Mechanisms of ammonia and ammonium transport by rhesus-associated glycoproteins

Abstract

In this study we characterized ammonia and ammonium (NH₃/NH₄⁺) transport by the rhesus-associated (Rh) glycoproteins RhAG, Rhbg, and Rhcg expressed in Xenopus oocytes. We used ion-selective microelectrodes and two-electrode voltage clamp to measure changes in intracellular pH, surface pH, and whole cell currents induced by NH₃/NH₄⁺ and methyl amine/ammonium (MA/MA⁺). These measurements allowed us to define signal-specific signatures to distinguish NH₃ from NH₄⁺ transport and to determine how transport of NH₃ and NH₄⁺ differs among RhAG, Rhbg, and Rhcg. Our data indicate that expression of Rh glycoproteins in oocytes generally enhanced NH₃/NH₄⁺ transport and that cellular changes induced by transport of MA/MA⁺ by Rh proteins were different from those induced by transport of NH₃/NH₄⁺. Our results support the following conclusions: 1) RhAG and Rhbg transport both the ionic NH₄⁺ and neutral NH₃ species; 2) transport of NH₄⁺ is electrogenic; 3) like Rhbg, RhAG transport of NH₄⁺ masks NH₃ transport; and 4) Rhcg is likely to be a predominantly NH₃ transporter, with no evidence of enhanced NH₄⁺ transport by this transporter. The dual role of Rh proteins as NH₃ and NH₄⁺ transporters is a unique property and may be critical in understanding how transepithelial secretion of NH₃/NH₄⁺ occurs in the renal collecting duct.

the mammalian rhesus-associated (Rh) glycoproteins are members of the solute transporter family SLC42 and include RhAG, which is present exclusively in erythrocytes, and two nonerythroid members, RhBG and RhCG. RhAG is one component of the erythrocyte “Rh complex” that is mostly known for its antigenicity but also has an important role in maintaining the stability and structure of the red cell membrane. RhBG and RhCG are expressed in several tissues, including kidney, liver, skin, and the gastrointestinal tract (5–8, 10, 25, 29). RhBG and RhCG (mouse Rhbg and Rhcg, respectively) are mostly expressed together in the same epithelial cell type but with opposing polarity. Several studies over the past decade have linked the function of Rh proteins to transport of NH₃ and/or NH₄⁺ (for review see Refs. 2, 9, 23, and 28). Rhbg and Rhcg are expressed in tissues where transport of NH₃/NH₄⁺ is likely; however, a definitive role of Rh glycoproteins as transporters is yet to be determined.

Initial studies of Amt, the bacterial homolog to the Rh proteins (14), and later of MEP, the yeast homolog (13), were the first to infer that Rh proteins may be involved in ammonia transport. In later studies on Rh proteins, four possible mechanisms of transport were proposed. Westhoff et al. (30) first proposed that these proteins served as electroneutral countertransporters of NH₄⁺ coupled to H⁺ efflux. These studies were based on their measurements of methyl amine/ammonium (MA/MA⁺) uptake (as surrogates for NH₃/NH₄⁺) in Xenopus laevis oocytes expressing RhAG (30). In effect, RhAG would be transporting net NH₃ equivalents without causing any changes in intracellular pH (ΔpH_i = 0). Several studies conducted in oocytes and mammalian cells expressing Rhbg were in agreement with the conclusion of Westhoff et al. (11, 12, 31). A second proposed mechanism was that Rh glycoproteins transport only the neutral NH₃ gas species and not NH₄⁺. Studies on liposomes in which RhCG was reconstituted demonstrated an increase in NH₃ permeability but no effect on NH₄⁺ permeability (15). Another study involving surface pH (pH_s) measurements in oocytes expressing RhCG also concluded that NH₃ was being transported (17). If this were indeed true, RhCG would be transporting net NH₃ across the oocyte membrane, causing alkalinization of intracellular pH (pH_i). A third proposed mechanism was that Rh glycoproteins promoted efflux of NH₄⁺. This was based on a study conducted by Marini et al. (13), who found a higher rate of extracellular accumulation of NH₄⁺ in yeast cells expressing RhAG or RhCG (initially referred to as RhGK) than in controls. Finally, Benjelloun et al. (4) reported that HeLa cells expressing RhAG transport both NH₃ and NH₄⁺. Other studies on RhCG (3) and Rhbg (18, 19, 21, 24) also proposed NH₃ and NH₄⁺ transport. The complexity of studying NH₃/NH₄⁺ transport was discussed by Musa-Aziz et al. (17), who analyzed five proposed models of NH₃/NH₄⁺ transport in oocytes expressing the bacterial Rh homolog AmtB. The uncertainty as to how exactly Rh glycoproteins transport NH₃/NH₄⁺ is evident in the numerous models proposed by various researchers. We conducted functional studies employing a group of simultaneous or parallel direct measurements (current, voltage, pH_i, and pH_s) to detect specific signatures that distinguish NH₃ from NH₄⁺ transport. This approach allowed us to characterize the mechanism of NH₃/NH₄⁺ transport by RhAG, Rhbg, and Rhcg and to show how the mechanisms each employ differ from one another.

MATERIALS AND METHODS

Solutions and Reagents

Solutions were prepared using salts and reagents purchased from Sigma (St. Louis, MO) unless otherwise noted. ND96 frog oocyte Ringer solution [in mM: 96 NaCl, 2 KCl, 1.8 CaCl₂, 1 MgCl₂, and 5 HEPES (N₂C₈H₁₈SO₄) buffer] was used as the standard medium that bathed the oocyte. In this solution, 5 mM NaCl was replaced with 5 mM NH₄Cl or 5 mM methylammonium hydrochloride (CH₃NH₄·HCl) to make 5 mM ammonium and methylammonium HEPES oocyte Ringer solution, respectively. Oocytes were stored in OR3 (Leibovitz) medium (GIBCO BRL) containing glutamate, 500 U each of penicillin and streptomycin, and 5 mM HEPES buffer, with pH adjusted to 7.5. The pH of all solutions used to perfuse the bath was adjusted to 7.5 at room temperature (22°C) using 5 and 1 N HCl and NaOH. The osmolality of all solutions was ∼200 ± 10 mosM (mmol/kg). Before it was used to anesthetize X. laevis frogs, 0.2% tricaine (C₉H₁₁NO₂·CH₄SO₃) solution (7.65 mM tricaine and 5 mM HEPES) was adjusted to pH 7.1–7.4 at room temperature. A buffer solution, pH 7.0, containing 0.015 M NaCl, 0.023 M NaOH, and 0.04 M KH₂PO₄ was used to backfill silanized pH-sensitive microelectrodes (1).

cRNA Synthesis

RhAG, Rhbg, and Rhcg cDNA constructs in pGH19 expression vector were used to prepare cRNA of the respective clones. cDNA was linearized with NotI or an appropriate restriction enzyme that could cleave downstream of the insert to produce a linear template. This was followed by proteinase K digestion (1 mg/ml). Linearized cDNA was purified with the QIAquick PCR purification kit (Qiagen). Capped cRNA was transcribed in vitro from the linearized cDNA constructs with T3, T7, or the appropriate RNA polymerase using the mMessage mMachine kit (Ambion, Austin, TX) according to the manufacturer's instructions. cRNA was purified and concentrated with the RNeasy MinElute RNA Cleanup Kit (Qiagen). The concentration of cRNA was determined by UV absorbance, and its quality was assessed by formaldehyde-MOPS-1% agarose gel electrophoresis.

Isolation of Oocytes

Oocytes in stages 5/6 were harvested as described by Nakhoul et al. (20). Briefly, X. laevis frogs were anesthetized in 0.2% tricaine. A 1-cm incision was made in the abdominal wall, first through the abdominal skin and then through the muscular plane of the peritoneum. A lobe of the ovary, which contains the oocytes, was externalized through the incisions, and the distal portion was cut. For wound closure, the lips of the incisions were joined by suturing the muscular plane of the peritoneum using 4-0 chromic gut and then suturing the abdominal skin using 6-0 silk. The excised lobe of the ovary was rinsed several times in Ca²⁺-free ND96 solution until the solution was clear. Separation of the oocytes was achieved by agitation of the ovary in ∼30 ml of sterile-filtered Ca²⁺-free solution containing 3–5 mg of collagenase (type IA) for 30–40 min. Free oocytes were rinsed multiple times with sterile OR3 medium, sorted to allow selection of the mature oocytes, and stored at 18°C. All animal procedures were performed in accordance with protocols approved by the Institutional Animal Care and Use Committee of Tulane University.

Injection of Oocytes

Sorted oocytes stored in OR3 medium were visualized under a dissecting microscope and injected with 50 nl of Rhbg, Rhcg, or RhAG cRNA (0.05 μg/μl, totaling 2.5 ng of cRNA per oocyte). Control oocytes were injected with 50 nl of sterile H₂O. Accurate and precise delivery of nanoliter volumes of cRNA was achieved using the Nanoject II Auto-Nanoliter injector (Drummond Scientific). Briefly, sterile injection pipettes with ∼20-μm tip diameters were backfilled with mineral oil and attached to the injector. The tip of the securely mounted micropipette was then filled with the sample cRNA. Injections were performed manually; during the injections, the selected volume of sample cRNA was dispensed into each individual oocyte under microscope visualization. Injected oocytes were stored in fresh sterile OR3 medium at 18°C and used in our experiments 3–5 days after injection with cRNA.

Electrophysiological Measurements in Oocytes

Electrophysiological measurements in oocytes involved measuring whole cell currents (I), pH_i, and pH_s. Two-electrode voltage (TEV) clamp (model OC-725, Warner Instruments) was used to measure I. Two voltage microelectrodes were pulled from sections of borosilicate glass capillary tubing with filament (1.5 mm OD × 0.86 mm ID; Warner Instruments) and filled with 3 M KCl and then simultaneously used to impale the oocytes during measurements. Resistance of all electrodes was 2–10 MΩ, and tip potential was <4 mV. Two bath electrodes, serving as free-flowing reference electrodes, were filled with 3 M KCl and directly immersed into the chamber in which the oocytes were placed. The membrane potential (V_m) was obtained by measurement of the voltage difference between the intracellular voltage microelectrode and the free-flowing 3 M KCl Ag-AgCl reference electrode in the bath. All oocytes were clamped at −60 mV, and long-term readings of current were sampled at a rate of once per second. Inward flow of cations (inward current) is defined by convention as negative current.

Single-barreled H⁺-selective microelectrodes that were manufactured according to the methods described by Sackin and Boulpaep (26) were used to measure pH_i and pH_s. The microelectrodes were made from borosilicate glass capillary tubing (1.5 mm OD × 0.86 mm ID; Harvard Apparatus) that was pulled to a tip diameter <0.2 μm and dried in an oven for 2 h at 200°C. The dried capillary tubing was then vapor-silanized with 50 μl of bis(dimethylamino)-dimethyl silane in a closed vessel (300 ml) as described by Siebens and Boron (27). A very fine glass capillary attached to a syringe was used to fill the tip of a section of silanized capillary tubing with H⁺ ionophore I-cocktail B (Fluka-Sigma) and backfill the remaining volume of the tubing with a pH 7.0 buffer solution. The filled tubing was then fitted with a holder containing an Ag-AgCl wire and connected to a high-impedance electrometer (model FD-223, World Precision Instruments). The tip diameter of the pH electrodes used for pH_s measurements (∼15 μm) was significantly greater than that of the pH electrodes used for pH_i measurements (<0.2 μm). We calibrated both types of electrodes in standard pH 6 and 8 buffer solutions and confirmed that they had slopes >58 mV/pH. When conducting pH_i measurements, we simultaneously impaled the oocytes with a pH electrode and a voltage electrode. A reference electrode filled with 3 M KCl was also immersed in the bath. However, when conducting pH_s measurements, we pushed the electrode tip against the surface of the oocyte (∼60 μm) without impaling it. The gross potentials of the pH microelectrodes were measured by calculating the difference between the pH microelectrode and the reference electrode. The pure pH_i voltage was obtained by electronically subtracting V_m from the gross potential of the pH electrode.

Oocytes were placed in a special perfusion chamber and held on a nylon mesh, where they were visualized with a dissecting microscope. We initialized our experiments by continuously flowing bath solutions of standard HEPES followed by a test solution usually containing 5 mM NH₄Cl (NH₃/NH₄⁺ solution) or 5 mM MA/MA⁺ and then recovery in standard HEPES. This enabled us to measure the effect of each solution on pH_i, pH_s, V_m, or I of a single oocyte. Bath solutions were rapidly switched using a combination of a six- and a four-way valve system that is activated pneumatically. All solutions had a pH of 7.5 and flowed at a constant rate of 3–5 ml/min. Experiments were conducted at room temperature.

We used pH_i vs. time tracings to calculate the initial rates of change of pH_i (dpH_i/dt). We obtained the dpH_i/dt for each pulse (NH₃/NH₄⁺ and MA/MA⁺ effects) by fitting the first 20 data points of the experimental pH_i vs. time curve to a linear regression line. The slope of this regression line was used to determine the recorded dpH_i/dt. We also used pH_i, pH_s, and I vs. time measurements to calculate the absolute ΔpH_i, ΔpH_s, and ΔI, respectively. Measurements of ΔpH_i were obtained at ∼10 min after solution change in the bath or, alternatively, when a new steady-state pH_i was reached. Measurements of ΔpH_s were obtained at peak values of pH_s. We did not compute rates of change of pH_s, because the changes occurred too fast (within seconds) to allow enough data points for reliable measurement of rates. The final calculated values for dpH_i/dt, ΔpH_i, ΔpH_s, and ΔI in experimental and H₂O-injected control oocytes were tabulated and analyzed for statistical significance.

Statistical Analysis

In all protocols, values are reported as means ± SE of the number of oocytes assayed. Statistical significance was determined using Student's t-test. P < 0.05 was considered significant.

RESULTS

NH₄⁺- and MA⁺-Induced Currents in Oocytes Expressing RhAG or Rhcg

In TEV-clamp experiments, we clamped oocytes at −60 mV and measured whole cell currents in response to 5 mM NH₄Cl (NH₃/NH₄⁺) or MA/MA⁺ in the bath. As seen in Fig. 1 (segment ab), an inward current was observed in both RhAG-expressing (red tracing) and control (H₂O-injected, blue tracing) oocytes upon change of the bath solution from HEPES to 5 mM NH₄Cl. The inward currents recovered instantaneously upon return of the bath solution to HEPES (segment bc). Figure 1 also shows that, in oocytes expressing RhAG, 5 mM MA/MA⁺ (a substitute for NH₃/NH₄⁺) caused an inward current (segment cd). Return of the bath solution to HEPES resulted in recovery of the inward current observed in RhAG-expressing oocytes (segment de). No change in current was observed in control oocytes (segment cd, blue tracing). Figure 2, which is similar to Fig. 1, shows the effect of NH₃/NH₄⁺ and MA/MA⁺ on H₂O-injected or Rhcg-expressing oocytes. In both cases (Rhcg and control oocytes), NH₃/NH₄⁺ caused an inward current (segment ab), but MA/MA⁺ did not (segment cd). The NH₃/NH₄⁺-induced currents were similar in Rhcg-expressing and H₂O-injected oocytes.

Fig. 1.

NH₃/NH₄⁺- and methyl amine/ammonium (MA/MA⁺)-induced currents (I) in H₂O-injected and glycoprotein (RhAG)-expressing oocytes. All oocytes were clamped at −60 mV and exposed to standard HEPES-Ringer (Std HEPES), 5 mM NH₄Cl HEPES-Ringer (NH₃/NH₄⁺), or 5 mM methylamine hydrochloride HEPES-Ringer (MA/MA⁺), pH 7.5, solution. In oocytes expressing RhAG (red tracing), an inward current was observed upon switching solution from standard HEPES to NH₃/NH₄⁺ (segment ab). Inward current recovered upon return to standard HEPES solution (segment bc). An inward current was also observed upon switching solution from standard HEPES to MA/MA⁺ (segment cd). Inward current also recovered upon return to standard HEPES solution (segment de). In H₂O-injected oocytes (blue tracing), NH₄⁺ influx caused an inward current (segment ab) that was not seen with MA⁺ (segment cd). Tracings are representative of results from 6 RhAG-expressing and for 10 H₂O-injected oocytes.

Fig. 2.

NH₃/NH₄⁺- and MA/MA⁺-induced currents in H₂O-injected and Rhcg-expressing oocytes. In oocytes expressing Rhcg (red tracing) or injected with H₂O (blue tracing), an inward current was observed upon switching solution from standard HEPES to NH₃/NH₄⁺ (segment ab). Inward current recovered upon return to standard HEPES solution (segment bc). Exposure to MA/MA⁺ did not cause a significant change in current (segment cd). Tracings are representative of results from 23 Rhcg-expressing and 22 H₂O-injected oocytes.

Figure 3A, a summary of the data, indicates that exposure of RhAG-expressing oocytes to 5 mM NH₃/NH₄⁺ bath solution caused an inward current of −68.4 ± 9.5 nA (n = 6). This was significantly greater (P < 0.05) than the current in control oocytes, which was −44.2 ± 5.7 nA (n = 10). On the other hand, exposure of RhAG-expressing oocytes to 5 mM MA/MA⁺ caused an inward current of −41 ± 4.2 nA (n = 9). This was significantly greater (P < 0.001) than the current measured in control oocytes, which was not different from 0 (+0.25 ± 0.2 nA, n = 10). Figure 3B, a summary of the data from Rhcg-expressing oocytes, indicates that inward currents in oocytes expressing Rhcg (−25.9 ± 3.8 nA, n = 23) were not significantly different (P > 0.05) from those in H₂O-injected control oocytes (−22.9 ± 3.4 nA, n = 22). MA/MA⁺ (5 mM) did not cause a significant inward current in Rhcg-expressing or H₂O-injected oocytes.

Fig. 3.

NH₄⁺ and MA⁺ currents in oocytes expressing glycoproteins. A: in oocytes expressing RhAG, exposure to NH₃/NH₄⁺ (5 mM) induced an inward current of −68.4 ± 9.5 nA (n = 6) that was significantly larger than in H₂O-injected (control) oocytes (P < 0.05). Exposure to MA/MA⁺ (5 mM) caused an inward current of −41 ± 4.2 nA (n = 9) that was also significantly larger than in H₂O-injected oocytes (P < 0.001). In H₂O-injected oocytes, exposure to NH₃/NH₄⁺ caused an inward current of −44 ± 5.7 nA (n = 10), whereas exposure to MA/MA⁺ did not cause a significant current. *Statistical significance. B: in oocytes expressing Rhcg, exposure to NH₃/NH₄⁺ (5 mM) induced an inward current of −25.9 ± 3.8 nA (n = 23) that was not significantly different −22.9 ± 3.4 nA (n = 22) in H₂O-injected (control) oocytes. Exposure to MA/MA⁺ (5 mM) did not cause a significant inward current in oocytes expressing Rhcg (−0.3 ± 0.3 nA, n = 23) or H₂O-injected oocytes (+0.2 ± 0.2 nA, n = 22).

NH₃/NH₄⁺ and MA/MA⁺ Effects on pH_i in Oocytes Expressing RhAG or Rhcg

In the above-described experiments (Figs. 1–3), the inward currents caused by exposure to NH₃/NH₄⁺ (and MA/MA⁺) are consistent with influx of NH₄⁺ (and MA⁺). Transport of NH₃ (or MA), which cannot be detected by current measurement, could be either negligible or masked by transport of NH₄⁺ (and MA⁺). To investigate this hypothesis further, we measured pH_i to verify that the charged component in solution responsible for the inward currents was, in fact, NH₄⁺ and MA⁺ ions that were being transported across the membrane. If this indeed were the case, both NH₄⁺ and MA⁺ would be expected to dissociate upon entry into the oocyte, releasing H⁺ ions, thereby causing a decrease in pH_i. If, on the other hand, electroneutral NH₃ or MA were being transported across the membrane, an increase in pH_i would be expected due to the consumption of intracellular H⁺ ions during the protonation of NH₃ and MA to form NH₄⁺ and MA⁺ intracellularly.

As shown in Fig. 4, in oocytes expressing RhAG, changing the bath solution from HEPES to 5 mM NH₃/NH₄⁺ Ringer solution caused a significant decrease in pH_i and depolarization of the cell (segment ab). Return of the solution to HEPES caused recovery of pH_i and V_m to steady state (segment bc). On the other hand, exposure of the oocytes to 5 mM MA/MA⁺ caused an increase (not a decrease) in pH_i but depolarized the cell as well (segment cd). Removal of MA/MA⁺ from the bath completely reversed these changes (segment de).

Fig. 4.

Effects of NH₃/NH₄⁺ and MA/MA⁺ on intracellular pH (pH_i) and membrane potential (V_m) in oocytes expressing RhAG. Exposure of RhAG-expressing oocytes to NH₃/NH₄⁺ (5 mM) caused a rapid decrease in pH_i (dpH_i/dt = −11 × 10⁻⁴ pH units/s) and a large depolarization of the cell membrane (segment ab). Removal of bath NH₃/NH₄⁺ resulted in full recovery of pH_i and V_m (segment bc). On the other hand, exposure of the oocyte to MA/MA⁺ (5 mM) caused a significant increase in pH_i (dpH_i/dt = +7 × 10⁻⁴ pH units/s) but also significantly depolarized the cell (segment cd). Removal of MA/MA⁺ reversed these changes (segment de). S, slope. Tracings are representative results from 20 experiments.

Similar experiments were conducted on H₂O-injected (control) oocytes. As shown in Fig. 5, changing the bath solution of control oocytes from HEPES to 5 mM NH₃/NH₄⁺ also caused intracellular acidification and depolarization of the cell membranes (segment ab), both of which recovered to steady state following return of the bath solution to HEPES (segment bc). However, contrary to observations in oocytes expressing RhAG, exposure to MA/MA⁺ or its removal did not cause changes in pH_i or V_m in control oocytes (segment cd).

Fig. 5.

Effects of NH₃/NH₄⁺ and MA/MA⁺ on pH_i and V_m in H₂O-injected oocytes. Exposure of control oocytes to NH₃/NH₄⁺ (5 mM) caused a small and slow decrease in pH_i (dpH_i/dt = −6 × 10⁻⁴ pH units/s) and depolarization of the cell membrane (segment ab). Both changes readily recovered upon removal of NH₃/NH₄⁺ from the bath (segment bc). However, exposure of the H₂O-injected oocyte to MA/MA⁺ (5 mM) caused only a small increase in pH_i and no depolarization (segment cd). Removal of MA/MA⁺ reversed these changes (segment de). Tracings are representative results from 13 experiments.

A third set of experiments was conducted to determine the effects of NH₃/NH₄⁺ (and MA/MA⁺) on pH_i in Rhcg-expressing oocytes. As shown in Fig. 6, exposure of Rhcg-expressing oocytes to 5 mM NH₃/NH₄⁺ caused a decrease in pH_i and depolarization of the cell (segment ab). Removal of NH₃/NH₄⁺ from the bath caused complete recovery of pH_i and V_m (segment bc). Exposure of the oocyte to 5 mM MA/MA⁺ did not cause a significant change in pH_i or V_m (segment cd). These experiments indicate that oocytes expressing Rhcg behaved qualitatively similarly to H₂O-injected oocytes and differently from RhAG-expressing oocytes (compare Fig. 6 with Figs. 4 and 5).

Fig. 6.

Effects of NH₃/NH₄⁺ (5 mM) and MA/MA⁺ (5 mM) on pH_i and V_m in oocytes expressing Rhcg. Exposure of Rhcg-expressing oocytes to NH₃/NH₄⁺ (5 mM) caused a decrease in pH_i and depolarization of the cell membrane (segment ab). Removal of bath NH₃/NH₄⁺ reversed these changes (segment bc). In contrast, exposure of the oocyte to MA/MA⁺ (5 mM) caused a very small increase in pH_i and no depolarization (segment cd). Removal of MA/MA⁺ reversed these changes (segment de). Tracings are representative of results from 6 experiments.

The results of these experiments are summarized in Fig. 7. According to our data (Fig. 7A), 5 mM NH₃/NH₄⁺ in the bath solution resulted in a decrease in pH_i of 0.31 ± 0.04 pH units (n = 20) in RhAG-expressing oocytes, 0.28 ± 0.06 pH units (n = 13) in H₂O-injected oocytes, and 0.24 ± 0.03 pH units (n = 6) in Rhcg-expressing oocytes. These changes in pH_i were not significantly different among the three groups. In contrast to NH₃/NH₄⁺, MA/MA⁺ generally caused an increase, rather than a decrease, in pH_i. The magnitude of change in pH_i (ΔpH_i) caused by MA/MA⁺ was 0.14 ± 0.02 pH units (n = 21) in RhAG-expressing oocytes, 0.04 ± 0.01 pH units in control oocytes (n = 13), and 0.03 ± 0.01 pH units (n = 7) in Rhcg-expressing oocytes. The increase in pH_i was significantly greater (P < 0.001) in RhAG-expressing than control or Rhcg-expressing oocytes. There was no statistical difference in ΔpH_i between Rhcg-expressing and control oocytes.

Fig. 7.

Effects of NH₃/NH₄⁺ and MA/MA⁺ on pH_i and V_m in oocytes expressing RhAG or Rhcg or H₂O-injected oocytes. A: effect on ΔpH_i. In age-matched oocytes, NH₃/NH₄⁺ (5 mM) caused pH_i to decrease in experimental (RhAG- and Rhcg-expressing) and control (H₂O-injected) oocytes. ΔpH_i values were not significantly different from each other. Decrease in pH_i in RhAG- and Rhcg-expressing and control oocytes was 0.31 ± 0.04 (n = 20), 0.24 ± 0.03 (n = 6), and 0.28 ± 0.06 (n = 13) pH units, respectively. On the other hand, MA/MA⁺ (5 mM) caused a slight increase in control and Rhcg-expressing oocytes, with a significant increase only in RhAG-expressing oocytes (P < 0.001 vs. control, P < 0.01 vs. Rhcg). The increase in pH_i of RhAG-expressing oocytes was 0.14 ± 0.02 pH units (n = 21), whereas the increase in Rhcg-expressing and control oocytes was 0.03 ± 0.01 (n = 7) and 0.04 ± 0.01 (n = 13) pH units, respectively. *Statistical significance. B: effect on rate of pH_i change. NH₃/NH₄⁺ (5 mM) in the bath caused a significantly faster (P < 0.05) decrease in pH_i in oocytes expressing RhAG (dpH_i/dt = −10 ± 0.95 × 10⁻⁴ pH units/s, n = 22) than in H₂O-injected control oocytes (dpH_i/dt = −6.6 ± 1.1 × 10⁻⁴ pH units/s, n = 15) or Rhcg-expressing oocytes (dpH_i/dt = −5.9 ± 0.47 × 10⁻⁴ pH units/s, n = 7). Rate of acidification in oocytes expressing Rhcg was not significantly different from that in control oocytes. In contrast, in oocytes expressing RhAG, MA/MA⁺ (5 mM) caused an increase in pH_i at a rate of +10.9 ± 2.0 × 10⁻⁴ pH units/s (n = 21). This rate of alkalinization was significantly higher than that in Rhcg-expressing oocytes (dpH_i/dt = +1.0 ± 0.35 × 10⁻⁴ pH units/s, n = 7, P < 0.01) or control oocytes (dpHi/dt = +1.5 ± 0.30 × 10⁻⁴ pH units/s, n = 13, P < 0.001). The difference in the rate of alkalinization between Rhcg-expressing and control oocytes caused by MA/MA⁺ was not statistically significant. *Statistical significance. C: effect on V_m. NH₃/NH₄⁺ (5 mM) caused experimental and control groups to depolarize, with no significant difference between Rhcg-expressing (20.4 ± 2.7 mV, n = 8), RhAG-expressing (15.5 ± 1.3 mV, n = 28), and control (17.7 ± 1.8 mV, n = 20) oocytes. In contrast, MA/MA⁺ (5 mM) caused significant depolarization in oocytes expressing RhAG (10.8 ± 1.1 mV, n = 26) compared with almost no depolarization in H₂O-injected oocytes (0.2 ± 0.6 mV, n = 19, P < 0.0001) or oocytes expressing Rhcg (−2.4 ± 0.8 mV, n = 8, P < 0.001). *Statistical significance.

Figure 7B shows that dpH_i/dt calculated for RhAG-expressing oocytes following exposure to 5 mM NH₃/NH₄⁺ bath solution was −10 ± 0.95 × 10⁻⁴ pH units/s (n = 22), which was significantly faster (P < 0.05) than −6.6 ± 1.1 × 10⁻⁴ pH units/s calculated for control oocytes (n = 15) or −5.9 ± 0.47 × 10⁻⁴ pH units/s (n = 7) for Rhcg-expressing oocytes. On the other hand, dpH_i/dt calculated for RhAG-expressing oocytes following exposure to 5 mM MA/MA⁺ was +10.9 ± 2.0 × 10⁻⁴ pH units/s (n = 21), which was significantly faster (P < 0.01) than +1.5 ± 0.30 × 10⁻⁴ pH units/s (n = 13) calculated for control oocytes or +1.0 ± 0.35 × 10⁻⁴ pH units/s (n = 7) for Rhcg-expressing oocytes. There was no statistical difference between H₂O-injected and Rhcg-expressing oocytes.

Figure 7C summarizes the effects of NH₃/NH₄⁺ and MA/MA⁺ on V_m. Exposure to 5 mM NH₃/NH₄⁺ in the bath depolarized the cell in RhAG-expressing, Rhcg-expressing, and control oocytes; however, no significant difference was detected among the three groups. The depolarization was 15.5 ± 1.3 mV in RhAG-expressing oocytes (n = 28), 17.7 ± 1.8 mV (n = 20) in control oocytes, and 20.4 ± 2.7 mV in Rhcg-expressing oocytes (n = 8). On the other hand, exposure to 5 mM MA/MA⁺ in the bath caused significant depolarization of the cell membranes in RhAG-expressing oocytes but almost no membrane depolarization was detected in control oocytes or Rhcg-expressing oocytes. MA/MA⁺ caused a 10.8 ± 1.1 mV membrane depolarization in RhAG-expressing oocytes (n = 26), which was significantly larger (P < 0.001) than 0.2 ± 0.6 mV (n = 19) in control oocytes or −2.4 ± 0.8 mV (n = 8) in Rhcg-expressing oocytes.

NH₃/NH₄⁺- and MA/MA⁺-Induced Changes in pH_s in Oocytes Expressing RhAG or Rhcg

To determine whether NH₄⁺ or NH₃ is being transported across the oocyte membrane, we measured pH_s using H⁺-selective electrodes. In these experiments the tip of the pH microelectrode was broken to a blunt end of ∼15 μm diameter. The electrode was pressed against the membrane surface (forming a slight dimple) without impaling the cell. As such, we measured the pH in the microenvironment between the microelectrode and the surface of the cell. In this configuration, if NH₃ is predominantly transported into the cell faster than NH₄⁺, we expect a decrease in pH_s, because as NH₃ moves intracellularly, NH₄⁺, which is in equilibrium with NH₃, dissociates to replenish NH₃, leading to H⁺ accumulation at the surface. As shown in Fig. 8, exposure of oocytes that express RhAG to 5 mM NH₃/NH₄⁺ caused initial rapid acidification at the surface (segment ab, red tracing) followed by partial recovery (segment bc, red tracing). Returning the bath solution to HEPES caused recovery of pH_s to steady state with a slight overshoot (segment cde). RhAG-expressing oocytes also acidified at the surface upon changing the solution from HEPES to 5 mM MA/MA⁺ (segment efg, red tracing). Returning the bath solution to HEPES caused recovery of pH_s to steady state (segment gh).

Fig. 8.

Surface pH (pH_s) changes induced by NH₃/NH₄⁺ and MA/MA⁺ in RhAG-expressing and H₂O-injected oocytes. Exposure of oocytes to NH₃/NH₄⁺ (5 mM) caused a rapid decrease in pH_s (segment ab) followed by partial recovery (segment bc) of RhAG-expressing (red tracing) and control (blue tracing) oocytes. Removal of NH₃/NH₄⁺ reversed these changes (segment cde). Exposure of oocytes to MA/MA⁺ (5 mM) also caused a decrease in pH_s in RhAG-expressing oocytes (segment efg, red tracing) but very little change in H₂O-injected oocytes (blue tracing). Removal of MA/MA⁺ reversed these changes (segment gh). The decrease in pH_s during NH₃/NH₄⁺ or MA/MA⁺ exposure was more pronounced in oocytes expressing RhAG than in H₂O-injected control oocytes (P < 0.001). Tracings are representative of results from 12 RhAG-expressing and 12 H₂O-injected oocytes.

Figure 8 also shows a superimposed tracing of an experiment conducted on H₂O-injected oocytes (blue tracing). In control oocytes, changing the bath solution from HEPES to 5 mM NH₃/NH₄⁺ also caused surface acidification (segment abc), which recovered to steady state following return of the bath solution to HEPES (segment cde, blue tracing). Contrary to observations in RhAG-expressing oocytes, bath MA/MA⁺ did not cause significant changes in pH_s (segment efg, blue tracing).

Similar experiments were also conducted on Rhbg- and Rhcg-expressing oocytes. Figure 9 shows superimposed tracings from representative experiments conducted on Rhbg-expressing (green tracing), Rhcg-expressing (red tracing), and H₂O-injected (blue tracing) oocytes. Exposure of all three (Rhbg, Rhcg, and control) oocytes to 5 mM NH₃/NH₄⁺ caused a decrease in pH_s (segment abc) that readily recovered upon removal of NH₃/NH₄⁺ from the bath (segment cde). However, it is important to note that the decrease in pH_s was characteristically different in each case. In Rhbg-expressing oocytes, NH₃/NH₄⁺ caused a rapid decrease in pH_s (segment ab, green tracing) that partially recovered (segment bc). In Rhcg-expressing oocytes, there was a sustained decrease in pH_s (segment abc, red tracing) that was smaller than in Rhbg-expressing oocytes (green tracing) but larger than in control oocytes (blue tracing). In the same experiments, exposure of the oocytes to MA/MA⁺ caused a decrease in pH_s of Rhbg-expressing oocytes (segment ef, green tracing), but there was almost no change in pH_s of Rhcg-expressing (red tracing) or control (blue tracing) oocytes. The MA/MA⁺-induced decrease in pH_s of Rhbg-expressing oocytes also showed some recovery (segment fg, green tracing), and the change was smaller than that caused by NH₃/NH₄⁺ (compare segment ef with ab). All changes were reversed upon removal of MA/MA⁺ from the bath.

Fig. 9.

Changes in pH_s induced by NH₃/NH₄⁺ and MA/MA⁺ in Rhbg-expressing, Rhcg-expressing, and H₂O-injected oocytes. Exposure of oocytes to NH₃/NH₄⁺ (5 mM) caused a rapid decrease in pH_s in Rhbg-expressing, Rhcg-expressing, and control oocytes (segment ab). In oocytes expressing Rhbg (green tracing) and control oocytes (blue tracing), decrease in pH_s was followed by a spontaneous and partial recovery (segment bc), whereas in Rhcg-expressing oocytes (red tracing), pH_s acidification was sustained. These changes were reversed upon removal of NH₃/NH₄⁺ from the bath (segment cde). MA/MA⁺ (5 mM) also caused a rapid and transient, but smaller, decrease in pH_s in oocytes expressing Rhbg (segment efg, green tracing) but very little change in Rhcg-expressing (segment efg, red tracing) or control (segment efg, blue tracing) oocytes. Tracings are representative of results from 12 H₂O-injected, 11 Rhbg-expressing, and 7 Rhcg-expressing oocytes.

The results of these experiments are summarized in Fig. 10. These data indicate that bath NH₃/NH₄⁺ caused a decrease in pH_s (peak values) of 0.70 ± 0.04, 0.32 ± 0.02, 0.45 ± 0.03, and 0.66 ± 0.11 pH units in RhAG-expressing oocytes (n = 12), control oocytes (n = 12), Rhcg-expressing oocytes (n = 7), and Rhbg-expressing oocytes (n = 8), respectively (RhAG > Rhbg > Rhcg > H₂O-injected). The acidification was significantly greater in oocytes expressing Rhbg (P < 0.01), Rhcg (P < 0.01), and RhAG (P < 0.001) than in control oocytes. Importantly, the difference in NH₃/NH₄⁺-induced pH_s acidification between RhAG- and Rhcg-expressing oocytes or between Rhbg- and Rhcg-expressing oocytes was also significant (P < 0.01); however, the difference between RhAG- and Rhbg-expressing oocytes was not statistically significant (P > 0.05).

Fig. 10.

Effects of NH₃/NH₄⁺ and MA/MA⁺ on pH_s in oocytes expressing RhAG, Rhbg, and Rhcg and H₂O-injected oocytes. Exposure of oocytes to NH₃/NH₄⁺ (5 mM) caused a decrease in pH_s of oocytes expressing Rhbg (−0.66 ± 0.11, n = 11), Rhcg (−0.45 ± 0.03, n = 7), and RhAG (−0.70 ± 0.045, n = 12) that was significantly greater (P < 0.01) than in H₂O-injected oocytes (−0.32 ± 0.02, n = 12). Also, the decrease in pH_s in oocytes expressing Rhbg and RhAG was significantly greater (P < 0.05) than in oocytes expressing Rhcg. On the other hand, the decrease in pH_s induced by MA/MA⁺ (5 mM) was significantly greater in oocytes expressing Rhbg (−0.22 ± 0.05, n = 8) and RhAG (−0.29 ± 0.02, n = 13) than H₂O-injected oocytes (0.01 ± 0.01, n = 12, P < 0.01) or oocytes expressing Rhcg (−0.01 ± 0.01, n = 11, P < 0.01). *Significantly different from control. #Significantly different from Rhcg.

Figure 10 also summarizes the results due to MA/MA⁺. pH_s decreased by 0.29 ± 0.02, 0.01 ± 0.01, and 0.22 ± 0.05 pH units in RhAG-expressing oocytes (n = 13), Rhcg-expressing oocytes (n = 11), and Rhbg-expressing oocytes (n = 8), respectively, and by only 0.01 ± 0.01 pH units in control oocytes (n = 12). The degree of acidification was statistically different between control oocytes and oocytes expressing RhAG (P < 0.001) or Rhbg (P < 0.001). Acidification of Rhcg-expressing oocytes was not statistically different from that of control oocytes (P > 0.05). Interestingly, RhAG-expressing oocytes, with regard to pH_s, acidified as much as Rhbg-expressing oocytes. The difference between both groups was not significant; however, surface acidification was significantly greater in RhAG- and Rhbg- than Rhcg-expressing oocytes (P < 0.01).

DISCUSSION

Overview

In this study we examined the transport characteristics of Rh glycoproteins with respect to NH₃/NH₄⁺ transport. We relied on measurements of NH₃/NH₄⁺ (and MA/MA⁺)-induced changes in I, pH_i, and pH_s. We measured these changes in oocytes expressing the Rh glycoproteins (RhAG, Rhcg, or Rhbg) using the TEV-clamp technique and ion-selective microelectrodes. Because total ammonia (NH₃/NH₄⁺) is NH₄⁺ in equilibrium with NH₃ as described in this reaction

$${\text{NH}}_4^{+}={\text{NH}}_{\text{3}}+{\text{H}}^{+}\hspace{0.17em}\hspace{0.17em}\text{p}{K}_{\text{a}}=9.25$$ (1)

it is critical to determine which substrate (NH₃ or NH₄⁺) is transported by the protein. Depending on the net transport of the ionic NH₄⁺ and electroneutral NH₃ component species, we would expect three possible outcomes. First, if Rh glycoproteins transport only the ionic NH₄⁺ species, but not the uncharged NH₃ species, then we can directly measure the charge being transported across the oocyte membrane with the TEV-clamp technique. In our experiments the net influx of NH₄⁺ would be detected as an NH₄⁺-induced inward current. Furthermore, results from pH_i measurements would show acidification caused by the dissociation of the entering NH₄⁺ into NH₃ and H⁺ intracellularly. In contrast, pH_s measurements would show surface alkalinization, because as NH₄⁺ enters the cell, the equilibrium reaction at the cell surface (reaction 1) shifts to the left, leading to consumption of surface H⁺ and, therefore, an increase in pH_s. We expect similar changes in I, pH_i, and pH_s if we expose the oocytes to MA/MA⁺ and only MA⁺ is being transported. The equilibrium reaction of MA/MA⁺ is similar to that of NH₃/NH₄⁺ as depicted in reaction 2

$${\text{MA}}^{\text{+}}=\text{MA}+{\text{H}}^{+}\hspace{0.17em}\hspace{0.17em}\text{p}{K}_{\text{a}}=10.2$$ (2)

Due to their similarity and because MA/MA⁺ can be easily radiolabeled, MA/MA⁺ has often been used as a replacement for NH₃/NH₄⁺ in uptake studies.

Second, if, however, only the uncharged NH₃ species (and not NH₄⁺) were being transported, the following changes would be expected. Inward currents would be absent or minimal in the I measurements, and net NH₃ influx would cause an increase in pH_i that can be detected by pH_i measurements. This would be due to the consumption of intracellular H⁺ by the influxed NH₃, forming NH₄⁺ intracellularly. On the other hand, at the cell surface, net NH₃ influx is expected to cause a decrease in pH_s due to the release of H⁺ ions at the surface during dissociation of NH₄⁺ to replenish NH₃ and maintain equilibrium at the cell surface caused by the entry of NH₃ into the oocyte (rightward shift of reaction 1).

Third, Rh glycoproteins may simultaneously transport both the ionic NH₄⁺ and the neutral NH₃ species. If this is indeed the case, then we expect to detect transient changes in pH_i or pH_s, depending on the rate of entry of NH₃ relative to NH₄⁺. A higher rate of NH₃ entry than NH₄⁺ influx would cause an initial pH_i increase (and an initial pH_s decrease) followed by at least partial recovery as NH₄⁺ influx follows. On the other hand, a higher relative rate of NH₄⁺ entry would cause an initial pH_i decrease (and an initial pH_s increase) also followed by recovery. To detect any transients in pH_i or pH_s, the flux of either NH₄⁺ or NH₃ would have to be greater than the other. In other words, if the ratio (i.e., flux of NH₃ to flux of NH₄⁺) remains constant, then no changes in pH_i or pH_s would be detected.

Evidence for NH₄⁺ Transport

The data in our study demonstrate that RhAG simultaneously transports both the ionic NH₄⁺ (MA⁺) and the neutral NH₃ (MA) species. Our data show that RhAG, like Rhbg (21), transports NH₄⁺ (and MA⁺) in an electrogenic manner. This is evident by the fact that NH₄Cl (NH₃/NH₄⁺) and CH₃NH₄·HCl (MA/MA⁺) induced inward currents, which can be directly attributed to net NH₄⁺ and MA⁺ influx, respectively. Moreover the magnitudes of these currents are larger in RhAG-expressing than H₂O-injected control oocytes (Figs. 1 and 3A). Although inward currents can be observed in control oocytes exposed to NH₃/NH₄⁺, these inward currents are due to native transport mechanisms that are smaller in magnitude, have different (slower) kinetics, and are unlikely related to Rh proteins (Fig. 1). Although it may be suggested that exposure to NH₃/NH₄⁺ (or MA/MA⁺) may lead to the activation of a native endogenous transporter or a channel that can transport NH₃/NH₄⁺, rather than transport through the expressed exogenous Rh transporter, our data do not support this possibility. A main observation here is that MA⁺ induces an inward current in RhAG-expressing oocytes but has no effect in H₂O-injected control oocytes. This confirms the substrate specificity of Rh glycoproteins for MA/MA⁺ and argues against activation of a native transporter. It is interesting to note that, regarding effects on current, no significant difference was observed between Rhcg-expressing and control oocytes (Figs. 2 and 3B), suggesting that expression of Rhcg does not lead to significant net transport of the charged NH₄⁺ (or MA⁺) species over that of NH₃ (or MA).

Evidence for NH₃ Transport From pH_s Measurements

In addition to the evidence of electrogenic transport by Rh glycoproteins (RhAG and Rhbg) demonstrated by our TEV measurement of current, our pH_s measurements provide evidence for transport of the neutral gas species as well. The surface acidification observed in oocytes expressing Rh glycoprotein (Figs. 8 and 9) can be indirectly attributed to net NH₃ (or MA) influx that dominates any NH₄⁺ (or MA⁺) influx. Surface acidification was also recorded in control oocytes exposed to NH₃/NH₄⁺, although significantly less than in Rh-expressing oocytes, indicating residual Rh-independent NH₃ influx. Upon exposure to NH₃/NH₄⁺ solution, RhAG-, Rhbg-, and Rhcg-expressing oocytes had varying and significant amounts of peak surface acidification compared with control oocytes (Figs. 8–10), demonstrating that RhAG, Rhbg, and Rhcg transport net NH₃ into the oocyte. The data also show more transport of NH₃ into the oocyte by RhAG and Rhbg than Rhcg, which, as described earlier in our voltage-clamp data, was not found to transport NH₄⁺ significantly.

It can be argued that the surface extracellular acidification that we measured may not be due to NH₃ (and MA) influx as we propose but, rather, that intracellular acidification caused by NH₄⁺ is triggering endogenous acid-extruding mechanisms in the oocyte to transport intracellular H⁺ to the surface and, therefore, is causing the surface extracellular acidification that we observe. This is unlikely, because it can be demonstrated that intracellular acidification caused by means other than exposure to NH₃/NH₄⁺ does not always cause pH_s acidification but, rather, an increase in pH_s. For example, we conducted experiments whereby exposure of oocytes to butyrate (HB) caused a decrease in pH_i of ∼0.3 units in oocytes expressing Rhbg, usually accompanied by a small depolarization. The intracellular acidification is caused by butyrate diffusion into the cell, leading to release of intracellular H⁺ (19), as shown in the following reaction

$$\text{HB}={\text{H}}^{\text{+}}+{\text{B}}^{-}$$ (3)

As HB is transported into the cell, reaction 3 is shifted to the left, causing consumption of surface H⁺ that we were able to detect as an increase in pH_s and not a decrease as with NH₃/NH₄⁺.

In another example, CO₂ exposure (switch from HEPES to CO₂/HCO₃⁻) causes intracellular acidification by an average of ∼0.28 in the oocyte due to diffusion of CO₂ across the cell membrane. The pH_s that accompanies this change is an alkalinization caused by a leftward shift of the following reaction

$${\text{CO}}_2+{\text{H}}_{\text{2}}\text{O}\hspace{0.17em}\hspace{0.17em}=\hspace{0.17em}\hspace{0.17em}{\text{HCO}}_{\text{3}}^{-}+{\text{H}}^{+}$$ (4)

and, thus, consumption of H⁺. This was confirmed by us and was published by others (16).

Evidence for Simultaneous NH₃ and NH₄⁺ Transport

Interestingly, although NH₃/NH₄⁺ produces an abrupt fall in pH_s in oocytes expressing RhAG, Rhbg, or Rhcg, only RhAG- and Rhbg-expressing oocytes exhibit a subsequent spontaneous recovery of pH_s that is not seen in Rhcg-expressing oocytes and is generally faster than that observed in control oocytes (Figs. 8 and 9). Further inspection of the tracings shows that 1) the spontaneous pH_s recovery in H₂O-injected control oocytes exposed to NH₃/NH₄⁺ is much slower than in oocytes expressing Rh proteins, 2) there is a complete lack of pH_s recovery in control oocytes exposed to MA/MA⁺, 3) there is no pH_s recovery in Rhcg-expressing oocytes exposed to either NH₃/NH₄⁺ or MA/MA⁺, 4) the pH_s recovery in RhAG- or Rhbg-expressing oocytes exposed to MA/MA⁺ is much slower than that observed when these same oocytes are exposed to NH₃/NH₄⁺ (Figs. 8 and 9), and 5) the pH_s recovery in Rhbg-expressing oocytes exposed to MA/MA⁺ is much faster than that in RhAG-expressing oocytes exposed to the same solution. The spontaneous pH_s recovery in RhAG- and Rhbg-expressing oocytes is rapid initially but slows with time. The rapid phase of pH_s recovery is consistent with subsequent cellular influx of NH₄⁺ (and MA⁺), which equilibrates across the plasma membrane during the course of a few minutes, resulting in the slowing of the pH_s recovery observed in our tracings. In essence, rapid NH₃ influx shifts the equilibrium reaction (reaction 1) to the right initially, whereas subsequent influx of NH₄⁺ consumes surface H⁺ by shifting the reaction to the left. Although it could be suggested that the spontaneous pH_s recovery may be due to diffusion of H⁺ at the surface of the oocyte out of the surface unstirred boundary layer (between the electrode and the cell) and into the bulk solution (rather than subsequent influx of NH₄⁺), the fact that the rates of pH_s recovery differ between experimental (RhAG- and Rhbg-expressing) and control oocytes argues against this possibility. Similarly, the absence of pH_s recovery in Rhcg-expressing oocytes also suggests that the spontaneous recovery is not due to H⁺ leak from the microdomain between the electrode tip and the oocyte membrane.

Surface acidification was also recorded in RhAG- and Rhbg-expressing oocytes exposed to MA/MA⁺ solution, suggesting that both transporters favor initial net MA influx over MA⁺ entry. Recall that these transporters (Rhbg and RhAG) were found to transport MA⁺ based on our TEV-clamp data. Exposure to MA/MA⁺ solution did not cause significant surface acidification in Rhcg-expressing and control oocytes. One possibility is that in Rhcg-expressing and control oocytes the relative influx of MA and MA⁺ may be equal, and, therefore, no change in pH_s can be detected. However, given that our TEV-clamp data show no evidence for MA⁺ influx (no inward current), we conclude that neither MA nor MA⁺ is being significantly transported by Rhcg-expressing or H₂O-injected oocytes. In the case of Rhcg-expressing oocytes, this may indicate lack of substrate selectivity by Rhcg for MA/MA⁺. In the case of control oocytes, this indicates that native transport mechanisms (those that can transport NH₃/NH₄⁺) are not able to transport MA/MA⁺.

Finally, our pH_i measurements provide further evidence that RhAG simultaneously transports both the ionic NH₄⁺ (and MA⁺) and the neutral NH₃ (and MA) species. Exposure of both experimental (RhAG- and Rhcg-expressing) and control oocytes to NH₃/NH₄⁺ caused a decrease in pH_i in all oocyte groups. In contrast, exposure of RhAG oocytes to MA/MA⁺ caused an increase in pH_i (Fig. 4), whereas exposure of Rhcg-expressing or control oocytes to MA/MA⁺ caused a very small, if any, increase in pH_i (Figs. 5 and 6). The dpH_i/dt calculated from the initial slope of our tracings is a measure of the relative flux of NH₃/NH₄⁺ or MA/MA⁺. As previously mentioned, we expect that an increase in NH₄⁺ influx would result in a faster rate of pH_i decrease, because NH₄⁺ entering the oocyte would dissociate instantaneously into NH₃ and H⁺. This H⁺ resulting from the dissociation of NH₄⁺ is depicted by the acidification rate (−dpH_i/dt) measurements and is a reflection of the NH₄⁺-induced accumulation of intracellular H⁺. Similarly, MA-induced H⁺ consumption can be defined as the H⁺ consumed during intracellular formation of MA⁺. Therefore, MA influx can be indirectly estimated from the rate of MA-induced pH_i increase. In our experiments the rate of NH₄⁺-induced pH_i decrease was significantly higher in RhAG-expressing than control and Rhcg-expressing oocytes. Hence, our data strongly support enhanced net NH₄⁺ transport by RhAG. The rate of MA-induced pH_i increase was also significantly higher in RhAG-expressing than control and Rhcg-expressing oocytes. Accordingly, our data strongly support enhanced MA transport by RhAG. The fact that RhAG is capable of transporting the NH₃ analog MA provides further evidence for electroneutral transport by RhAG. This also indicates that, in pH_i measurements, NH₃ influx by RhAG is masked by the NH₄⁺ influx, since our pH measurements are dependent on the relative flux of NH₃ to NH₄⁺ and are only a measure of the difference in these two fluxes. This is confirmed by the NH₃-induced pH_s measurements as discussed above.

Our results indicate that RhAG enhances the influx of net NH₃ (as suggested by greater acidification of pH_s), yet it also increases the influx of net NH₄⁺ [indicated by faster rate of NH₄⁺-induced intracellular acidification (−dpH_i/dt)]. However, paradoxically, the enhanced NH₃ influx is not detected by pH_i measurements, expected in this case as transient pH_i increase. This may be explained by the fact that our pH_s electrodes have a larger tip diameter, so as not to impale the oocyte during pH_s measurements. This causes our pH_s electrodes to have a lower resistance and correspondingly faster response time during measurements. As can be seen in our pH_s tracings, the −ΔpH_s upon exposure to NH₄Cl solution is instantaneous and followed by a subsequent rapid partial recovery of pH_s, which slows with time. We attribute the rapid pH_s recovery to net NH₄⁺ entry that decreases with time due to equilibration in this restricted unstirred domain.

In addition to pH_i measurements, simultaneous measurements of V_m showed NH₃/NH₄⁺-induced depolarization of the membrane in all oocyte groups, consistent with NH₄⁺ entry. However, the membrane depolarizations in RhAG- and Rhcg-expressing oocytes were not significantly larger than those in control oocytes (Fig. 7C). After exposure to MA/MA⁺, changes in V_m were observed only in RhAG-expressing oocytes, and not in Rhcg-expressing or control oocytes. This is consistent with our observation that MA⁺ entry occurred only in RhAG-expressing oocytes, and not in Rhcg-expressing and control oocytes, and is, therefore, the likely reason for the significant depolarization.

Factors To Be Considered in Characterizing NH₃/NH₄⁺ Transport

In our study we characterized NH₃/NH₄⁺ transport by Rh glycoproteins on the basis of measurements of I, pH_s, and pH_i. These measurements provide multiple parameters for an accurate evaluation of the complex process of NH₃ and NH₄⁺ transport. However, other points need to be considered as well. Among those are the following.

Does the NH₄⁺-induced inward current completely explain the intracellular acidification?

In other words, to what degree is the charge change coupled to the pH_i change? To answer this question, we estimated intracellular NH₄⁺ concentration based on short-circuit current measurements (see appendix). By our calculations, the steady-state intracellular NH₄⁺ concentration corresponding to the measured ΔI was 0.197 mM. This steady-state value is too small to explain the measured ΔpH_i of 0.28. In fact, a ΔpH_i of 0.28 would require an accumulation of 646 mM NH₄⁺ at a steady state to cause this change in pH_i. However, as intracellular NH₄⁺ accumulates inside the cell, a loss of intracellular NH₃ (by sequestration, metabolism, or transport) would shift the reaction (NH₄⁺ → NH₃ + H⁺) to the right, leading to substantial pH_i decrease. We discussed this issue in a previous publication (22), and similar results were recently reported and published by others who reached essentially the same conclusion (17).

Possible coupling of Rh proteins to endogenous acid-extruding mechanisms in the oocyte.

In the oocyte system, this scenario is unlikely for the following reasons. The two most likely non-HCO₃⁻ acid-extruding mechanisms are the Na/H exchanger and H⁺-ATPase. Both mechanisms, if present in the oocyte, seem to work at a very low rate. For example, removal of bath Na⁺, which reverses Na/H exchange and should, in principle, substantially decrease pH_i, causes very little change in pH_i (22), and amiloride has no effect on pH_i. Moreover, changing bath pH from 7.5 to 8.2 or 6.8 also causes no visible change in pH_i, suggesting that H⁺-ATPase is also not effective (19). More importantly, after an acid load (by butyrate, CO₂, or even NH₃/NH₄⁺), there is no spontaneous pH_i recovery (no undershoot). Recovery of pH_i occurs only after removal of butyrate, CO₂, or NH₃/NH₄⁺, and the recovery rate is very slow, suggesting that acid extrusion is absent or sluggish. However, in native tissues such as the kidney collecting duct, the coupling of Rhbg and/or Rhcg to endogenous acid-base transporters is very plausible.

Is Rhcg expressed properly in the oocyte?

In oocytes expressing Rhcg, exposure to NH₃/NH₄⁺ induced changes in pH_i (dpH_i/dt and V_m) that closely resembled those in H₂O-injected control oocytes. This could imply that Rhcg may not be expressed in the oocyte or that it is not inserted in the membrane. We therefore confirmed by immunohistochemistry that Rhcg was properly expressed in the oocyte membrane, as shown in Fig. 11. Furthermore, measurements of pH_s indicated a significant difference between H₂O-injected and Rhcg-expressing oocytes, consistent with a role of Rhcg in electroneutral transport of NH₃ as discussed above.

Fig. 11.

Expression of Rhcg in oocytes. Left: immunolabeling of Rhcg (green) at the membrane of oocytes injected with cRNA for Rhcg. Right: no labeling of Rhcg in H₂O-injected oocytes.

Conclusions

We discussed four possible mechanisms of transport that had been proposed by various groups. The first proposal was put forth by Westhoff et al. (30), who suggested that these proteins served as electroneutral countertransporters of MA⁺ (or NH₄⁺) coupled to H⁺ efflux. In effect, they concluded that RhAG transported net NH₃ equivalents without causing changes in pH_i. Contrary to this proposal, our pH_i studies demonstrate that RhAG and Rhbg serve as electrogenic transporters of NH₄⁺ and that NH₄⁺ transport is not likely coupled to H⁺ efflux. This is evident in our pH_i tracings, where NH₄⁺ entry causes a decrease in pH_i. An electroneutral NH₄⁺/H⁺ exchanger is not expected to cause significant pH_i decrease. A second proposed mechanism was that Rh glycoproteins transport only the neutral NH₃ gas species (15, 17). Our pH_i and I data clearly show that some Rh glycoproteins also transport the ionic NH₄⁺ species in an electrogenic manner. This is especially evident in the inward currents observed in oocytes expressing RhAG and Rhbg. A third proposal suggested that RhCG transported NH₃ without affecting NH₄⁺ permeability (15). Our studies indicate that Rhcg is likely an NH₃ transporter. This is evident in our pH_s measurements, where we observe significant surface acidification, caused by influx of NH₃, in Rhcg-expressing oocytes upon NH₄Cl exposure. The lack of a subsequent pH_s recovery and the absence of inward currents, both of which would indicate NH₄⁺ transport, provide further evidence for the role of Rhcg as a predominantly NH₃ transporter. A fourth proposed mechanism was that Rh glycoproteins promote only efflux of NH₄⁺ (13). Our pH_i and I measurements show intracellular acidification, not alkalinization, and an inward current consistent with NH₄⁺ influx. However, this does not preclude the possibility that bidirectional movement of NH₄⁺ is likely.

Cellular transport of NH₃/NH₄⁺ is important in tissues and, particularly, in the kidney, where renal excretion of NH₄⁺ in the urine is critical for regulation of acid-base balance. In the collecting duct, secretion of total ammonia (NH₃/NH₄⁺) occurs along the length of the tubule, where it contributes to net acid excretion by the kidney. The secretion of NH₃/NH₄⁺ in this segment is a complex process that involves transcellular NH₃ and NH₄⁺ transport. The apical membrane of the collecting duct cells (where Rhcg is expressed) is the site where rapid NH₃ diffusion has to occur, yet permeability of this membrane to NH₃ is rate-limiting. On the other hand, the basolateral membrane (where Rhbg is expressed) is the site where NH₄⁺ transport occurs, presumably by NH₄⁺ transporters. Our data showing net NH₃ and NH₄⁺ transport by Rhcg and Rhbg are consistent with a role for these transporters in renal NH₃/NH₄⁺ transport. It is likely, for example, that Rhbg transports NH₄⁺ (and possibly NH₃) at the basolateral membrane and that Rhcg predominantly transports NH₃ at the apical membrane. As such, Rhbg and Rhcg, working in tandem, function as the elusive NH₃/NH₄⁺ transporters in the kidney to explain how the distal nephron achieves net transepithelial NH₃/NH₄⁺ transport.

In summary, our data support the following conclusions. 1) RhAG and Rhbg transport the ionic NH₄⁺ and the neutral NH₃ species. 2) Transport of NH₄⁺ is electrogenic. 3) RhAG and Rhbg expression enhance MA transport, an electroneutral component. 4) Like Rhbg, RhAG also transports MA⁺, an electrogenic component. The charged MA⁺ seems to be a direct substrate for RhAG. 5) Rhcg is likely a predominantly NH₃ transporter. 6) RhAG and Rhbg are unlikely to be NH₄⁺/H⁺ exchangers. These data, which clearly show robust NH₄⁺ transport (in addition to NH₃) by Rhbg and, predominantly, NH₃ transport by Rhcg, support an important role for these transporters in transepithelial NH₃/NH₄⁺ transport in the kidney. The dual role of Rh proteins as NH₃ and NH₄⁺ transporters is a unique property and may be critical in understanding renal handling of acid-base homeostasis.

GRANTS

This work was supported by Veterans Affairs Merit Grant BX 001513 (Southeast Louisiana Veterans Health Care System), National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases Grant RO1-DK-6225, and Tulane institutional funds. The contents of this paper do not represent the views of the US Department of Veterans Affairs or the United States Government.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

T.C., K.B., M.T.I., and N.L.N. performed the experiments; T.C., S.A.-N., and N.L.N. analyzed the data; T.C., S.A.-N., L.L.H., and N.L.N. interpreted the results of the experiments; T.C. and N.L.N. prepared the figures; T.C. drafted the manuscript; T.C., S.A.-N., L.L.H., and N.L.N. approved the final version of the manuscript; S.A.-N., L.L.H., and N.L.N. developed the concept and designed the research; S.A.-N. and N.L.N. edited and revised the manuscript.

Appendix

Estimating intracellular NH₄⁺ concentration from whole cell current measurements.

If it is assumed that all the measured inward current is caused by influx of bath NH₄⁺, it is possible to theoretically estimate intracellular NH₄⁺ from the measured change in inward current. In RhAG-expressing oocytes, exposure to 5 mM NH₄⁺ caused a change in whole cell current (ΔI) of −68.4 ± 9.5 nA over an average of ∼250 s. The number of coulombs (Q) that are presumably carried by movement of NH₄⁺ can be calculated as

$$\begin{array}{c}Q=I\left(\text{amp}\right)\times t\left(\text{seconds}\right)\\ =68.4\times {10}^{-9}\times 250\\ =1.71\times {10}^{-5}\hspace{0.17em}\text{coulombs}\end{array}$$

If NH₄⁺ transfer is responsible for all this charge, then the number of moles of NH₄⁺ that would cause this change can be calculated by dividing by Faraday's constant

$$1.71\times {10}^{-5}/96.500\hspace{0.17em}\text{mol}\hspace{0.17em}{\text{NH}}_4^{+}$$

$$1.77\times {10}^{-10}\hspace{0.17em}\text{mol}\hspace{0.17em}{\text{NH}}_4^{+}$$

For an oocyte volume of 0.9 μl (assuming a spherical volume with a diameter of 1.2 mm), the calculated change in intracellular NH₄⁺ concentration is

$$1.77\times {10}^{-10}\hspace{0.17em}\text{mol}/0.9\times {10}^{-6}\hspace{0.17em}\text{liters}\hspace{0.17em}=1.97\times {10}^{-4}\text{M}=197\hspace{0.17em}\mu \text{M}$$

Calculating concentration of intracellular NH₄⁺ from changes in pH_i assuming that all the pH_i change was caused by accumulating NH₄⁺ intracellularly.

If it is assumed that the NH₃/NH₄⁺-induced change in pH_i of 0.28 ± 0.03 (measured) in oocytes expressing RhAG is caused solely by NH₄⁺ influx and that there is no permeability to NH₃, it is possible to calculate the concentration of intracellular NH₄⁺ from the change in pH_i and the buffering power (β). The intrinsic buffering power in the oocyte is calculated to be 12.4 ± 1.6 mM/pH unit from CO₂-induced pH_i changes as described by Boron and De Weer (4a). Accordingly,

$$\left[{\text{H}}^{+}\right]=\beta \times \Delta {\text{pH}}_{\text{i}}=3.47\hspace{0.17em}\text{mM}$$

Because H⁺ is formed from dissociation of NH₄⁺, an equivalent amount of NH₃ of 3.47 mM is formed. With the assumptions of no loss of NH₃ and a pK_a of 9.25 for the NH₄⁺/NH₃ equilibrium, then at pH_i 6.98 (pH_i after exposure to NH₃/NH₄⁺)

$$\begin{array}{c}\left[{\text{NH}}_{\text{4}}^{+}\right]=\left[{\text{NH}}_{\text{3}}\right]\times {10}^{{\text{pK-pH}}_{\text{i}}}\\ =3.47\times {10}^{9.25-6.98}\\ =646.14\hspace{0.17em}\text{mM}\end{array}$$

Obviously, this is an unrealistically large amount to accumulate intracellularly and is many orders of magnitude greater than the calculated intracellular NH₄⁺ from current measurements. It is likely, however, that there is a loss of intracellular NH₃ (by sequestration, transport, or metabolism), which can lead to intracellular acidification by rightward shift of the reaction NH₄⁺ = NH₃ + H⁺ and generation of H⁺. We discussed this possibility previously (22), and others (17) proposed similar conclusions.