Renal intercalated cells are rather energized by a proton than a sodium pump

Abstract

The Na(+) concentration of the intracellular milieu is very low compared with the extracellular medium. Transport of Na(+) along this gradient is used to fuel secondary transport of many solutes, and thus plays a major role for most cell functions including the control of cell volume and resting membrane potential. Because of a continuous leak, Na(+) has to be permanently removed from the intracellular milieu, a process that is thought to be exclusively mediated by the Na(+)/K(+)-ATPase in animal cells. Here, we show that intercalated cells of the mouse kidney are an exception to this general rule. By an approach combining two-photon imaging of isolated renal tubules, physiological studies, and genetically engineered animals, we demonstrate that inhibition of the H(+) vacuolar-type ATPase (V-ATPase) caused drastic cell swelling and depolarization, and also inhibited the NaCl absorption pathway that we recently discovered in intercalated cells. In contrast, pharmacological blockade of the Na(+)/K(+)-ATPase had no effects. Basolateral NaCl exit from β-intercalated cells was independent of the Na(+)/K(+)-ATPase but critically relied on the presence of the basolateral ion transporter anion exchanger 4. We conclude that not all animal cells critically rely on the sodium pump as the unique bioenergizer, but can be replaced by the H(+) V-ATPase in renal intercalated cells. This concept is likely to apply to other animal cell types characterized by plasma membrane expression of the H(+) V-ATPase.

Keywords: ion transporter; plasma membrane; proton pump.

Fig. 1.

Effects of inhibition of the H⁺ V-ATPase on intercalated cell (IC) volume and membrane voltage in the isolated microperfused cortical collecting duct (CCD). (A) Cells were perfusion loaded with calcein-AM (green) and Alexa 594-conjugated peanut lectin (red) to identify ICs (red). A differential interference contrast overlay is shown. (B) Addition of 40 nM bafilomycin to the bathing solution caused a significant reduction of intracellular calcein fluorescence in ICs, indicating cell swelling, but not in principal cells (PCs). (C) Summary of bafilomycin or ouabain-induced cell volume changes in PCs vs. ICs. Ouabain (100 µM) was added to the bathing solution. *P < 0.05, IC or PC vs. baseline. (D) ANNINE-6 (green) was loaded from the bath; note its membrane-specific fluorescence along the basolateral cell membranes. The specific apical membrane binding of Alexa 594-conjugated peanut lectin (red) identified ICs. The image was taken in the presence of 40 nM bafilomycin in the bath. (E) In contrast to PCs that are intensely green fluorescent, ICs show diminished ANNINE-6 fluorescence indicating membrane depolarization. *P < 0.05, IC vs. baseline.

Fig. 2.

Effects of amiloride (10⁻⁵ M), ouabain (10⁻⁴ M), and bafilomycin (4.0 × 10⁻⁸ M) on Na⁺, Cl⁻, K⁺ transepithelial fluxes, and on transepithelial voltage (V_te) in CCDs isolated from Na⁺-restricted mice. (A) CCDs isolated from mice on a Na⁺-depleted diet develop a lumen negative V_te, which is completely abolished by either amiloride 10⁻⁵ M or ouabain at a concentration ≥10⁻⁴ M. (B) K⁺ secretion is fully abolished by either amiloride 10⁻⁵ M or ouabain at a concentration ≥10⁻⁴ M. J_K, rate of K⁺ secretion. (C) Na⁺ absorption is inhibited by ∼60% by either amiloride at 10⁻⁵ M or ouabain at 10⁻⁴ M, but the effects of the drugs are not additive indicating that the amiloride-resistant component of J_na is ouabain-resistant. Cl⁻ absorption is not affected by either amiloride at 10⁻⁵ M or ouabain at 10⁻⁴ M. (D) The amiloride-resistant component of NaCl absorption is abolished by addition of 10⁻⁸ M bafilomycin to the basolateral solution. n = 5–6 independent tubules per group. Statistical significance was tested by ANOVA followed by Bonferroni’s post hoc test when appropriate. *P < 0.05, and **P < 0.01 vs. control (no inhibitor) group.

Fig. 3.

Characterization of Slc4a9 (Ae4) expression. (A and B) Immunohistochemical detection of Slc4a9 (Ae4) protein abundance in the mouse kidney cortex using an anti-peptide antibody specific for the murine Slc4a9 protein in wild-type animals. (A) Low-magnification view of the whole renal cortex. (Scale bar, 200 µm.) (B) High-magnification view centered on a cortical collecting duct (CCD). (Scale bar, 40 µm.) (C) Absence of Slc4a9/Ae4 protein by immunohistochemistry in Slc4a9^−/− mice. (Because of the absence of staining, the CCD is indicated by a dotted line.) (Scale bar, 40 µm.) (D) Localization of Slc26a4/pendrin (red) in the kidney cortex of Slc4a9^−/− mice. (Scale bar, 20 µm.) (E) Localization of Slc26a4/pendrin (green) and Slc4a9/Ae4 (red) on opposite sides of type-B ICs in mice. (Scale bar, 20 µm.) (F) Localization of Slc4a1/Ae1 (green) and Slc4a9/Ae4 (red) within distal tubuli of the mouse kidney showing that AE1 and AE4 are expressed in different cells. (Scale bar, 40 µm.) (G) Immunogold labeling of mouse kidney sections with an anti-Ae4 antibody shows predominant basolateral staining, 3,000× magnification. (Scale bar, 5 µm.) (H) A 12,000× magnification of Inset in G. (Scale bar, 1 µm.) (I) CCDs microdissected from normal rabbit kidney were stained for Ae4 (red) and either peanut lectin (green in I, J, and K) or the tight junction protein ZO-1 (green in L and M). (Scale bar, 20 µm.) I reveals a Z-stack image of an individual CCD that was obtained via confocal microscopy. (Scale bar, 2 µm.) (J–M) Images are obtained via 3D reconstruction of individual β-ICs that have been rotated to view the lateral (J and L) and vertical (K and M) perspectives. (Scale bar, 2 µm.)

Fig. 4.

Characterization of Ae4 activity in the mouse kidney. (A) Na⁺ dependence of pH_i changes in ICs of CCDs isolated from Slc4a9^+/+ and Slc4a9^−/− mice. In the experiments presented here, CCDs were isolated from mice fed on a Na⁺-restricted diet. Traces shows the average of pH_i changes recorded in the presence of HCO₃⁻/CO₂ when luminal Na⁺ is initially removed from and then readded to the peritubular solution. Mean starting pH_i (immediately before luminal Na⁺ removal) was 7.31 ± 0.05 in Slc4a9^+/+ mice in the presence of HCO₃⁻/CO₂, 7.41 ± 0.12 in Slc4a9^−/− mice in the presence of HCO₃⁻/CO₂, 7.39 ± 0.10 in Slc4a9^+/+ mice in the absence of HCO₃⁻/CO₂, and 7.38 ± 0.05 in Slc4a9^−/− mice in the absence of HCO₃⁻/CO₂, respectively. (B and C) Initial rates of base-equivalent fluxes during sodium removal (B) or addition (C) to the peritubular solutions in either the presence or the absence of HCO₃⁻/CO₂. Values are means ± SE of values obtained in four to five independent ICs from three to six independent tubules; each tubule was isolated from independent animals. Statistical significance was tested by ANOVA followed by Bonferroni's post hoc test. *P < 0.05 and **P < 0.01 vs. all other groups. (D) Analyses of amiloride-resistant J_Na, and J_Cl in CCDs isolated from either Slc4a9^+/+ or Slc4a9^−/− mice maintained on a Na⁺-depleted diet. *P < 0.05, and **P < 0.01 vs. Slc4a9^+/+. Note that the magnitude of fluxes shown here cannot be compared with fluxes obtained in experiments shown in Fig. 2 or Fig. 4E because the Slc4a9 line has a different genetic background. (E) Analyses of amiloride-resistant J_Na, and J_Cl in CCDs isolated from either Slc4a8^+/+ or Slc4a8^−/− maintained on a Na⁺-depleted diet. *P < 0.05, and **P < 0.01 vs. Slc4a8^+/+. (F) Schematic description of electroneutral NaCl absorption energized by the H⁺ V-ATPase. Two cycles of pendrin coupled with one cycle of Ndcbe result in the net uptake of one Na⁺, one Cl⁻, and two HCO₃⁻ ions, whereas one Cl⁻ ion is recycled across the apical membrane. Then Cl⁻ ion exits the cell through a basolateral chloride channel, while Na⁺ and bicarbonate exit the cell at the basolateral membrane via Ae4. All these transporters are indirectly energized by the H⁺ V-ATPase. C. A., carbonic anhydrase.