Dissecting the proton transport pathway in electrogenic Na+/H+ antiporters

Abstract

Sodium/proton exchangers of the SLC9 family mediate the transport of protons in exchange for sodium to help regulate intracellular pH, sodium levels, and cell volume. In electrogenic Na⁺/H⁺ antiporters, it has been assumed that two ion-binding aspartate residues transport the two protons that are later exchanged for one sodium ion. However, here we show that we can switch the antiport activity of the bacterial Na⁺/H⁺ antiporter NapA from being electrogenic to electroneutral by the mutation of a single lysine residue (K305). Electroneutral lysine mutants show similar ion affinities when driven by [Formula: see text]pH, but no longer respond to either an electrochemical potential ([Formula: see text]) or could generate one when driven by ion gradients. We further show that the exchange activity of the human Na⁺/H⁺ exchanger NHA2 (SLC9B2) is electroneutral, despite harboring the two conserved aspartic acid residues found in NapA and other bacterial homologues. Consistently, the equivalent residue to K305 in human NHA2 has been replaced with arginine, which is a mutation that makes NapA electroneutral. We conclude that a transmembrane embedded lysine residue is essential for electrogenic transport in Na⁺/H⁺ antiporters.

Keywords: Na+/H+exchangers; energetics; membrane protein; proton transport; secondary active transporters.

Fig. 1.

The pH dependence of NapA and its putative ion binding site. (A) Cartoon showing the experimental setup for determination of Na⁺/Li⁺ affinity. The ATP synthase and NapA are coreconstituted into liposomes. Free K⁺ diffusion by valinomycin suppresses the effect of $\mathrm{\Delta }\mathrm{\Psi }$. ACMA, 9-amino-6-chloro-2-methoxyacridine. (B) Representative ACMA fluorescence traces for Na⁺/H⁺ antiporter activity. ATP-driven H⁺ pumping (brown arrow) establishes a $\mathrm{\Delta }$pH (acidic inside) as monitored by quenching of fluorescence. H⁺ efflux is initiated by the addition of NaCl/LiCl (black arrow), and further NH₄Cl addition (gray arrow) collapses the proton gradient. (C) The pH dependence of Na⁺/H⁺ (5 mM) and Li⁺/H⁺(5 mM) antiport activities of NapA WT. Apparent affinity values (K_m) are shown at certain pH values. (D) Comparison of Li⁺/H⁺ antiport activities of NapA WT (5 mM) and mutants (10 mM) with apparent K_m values at certain pH values.

Fig. S1.

(A) Degree of maleimide-PEG-5k (black circle) conjugation, to the NapA mutant (V31C), in detergent or proteoliposomes, was used to evaluate the directionality of the protein in liposomes. MW, molecular weight. (B) The pH dependence comparison between NapA WT (circles, 5 mM) and K305H (squares, 10 mM) mutant. (C) NapA mutants react differently to the addition of valinomycin (100 $\mu$M). NapA WT, K305H, S127A, and T126A show the transport activity (pH 8.0) after the addition of valinomycin (100 $\mu$M), which means the $\mathrm{\Delta }\psi$ was built up after addition of LiCl (40 mM), whereas NapA K305A, K305R and K305Q didn’t react to the addition of valinomycin, showing that $\mathrm{\Delta }\psi$ wasn’t built up after the addition of LiCl. (D) NapA WT responds to different membrane potentials. (E) pH dependence of pyranine fluorescence in empty liposomes containing K₂SO₄ or Na₂SO₄. (F) Pyranine fluorescence dependence of proton (H⁺) concentration between pH 6.8 and 8.4. The function is Y = Y0*exp(k*X), where Y0 = 0.1996 and k = 2,196 for K₂SO₄ and Y0 = 0.2340 and k = 2,162 for Na₂SO₄.

Fig. 2.

Switching NapA WT activity from being electrogenic to electroneutral. (A) Cartoon representation of dimeric NapA. Ion translocation domain (core) is colored in blue, dimerization domain (dimer) in wheat, and the linker transmembrane helix (TM6) in gray. Residues used for mutagenesis are depicted in stick form in the right protomer and color-coded (S127 yellow, T126 purple, D156 green, D157 magenta and K305 orange). (Inset) The ion binding site extracted from an Na⁺ (purple sphere) bound state of a molecular dynamic simulation (16). The salt bridge between K305 and D156 is depicted by black dashed line. Water molecules coordinating the ion are represented by red spheres. (B) Representative ACMA fluorescence traces for NapA WT and its mutants at pH 8.0 for Li⁺/H⁺ antiport activity (20 mM). Brown arrow, addition point of ATP (120 $\mu$M); black arrow, addition point of LiCl (20 mM); gray arrow, addition point of NH₄Cl (20 mM). (C) Cartoon showing the experimental setup for Li⁺-driven proton efflux experiments. In the starting conditions, pH values on the inside and outside are identical. Lithium addition drives H⁺ efflux, and alkalinization of the liposome volume is monitored by the pH-sensitive dye pyranine. Valinomycin is added to extinguish the membrane potential. (D) NapA activity monitored with liposome-entrapped pH-sensitive fluorophore dye pyranine recorded at 510 nm (excitation 406 and 460 nm). In WT NapA, addition of LiCl (40 mM) at pH 8.0 builds up a membrane potential (distance a) that was dissipated upon addition of valinomycin at 4 min (distance b) and leads to further H⁺ efflux; this releases the inhibitory membrane potential established during the first transport phase. (E) Valinomycin sensitivity for NapA WT and mutants was calculated by measuring the ratio between distance b and distance a; see D.

Fig. 3.

NapA WT and mutant antiport activity when driven by a membrane potential. (A) Cartoon showing the experimental setup for $\mathrm{\Delta }\mathrm{\Psi }$-driven proton influx experiments. In the starting conditions, H⁺ (pH 7.8) and Li⁺ (100 mM) concentrations are identical. Addition of valinomycin generates a Nernst potential (negative inside) that energizes proton influx in electrogenic transporters that is monitored by the entrapped pyranine. (B) Pyranine traces depicting the amount of WT NapA and mutant activity when transport was driven solely by an electrical potential of $-$116 mV. (C and D) NapA WT and mutants’ response to increasing membrane potentials ($\mathrm{\Psi }$), which was established by a K⁺-diffusion potential, using different KCl concentrations in the measuring buffer.

Fig. 4.

Membrane potential generation in NapA WT and mutants. (A) Cartoon showing the experimental setup for Li⁺-driven proton efflux experiments. In the starting conditions, pH values on the inside and outside are identical. Lithium addition drives H⁺ efflux, and if a membrane potential is generated it was detected by the potential-sensitive dye DiSC₃(5). (B) DiSC₃(5) fluorescence traces showing the generation of a membrane potential in NapA WT (black trace). Valinomycin is added to extinguish the membrane potential. Negative controls with either empty liposomes or liposomes containing D157N are also shown. (C) Similar to B, but electrogenic variants K305H and S127A are shown. (D) Similar to B, but electroneutral variants K305A, K305Q, and K305R are shown.

Fig. 5.

$\mathrm{\Delta }$Li⁺-driven proton efflux of NapA WT and mutants at different pH values. (A) In the absence of $\mathrm{\Delta }$pH, upon the addition of substrate generating an Li⁺ concentration gradient (30 mM), WT NapA and mutants are more active at pH 6.6 and less active at pH 8.0; $\mathrm{\Delta }$H⁺ was calculated as described in Materials and Methods. (B) Representative AMCA traces for Li⁺ (black) and Na⁺ (red) showing that an inactive D156N mutant was rescued by the additional mutation of K305 to glutamine; K_m = 106 ± 22 mM for Li⁺ and K_m = 360 ± 112 mM for Na⁺. In contrast, activity of the D156N mutant could not be rescued by the additional mutation K305N [green traces (Na⁺)]; black arrow, addition of LiCl or NaCl; gray arrow, addition of NH₄Cl (20 mM). (C) As described in Fig. 2D, the valinomycin sensitivity of the NapA D156N-D305Q mutant was measured upon addition of Li⁺ (pH 6.6).

Fig. S2.

Sequence comparison of NapA to human and bacterial Na/H⁺ antiporters. Sequence alignment of T. thermophilus NapA (UniProt: Q72IM4) to Oceanithermus profundus NapA (E4U6Q4; 65% sequence identity), Bacillus cereus NapA (C2MWQ1; 35% sequence identity), Enterococcus hirae NapA (P26235; 30% sequence identity), Methanococcus jannaschii NhaP1 (Q60362; 23% sequence identity), Homo sapiens NHA2 (SLC9B2; Q86UD5; 21% sequence identity), Homo sapiens NHE1 (SLC9A1; P19634; (<15% sequence identity), and E. coli NhaA (P13738; <15% sequence identity) using ClustalW (www.ebi.ac.uk/clustalw/) and MAFFT multiple sequence alignment in Jalview. Highlighted in bold are the conserved ion-binding aspartates in TM5. In Human NHA2, the two aspartates are conserved, but, like electroneutral transporters NHE1 and MjNhap1, the positively charged residue in TM10 is arginine (shown in red) and not lysine as in the electrogenic transporters.

Fig. 6.

The energetics of human NHA2 (hNHA2) antiport activity. (A) Representative ACMA fluorescence traces for hNHA2 antiport activity using the setup described for NapA in the legend of Fig. 1A. (B) The pH-dependent sodium (150 mM) and lithium (60 mM) activity for hNHA2. (C) The hNHA2 activity was monitored with liposome-entrapped pH-sensitive fluorophore dye pyranine recorded at 510 nm (excitation 406 and 460 nm). In hNHA2, addition of LiCl (60 mM) at pH 7.3 does not seem to build up a membrane potential, as no additional increase in H⁺ efflux was observed following addition of valinomycin at 4 min (black trace). As a comparison, the inactive aspartate mutant D279N is shown (green trace). (D) The dependency of hNHA2 antiporter activity in the presence of a $-$116 mV potassium diffusion potential, displaying an absence of $\mathrm{\Delta }\mathrm{\Psi }$-driven proton influx. The pyranine traces depicting the relative activities are shown for WT NapA (black) and the electroneutral NapA mutant K305A (red), in comparison with hNHA2 (green) and empty liposomes (blue).

Fig. S3.

(A) Gel filtration profile of hNHA2 with the purified protein and coomassie-stained SDS gel shown in Inset. (B) The apparent binding affinity (K_M) of hNHA2 for Na⁺ (black circles) or Li⁺ (white circles) at pH 8.0. (C) Sensitivity of hNHA2 WT, D279N mutant, and electroneutral NapA mutations to a membrane potential as assessed by the addition of valinomycin at pH 7.3. (D) Sensitivity of hNHA2 WT and D279N to any membrane potentially generated after LiCl addition, as assessed by the addition of valinomycin after 4 min at pH 6.6.

Fig. 7.

Schematic of the NapA transport cycle. See last paragraph of Discussion for details.