. As will be shown below, the Na+ loss in 2 mM [Na+]o in 15 min is a little less than 10 mmol/l cells. A corresponding H+ influx would mean about two protons per hemoglobin molecule (∼5 mM hemoglobin in mammalian erythrocytes). Results with amiloride were indistinguishable from the controls.
Activation of NHE by cell shrinkage at low [Na+]o?
If there is net Na+ efflux at low [Na+]o, resultant cell shrinkage could activate NHE and be responsible for the Na+-activated NHE at low [Na+]o. We first looked for net Na+ effluxes at 100, 40, and 2 mM [Na+]o. Figure 4 shows [Na+]c at 40 and 2 [Na+]o during a 15-min incubation. There was no significant loss of Na+ at 100 or 40 mM [Na+]o in 15 min (100 mM results not shown). In 2 mM [Na+]o, there was a loss of ∼10 mmol/l cells of cell Na+ in 15 min. The efflux was completely inhibited by amiloride and, therefore, was probably NHE running in reverse.
We next measured percent cell waters after incubation for 10 min at 100, 50, and 2 mM [Na+]o. Table 2 (constant total solute) shows percent cell waters in cells in isotonic (so-called) medium with 100, 50, and 2 mM [Na+]o after incubation for 10 min. (No significance is to be attached to the use of 50 mM Na+ in these experiments and 40 mM in Fig. 4; the results would have been essentially the same if either concentration had been used in both experiments.) Cells shrank 2.5% at 2 mM [Na+]o compared with cells at 100 mM [Na+]o; the volume decrease was statistically significant. This extent of shrinkage may account for the activated NHE at low [Na+]o. To test this, true isotonic media were made by finding the NMDG Cl concentrations at 50 and 2 mM [Na+]o that cause no cell shrinkage. Table 2 (varying total solute) also shows the composition of these media and the percent cell waters in cells incubated in them. Cell volume was constant from 100 to 2 mM [Na+]o. The percent cell waters were highly reproducible within and among experiments, thus the low SEs in Table 2. Next, Na+ influxes were compared in the two kinds of media: one with fixed total solute concentration and the other with total solute concentration reduced as [Na+]o was reduced to maintain cell volume constant.
Unidirectional Na+ influxes in low [Na+]o media with fixed and varied total solute.
Amiloride-inhibitable unidirectional Na+ influxes were compared in the media with constant solute concentration and in the true isotonic media in which percent cell water was constant over a range of Na o+ values. These results, in Fig. 5, show that the NHE flux was the same in the two media.
Therefore, the Na+-activated Na+ influx at low [Na+]o mediated by NHE is not due to shrinkage, resulting from Na+ loss. As just shown, it was also not due to acidification (Table 1). It is simple activation of NHE by [Na+]o. Therefore, the decline in Na+ influx >40 mM [Na+]o is inhibition of NHE by Na+.
NHE is inhibited by external Na+.
Figure 4 showed that there is no loss of Na+ in 15 min from cells in 40 mM [Na+]o. Therefore, as [Na+]o is raised from 40 mM and NHE is inhibited, [Na+]c remains constant (it did not decrease as [Na+]o was lowered). Therefore, the inhibition of exchange with the increase in [Na+]o is due to Na+ acting on the exchanger at external sites.
Na+ influx in cells in hypertonic media.
Figure 6A shows unidirectional Na+ influxes in cells shrunken in hypertonic medium, 415 mosmol/kgH2O (isotonic medium and 120 mM sucrose) over a range of [Na+]o values, total influx (solid circles), and influx with amiloride (solid squares). The fluxes in isotonic media from Fig. 2 are included for comparison (open circles and squares). Osmotic shrinkage greatly stimulated total Na+ influx. Shrinkage was without effect on the amiloride-insensitive Na+ influx.
NHE in isotonic and hypertonic media.
Figure 6B shows the amiloride-inhibitable fluxes (solid symbols) calculated from the results in Fig. 6A. For cells in isotonic media, there was both activation of NHE and, at higher [Na+]o, inhibition of NHE. The NHE flux is not zero at 150 mM [Na+]o. The curve for hypertonic media was fitted using the Hill equation; there was a good fit to the data. The K1/2 for Na+ was 58.5 ± 4.4 mM, and the nH was 0.968 ± 0.038 (±asymptotic SE). The nH is not significantly different from one, and the curve is hyperbolic. There is no hint of inhibition of Na+ influx at high [Na+]o. The K1/2 is at the high end of the range for NHEs, 3–50 mM (2).
Almost all of the increase in influx caused by shrinkage can be accounted for by relief of the inhibition of NHE by Na o+. The part of flux in hypertonic media that is not due to the relief of inhibition is the flux in isotonic medium. To obtain the Na+ flux that is due to relief of inhibition, the individual fluxes in isotonic medium were subtracted from the calculated fluxes in hypertonic media. The difference curve for relief of inhibition is given by the open symbols. The difference curve was fitted using the Hill equation, giving a good fit to the data. The K1/2 was 82.3 ± 2.2, and the nH was 1.43 ± 0.02. This K1/2 for inhibition (K1/2 i) was greater than the K1/2 for activation (K1/2 a), 58.5 mM, and Na+ inhibits at sites with a lower apparent affinity than that of the sites at which it activates. The nH was greater than unity, indicating more than one inhibitory site for Na+ on each transporter. The Jmax for activation by 120 mM sucrose was 161 ± 8 mmol·original liter cells−1·h−1. The Jmax for the difference curve was 146 ± 2 mmol·original liter cells−1·h−1. The Jmax for the difference curve was less than the Jmax for hyperosmotic activation because of the positive NHE in isotonic medium at 150 mM [Na+]o. If the NHE flux in isotonic medium had decreased to zero at 150 mM [Na+]o, then the two Jmax values would likely have been the same. The difference between the Jmax values suggests that NHE in isotonic medium does not go to zero as [Na+]o is raised above 150 mM.
The curve for NHE in isotonic media fitted by using constants for activation and inhibition.
The following equation, with components of both activation and inhibition, was used to fit the data for NHE in isotonic media: (2) where J is NHE influx, Jmax a is the maximum NHE, K1/2 a is for NHE, Jmax i is the Jmax of the difference curve for inhibition (Fig. 6B), and Ki is a constant from Eq. 1 for inhibition comprising a dissociation constant for Na+, the nH, and interaction factors between the Na+ sites. Ki1/nH = K1/2 i (30). With the constants obtained from the fits to the curves in Fig. 6B, the equation becomes (3) where Jmax a and Jmax i are unknowns. Equation 3 was solved for the Na+ fluxes in isotonic media in Fig. 6B. The results of the fit are shown in Fig. 7. The fit to the data was good. Jmaxa and Jmax i, 145.3 ± 1.5 and 138.0 ± 1.8 mmol·original liter cells−1·h−1, respectively, were within 11 and 6%, respectively, of the Jmax values determined in Fig. 6B. The good fit of the curve to the data supports the conclusion that the activation of NHE by shrinkage is due to relief of inhibition of exchange by Na+ in isotonic media. This is a central finding of this study.
Activation of NHE in hypertonic media with 25–100 mM sucrose.
NHE was determined in media with osmolalities intermediate between those in Fig. 6B. The intent was to determine the effect of shrinkage on K1/2i. NHE fluxes in media made hypertonic with 25, 50, 75, and 100 mM sucrose are shown in Fig. 8. The results for 0 and 120 mM sucrose from Fig. 6B are included. There was inhibition of NHE by [Na+]o at all four new sucrose concentrations indicated by the decreased fluxes at high [Na+]o. The [Na+]o at which the inhibition becomes apparent occurs at higher [Na+]o values with increased sucrose concentration. This is consistent with an increase in K1/2 i provoked by increased shrinkage.
Difference curves for fluxes in 25–100 mM sucrose media.
The curve for isotonic media in Fig. 6B was subtracted from the curve in 120 mM sucrose media to give a curve, called a difference curve, for the flux inhibited by Na o+ in isotonic media and activated by shrinkage in the hypertonic media (Fig. 7). Figure 9 shows the corresponding difference curves for 25, 50, 75, and 100 mM sucrose media, obtained by subtraction of the curves in Fig. 8 for 25–100 mM sucrose from the calculated curve for 120 mM sucrose media. The difference curve for zero sucrose from Fig. 7 was included for comparison. The curves drawn through the points in Fig. 9 are from Hill equation fits to the data, assuming a Jmax of 146 mmol·liter cells−1·h−1, calculated for the zero sucrose difference curve. This assumes that Jmax did not change with shrinkage, the assumption that led to the good fit in Fig. 7. The K1/2i values for the 25 and 50 mM sucrose curves are given in Table 3, along with that determined above for zero sucrose, isotonic media. Reasonable estimates for the constants for the 75 and 100 mM sucrose curves could not be made because the slopes of these curves are still increasing at 150 mM [Na+]o. The K1/2 i values were 15 and 50% higher at 25 and 50 mM, respectively, than in isotonic media. The differences were statistically significant. Therefore, shrinkage activates NHE by reducing the apparent affinity for Na+ at the inhibitory sites on NHE.
NHE vs. osmolality at 150 mM [Na+]o.
Figure 10 shows NHE-mediated Na+ influxes at 150 mM [Na+]o in media with 0–120 mM sucrose taken from Fig. 8. The increased flux with increasing sucrose concentration is due to the shrinkage activation of NHE and the decrease in K1/2 i. Maximum slope of the curve is at the higher sucrose concentrations, consistent with a sigmoid curve and with the fact that there is more than one inhibitory site for Na+ on each exchanger. The curve could not be fitted to the Hill equation because the slope is near maximal at the highest sucrose concentration. Corresponding curves at lower [Na+]o values from Fig. 8 are also sigmoid and also could not be fitted using the Hill equation.
The rate of Na+/H+ exchange in dog red blood cells is relatively low in isotonic media and is activated >10-fold by osmotic shrinkage in hypertonic medium (isotonic medium + 120 mM sucrose) at physiological [Na+]o. The rate of NHE is low in isotonic medium in part because of inhibition by Na+ at external sites on the exchanger. Osmotic shrinkage activates NHE by reducing this inhibition and abolishing it in 120 mM sucrose media. There is more than one inhibitory site interacting with each transporter. The K1/2 i for the interaction of Na+ with inhibitory sites increases significantly with cell shrinkage: 15% in 25 mM sucrose and 55% in 50 mM sucrose media. In 120 mM sucrose media (415 compared with 290 mosmol/kgH2O of isotonic media), the K1/2 i increases to the extent that inhibition by Na o+ is not detectable. Therefore, shrinkage inactivates inhibition by Na+ by reducing the affinity for Na+ at the inhibitory sites. This is the first report of a role for inhibition by Na+ in the regulation of NHE. Also involved in this regulation is the well-known stimulation of NHE by the shrinkage-induced shift in the affinity for intracellular H+ (3).
The hyperbolic kinetics of NHE in 120 mM sucrose media are consistent with activation of NHE by Na+ at a single site on each exchanger. Hyperbolic kinetics for Na+ influx via NHE activated by acidification or shrinkage have been reported in several studies on the NHE-1, -2, and -3 isoforms (16, 21, 34, 36). The K1/2 a values (K1/2 for activation) ranged from 10 to 50 mM [Na+]o. The K1/2 a in dog red blood cells is at the high end of this range.
The Jmax measured in 120 mM sucrose media may not be the maximum velocity of the transporter. Measurements in more hypertonic media were attempted, but acceptable results could not be obtained, owing to the high rates of transport and distortion of the curve at the high [Na+]o values. The kinetic constants might be quantitatively different at higher Jmax values, but the general conclusions would be the same.
There are few results in the literature on the kinetics of Na+ influx through unstimulated NHE. In rat thymocytes, total Na+ influx was not much greater than the amiloride-insensitive flux, so the kinetics of unstimulated NHE could not be characterized (8). NHE was measured in unstimulated Chinese hamster lung fibroblasts at 130, 15, and 1 mM [Na+]o. There was measurable NHE at 1 mM [Na+]o, but not at the higher [Na+]o values (24). This effect at 1 mM Na+ was attributed to an intracellular acidification by 0.15 pH units in cells at low [Na+]o. Intracellular buffering to pH 7.35 abolished the stimulation of NHE at 1 mM [Na+]o. In the present work on dog red blood cells, it is clear that the stimulation of NHE at low [Na+]o is not due to acidification.
Extracellular Li+ interacts in a complex manner with NHE of brush-border membrane vesicles from rabbit renal cortex (11). Li+ inhibits at the extracellular surface by competing with Na+; Li+ has a lower K1/2 than Na+ (by ∼12-fold) and a lower Jmax (by ∼3-fold). In addition, Li+ interacts with a separate modifier site, where it inhibits NHE noncompetitively with a very high affinity (Ki ∼ 50 μM). Amiloride also binds to this site, resulting in noncompetitive inhibition of NHE. There was no evidence for Na+ binding to this modifier site (11), making it unlikely that this site is the same as the Na+ inhibitory site of the present study. K+ is a competitive, dead-end inhibitor of NHE-1 with a very low affinity (K1/2 ∼ 180 mM) (21). There is an early report of slight inhibition of NHE by Cs+ (14) and another report of no effect of Cs+ on NHE (12). There is evidence for inhibition of NHE by extracellular Ca2+ (26, 32), although at which site(s) is not known.
When the intracellular allosteric H+ sites are unoccupied, the NHE flux is low. Acidification or shrinkage caused H+ binding at these sites and stimulation of NHE (3, 9, 19, 24). The newly described allosteric inhibitory Na+ sites also maintain NHE low in isotonic medium but when they are occupied by Na+. Cell shrinkage reduces Na+ affinity at these sites, and NHE is stimulated. Therefore, two allosteric sites participate in shrinkage activation of NHE. The H+ site leads to activation when it is occupied, and the Na+ site does so when it is not. The relationship between the allosteric H+ and Na+ sites remains to be elucidated.
This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grant R37 DK-33640.
We thank Dr. John M. Russell for a careful reading of the manuscript. A preliminary report of this work was presented at a meeting of The Red Cell Club in New Haven, CT, on October 24, 2003.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2004 the American Physiological Society