NHE4 is critical for the renal handling of ammonia in rodents

Abstract

Ammonia absorption by the medullary thick ascending limb of Henle's loop (MTALH) is thought to be a critical step in renal ammonia handling and excretion in urine, in which it is the main acid component. Basolateral Na+/H+ exchangers have been proposed to play a role in ammonia efflux out of MTALH cells, which express 2 exchanger isoforms: Na+/H+ exchanger 1 (NHE1) and NHE4. Here, we investigated the role of NHE4 in urinary acid excretion and found that NHE4-/- mice exhibited compensated hyperchloremic metabolic acidosis, together with inappropriate urinary net acid excretion. When challenged with a 7-day HCl load, NHE4-/- mice were unable to increase their urinary ammonium and net acid excretion and displayed reduced ammonium medulla content compared with wild-type littermates. Both pharmacologic inhibition and genetic disruption of NHE4 caused a marked decrease in ammonia absorption by the MTALH. Finally, dietary induction of metabolic acidosis increased NHE4 mRNA expression in mouse MTALH cells and enhanced renal NHE4 activity in rats, as measured by in vitro microperfusion of MTALH. We therefore conclude that ammonia absorption by the MTALH requires the presence of NHE4 and that lack of NHE4 reduces the ability of MTALH epithelial cells to create the cortico-papillary gradient of NH3/NH4+ needed to excrete an acid load, contributing to systemic metabolic acidosis.

Figure 1. Urinary pH, ammonia and net acid excretions, and blood bicarbonate and pH levels in normal and NHE4^–/– littermate mice on a normal diet and during HCl loading.

In response to HCl-induced metabolic acidosis, both NHE4^+/+ (filled circles) and NHE4^–/– mice (open circles) exhibit a rapid increase in (A) urinary ammonium and (C) net acid excretions. (B) A rapid decrease in urinary pH was also seen in NHE4^+/+ mice but not NHE4^–/– mice. (D) Blood pH and bicarbonate were significantly lower after 7 days HCl load in NHE4^–/– mice. During acid loading, urinary ammonium and net acid excretions were significantly lower in NHE4^–/– mice than in NHE4^+/+ mice. Values are mean ± SEM (n = 20 and n = 18 at baseline and n = 12 and n = 7 during HCl loading in NHE4^+/+ and NHE4^–/– mice, respectively). The x axis labels indicate the day of treatment. *P < 0.01; **P < 0.05.

Figure 2. Tissue ammonia content in kidneys from NHE4^+/+ and NHE4^–/– mice.

Figure 3. Measurement of basolateral Na⁺/H⁺ exchange activity in rat MTALH cells.

Tubules were initially perfused and bathed with a Na⁺-free, CO₂/HCO₃^–-free, HEPES-buffered solution (solution A). Luminal fluid also contained 0.1 mM furosemide and 1 mM amiloride to prevent apical Na⁺ entry and proton efflux. After a 2-minute recording, the peritubular (PT) solution was changed to a 144 mM Na⁺, CO₂/HCO₃^–-free, HEPES-buffered solution (solution B), containing EIPA at various concentrations. The initial rate of pH_i recovery (dpH_i/dt) was calculated on the linear part of the curve after peritubular Na⁺ addition. (A) The effect on pH_i of peritubular Na⁺ addition under control conditions. The figure shows an example of the pH_i time course, before and after Na⁺ addition. The arrow shows the initial, linear change in pH_i, allowing the calculation of dpH_i/dt. In the absence of external Na⁺, pH_i was close to 6.4 and was identical in all experiments regardless to the EIPA concentration. (B) The effect of various peritubular concentrations of EIPA on Na⁺/H⁺ exchange activity in the rat MTALH. The inhibition elicited by the various concentrations of EIPA was consistent with the presence of 2 components of Na⁺/H⁺ exchange activity with distinct EIPA sensitivities: 0.3 to 3 μM EIPA inhibited approximately 80% of the initial rate of pH_i recovery, corresponding to NHE1 activity; 100 μM EIPA was required to completely inhibit pH_i recovery (NHE1 and NHE4 activities, shown by arrows).

Figure 4. Effect of peritubular EIPA on transepithelial ammonia absorption in the rat MTALH.

(A) The effect of 1 μM peritubular EIPA. Tubules were perfused and bathed with a standard solution, containing 139 mM Na⁺ and 4 mM NH₄⁺ (solution C). The addition of 1 μM EIPA to the peritubular solution had no effect on Vte (Vte: control [C], 12.1 ± 1.6 mV; 1 μM EIPA, 12.7 ± 1.3 mV; recovery [R], 11.8 ± 1.9 mV; NS; flow rate: control, 3.2 ± 0.4 nl/min; 1 μM EIPA, 3.1 ± 0.3 nl/min; recovery, 3.1 ± 0.3 nl/min; NS) or transepithelial ammonia flux. Individual results obtained with 5 independent tubules are displayed. (B) The effect of 10 μM peritubular EIPA. Addition of 10 μM EIPA to the peritubular solution elicited a decrease in transepithelial ammonia flux. Vte and flow rate were not affected by EIPA (Vte: control, 7.9 ± 1.4 mV; 10 μM EIPA, 9.2 ± 1.4 mV; recovery, 8.7 ± 1.2 mV; NS; flow rate: control, 3.0 ± 0.2 nl/min; 10 μM EIPA, 3.1 ± 0.2 nl/min; recovery, 3.8 ± 0.3 nl/min; NS). Individual results obtained with 7 independent tubules are displayed.

Figure 5. Transepithelial ammonia absorption in the MTALH harvested from NHE4^+/+ and NHE4^–/– littermate mice on a normal diet.

Tubules were perfused and bathed with a standard solution, containing 139 mM Na⁺ and 4 mM NH₄⁺ (solution C). Transepithelial ammonia fluxes were 15.6 ± 2.2 pmol/min/mm in NHE4^+/+ mice and 8.7 ± 1.7 pmol/min/mm in NHE4^–/– littermates (P < 0.05) (flow rate: NHE4^+/+, 4.7 ± 0.4 nl/min; NHE4^–/–, 4.0 ± 0.5 nl/min [NS]). Horizontal lines indicate the mean level of flux, and individual symbols each represent an individual animal.

Figure 6. Effect of CMA on NHE4 activity and NHE4 mRNA expression.

(A) NHE4 and Hprt mRNAs were quantified using a Real-Time RT-PCR experiment (see the Methods section). The average results obtained with 10 kidneys homogenates per group are displayed. (B) Basolateral membrane hydrogen ion flux was calculated from the MTALH of control and acidotic rats, in the presence of 1 μM peritubular EIPA to inhibit NHE1 activity (see the Methods section). The average results obtained with 7 independent tubules for the control group and 8 for the acidosis group are displayed.

Figure 7. Effect of CMA on ammonia transport.

(A) The effect of 1 μM peritubular EIPA on transepithelial ammonia absorption in the rat MTALH. Addition of 1 μM EIPA to the peritubular solution did not change transepithelial ammonia flux, flow rate (flow rate: control, 3.8 ± 0.2 nl/min; 1 μM EIPA, 3.8 ± 0.2 nl/min; NS) or Vte (Vte: control, 5.4 ± 2.2 mV; 1 μM EIPA, 2.1 ± 2.1 mV; NS). The individual results obtained with 3 independent tubules are displayed. (B) The effect of 10 μM peritubular EIPA on transepithelial ammonia absorption in the acidotic rat MTALH. Addition of 10 μM EIPA to the peritubular solution elicited a decrease in transepithelial ammonia flux. Vte and flow rate were not affected by EIPA (Vte: control, 4.5 ± 1.3 mV; 10 μM EIPA, 4.2 ± 1.0 mV; NS; flow rate: control, 3.7 ± 0.1 nl/min; 10 μM EIPA, 3.8 ± 0.3 nl/min; NS). The individual results obtained with 4 independent tubules are displayed.