Conformational changes in NhaA Na⁺/H⁺ antiporter

Abstract

Na⁺/H⁺ antiporters play a primary role in Na⁺/H⁺ homeostasis in cells and many organelles and have long been drug targets. The X-ray structure of NhaA, the main antiporter of Escherichia coli, provided structural insights into the antiport mechanism and its pH regulation and revealed a novel fold; six of the 12 TMs (Trans membrane segments) are organized in two topologically inverted repeats, each with one TM interrupted by an extended chain creating a unique electrostatic environment in the middle of the membrane at the cation binding site. Remarkably, inverted repeats containing interrupted helices with similar functional implications have since been observed in structures of other bacterial secondary transporters with almost no sequence homology. Finally, the structure reveals that NhaA is organized into two functional regions: a ‘pH sensor' – a cluster of amino acyl side chains that are involved in pH regulation; and a catalytic region that is 9 Å removed from the pH sensor. Alternative accessibility of the binding site to either side of the membrane, i.e., functional-dynamics, is the essence of secondary transport mechanism. Because NhaA is tightly pH regulated, structures of the pH-activated and ligand-activated NhaA conformations are needed to identify its functional-dynamics. However, as these are static snapshots of a dynamic protein, the dynamics of the protein both in vitro and in situ in the membrane are also required as reviewed here in detail. The results reveal two different conformational changes characterizing NhaA: One is pH-induced for NhaA activation; the other is ligand-induced for antiport activity.

Keywords::

• Membrane-transport protein
• NhaA
• Na⁺/H⁺ antiporter
• structural-functional dynamics

Introduction

In 1974, P. Mitchell and his colleagues (West and Mitchell 1974) discovered sodium proton antiport activity in bacterial cells and suggested that Na⁺/H⁺ antiporter proteins play primary roles in cell homeostasis. Today, 40 years later, sodium proton antiporters are identified in the cytoplasmic and organelle membranes of almost all cells, including plants, animals and microorganisms (Speelmans et al. 1993, Padan et al. 2004, Orlowski and Grinstein 2007) and they have long been human drug targets (Fliegel 2008). Na⁺/H⁺ antiporters are actively involved in cell energetic, and homeostasis of intracellular pH, cellular Na⁺ content and cell volume (recent reviews in Padan et al. 2004, 2009, Padan 2008, Krulwich et al. 2011). The genome project has yielded a multiplicity of genes that encode putative Na⁺/H⁺ antiporters. They were classified into the monovalent cation proton antiporter (CPA) superfamily (http://www.tcdb.org/). Two subfamilies NHE (Orlowski and Grinstein 2004) and NHA (Brett et al. 2005) contain orthologues ranging from bacterial to human, and possibly share a similar structural fold (Brett et al. 2005, Landau et al. 2007, Slepkov et al. 2007, Schushan et al. 2010).

EcNhaA (or otherwise NhaA) is the main Na⁺/H⁺ antiporter in Escherichia coli which is indispensable for pH and Na⁺ homeostasis (Padan et al. 2005). NhaA is an electrogenic antiporter with a stoichiometry of 2H⁺/Na⁺ and a high turnover rate of 10⁵ per min (Taglicht et al. 1993, Padan et al. 2005). Similar to many other prokaryotic (Padan et al. 2005) and eukaryotic (Putney et al. 2002, Wakabayashi et al. 2003, Orlowski and Grinstein 2004) Na⁺/H⁺ antiporters, NhaA is strictly dependent on pH (Taglicht et al. 1991). It is inactive below pH 6 and its activity rate rises by three orders of magnitude between pHs 7 and 8.5.

The X-ray structure of down-regulated NhaA, crystallized at pH 4 (Hunte et al. 2005, Screpanti et al. 2006; Figure 1A), has provided the first structural insight into the mechanism of antiport and pH regulation (Padan 2008). Although NhaA is a dimer in the membrane (Williams et al. 1999, Gerchman et al. 2001, Hilger et al. 2005, Hilger et al. 2007), the monomer is its functional unit (Rimon et al. 2007, Herz et al. 2009). NhaA consists of 12 trans-membrane helices (TMs). A cytoplasmic funnel opens to the cytoplasm and ends in the middle of the membrane at the putative ion-binding site (Figure 1A, 1B; Hunte et al. 2005). There is also a periplasmic funnel open to the periplasm and separated from the cytoplasmic funnel by a group of densely packed hydrophobic residues.

Figure 1. General architecture of NhaA Na⁺/H⁺ antiporter. (A) Ribbon representation of the crystal structure of NhaA (Hunte et al. 2005) viewed parallel to the membrane (broken line). The 12 TMs are labeled with Roman numerals. The cytoplasmic and periplasmic funnels are marked (black line). Cytoplasmic- and periplasmic-oriented TMs are denoted c and p, respectively. (B) Functional organization in NhaA (Padan 2008). A stick-and-ball representation of functionally important residues: in the putative active site (red or black) and the pH sensor (yellow or magenta). (C) The TM IV/XI assembly with the interrupted TMs and the putative active site (red circle). The Figures were generated by PyMOL.

Remarkably, the NhaA structure revealed a novel fold in which TMs III, IV and V are topologically inverted in respect to TMs X, XI and XII, and in each repeat, one TM (IV and XI, respectively) is interrupted by an extended chain in the middle of the membrane (Figure 1C). This non-canonical TM (IV/XI) assembly creates a delicately balanced electrostatic environment in the middle of the membrane at the putative ion-binding site(s) and has been suggested to play a critical role in the cation exchange activity of the antiporter (Hunte et al. 2005, Screpanti and Hunte 2007). Following NhaA, structures of other bacterial ion-coupled secondary transporters, that share little or no sequence homology with NhaA, have been determined. Interestingly, similar to NhaA, their structural fold revealed inverted topological repeats containing helices, interrupted by extended chain with similar functional implications (reviewed in Padan et al. 2009). Recently, a bacterial homologue of the bile acid sodium symporter, ASBT, has been determined at 2.2 Å resolution. Overall, the architecture of this protein is remarkably similar to that of NhaA, despite having no detectable sequence homology (Hu et al. 2011). The two structures superimpose with an r.m.s.d. of 2.9 Å over 202 C atoms and represent the only two examples where the interrupted helices cross each other. In proximity to this crossing, the sodium-binding site was identified in the ASBT homologue and predicted for NhaA (Hu et al. 2011), in line with previouse deduction on the basis of the NhaA structure, evolutionary conservation and mutation analysis (Padan 2008). It is apparent that in prokaryotic and eukaryotic secondary transporters, conservation of structural fold and functional mechanism exists rather than primary sequence (Jiang et al. 2003, Gouaux 2009, Krishnamurthy et al. 2009, Hu et al. 2011). Hence, the prokaryotic transporters, which are much easier to express and manipulate than their eukaryotic counterparts, have been adopted as model systems. Indeed, using the NhaA crystal structure as a basis, we have recently constructed model structures of the human NHE1 (Landau et al. 2007) and NHA2 (Schushan et al. 2010) (See also the section entitled Evolutionary based computation).

The structure reveals that NhaA is organized into two functional regions (Padan et al. 2009; Figure 1B): (i) The ‘pH sensor', a cluster of amino acyl side chains that are involved in pH regulation, and (ii) a catalytic region containing the ion-binding site that is about 9 Å removed from the pH sensor. Nevertheless, since the structure has been determined at pH 4 and NhaA is down-regulated at this pH, many questions regarding the active conformation(s) remain: How does NhaA translocate the cations? How is energy coupled to translocation? How does pH regulate NhaA?

NhaA catalyzes a coupled stoichiometric translocation of 1 Na⁺ in-exchange for 2 protons across the membrane, in opposite-directions. Therefore, the uphill export of Na⁺ from the bacterial cell against a concentration gradient is achieved by transduction of free energy released from the downhill movement of H⁺ with the electrochemical gradient maintained across the bacterial membrane (ΔμH⁺, interior negative). The essence of the NhaA transport mechanism (Mager et al. 2011) – like that of other secondary transporters – lies in alternative accessibility of the binding site to either sides of the membrane (Jardetzky 1966, Kaback et al. 2011). Hence, NhaA is a ‘nano-machine' that undergoes conformational changes and converts energy forms (ΔµH⁺ ↔ ΔµNa⁺/Li⁺) during turnover. Furthermore, because NhaA is tightly pH regulated, it is expected to undergo both pH-induced and ligand-induced conformational changes. Therefore, to understand the functionality of NhaA in molecular/atomic terms, we need to obtain: (1) The crystal structure of the active NhaA conformations, and (2) the dynamics of the protein both in vitro and in situ in the membrane.

Identifying conformational changes by 3D and 2D crystallization

X-ray crystallography

To determine the structures of the active NhaA conformations, we must obtain crystals that diffract X-rays at alkaline pH, because NhaA is pH-activated between pH 7 and pH 8.5 (Taglicht et al. 1991). We have already obtained 3D crystals at pH 7.5 that diffract X-rays at 5.1 Å resolution (in collaboration with Tsuyoshi Nonaka and Hartmut Michel, Max Planck Institute of Biophysics, Frankfurt, Germany, unpublished results). The structure revealed the backbone of the NhaA dimer which accords with the projection map obtained by cryo electron microscopy of 2D crystals (Williams et al. 1999, Williams 2000). The monomeric helix packing of the dimer crystal structure is very similar to that observed in the crystal structure at pH 4 (Screpanti et al. 2006). Importantly, the results show that NhaA crystallizes at alkaline pH, although, as yet, at low resolution.

Cryo-electron microscopy of 2D crystals

Cryo-electron microscopy of 2D crystals is a very good approach to study conformational changes of membrane proteins because in these crystals the protein is embedded in lipids. However, as yet, high resolution structures have rarely been obtained by this technique. Cryoelectron-microscopy of NhaA 2D crystals (projection map and three-dimensional structure, Williams et al. 1999, Williams 2000, respectively) discovered the NhaA dimer and number of helices but only when the NhaA 3D crystal structure was obtained, helix assignment, topology and side chains were revealed.

Recently (Appel et al. 2009) two sequential conformational changes in NhaA were identified by cryo-electron microscopy of 2D NhaA crystals grown at acidic pH and then incubated at alkaline pH. The first change has been suggested to mark the pH activation of NhaA. It was induced by a rise in pH from 6 to 7 and involved a local ordering of the N-terminus. The second change has been suggested to reflect the cation translocation. It was induced by the ligands, Na⁺ and Li⁺ at pH above 7 and involved displacement of helix IVp toward helix V.

Elucidating whether the crystal structure is native

A crystal structure represents a conformation which is thermodynamically favorable under the crystallization conditions and its relevance to the protein native conformations may not be readily apparent. For crystallization, a membrane protein is extracted from the membrane with detergents; these may or may not correctly replace the membrane and/or change the lipids composition, occasionally needed for stabilizing a native conformation (Palsdottir and Hunte 2004). The crystal lattice can also modify the protein native conformation. Importantly, there are reliable biochemical and biophysical procedures to evaluate whether a crystal structure is native because they give structural/functional information both, in vitro, of the purified protein and in situ in the membrane (See Identifying conformational changes in situ in the membrane section).

In silico predictions of conformational changes

The crystal structure paved the way to structure-based computation.

MCCE analysis and MD simulation

Multiconformation Continuum Electrostatics analysis (MCCE) allows multiple positions of side chain rotamers, hydroxyl protons and water protons in the calculation of the pH dependence of the ionization equilibrium of titratable groups providing the pK_a values of the ioniziable residues. The results of the study with NhaA revealed clusters of electrostatically tightly interacting residues in trans-membrane arrangements (Olkhova et al. 2006). Clusters I and II encompass the pH sensor/transducer at the orifice of the cytoplasmic funnel, cluster III and IV are located at the active site and at the rim of the periplasmic funnel respectively (Figure 1A and B). Importantly, all clusters are electrostatically connected via His256 and Asp133, the former forms tight connections between clusters I and II the latter couples cluster II and III. Many residues in the pH sensor/transducer (cluster I and II) have an extreme pK_a suggesting no change in the protonation state of these residues within the physiological pH range, unless the NhaA structure alters. Glu78 in the pH sensor, with pK_a near the physiological range is suggested to trigger such pH-dependent structural change. Furthermore, the two essential aspartates in the active site (Asp163 and Asp164, Figure 1B) remain protonated until pH activates NhaA.

The physical separation between the pH sensor/transducer and the active site revealed by the structure and computation entails long-range pH induced conformational changes for pH activation as suggested for both prokaryotic and eukaryotic antiporters (Putney et al. 2002, Wakabayashi et al. 2003, Padan et al. 2005, 2009). The importance of the electrostatic network in pH activation of NhaA has been validated experimentally (see section entitled Validation of the computational predictions in the membrane).

Molecular dynamics (MD) simulation provides a powerful tool for computational investigation of protein dynamics (Olkhova et al. 2007). The dynamic behavior of the hydrogen-bonded network in NhaA was studied on shifting the pH from 4–8. The helical regions preserved the general architecture of NhaA throughout the pH change as was experimentally validated (Tzubery 2008). In contrast, upon alkaline shift to pH 8, large conformational drifts occurred in the loop regions as observed biochemically (Gerchman et al. 1999, Venturi et al. 2000) and biophysically (Tzubery et al. 2003, 2008). An increased flexibility of TM IV was also predicted (Olkhova et al. 2007) and verified by cryo-electron microscopy of 2D crystals (Appel et al. 2009). The barrier observed between the funnels at acidic pH was penetrated by water (Olkhova et al. 2007) and a remarkable conformational reorganization was identified in TM X (Olkhova et al. 2009). Accordingly, a pH-dependent movement of Lys300 was verified by cross-linking (Kozachkov et al. 2007). MD simulation was used to advance a model for the dynamics of NhaA during pH regulation and exchange activity (Arkin et al. 2007).

Evolutionary based computation

Using the crystal structure of the EcNhaA as a template and evolutionary conservation, the 3D structures of human NHE1 and NHA2 were predicted (Landau et al. 2007, Schushan et al. 2010). These Na⁺/H⁺ exchangers are vital for cellular homeostasis and play key roles in pathological conditions such as cancer and heart diseases (NHE1) (Fliegel 2008) and essential hypertension (NHA2) (Xiang et al. 2007).

Modeling of NHE1 was particularly challenging because of the extremely low sequence identity between NhaA and NHE1 (Landau et al. 2007). Nevertheless, the model-structure was supported by evolutionary conservation analysis and empirical data. It also revealed the location of the binding site of NHE inhibitors; which we validated by conducting mutagenesis studies with EcNhaA and its specific inhibitor 2-aminoperimidine. The model structure features a cluster of titratable residues that are evolutionarily conserved and are located in a conserved region in the center of the membrane implying that they are involved in the cation binding and translocation.

Human NHA2 is a poorly characterized Na⁺/H⁺ antiporter that only recently has been implicated in essential hypertension (Xiang et al. 2007). We used a range of computational tools and evolutionary conservation analysis to build and validate a three-dimensional model of NHA2 based on the crystal structure of the distantly evolutionary related EcNhaA (Schushan et al. 2010). The model guided mutagenic evaluation of transport function, ion selectivity, and pH dependence of NHA2 by phenotype screening in yeast. A cluster of essential, highly conserved titratable residues were found located in an assembly region made up of two discontinuous helices of inverted topology, each interrupted by extended chain. Whereas in NhaA, oppositely charged residues compensate for partial dipoles generated within this assembly, in NHA2, polar but non-charged residues suffice. This study establishes NHA2 as a prototype for the poorly understood, yet ubiquitous, CPA2 antiporter family recently recognized in plants and metazoans and illustrates a structure-driven approach to get functional information on a newly discovered transporter.

Computation based on pseudo-symmetry of the protein structure and elastic network analysis predicts a pH-induced conformational change in NhaA

Using two computational approaches, we have recently explored (Schuchan et al. 2012) the conformational change undertaken by NhaA following pH elevation to alkaline ranges. Based on pseudo-symmetric features of the crystal structure (Radestock and Forrest 2011), a conformation of NhaA at alkaline pH was modeled. It presents a closed cytoplasmic funnel, and a periplasmic funnel of extended volume in contrast to the cytoplasmic open conformation revealed by the crystal structure (Hunte et al. 2005). To examine the transporter's functional direction of motion, we conducted elastic network analysis of the crystal structure and detected two main normal modes of motion. Notably, the results of both analyses predicted similar trends of conformational changes, consisting of an overall rotational, twisting motion of the two domains around a putative symmetry axis passing approximately through the funnel centers, perpendicular to the membrane plane. This motion, along with conformational changes within specific helices, results in closure at the cytoplasmic- end and opening in the periplasmic-end. The predictions validated by cross-linking experiments suggest that the model-structure and motion depict alkaline pH-induced conformational changes, mediated by a cluster of evolutionarily-conserved, titratable residues, at the cytoplasmic ends of TMs II, V and IX.

Validation of the computational predictions in the membrane

It is most important to validate experimentally that the computational predictions are realized in situ in the membrane. As described above, this was achieved with all computations applied to NhaA. Therefore only one example is presented here.

Combined computational and biochemical study revealed the importance of the electrostatic interactions between the pH sensor/transducer and the cation binding site (Olkhova et al. 2009). Functional and computational studies were conducted on several variants with mutations in the ion binding sites (Asp163, Asp164). The most significant difference found computationally between the wild-type antiporter and the active site variants D163E, D163N and D164N was either a low pK_a value or an impaired electrostatic connection of Glu78 making the active site variants insensitive to pH and inactive. In marked contrast to the dead variants, D164E was predicted to remain active and sensitive to pH. Both predictions were validated experimentally by determining the growth of the variants under selective conditions and their Na⁺/H⁺ antiporter activity in isolated membrane vesicles.

Identifying conformational changes in situ in the membrane

Test of accessibility of a membrane protein to various probes as a function of changes of a specific experimental condition

Accessibility to trypsin, or to other proteases, is monitored by digestion of the protein. When the digestion pattern changes as a function of a change in a particular experimental condition, the specific change, most likely, induces a conformational change in the protein. For example, NhaA has many putative trypsin recognition sites according to the primary structure, but only one of them at Lys249 in TM IX is cleaved by trypsin and only at alkaline pH (Figure 1B). Importantly, the pH profile of the digestion reflects the pH dependence of the NhaA antiport activity (Rothman et al. 1997, Gerchman et al. 1999). Hence, digestion of NhaA by trypsin revealed a pH-induced conformational change in NhaA at TM IX.

Accessibility to monoclonal antibodies (mAb) directed against the target protein is monitored by assaying binding of the mAb to the protein. Several mAbs specific to NhaA were isolated (Padan et al. 1998) and their epitopes were mapped (Rimon et al. 2008). The binding of mAb 1F6 to NhaA revealed a pH-induced conformational change of NhaA at the N-terminus (Venturi et al. 2000). This finding has recently been verified by cryo-electron microscopy of 2D crystals (Appel et al. 2009).

Test of accessibility of site directed Cys replacements to various probes as a function of changes in a specific experimental condition

Many biophysical and biochemical approaches in the study of structure-function relationship in membrane proteins in situ are based on replacement of an amino acid with Cys in a target protein which is engineered as Cys-less (CL). NhaA cysteines, as those of many other membrane transport proteins (Guan and Kaback 2007), can be replaced with other amino acids, with minor effect on activity (Olami et al. 1997). Cys-replacements are inserted site-specifically in CL-NhaA and the effect of the replacement itself is studied on the functionality of the protein and its pH dependence (Cys-scanning). Furthermore, the Cys replacement in CL-NhaA allows the application of various SH-reagents to probe the protein site-specifically for different properties without a background from the native cysteines. Test of accessibility of the Cys replacements to various membrane permeant probes is often done with NEM (N-ethylmaleimide) that ethylates Cys in the presence of water. Therefore, NEM-accessibility traces water filled cavities in the protein. Charged and membrane impermeant probes ca. MTSES^- (2-sulfonatoethyl methanethiosulfonate) and MTSET⁺ ((2-(trimethylammonium)-ethyl)methanethiosulfonate) are used to identify residues that are exposed on the protein surface or connected via water-filled funnels to the environment water space (Guan and Kaback 2007).

Test of accessibility of CL-NhaA Cys replacements to various probes as a function of pH identified locations on NhaA that change conformation with pH. Accessibility test of Cys replacement E252C on TM IX to the membrane permeant fluorescence- probe MIANS (2-(4′-maleimidlanilino)-naphthalnene-6-sulfonic acid) showed that E252C changes conformation with pH (Tzubery et al. 2003) in a pattern very similar to the pH profile of NhaA activity. The ligand (Na⁺ and Li⁺) also induced conformational change at E252C. Test of accessibility of the Cys replacements on TM IX and II to various probes revealed that the cytoplasmic funnel deepens at alkaline pH as compared to acidic pH (Tzubery et al. 2008) and that E65C on TM II changes conformation with pH (Herz et al. 2010). Remarkably, similar pH profiles were found for the identified pH induced conformational changes and for the activity of NhaA in situ.

Estimation of intra-molecular distances between pairs of Cys replacements as a function of changes in a specific experimental condition

Distances between two Cys replacements can be estimated by various probes: bifunctional cross linking reagents of known length, fluorescent (FRET, Fluorescence Energy Transfer) or EPR (Electron Paramagnetic Resonance) probes. Determining of the distance is performed as a function of changes in a particular experimental condition and a distance change implies that the change induces a conformational change in the protein. To identify a pH-induced conformational change in NhaA cross linking was determined as a function of pH. It revealed that at alkaline pH the distance between the conserved center of helix X and E78 of TM II is drastically decreased as compared to the crystal structure at pH 4, implying a pH-induced conformational change in one or both helices (Kozachkov et al. 2007). Singly spin-labeled mutants of NhaA were used to determine distances between NhaA monomers as a function of pH by EPR techniques. The structure at the monomer-monomer interaction changes only moderately with pH (Hilger et al. 2005).

Site-specific Trp fluorescence identifies conformational changes in proteins and is a tool to study their dynamics

It is practically impossible to assess the dynamics (kinetics) of conformational changes using any of the approaches described above, because the amount of time needed for these experiments is greater than the turnover rate of the transporter, and specifically that of NhaA, which is one of the fastest transporters (Taglicht et al. 1991). For this reason, we pursued in parallel the use of site-specific tryptophan fluorescence. This approach enables identification of conformational changes (Menezes et al. 1990, Weitzman et al. 1995, Smirnova et al. 2009) on the basis of fluorescence changes that can be easily recorded. However, the technique is often complicated by high background fluorescence of native Trps that do not change conformation.

Paving the way to site-directed Trp fluorescence with NhaA

NhaA has eight tryptophans with the fluorescence emission spectrum characteristic to Trp (Lackowicz 1999) (Figure 2A). Hardly any change in the Trp fluorescence emission spectrum of wild type NhaA was observed after changing either Li⁺/Na⁺ concentration or pH (Kozachkov and Padan 2011) implying that many of the native Trp do not respond to the changes and their fluorescence background interferes with the use of Trp fluorescence as a probe. We therefore constructed functional Trp-less NhaA by replacing each Trp with Phe.

Figure 2. Identifying conformational changes of NhaA by site-directed Trp fluorescence. (A) The Figure shows the NhaA α-chain and the native Trp (green), the Trp replacements of Phe (orange), and the essential Asp163 and Asp164 in the active site (blue) in ball representation. (B, D, E, G, H). The steady-state fluorescence emission spectra of purified WT NhaA and its variants as indicated in DDM micells and the effect of pH or Li⁺. The proteins were affinity purified and incubated at protein concentration of 0.3–1 μM in a reaction mixture containing 20 mM 1,3-bis[tris(hydroxymethyl)methylamino]propaneat the indicated pH values, 150 mM choline chloride, 5 mM MgCl₂ 10% sucrose, and 0.015% DDM at pH 8.5 in the presence (red) or absence (black) of 10 mM LiCl or at pH 6 (green) without LiCl. The fluorescence emission spectra were measured with an excitation at 295 nm (B, D, E, H) or at 290 nm (G). For further details see (Kozachkov and Padan 2011). (C, F, I) The two different conformational changes, a pH-induced and a Li⁺-induced, are reflected by the fluorescence changes of single Trp/F136W and Trp/F339W, respectively. The reaction mixtures were as above and the fluorescence emission was measured with the time-dependent mode of the spectro-fluorimeter. The excitation was at 295 nm and the emission was at at the peak of each variant, 344, 335 and 338 nm for Trp-less NhaA, single Trp/F136W, and single Trp/F339W, respectively.

Purified Trp-less NhaA lacks the Trp fluorescence emission characteristic of wild-type NhaA (Figure 2B) and its residual fluorescence emission did not respond to changes either in the ligand concentration or pH (Figure 2B and 2C). Hence, Trp-less NhaA provides a Trp-fluorescence-less background for studying structure-function relationships in NhaA by site-directed Trp fluorescence. Guided by the NhaA structure and in order to minimize structural disturbance, two native Phe residues were replaced with Trp in Trp-less NhaA (F136W and F339W) at sites strategically located in respect to NhaA function: At position 136 (F136W) as a fluorescenc probe near the pH sensor and at position 339 (F339W) at the active site (Figure 2A). Both mutants were found functional with a pH profile similar to that of the wild type (Kozachkov and Padan 2011).

Single Trp/F136W is a reporter of the pH-induced conformational change of NhaA

The single Trp variants exhibited characteristic emission spectra with a typical Trp fluorescence λ_max at around 340 nm (Lackowicz 1999) (Figure 2D and 2G). While in the variants single Trp/F339W (Figure 2G) the λ_max of the emission spectra was insensitive to the pH change, the variant single Trp/F136W showed a red shift of 5 nm (from 338 nm at low pH to 443 nm at high pH) when pH was raised from 6–8.5 (Figure 2D). Hence, the effect of pH on the fluorescence emission spectrum of single Trp/F136W is variant specific.

It was previously shown that the Trp fluorescence spectrum in solution is unaffected by pH at this pH range but it is highly sensitive to the solvent polarity (Lackowicz 1999). Based on previous experimental and computational analysis the cytoplasmic funnel, where Phe136 is located, becomes exposed to water at physiological pH. Therefore, we attribute this red shift in the single Trp/F136W to the change in the polarity of its local environment, most probably from one with a low dielectric constant to a higher one.

In addition to the change in the emission spectrum of single Trp/F136W, (Figure 2D and F), a pH shift from pH 6.0–8.5 induces a red shift and dramatically increases fluorescence in a reversible fashion, and no effect is observed when either Na⁺ or Li⁺ is added. This phenomena is specific to the active NhaA variant, no change in fluorescence was observed when a lethal mutation (D164N) was introduced to this variant to get a single Trp/F136W/D164N (Figure 2E). Remarkably, the pH-dependent fluorescence response of variant single Trp/F136W was not affected at all by either Na⁺ or Li⁺ at any pH. This means that the pH-induced conformational change is unrelated to the ligand binding and most probably reflects the pH activation of the antiporter.

Previous results strongly support the contention that a conformational change induced solely by increased pH activates NhaA (reviewed in Tzubery et al. 2003, Krulwich et al. 2011, Mager et al. 2011). Thus, here we show pH-induced conformational changes which are independent of the presence of Na⁺/Li⁺. The Li⁺/Na⁺-independent pH-induced conformational change has a pH profile very similar to the Na⁺/H⁺ antiporter activity of single Trp/F136W. Similar pH profile has been identified recently in helix I by cryo-electron microscopy (Appel et al. 2009). Remarkably, the crystal structure shows that Phe136 is in close proximity to TM I, where the pH sensor is located (Padan 2008). The distance between the C-α of Gly15 (TM I) and Phe136 (TM IVc) is ∼5 Å.

Single Trp/F339W reports upon the ligand-induced conformational change of NhaA

Single Trp/F339W monitors conformational changes initiated by ligand binding (Figure 2G and I). The fluorescence of F339W is quenched in a concentration-dependent manner and the response is specific to Na⁺ and Li⁺, NhaA ligands. The effect is observed only in a pH range at which NhaA is active. Moreover, a pH profile very similar to that shown for pH activation of NhaA is observed with everted membrane vesicles. The fluorescence quenching is completely reversible when pH is reduced to pH 6 at which NhaA is down regulated (Figure 2I). The signal was completely abolished when the lethal D164N mutation was introduced to the variant (single Trp/F339W-D164N and Figure 2H). Hence, a functional NhaA is needed to obtain the fluorescent signal. Based on these facts, we conclude that single Trp/F339W reports a ligand-induced conformational change. In strong support of this conclusion, cryo-electron microscopy shows directly that there is a ligand-induced conformational change in TM IVp between pH 7.0 and 8.5 (Appel et al. 2009), a pH range very similar to that used here with single Trp/F339W. The study also shows that TM IVp moves towards the active site and therefore towards F339W as well. pH-dependent Li⁺/Na⁺-induced conformational change were also observed at the N-terminal portion of helix IX where W258 (Kozachkov and Padan 2011) and E252 (Tzubery et al. 2003) are located.

Hence, the site-directed Trp fluorescence assay opens new avenues in NhaA research: (i) Identifying additional locations of conformational changes; (ii) identifying indirectly residues in the cation-binding site and in the pH sensor by studying the effects of the corresponding mutations on the fluorescence emission of single-Trp/F339W and single-Trp/F136W, respectively; and finally, (iii) determining the kinetics of the NhaA conformational changes as described below.

Functional dynamics of NhaA

Because of the rapid kinetics of NhaA (Taglicht et al. 1991), many questions regarding NhaA dynamics have remained unanswered; Neither the order of events or the partial reactions of the NhaA transport cycle and their pH regulation are known. It is clear that in addition to the conformational changes (see above), these parameters are needed to understand the molecular mechanism of a transporter. Fortunately, experimental approaches to tackle transporter's dynamics have been initiated.

Site specific fluorescence

Using spectrofluorimeter equipped with stopped flow site directed fluorescence of Trp has been used to explore conformational changes and different kinetic aspects of transport protein activity (Smirnova et al. 2009, 2011, Rasmussen et al. 2010, Imhof et al. 2011). For example, in LacY this approach brought a line of independent evidences which reinforce the alternating access mechanism (Smirnova et al. 2009) and determined the rate limiting step of the transport cycle (Smirnova et al. 2011). Recently, single-molecule fluorescence resonance energy transfer imaging has been used to reveal the dynamics of conformational changes during Na⁺ coupled transport in LeuT, a bacterial homologue of neurotransmitter/Na⁺ symporters (Zhao et al. 2011).

We are using a spectrofluorimeter with stopped flow to determine the kinetics of the pH-induced and ligand-induced fluorescence changes, with the aim of elucidating the functional dynamics of NhaA and answering the following critical questions regarding pH regulation and antiport activity: (a) Does the pH-induced conformational change precede ligand-induced conformational changes? Notably, in the eukaryotic antiporter NHE1, the pH-induced conformational change is much slower than the exchange rate (Hayashi et al. 2002). We expect an even more drastic difference with NhaA, owing to its rapid kinetics (Taglicht et al. 1991). (b) What is the rate-limiting step of the cation translocation?

Electrophysiology

Electrophysiology is an established method to characterize eukaryotic transporters-functionality using voltage clamp or patch clamp methods. However, apart from few cases, bacterial transporters cannot be investigated with these techniques because of the small size of bacteria and because bacterial transporters are difficult to express in mammalian cells or oocytes. Electrophysiology based on solid supported membranes (SSM) has been found extremely useful for applying electrophysiology to bacterial transporters (Ganea and Fendler 2009); it has been used to characterize transport of a charged substrate, identify electrogenic partial reactions, determine rate constants and substrate affinities and develop kinetic models for transport reactions.

SSM based-electrophysiology has been applied to the study of the kinetic parameters of NhaA (Zuber et al. 2005, Mager et al. 2011). For the first time, forward and reversed transport reactions were investigated at zero membrane potential using right side out and inside-out membrane vesicles (Mager et al. 2011). Although partial reactions of NhaA have not yet been identified, the electrophysiological results reveal that the Na⁺/H⁺ antiport activity is highly symmetrical and a kinetic model has been advanced in line with the alternating accessibility mechanism (Mager et al. 2011); during ion translocation, Na⁺ and H⁺ compete at the active site. This competition has also been suggested to underlie the pH regulation of NhaA. However, experimental results as well as structural results (this review) imply that in addition to the exchange activity at the active site there is a pH sensor/transducer site that regulates the pH response of NhaA. For example, the pH-induced conformational change revealed by monitoring the fluorescence of single Trp F136W was not affected by Na⁺ concentration (Kozachkov and Padan 2011).

Summary

On the basis of the crystal structure (Hunte et al. 2005) and experimental data (Padan 2008), we have suggested that the TM IV/XI assembly, with its unique and delicate electrostatic balance in the middle of the membrane, allows the rapid conformational changes expected for NhaA. Based on the crystal structure, many computational approaches have been opened providing predictions of the NhaA conformational changes. Experimental approaches to follow the dynamics of NhaA in situ in the membrane have been developed. These have revealed residues that are crucial for NhaA activity and regulation, as well as elucidated the pH- and ligand-induced conformational changes. Ultimately, integrating results of all approaches will shed light on the mechanism of activity and pH regulation of NhaA, a prototype of the CPA2 family of transporters.

Acknowledgements

The review was supported by EDICT (European Drug Initiative on Channels and Transporters); Grant Number: EU EP7 and The USA-Israel Binational Science Foundation Grant Number: BSF#20050130.

Declaration of interest : The author reports no conflicts of interest. The author alone is responsible for the content and writing of the paper.