Evidence from renal proximal tubules that and solute reabsorption are acutely regulated not by pH but by basolateral and CO₂

Abstract

Respiratory acidosis, a decrease in blood pH caused by a rise in [CO₂], rapidly triggers a compensatory response in which the kidney markedly increases its secretion of H⁺ from blood to urine. However, in this and other acid-base disturbances, the equilibrium CO₂ + H₂O ⇄ + H⁺ makes it impossible to determine whether the critical parameter is [CO₂], , and/or pH. Here, we used out-of-equilibrium solutions to alter basolateral (BL) , [CO₂], or pH, systematically and one at a time, on isolated perfused S2 rabbit proximal tubules. We found that increasing from 0 to 44 mM, at a fixed [CO₂]_BL of 5% and a fixed pH_BL of 7.40, caused reabsorption (J_HCO3) to fall by half but did not significantly affect volume reabsorption (J_V). Increasing [CO₂]_BL from 0% to 20%, at a fixed of 22 mM and pH_BL of 7.40, caused J_HCO3 to rise 2.5-fold but did not significantly affect J_V. Finally, increasing pH_BL from 6.80 to 8.00, at a fixed of 22 mM and [CO₂]_BL of 5%, did not affect either J_HCO3 or J_V. Analysis of the J_HCO3 and J_V data implies that, as the tubule alters J_HCO3, it compensates the reabsorption of other solutes to keep J_V approximately constant. Because the cells cannot respond acutely to pH changes, we propose that the responses of J_HCO3 and the reabsorption of other solutes to changes in or [CO₂]_BL involve sensors for basolateral and CO₂.

A major task of the kidney is to secrete H⁺ into the urine. Inadequate renal H⁺ secretion caused, for example, by mutations to acid-base transporters (1–4) or carbonic anhydrases (5), or by renal failure (6, 7), can lead to a life-threatening decrease in blood pH. Moreover, to maintain a stable blood pH, the kidney must appropriately increase H⁺ secretion in response to metabolic acidosis (a decrease in blood pH caused by a decrease in at a fixed CO₂) or to respiratory acidosis. However, a half century after the classical observation that respiratory acidosis rapidly stimulates renal H⁺ secretion (8, 9), we still have little insight into how the kidney senses acute acid-base disturbances.

The renal proximal tubule (PT) reabsorbs (from lumen to blood) a liquid that contains ≈80% of the filtered by the glomerulus. The PT cell does this reabsorbtion by secreting H⁺ into the PT lumen and using this H⁺ to titrate luminal to CO₂ and H₂O. After entering the cell across the apical membrane, the CO₂ and H₂O recombine to produce H⁺ and . The cell extrudes the H⁺ into the lumen across the apical membrane through Na-H exchangers (10–12) and H⁺ pumps (13) and moves the out across the basolateral membrane via the electrogenic Na/HCO₃ cotransporter NBCe1-A (14, 15). Carbonic anhydrases catalyze the interconversions between CO₂ and H₂O on the one hand and and H⁺ on the other (5). Because blood pH is the parameter regulated by these acid-base transport processes, blood pH or, more likely, intracellular pH (pH_i), has been thought to be the parameter sensed by renal cells. However, because of the interconversion CO₂ + H₂O ⇄ + H⁺, it had been impossible to distinguish pH unambiguously from and CO₂ as potential signals. Using the method that our laboratory developed for generating out-of-equilibrium (OOE) solutions (16), we can now approach the problem by independently varying basolateral , [CO₂], and pH.

In the present study, we perfused single, isolated S2 segments of rabbit PTs and collected the fluid that had passed along the PT lumen. Analysis of this collected fluid allowed us to compute volume reabsorption (J_V); reabsorption (J_HCO3), which is virtually the same as the H⁺-secretion rate under the conditions of our experiments; and the reabsorption of solutes other than NaHCO₃ (J_Other). We made the surprising observation that, at least in the short term, J_HCO3 and J_Other do not respond to changes in basolateral or intracellular pH. The most straightforward hypothesis is that PT cells have sensors for basolateral and a parameter related to CO₂.

Materials and Methods

Except for the compositions of most of the OOE solutions, our methods are described in detail in ref. 17.

Tubule Perfusion. We hand dissected kidneys from “pathogen-free” female New Zealand White rabbits (1.4–2.0 kg) to yield individual segments of midcortical S2 PTs in accordance with an approved animal protocol. We used two assemblies of concentric glass pipettes to perfuse lumens of single isolated PTs at 37°C, as described by Burg and coworkers (18) and modified by Quigley and Baum (19). We used a calibrated collection pipette (volume ≅ 55 nl) to collect samples of fluid that had flowed down the PT lumen; the luminal collection rate was 12.6 ± 0.3 nl/min (n = 50 J_V/J_HCO3 measurements). We superfused the basolateral surface of the PT at 7 ml/min.

Solutions. Table 1 lists the compositions of the solutions. We dissected PTs in Hanks' solution (solution 1) at 4°C (20). During tubule perfusion, the luminal perfusate always was solution 2, which contained [³H]methoxyinulin (molecular mass ≈ 7,146 Da; catalog no. NET-086L, PerkinElmer). We dialyzed the [³H]methoxyinulin for 72 h by using a membrane with a molecular mass cutoff of 3,500 Da. After establishing luminal perfusion, we allowed a 20- to 30-min warmup period in which solution 3 flowed through the bath (i.e., basolateral solution) at 37°C. We then changed the bath to either solution 4 (Fig. 1A) or to solution 7 (Fig. 1 B and C), which differed only in [Cl^-], achieved by replacing NaCl with Na gluconate. In control experiments (data not shown), we found that J_V and J_HCO3 were indistinguishable in solutions 4 and 7. All solutions had osmolalities of 300 ± 2 mosM. Note that solution 7 was identical to luminal solution 2 except for the addition of 32.5 mM Hepes (titrated with NaOH) in solution 7, which replaced an osmotically equivalent amount of NaCl in solution 2.

Table 1. Physiological solutions.

	Solution
Component	1	2	3	4	5A	5B	6A	6B	7	8A	8B	9A	9B	10A	10B	11A	11B	12A	12B	13A	13B	14
NaCl	137	124.7	115	79	137.3	20.7	137.3	20.7	101	91	113	113	91	113	91	113	91	54.7	142.1	54.7	132	130.7
KCl	5	5	5	5	0	10	0	10	5	0	10	10	0	10	0	10	0	10	0	10	0	5
Na2HPO4	0.3	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
NaH2PO4	0	2	2.3	2	4	0	0	0	2	4	0	0	4	0	4	0	4	0	4	0	4	2
CaCl2	0.2	1	3.6	1	2	0	2	0	1	2	0	0	2	2	0	2	0	2	0	0	2	1
MgSO4	0.8	1.2	1	1.2	2.4	0	2.4	0	1.2	2.4	0	0	2.4	0	2.4	0	2.4	0	2.4	0	2.4	1.2
MgCl2	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
Glucose	0	10.5	8.3	10.5	21	0	21	0	10.5	21	0	0	21	0	21	0	21	0	21	0	21	10.5
Glutamine	2	2	0	2	4	0	4	0	2	4	0	0	4	0	4	0	4	0	4	0	4	2
l-alanine	0	0	5	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
l-lactic acid	2	2	2	2	4	0	4	0	2	4	0	0	4	0	4	0	4	0	4	0	4	2
CO2, mM (%)	0	1.2 (5)	1.2 (5)	1.2 (5)	2.4 (10)	0	0	2.4 (10)	1.2 (5)	0	0	1.2 (5)	0	4.8 (20)	0	9.6 (40)	0	2.4 (10)	0	2.4 (10)	0	0
NaHCO3	0	22	22	22	0	0	0	88	22	0	44	44	0	44	0	44	0	44	0	44	0	0
Na gluconate	0	0	0	22	0	88	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
Tris·HCl	10	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
Hepes	0	0	0	32.5	0	65	65	0	32.5	65	0	0	65	0	65	0	65	0	65	0	65	32.5
Albumin, g/liter	0	0	20	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
pH	7.40	7.40	7.40	7.40	5.40	7.55	7.29	7.70	7.40	6.99	9.40	7.70	7.34	7.10	7.46	6.80	7.52	7.40	6.64	7.40	8.11	7.40

The concentrations are in mM except for CO₂ (given both in mM and %) and albumin (g/liter). Except for solution 1 (4°C), all solutions were titrated to the indicated pH at 37°C. Tris·HCl and Hepes were titrated with NaOH. Solutions: 1, Hanks' dissection; 2, lumen, equilibriated 5% CO₂, 22 mM HCO₂^-; 3, warmup bath, equilibrated 5% CO₂ 22 mM HCO₂; 4, 86 mM Cl^- bath, equilibrated 5% CO₂ 22 mM HCO₃^-; 5, 86 mM Cl^- bath, OOE solution 5% CO₂ 0 mM HCO₃^-; 6, 86 mM Cl^- bath, OOE solution 5% CO₂ 44 mM HCO₃^-; 7, standard bath, equilibrated 5% CO₂ 22 mM HCO₃^-; 8, bath, OOE solution 0% CO₂ 22 mM HCO₃^-; 9, bath, OOE solution 2.5% CO₂ 22 mM HCO₃^-; 10, bath, OOE solution 10% CO₂ 22 mM HCO₃^-; 11, bath, OOE solution 20% CO₂ 22 mM HCO₃^-; 12, pH 6.8 bath, OOE solution 5% CO₂ 22 mM HCO₃^-; 13, pH 8.0 bath, OOE solution 5% CO₂ 22 mM HCO₃^-; 14, Hepes lumen/bath. Solution 2 was used as a luminal perfusate. Solution 3 was used as a basolateral perfusate. OOE solutions were generated by rapidly mixing their respective A and B components in 1:1 ratio. Solutions 5B, 6A, 8A, 8B, 9B, 10B, 11B, 12B, and 13B were vigorously gassed with 100% O₂ to render them free of CO₂. Mixed pH was 7.40 for solutions 5, 6, 8—11, 6.80 for solution 12, and 8.00 for solution 13.

Fig. 1.

Effect of isolated changes in basolateral acid-base parameters on J_HCO3, J_V, and pH_i. (A) Effect of changing at a fixed [CO₂]_BL of 5% and a fixed pH_BL of 7.40. Data represent mean ± SEM; error bars are omitted when they are smaller than the size of the symbol. For Top and Middle, in each of the 13 experiments we paired an equilibrated bath solution (▴) with one of two OOE solutions (circles; see Materials and Methods). n = 7, = 0; n = 6, = 44 mM. For Bottom, each pH_i experiment (n = 6) included bath exposures to each of the three solutions in Top.(B) Effect of changing [CO₂]_BL at a fixed of 22 mM and a fixed pH_BL of 7.40. For Top and Middle, each of the 25 experiments paired an equilibrated bath solution (▴) with one of four OOE solutions (▪). n = 6, [CO₂]_BL values of 0, 2.5, and 10%; n = 7, [CO₂]_BL = 20%. For Bottom, each of the 20 pH_i experiments included bath exposures to the equilibrated and one or two OOE solutions. n = 12, [CO₂]_BL = 0; n = 6, [CO₂]_BL = 2.5%; n = 8, [CO₂]_BL values of 10% and 20%. (C) Effect of changing pH_BL at a fixed [CO₂]_BL of 5% and a fixed of 22 mM. For Top and Middle, each of the 12 experiments paired an equilibrated bath solution (▴) with one of two OOE solutions (♦). n = 6, pH_BL values of 6.80 and 8.00. For Bottom, each pH_i experiment included bath exposures to each of the three solutions in Top (n = 6). *, P < 0.01; **, P < 0.001; absence of * or ** indicates P > 0.05 in paired two-tailed t tests comparing OOE data with equilibrated data.

We generated OOE solutions by rapidly mixing streams of two dissimilar solutions (17), delivering the newly mixed solution to the PT within ≈200 ms. Fig. 1C in ref. 17 shows the detailed compositions of two newly mixed OOE solutions and the minute degree of equilibration that occurs during the ≈200-ms interval. The final compositions of OOE solutions in Fig. 1A (solutions 5 and 6) were the same as for solution 4 except that we replaced with gluconate or vice versa to keep [Cl^-] constant. The final compositions of the OOE solutions in Fig. 1B (solutions 7–11) were the same as for solution 7 except for [CO₂]. The final compositions of the OOE solutions in Fig. 1C (solutions 12 and 13) were the same as for solution 7 except for pH, the concentrations of neutral and anionic Hepes, and the [NaCl] (which we adjusted to keep osmolality constant).

We used solution 14 only in experiments in which we measured pH_i. Here, solution 14 was present in the lumen before we switched to solution 2. In addition, solution 14 was present in the bath when a -containing solution (i.e., solutions 4–13) was not present.

Measurement of J_V and J_HCO3. We measured J_HCO3 (pmol/min per mm tubule length) and J_V (nl·min^-1·mm^-1) by using an approach similar to that of McKinney and Burg (21). We assayed total CO₂ in aliquots of the perfusate and collected fluid by using a NanoFlo microfluorometer (World Precision Instruments, Sarasota, FL) and reagents (Diagnostic Kit 132-A) from Sigma-Aldrich (22, 23). Luminal [³H]methoxyinulin served as a volume marker for calculating J_V.

Each experiment consisted of two data collection periods. The bath contained equilibrated (solutions 4 or 7) during the first period and an OOE solution (solutions 5, 6, 8, 9, 10, 11, 12, or 13) during the second. In control experiments in which identical equilibrated solutions (solution 7) were present during both periods, J_HCO3 values were identical, as were J_V values (17). In each experiment, we divided the J_HCO3 (and J_V) value obtained during the second (OOE) collection period by the comparable values obtained during the first (control) period to generate OOE/control ratios for J_HCO3 (and J_V). In Fig. 1, J_HCO3 (and J_V) values for control conditions (triangles) are raw mean values; each OOE value (circles, squares, and diamonds) is the product of the raw mean J_HCO3 (or J_V) value and the average OOE/control ratio for J_HCO3 (or J_V).

Calculation of in the Fluid Reabsorbed by the PT. In experiments in which we varied (Fig. 1A) or [CO₂]_BL (Fig. 1B), we computed J_HCO3 (Fig. 1 A and B Top) and J_V (Fig. 1 A and B Middle). From these values, we calculated the in the reabsorbate as the ratio J_HCO3/J_V for each (Fig. 2A Lower) and each [CO₂]_BL (Fig. 2B Lower). Assuming that the PT reabsorbs fluid isosmotically (300 mosM),

[1]

Here, 2 × J_HCO3 is the reabsorption of NaHCO₃, and J_Other is the reabsorption of solutes other than NaHCO₃. We used this equation to compute J_Other for each value of and [CO₂]_BL (Fig. 2 Upper). Finally, knowing J_Other and J_V, we computed the concentration of all Other solutes in the reabsorbate as the ratio J_Other/J_V (Fig. 2 Lower).

Fig. 2.

Effect of changes in basolateral or [CO₂] on the calculated reabsorption of solutes other than NaHCO₃. (A) Changes in . (A Upper) the J_HCO3 data are replotted from Fig. 1 A Top. The J_Other values were calculated as described in Materials and Methods, assuming (J_HCO3 + J_Other)/J_V = 300 mosM. (A Lower) the calculated reabsorbate [NaHCO₃] values were obtained by dividing J_HCO3 values in A by the corresponding J_V values in B. The calculated concentrations of all other solutes (Other) was obtained by dividing J_Other in A by the corresponding J_V in B.(B) Isolated changes in [CO₂]_BL. The J_HCO3 data in Top are replotted from Fig. 1B Top. We computed the other values as described for Fig. 2 A.

Measurement of pH_i. We calculated pH_i from the fluorescence excitation ratio of 2′,7′-bis-(2-carboxyethyl)-5(and-6)carboxyfluorescein (24), loaded into cells as the acetoxymethyl ester (no. B-1170, Molecular Probes) (17). The inverted microscope was a Zeiss IM-35, equipped with apparatus for epiillumination, a 40×/NA 0.85 objective, dual filter wheels (Ludl Electronic Products, Hawthorne, NY) for alternating between 495 ± 5 nm and 440 ± 5 nm excitation filters (Thermo Oriel, Stratford, CT), a 510-nm long-pass dichroic mirror, a 530-nm long-pass filter, an image intensifier (KS-1381 intensifier, Videoscope, Dulles, VA), and a camera (CCD 72, Dage–M.T.I., Michigan City, IN). We converted the I₄₉₀/I₄₄₀ ratios to pH_i values by using the high-K⁺/nigericin technique (25), as modified for one-point calibrations (26).

Supporting Information. For additional information on results and discussion, see Figs. 5 and 6, Table 2, and Supporting Text, which are published as supporting information on the PNAS web site.

Results

Effect of Variations in . In the Fig. 1 A Top, the triangle ( = 22 mM) represents J_HCO3 with all three basolateral acid-base parameters at their equilibrated, physiological values. Switching to an OOE basolateral (i.e., bath) solution that was nominally -free, while holding [CO₂]_BL and pH_BL at their physiological values, caused J_HCO3 to increase by ≈50%. This maneuver approximates the “metabolic” half of metabolic acidosis. Conversely, switching from the equilibrated 5% CO₂/22 mM at pH 7.40 solution to an OOE bath solution with twice the normal , but with [CO₂]_BL and pH_BL at their physiological values, caused J_HCO3 to fall by ≈30%. Thus, changes in cause J_HCO3 to change in a direction that would help the PT to respond appropriately to stabilize blood pH.

The fluid that the PT reabsorbs is nearly isosmotic, and the calculated in this reabsorbed fluid is the ratio J_HCO3/J_V. Fig. 1A Middle shows that changes in do not cause significant changes in J_V. Thus, J_HCO3/J_V (i.e., in the reabsorbate) must vary considerably. In Fig. 1A, this calculated was 159 mM at a of 0 mM (i.e., the reabsorbate was isotonic NaHCO₃). In Fig. 2A Upper, we replot the J_HCO3 data from Fig. 1A Top and also plot the flux of all Other solutes, as described in Materials and Methods. Fig. 2A Lower summarizes the computed reabsorbate values of [NaHCO₃] and all Other solutes. The analysis in Fig. 2A illustrates that increasing not only lowers J_HCO3 but reciprocally raises the reabsorption of other solutes (J_Other), the appropriate response for maintaining a constant J_V and, thus, a constant blood pressure.

Because extracellular acid-base disturbances generally cause pH_i to change (27), we examined how changes in affect the steady-state pH_i of PT cells under the same conditions that prevailed for Fig. 1A Top and Middle. As shown in the Fig. 1A Bottom, an increase in from 0 to 22 mM caused steady-state pH_i to increase by 0.32. However, further increasing to 44 mM did not cause a statistically significant increase in steady-state pH_i.

Effect of Isolated Variations in [CO₂]_BL. In Fig. 1B Top, the triangle ([CO₂]_BL = 5%) represents virtually the same conditions as the triangle in Fig. 1A Top. Switching from this equilibrated, physiological solution to one in which we increased [CO₂]_BL to four times its physiological value, while holding and pH_BL at their physiological values, caused J_HCO3 to rise by ≈50% This maneuver approximates the “respiratory” half of respiratory acidosis. If we instead exposed the PT to a nominally CO₂-free OOE bath solution, while holding and pH_BL at their physiological values, J_HCO3 fell by 40% compared with normal. The small J_HCO3 observed in the nominal absence of basolateral CO₂ may be due, in part, to CO₂ from the lumen reaching the basolateral membrane. The midpoint of the J_HCO3 response to CO₂ is at a [CO₂]_BL of ≈6%, which is somewhat above the physiological [CO₂] of arterial blood and somewhat below that of the renal cortex (28, 29). Thus, isolated changes in [CO₂]_BL cause J_HCO3 to change in a direction that would help the PT to respond appropriately to stabilize blood pH.

Fig. 1B Middle shows that isolated changes in [CO₂]_BL tended to cause J_V to increase. However, none of the differences between the J_V values obtained in OOE solutions and the J_V value obtained in the equilibrated solution (solution 7) were statistically significant. As [CO₂]_BL rose from 0% to 20%, the calculated reabsorbate [NaHCO₃] rose from 70 mM (i.e., the isosmotic reabsorbate was ≈50% NaHCO₃) to 129 mM (i.e., the reabsorbate was mainly isotonic NaHCO₃). As summarized in Fig. 2B, increases in [CO₂]_BL not only raise J_HCO3 and reabsorbate [NaHCO₃] but also reciprocally lower J_Other and the total concentration of other solutes in the reabsorbate.

We also measured pH_i of the PT cells under the conditions of Fig. 1B Top and Middle. Fig. 1B Bottom shows that increases in [CO₂]_BL cause graded decreases in steady-state pH_i. Thus, the data in Fig. 1B are consistent with the hypothesis that elevations in [CO₂]_BL increase J_HCO3 indirectly by lowering pH_i. According to this hypothesis, the increased J_HCO3 that we observed when decreasing (Fig. 1A Top) also would have been caused by the attendant decrease in pH_i (Fig. 1A Bottom).

Effect of Variations in pH_BL. If pH_i changes determine J_HCO3, we should be able to increase J_HCO3 by lowering pH_i, even without changing or [CO₂]_BL. In Fig. 1C Top, the triangle (pH_BL = 7.40) represents the same conditions as the triangle in Fig. 1B Top. Surprisingly, switching to either a pH-6.80 or pH-8.00 OOE bath solution, while holding and [CO₂]_BL at their physiological values, caused no change in either J_HCO3 (Top) or J_V (Middle). As pH_BL rose from 6.80 to 8.00, the calculated in the reabsorbate ranged between 96 mM and 99 mM. On the other hand, Fig. 1C Middle shows that raising pH_BL from 6.80 to 8.00 caused substantial, graded increases in steady-state pH_i. In fact, the ΔpH_i/ΔpH_o ratio for these PT cells (≈60%) is among the highest recorded for any cell (30–33).

Discussion

What Contributes to J_Other? One of our most striking observations is that J_HCO3 and J_Other (i.e., reabsorption rate of solutes other than and its obligated Na⁺) change reciprocally to keep J_V relatively constant. The Other solutes include (i) Cl^-, (ii) organic solutes, and (iii) Na⁺ in excess of that accompanying . Diffusion and solvent drag through tight junctions, possibly augmented by apical Cl-base exchange (34), could contribute to Cl^- reabsorption (Fig. 3). Apical cotransport of Na⁺ with glucose, lactate, and glutamine could contribute to the reabsorption of organics. The near constancy of J_V implies that J_Na and, thus, the Na-K pump rate, must also be relatively stable.

Fig. 3.

Model of reabsorption by the proximal tubule. Changes in pH_BL cause large changes in pH_i in the same direction but have no short-term effect on J_HCO3. Increases in could moderate J_HCO3 (red arrows) by increasing backleak into the lumen and/or by directly inhibiting the electrogenic Na/HCO₃ cotransporter NBCe1-A. A sensor would contribute to the decrease in J_HCO3 (red arrows) and is necessary to account for the increased reabsorption (green arrow) of solutes other than NaHCO₃ (J_Other). At the apical membrane, this reabsorption of other solutes (enclosed by the large blue arrow) would be mediated by cotransporters mediating the uptake of Na/glucose, Na/lactate, and Na/glutamine; obligated Cl^- would pass through tight junctions by diffusion and solvent drag. Apical Cl-base exchange (in parallel with Na-H exchange) could also contribute to J_Other. Increases in CO₂ would increase J_HCO3 by stimulating a CO₂ sensor at or near the basolateral membrane. This sensor would, in turn, stimulate three acid-base transporters (green arrows) but reduce J_Other (red arrow). If the CO₂ sensor is a membrane protein, the CO₂ binding site is not necessarily facing outward as shown.

The PT could produce the highest J_Other observed in the present study (≈143 pmol·min^-1·mm^-1 in Fig. 2A) by reabsorbing ≈4.25 mM or ≈30% of the 14.5 mM of organic solutes initially present in the lumen (i.e., 10.5 mM glucose, 2 mM lactate, and 2 mM glutamine), along with the coupled Na⁺ and obligated Cl^-. To the extent that organic-solute transport contributes to J_Other, PT cells could reciprocally alter J_HCO3 and J_Other by changing the rate of the apical Na-H exchanger, predominantly NHE3 (35). To the extent that apical Cl-base exchange in parallel with Na-H exchange contributes to J_Other, the cell could reciprocally alter J_HCO3 and J_Other by changing J_Cl-base while keeping J_Na-H constant.

Do Proximal Tubules Sense Acute pH_i Changes? The conventional wisdom is that acid-base chemosensitive cells, including central chemoreceptor neurons, peripheral chemoreceptor glomus cells, and certain epithelial cells (e.g., PT cells), sense acute acid-base disturbances through pH_i changes (36). Although the evidence is consistent with a major role for pH_i in neurons and glomus cells, our data suggest that this conclusion is not the case in the proximal tubule. The circles in Fig. 4 are replots of the J_HCO3 data from Fig. 1A Top vs. the corresponding pH_i data from Fig. 1A Bottom. Similarly, the squares are replots of the J_HCO3 vs. the pH_i data from Fig. 1B. For both data sets, increases in pH_i, whether caused by a rising or a falling [CO₂]_BL, are associated with decreases in J_HCO3. However, the replot of the J_HCO3 vs. pH_i data from Fig. 1C, plotted as diamonds in Fig. 4, shows that even large pH_i changes are not sufficient to alter J_HCO3, provided both and [CO₂]_BL are constant. Supporting Text shows that large pH_i changes are likewise insufficient to alter J_Other (Fig. 5). Thus, pH_i cannot be the parameter triggering reciprocal changes in J_HCO3 and J_Other in Fig. 2 A and B.

Fig. 4.

Relationship between J_HCO3 and pH_i. The circles are replots of the J_HCO3 and pH_i data from Fig. 1 A Top and Bottom, values ranging from 0 mM on the left to 44 mM on the right (fixed [CO₂]_BL = 5%, fixed pH_BL = 7.40). The squares are replots of data from Fig. 1 B, [CO₂]_BL values ranging from 20% on the left to 0% on the right (fixed = 22 mM, fixed pH_BL = 7.40). The diamonds are replots of data from Fig. 1 C, pH_BL values ranging from 6.80 on the left to 8.00 on the right (fixed [CO₂]_BL = 5%, fixed = 22 mM).

In retrospect, the insensitivity of J_HCO3 to acute pH_i changes could have been anticipated for both theoretical and experimental reasons. First, theory predicts that a fall in pH_i could trigger an increase in J_HCO3 only by stimulating both H⁺ extrusion across the apical membrane and exit across the basolateral membrane. However, a fall in pH_i ought to inhibit NBCe1-A (Fig. 3), either directly or indirectly by lowering and , just as a fall in pH_i inhibits Cl-HCO₃ exchange in other cell types (37–40).

Second, previous experiments showed that, with absent from both lumen and bath or present only in the lumen, pH_i recovers (i.e., increases) slowly from acute intracellular acid loads (41). Just as important, the pH_i recovery rates are only modestly pH_i sensitive. On the other hand, with present both in lumen and bath or only in the bath, pH_i recovers far more rapidly but still with only modest pH_i sensitivity (see fig. 4 in ref. 41). In other experiments (42), in which was either absent from both lumen and bath (bilateral Hepes) or present in both lumen and bath (bilateral ), a more quantitative analysis of acid-extrusion rates (i.e., the sum of apical Na-H exchange and H⁺ pumping) was possible. In bilateral Hepes, acid extrusion was low and only modestly pH_i dependent. In bilateral , acid extrusion was again only modestly pH_i dependent but ≈8-fold higher at identical pH_i values (see fig. 5 in ref. 42). Thus, the main determinant of acid extrusion in intact PTs is not pH_i but parameters related to and/or [CO₂]_BL.

What, then, is the primary trigger for reciprocal changes in J_HCO3 and J_Other? A fall in could lower backleak through tight junctions (43, 44) and, thereby, raise J_HCO3. NBCe1-A (Fig. 3) appears to transport when operating with a stoichiometry of 1:2 (I. Grichtchenko and W.F.B., unpublished data) and may well transport both and when operating with a 1:3 stoichiometry in the PT. Thus, increases in and or decreases in and could stimulate NBCe1-A and thereby raise J_HCO3. However, although effects on backleak or NBCe1-A might serve as secondary modulators of J_HCO3, it is unclear how these effects could serve as primary triggers for changes in J_Other, let alone closely coordinated, reciprocal changes in J_HCO3 and J_Other (Fig. 2 Upper).

Even though our data indicate that PT cells do not respond to acute pH_i changes by modulating J_HCO3 or J_Other, our data do not address the issue of how prolonged pH_i changes might affect cells. In opossum-kidney cells, chronic metabolic or respiratory acidosis (pH_o 7.0 vs. 7.4) over a period of 2 days increases Na-H exchange by ≈30% (45). This stimulation is accompanied by increased levels of mRNA encoding the Na-H exchanger NHE3 and abrogated either by herbimycin A (inhibitor of certain soluble tyrosine kinases) or by overexpressing csk (46), a physiological inhibitor of c-src.

What Does the PT Sense? Supporting Text includes an analysis of how the experimental maneuvers in Fig. 1 influence the intracellular and basolateral values of [CO₂], pH, [H₂CO₃], , and (Fig. 6). The analysis reveals that no single parameter exhibits a pattern that consistently correlates, positively or negatively, with the patterns of J_HCO3 and J_Other in Fig. 2 Upper. Moreover, the analysis reveals that the next most parsimonious hypothesis, that the patterns of J_HCO3 and J_Other in Fig. 2 Upper reflect the sensing of two parameters, is straightforward for only three of 45 parameter pairs: (i) and [CO₂]_BL, (ii) and [CO₂]_i, and (iii) and [H₂CO₃]_i (Table 2). As summarized in Fig. 3, we propose that PT cells respond to increases in by reducing J_HCO3 and raising J_Other and respond to increases in [CO₂]_BL (or a related parameter) by raising J_HCO3 and reducing J_Other.

We already noted that adding to the basolateral, but not the luminal, side of the PT triggers a large increase in H⁺ extrusion (41, 42). The simplest explanation for these earlier data are that the basolateral CO₂ strongly enhanced acid-base transporters, overwhelming the inhibition imposed by the basolateral . Because adding to the lumen fails to stimulate acid extrusion (41), we propose that CO₂ stimulates a sensor at or near the basolateral membrane. In principle, the sensor could face the basolateral fluid and sense [CO₂]_BL per se, or the sensor could be inside the cell near its basolateral membrane and sense [CO₂]_i or [H₂CO₃]_i.

In summary, at least during the acute response to acid-base disturbances, the kidney appears to regulate whole-body acid-base balance not by monitoring blood pH per se but by monitoring the levels of the two major buffer components that determine pH: and CO₂. Moreover, it appears that the and CO₂ sensors operate in a push–pull fashion to diminish J_HCO3 and augment J_Other in the case of and to augment J_HCO3 but diminish J_Other in the case of CO₂. The net effects are not only to trigger changes in J_HCO3 that tend to restore blood pH but also to trigger compensatory alterations in the reabsorption of other solutes, thereby stabilizing J_V. Understanding how the PT transduces the and CO₂ signals may provide important clues for treating both acid-base disturbances and hypertension.