Model Structure of the Na+/H+ Exchanger 1 (NHE1)
2007; Elsevier BV; Volume: 282; Issue: 52 Linguagem: Inglês
10.1074/jbc.m705460200
ISSN1083-351X
AutoresMeytal Landau, Katia Herz, Etana Padan, Nir Ben‐Tal,
Tópico(s)Photoreceptor and optogenetics research
ResumoEukaryotic Na+/H+ exchangers are transmembrane proteins that are vital for cellular homeostasis and play key roles in pathological conditions such as cancer and heart diseases. Using the crystal structure of the Na+/H+ antiporter from Escherichia coli (EcNhaA) as a template, we predicted the three-dimensional structure of human Na+/H+ exchanger 1 (NHE1). Modeling was particularly challenging because of the extremely low sequence identity between these proteins, but the model structure is 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; we suggest that they are involved in the cation binding and translocation. We also suggest a hypothetical alternating-access mechanism that involves conformational changes. Eukaryotic Na+/H+ exchangers are transmembrane proteins that are vital for cellular homeostasis and play key roles in pathological conditions such as cancer and heart diseases. Using the crystal structure of the Na+/H+ antiporter from Escherichia coli (EcNhaA) as a template, we predicted the three-dimensional structure of human Na+/H+ exchanger 1 (NHE1). Modeling was particularly challenging because of the extremely low sequence identity between these proteins, but the model structure is 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; we suggest that they are involved in the cation binding and translocation. We also suggest a hypothetical alternating-access mechanism that involves conformational changes. Sodium/hydrogen transporters are ubiquitous transmembrane (TM) 3The abbreviations used are: TMtransmembraneEcNhaANa+/H+ antiporter from Escherichia coliNHE1Na+/H+ exchanger isoform 1APaminoperimidineMSAmultiple sequence alignment 3The abbreviations used are: TMtransmembraneEcNhaANa+/H+ antiporter from Escherichia coliNHE1Na+/H+ exchanger isoform 1APaminoperimidineMSAmultiple sequence alignment proteins that transport Na+ and H+ ions across the membrane, and are therefore imperative for vital cellular processes such as regulation of cellular pH, cell volume, and ion composition (1Slepkov E.R. Rainey J.K. Sykes B.D. Fliegel L. Biochem. J. 2007; 401: 623-633Crossref PubMed Scopus (194) Google Scholar). The mammalian Na+/H+ exchanger (NHE) family of transporters includes nine isoforms (NHE1 through NHE9), of which NHE1 is the most widely expressed. Following allosteric activation by intracellular acidification, NHE1 exchanges extracellular Na+ for intracellular H+ with Na+:H+ stoichiometry of 1:1 (2Orlowski J. Grinstein S. Pflugers Arch. 2004; 447: 549-565Crossref PubMed Scopus (547) Google Scholar). NHE1 is inhibited by amiloride and its derivatives and by benzoyl guanidium compounds such as cariporide (1Slepkov E.R. Rainey J.K. Sykes B.D. Fliegel L. Biochem. J. 2007; 401: 623-633Crossref PubMed Scopus (194) Google Scholar). Structurally, NHE1 is predicted to include two distinct domains: a TM N-terminal region of ∼500 amino acids that is involved in ion translocation and drug recognition, and a cytoplasmic regulatory C-terminal domain of nearly 300 residues (3Wakabayashi S. Fafournoux P. Sardet C. Pouyssegur J. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 2424-2428Crossref PubMed Scopus (238) Google Scholar). The cytoplasmic domain includes the H+ sensor and also serves to mediate regulation by other molecules or ions. transmembrane Na+/H+ antiporter from Escherichia coli Na+/H+ exchanger isoform 1 aminoperimidine multiple sequence alignment transmembrane Na+/H+ antiporter from Escherichia coli Na+/H+ exchanger isoform 1 aminoperimidine multiple sequence alignment NHE1 is associated with many pathological conditions that include cancer as well as heart, vascular, gastric, and kidney diseases (1Slepkov E.R. Rainey J.K. Sykes B.D. Fliegel L. Biochem. J. 2007; 401: 623-633Crossref PubMed Scopus (194) Google Scholar, 2Orlowski J. Grinstein S. Pflugers Arch. 2004; 447: 549-565Crossref PubMed Scopus (547) Google Scholar). For example, the activity of NHE1 is primarily involved in the damage inflicted on the human myocardium during and following a myocardial infarction, and accordingly, NHE1 inhibitors were shown to be beneficial during ischemia and reperfusion (1Slepkov E.R. Rainey J.K. Sykes B.D. Fliegel L. Biochem. J. 2007; 401: 623-633Crossref PubMed Scopus (194) Google Scholar). In addition, NHE1 plays a role in tumor growth by reversing the pH gradient in malignant cells, a phenomenon known as “malignant acidosis,” which is a key step in oncogenic transformation. Therefore, NHE1 inhibitors can potentially serve as anti-cancer drugs (1Slepkov E.R. Rainey J.K. Sykes B.D. Fliegel L. Biochem. J. 2007; 401: 623-633Crossref PubMed Scopus (194) Google Scholar). NhaA, the main Na+/H+ antiporter in Escherichia coli (EcNhaA), is indispensable for bacterial growth in alkaline pH (in the presence of Na+) and for adaptation to high salinity (4Padan E. Bibi E. Ito M. Krulwich T.A. Biochim. Biophys. Acta. 2005; 1717: 67-88Crossref PubMed Scopus (536) Google Scholar). EcNhaA is an electrogenic antiporter extracting one Na+ ion from the cell in return for inward current of two protons following cellular alkalization (4Padan E. Bibi E. Ito M. Krulwich T.A. Biochim. Biophys. Acta. 2005; 1717: 67-88Crossref PubMed Scopus (536) Google Scholar). The function of EcNhaA is specifically inhibited by 2-aminoperimidine (AP), a guanidine-containing naphthalene derivative with some similarity to the NHE1 inhibitor amiloride (5Dibrov P. Rimon A. Dzioba J. Winogrodzki A. Shalitin Y. Padan E. FEBS Lett. 2005; 579: 373-378Crossref PubMed Scopus (19) Google Scholar). The three-dimensional structure of EcNhaA was recently determined, and found to comprise 12 TM segments (6Hunte C. Screpanti E. Venturi M. Rimon A. Padan E. Michel H. Nature. 2005; 435: 1197-1202Crossref PubMed Scopus (521) Google Scholar). The bacterial EcNhaA and eukaryotic Na+/H+ exchangers play similar roles in controlling pH and electrolyte homeostasis, and have been suggested to share a common ancestor and a similar structural fold (1Slepkov E.R. Rainey J.K. Sykes B.D. Fliegel L. Biochem. J. 2007; 401: 623-633Crossref PubMed Scopus (194) Google Scholar, 7Brett C.L. Donowitz M. Rao R. Am. J. Physiol. 2005; 288: C223-C239Crossref PubMed Scopus (433) Google Scholar). Thus, our working hypothesis was that EcNhaA can be utilized as a template to predict the structure of the TM domain of NHE1. However, the proteins share very low sequence identity of about 10%, and it is not a simple matter to align their sequences and to predict the structure of NHE1 based on that of EcNhaA (8Forrest L.R. Tang C.L. Honig B. Biophys. J. 2006; 91: 508-517Abstract Full Text Full Text PDF PubMed Scopus (197) Google Scholar). In this study, by using a fold-recognition approach, we obtained a three-dimensional model of NHE1. Notably, the membrane topology of this model structure differs from the one that was suggested on the basis of hydrophobicity scales and cysteine accessibility analysis (9Wakabayashi S. Pang T. Su X. Shigekawa M. J. Biol. Chem. 2000; 275: 7942-7949Abstract Full Text Full Text PDF PubMed Scopus (197) Google Scholar). Reasons for the differences are discussed below. Our model of NHE1, which is supported by both phylogenetic and empirical data, incorporates the binding pocket of clinically important NHE inhibitors. This allowed us to locate the binding site of the AP inhibitor within the EcNhaA structure by site-directed mutagenesis. Finally, the integration of empirical data with the new structural model allowed us to suggest an alternating-access mechanism of the Na+/H+ exchange in molecular detail (Fig. 1A). Evolutionary Conservation Analysis of the NhaA Na+/H+ Antiporter Family—Calculation of evolutionary conservation scores was based on a multiple sequence alignment (MSA) of 94 sequences of bacterial NhaA Na+/H+ antiporters using a Bayesian method (10Mayrose I. Mitchell A. Pupko T. J. Mol. Evol. 2005; 60: 345-353Crossref PubMed Scopus (27) Google Scholar). The scores were mapped onto the three-dimensional structure of EcNhaA (Protein Data Bank entry 1ZCD (6Hunte C. Screpanti E. Venturi M. Rimon A. Padan E. Michel H. Nature. 2005; 435: 1197-1202Crossref PubMed Scopus (521) Google Scholar)) using the ConSurf web server (consurf.tau.ac.il/) (11Landau M. Mayrose I. Rosenberg Y. Glaser F. Martz E. Pupko T. Ben-Tal N. Nucleic Acids Res. 2005; 33: W299-W302Crossref PubMed Scopus (1058) Google Scholar). The procedure used to construct the MSA is described in the supplemental data. Evolutionary Conservation Analysis of NHE1-related Na+/H+ Exchangers—Calculation of evolutionary conservation scores was based on an MSA of 305 sequences of Na+/H+ exchangers using a Bayesian method (10Mayrose I. Mitchell A. Pupko T. J. Mol. Evol. 2005; 60: 345-353Crossref PubMed Scopus (27) Google Scholar). Scores were mapped onto the three-dimensional model of NHE1 using the ConSurf web server (consurf.tau.ac.il/) (11Landau M. Mayrose I. Rosenberg Y. Glaser F. Martz E. Pupko T. Ben-Tal N. Nucleic Acids Res. 2005; 33: W299-W302Crossref PubMed Scopus (1058) Google Scholar). The procedure used for MSA construction is described in the supplemental data. Construction of Three-dimensional Model Structure—Modeling of the structure of NHE1 (SwissProt entry SL9A1_HUMAN), residues 126-505, was based on the template structure of EcNhaA (6Hunte C. Screpanti E. Venturi M. Rimon A. Padan E. Michel H. Nature. 2005; 435: 1197-1202Crossref PubMed Scopus (521) Google Scholar), using the homology modeling program NEST (12Petrey D. Xiang Z. Tang C.L. Xie L. Gimpelev M. Mitros T. Soto C.S. Goldsmith-Fischman S. Kernytsky A. Schlessinger A. Koh I.Y. Alexov E. Honig B. Proteins. 2003; 53: 430-435Crossref PubMed Scopus (270) Google Scholar) with default parameters. The final model was based on the pairwise alignments constructed as described under “Results.” Experimental Procedure—EP432 cells transformed with plasmids encoding the various EcNhaA variants were grown. Everted membrane vesicles were prepared and used to determine the Na+/H+ antiporter activity. The procedures are described in the supplemental data. Figures—Figs. 1B, 3B, 4, 5, and 6B were drawn with PyMol (39DeLano W.L. The PyMOL Molecular Graphics System. DeLano Scientific, San Carlos, CA2002Google Scholar) (www.pymol.org).FIGURE 4Titratable residues in the NHE1 and EcNhaA transporters. A side view of the crystal structure of EcNhaA (6Hunte C. Screpanti E. Venturi M. Rimon A. Padan E. Michel H. Nature. 2005; 435: 1197-1202Crossref PubMed Scopus (521) Google Scholar) (A and C) and our model structure of NHE1 (B and D), which are displayed in a ribbon representation with the intracellular region in the upward direction. TM1 and the extra-membranal loops were omitted for clarity. The horizontal green lines mark the approximate boundaries of the hydrocarbon region of the membrane. In panels A and B, the transporters are colored gray, and the locations of the Cα atoms of the titratable residues are depicted as spheres. The red spheres correspond to aspartate and glutamate residues, and the blue to arginines and lysines. In panels C and D the amino acids are colored by their conservation grades using the color-coding bar, with turquoise through maroon indicating variable through conserved. Again, the locations of the Cα atoms of the titratable residues are depicted by spheres. It is evident that a central cluster of titratable residues is located in the conserved protein core, suggesting that it plays important functional roles in the transporters.View Large Image Figure ViewerDownload Hi-res image Download (PPT)FIGURE 5Mutagenesis studies in eukaryotic Na+/H+ exchangers. The three-dimensional model structure of NHE1 is displayed with gray ribbon and the residues that were mutated are presented using space-filled atoms using colors to represent the experimental outcome. Residues that were implicatedinion translocation (Pro167, Pro168, Ser235, Asp238, Pro239, Ala244, Leu255, Ile257, Val259, Phe260, Gly261, Glu262, Asn266, Asp267, Thr270, Ser351, Glu391, Cys421, and Tyr454) are colored red, residues that are involved in pH regulation (Arg180, Arg327, Glu330, Arg440, Gly455, and Gly456) in magenta, residues comprising the NHE-inhibitors binding site (Phe161, Phe162, Leu163, Glu346, and Gly352) in green, and unessential residues (Cys133, Gln157, Pro178, Glu184, Cys212, Glu248, His250, Leu254, His256, Ser263, Val269, Val271, Phe322, His325, Ser359, Asn370, Ser387, Ser388, Ser390, Thr392, Ser401, Thr402, Ser406, Asn410, Lys438, Lys443, Ccys477, Gln495, and Arg500) in yellow (for details see Tables 1S and 2S). A, a top view from the cytoplasmic side of the membrane. B, a side view parallel to the membrane with the intracellular side facing upward; the TM segments are numbered. C, a view from the extracellular side.View Large Image Figure ViewerDownload Hi-res image Download (PPT)FIGURE 6The binding site of the AP inhibitor of EcNhaA. A, the effect of AP inhibition on the antiporter activity of EcNhaA mutants compared with wild-type. The results obtained for the F71C and N64C mutations showing the most drastic AP effect are displayed. Everted membrane vesicles were isolated from EP432 cells expressing wild-type NhaA or the indicated mutants, and the Na+/H+ or Li+/H+ antiporter activity was measured at pH 7.5. At the onset of the reaction, membranes were added first and then Tris d-lactate (2 mm) (↓), and the fluorescent quenching (Q) was recorded until a steady state level of ΔpH (100% quenching) was reached. NaCl or LiCl, at the indicated concentrations, was then added (↑), and the new steady state of fluorescence obtained (dequenching) after each addition was monitored. Where indicated, AP, at the indicated concentrations, was added (↑), to the reaction mixture following the addition of the membranes. The experiments were repeated at least three times with practically identical results. Calculated IC50 is shown for each experiment. B, the crystal structure of EcNhaA (6Hunte C. Screpanti E. Venturi M. Rimon A. Padan E. Michel H. Nature. 2005; 435: 1197-1202Crossref PubMed Scopus (521) Google Scholar) is displayed in a gray ribbon representation. Space-filled atoms represent residues that were examined for their involvement in AP binding (Table 1). Residues that play a role in mediating AP inhibition (Asn64, Phe71, and His225) are colored green, whereas naïve residues (Trp62, Phe72, and Gly76) are colored yellow.View Large Image Figure ViewerDownload Hi-res image Download (PPT) Using the sequence of NHE1 as a target, we detected EcNhaA as the closest homologue according to the fold-recognition FFAS03 server (13Jaroszewski L. Rychlewski L. Li Z. Li W. Godzik A. Nucleic Acids Res. 2005; 33: W284-W288Crossref PubMed Scopus (417) Google Scholar). This finding strengthened our working hypothesis that the TM domains of the two exchangers share a similar fold. Use of Multiple Approaches to Align the TM Domains of NHE1 and EcNhaA—The sequence identity between EcNhaA and human NHE1 is only ∼10%, and we were unable to align their sequences using standard methods (data not shown). We therefore used several state-of-the-art approaches to construct alignments, and integrated the results. First, we extracted the pairwise alignment between NHE1 and EcNhaA, which displays 12.4% sequence identity, from a multiple-sequence alignment of a clan of transporters from the Pfam data base (14Bateman A. Coin L. Durbin R. Finn R.D. Hollich V. Griffiths-Jones S. Khanna A. Marshall M. Moxon S. Sonnhammer E.L. Studholme D.J. Yeats C. Eddy S.R. Nucleic Acids Res. 2004; 32: D138-D141Crossref PubMed Google Scholar). Two additional pairwise alignments were calculated using the FFAS03 (13Jaroszewski L. Rychlewski L. Li Z. Li W. Godzik A. Nucleic Acids Res. 2005; 33: W284-W288Crossref PubMed Scopus (417) Google Scholar) and HMAP (15Tang C.L. Xie L. Koh I.Y. Posy S. Alexov E. Honig B. J. Mol. Biol. 2003; 334: 1043-1062Crossref PubMed Scopus (76) Google Scholar) servers, which display 9.2 and 10.4% sequence identity, respectively. The procedures are described in the supplemental data. TM Helix Assignment—We used each of the above alignments to assign the boundaries of 12 TM segments (TM1-TM12) of NHE1, based on corresponding segments of the crystal structure of EcNhaA. Fig. 2 exemplifies the significant similarity between most of the TM segments predicted by the Pfam, FFAS03, and HMAP alignments. Using the iterative process described below, we predicted the final membrane topology (highlighted in yellow in Fig. 2 and illustrated in Fig. 3A). Initially, the three different alignments were manually adjusted to reduce gaps in the TM helices of EcNhaA, and used to build three-dimensional models of NHE1. The main dissimilarity between the different alignment methods appeared to be in the prediction of the TM6 and TM7 segments. The model structures provided additional information that was used to favor a specific assignment and improve it further; model structures that were favored were those with least polar residues facing the lipid bilayer. Such considerations favored adaptation of the Pfam assignment of TM6; they were not helpful, however, in assigning TM7, for which we therefore used information from a multiple-sequence alignment of homologous eukaryotic Na+/H+ exchangers. Because TM helices are expected not to include insertions and deletions of amino acids (8Forrest L.R. Tang C.L. Honig B. Biophys. J. 2006; 91: 508-517Abstract Full Text Full Text PDF PubMed Scopus (197) Google Scholar), we favored the assignment of TM7 to a gap-free region, as predicted by FFAS03 and HMAP alignments but not by Pfam. Similar reasoning led us to reject the assignment of the first TM segment to residues 103-127, although that was the assignment predicted by all three methods (Fig. 2), because this segment is highly variable and includes many insertions and deletions. In contrast, the next segment (residues 129-150), which was predicted by hydrophobicity analysis (9Wakabayashi S. Pang T. Su X. Shigekawa M. J. Biol. Chem. 2000; 275: 7942-7949Abstract Full Text Full Text PDF PubMed Scopus (197) Google Scholar) to be a TM segment, is devoid of gaps. Interestingly, the conservation pattern in this region is compatible with the periodicity of a helix, i.e. a conserved residue appears at every fourth position, resulting in a conserved helical face (Fig. 3A). Accordingly, this was the region to which we assigned TM1. The above helix assignment of NHE1 was used to refine the pairwise alignment between NHE1 and EcNhaA in the TM regions. The final pairwise alignment displays 10.6% sequence identity (supplemental data Fig. 1S). A three-dimensional model of NHE1 was subsequently constructed on the basis of this alignment and the EcNhaA template. An analysis pertaining to the three-dimensional location of the identical residues between NHE1 and EcNhaA is presented in the supplemental data. The Three-dimensional Model of NHE1 Is Compatible with Evolutionary Conservation Analyses of Na+/H+ Exchangers—In helical proteins, evolutionarily conserved amino acids are typically located in strategic regions at the interfaces between the TM segments, whereas variable residues face the membrane lipids. The extra-membranal loops are also enriched in variable amino acids (16Briggs J.A. Torres J. Arkin I. Proteins. 2001; 44: 370-375Crossref PubMed Scopus (44) Google Scholar, 17Hurwitz N. Pellegrini-Calace M. Jones D.T. Philos. Trans. R. Soc. Lond. B Biol. Sci. 2006; 361: 465-475Crossref PubMed Scopus (34) Google Scholar, 18Fleishman S.J. Unger V.M. Ben-Tal N. Trends Biochem. Sci. 2006; 31: 106-113Abstract Full Text Full Text PDF PubMed Scopus (76) Google Scholar). Accordingly, analyses of evolutionary conservation have been used to predict the structures of membrane proteins (19Fleishman S.J. Unger V.M. Yeager M. Ben-Tal N. Mol. Cell. 2004; 15: 879-888Abstract Full Text Full Text PDF PubMed Scopus (97) Google Scholar, 20Baldwin J.M. Schertler G.F. Unger V.M. J. Mol. Biol. 1997; 272: 144-164Crossref PubMed Scopus (632) Google Scholar, 21Adamian L. Liang J. BMC Struct. Biol. 2006; 6: 13Crossref PubMed Scopus (72) Google Scholar, 22Fleishman S.J. Harrington S.E. Enosh A. Halperin D. Tate C.G. Ben-Tal N. J. Mol. Biol. 2006; 364: 54-67Crossref PubMed Scopus (97) Google Scholar). They have also been exploited to validate model structures (23Fleishman S.J. Ben-Tal N. Curr. Opin. Struc. Biol. 2006; 16: 496-504Crossref PubMed Scopus (57) Google Scholar), as in the present study. We mapped the conservation scores calculated on the basis of the alignment of 94 sequences comprising the bacterial NhaA Na+/H+ antiporter family on the crystal structure of EcNhaA (6Hunte C. Screpanti E. Venturi M. Rimon A. Padan E. Michel H. Nature. 2005; 435: 1197-1202Crossref PubMed Scopus (521) Google Scholar) (Fig. 3B, a and b). As expected, the most highly conserved residues are at the inter-helix interfaces within the TM region, whereas the most variable residues are located in the periphery; where they face the lipid membrane and populate the extra-membranal loops. Reassuringly, a very similar pattern was observed for our model structure of NHE1 (Fig. 3B, c and d). The results, obtained using an alignment of 305 Na+/H+ exchangers related to NHE1, strongly support our NHE1 model structure. Interestingly, a cluster of titratable residues (Fig. 4, A and B), all evolutionarily conserved (Fig. 4, C and D), is located within the conserved core in the center of the membrane in each of the structures. Titratable residues are very rare in the membrane, presumably because of the large desolvation free energy associated with their transfer from the aqueous phase into the membrane (24Bowie J.U. Nature. 2005; 438: 581-589Crossref PubMed Scopus (364) Google Scholar, 25von Heijne G. Nat. Rev. Mol. Cell. Biol. 2006; 7: 909-918Crossref PubMed Scopus (376) Google Scholar). Their presence in the membrane is often associated with function (26Murtazina R. Booth B.J. Bullis B.L. Singh D.N. Fliegel L. Eur. J. Biochem. 2001; 268: 4674-4685Crossref PubMed Scopus (86) Google Scholar). These titratable residues were indeed shown to be essential for the activity of both transporters (27Galili L. Herz K. Dym O. Padan E. J. Biol. Chem. 2004; 279: 23104-23113Abstract Full Text Full Text PDF PubMed Scopus (53) Google Scholar, 28Inoue H. Noumi T. Tsuchiya T. Kanazawa H. FEBS Lett. 1995; 363: 264-268Crossref PubMed Scopus (116) Google Scholar, 29Kozachkov L. Herz K. Padan E. Biochemistry. 2007; 46: 2419-2430Crossref PubMed Scopus (42) Google Scholar) (supplemental data Tables 1S and 2S) and arguably are involved in conformational changes and cation translocation (please see “Discussion” and Fig. 1 for a detailed description of the role of these residues). The NHE1 Model Structure Is Consistent with the Positive-Inside Rule—Gunnar von Heijne and his co-workers (25von Heijne G. Nat. Rev. Mol. Cell. Biol. 2006; 7: 909-918Crossref PubMed Scopus (376) Google Scholar) showed that the topology of the vast majority of TM proteins is such that amino acid positions at the intracellular ends are enriched in the positively charged residues, lysine and arginine, relative to the extracellular side. This observation, termed the positive-inside rule, can be used to predict and evaluate the topology of membrane proteins. Analysis of the NHE1 three-dimensional model (incorporating residues 126-505) revealed 12 lysine/arginine residues on the cytoplasmic side and only 3 lysine residues on the extracellular side (Fig. 3A). For reference, EcNhaA includes 16 lysine/arginine residues on the cytoplasmic side and 5 on the extracellular side (6Hunte C. Screpanti E. Venturi M. Rimon A. Padan E. Michel H. Nature. 2005; 435: 1197-1202Crossref PubMed Scopus (521) Google Scholar). The NHE1 Model Structure Is Consistent with Mutagenesis Studies—Classical genetic and biochemical experiments and site-directed mutagenesis studies of eukaryotic Na+/H+ exchangers (supplemental data Tables 1S and 2S) have yielded abundant data. For simplicity, we divided these data into two main groups: residues that are essential for function versus those that are unessential. Residues were considered essential if their replacement resulted in loss or change of function (e.g. ion-translocation and pH-regulation), or if they were shown to be involved in binding of inhibitors. When these mutagenesis data are projected on the NHE1 model structure, it can be seen that most of the residues defined as essential for activity are located in the core of the TM domain (Fig. 5), which is consistent with their role in maintaining the architecture and function of the transporter. On the other hand, most of the unessential residues face the membrane or are located in the extra-membranal loops. One essential residue, Ser351, unexpectedly faces the membrane lipids, and its functional relevance will be discussed below. Residues that participate in pH regulation, and thus mediate cellular signals, are located both on a cytoplasmic loop and within the protein core. Mutagenesis studies point to 14 residues whose replacement affects the sensitivity of NHE1 to its inhibitors (Table 1S). Some of these mutations do not affect Na+ affinity, implying that the inhibitor-binding site is physically distinct and suggesting that the inhibitors induce allosteric regulation (30Ding J. Rainey J.K. Xu C. Sykes B.D. Fliegel L. J. Biol. Chem. 2006; 281: 29817-29829Abstract Full Text Full Text PDF PubMed Scopus (63) Google Scholar). We focused on residues whose replacement significantly alters sensitivity to NHE inhibitors (i.e. by more than 10-fold), and which are likely to be directly involved in the binding. Specifically, mutagenesis implies that the binding site incorporates residues Phe161, Phe162, and Leu163, all located in TM2, and a second region comprising Gly352 of TM8 and Glu346 on its preceding loop (Fig. 5 and Table 1S). These two regions are located close to each other in our model, and Leu163 (TM2) is in direct contact with Glu346 and Gly352 (Fig. 5). Moreover, this binding site is situated at the extracellular side of NHE1, in accordance with the location of the inhibitors (31Counillon L. Noel J. Reithmeier R.A. Pouyssegur J. Biochemistry. 1997; 36: 2951-2959Crossref PubMed Scopus (65) Google Scholar). All in all, our NHE1 model structure is in excellent agreement with the mutagenesis data. We note that our model was built independently of the mutagenesis data. The final alignment that was used to construct the structural model is very similar to the initial alignments that were obtained using purely computational methods (presented in Fig. 2). Therefore, the projection of the mutagenesis data on models constructed based on these alignments, which are totally independent of prior knowledge regarding the mutations, show similar results (an example is provided in supplemental Fig. 4S). Mutations that alter the binding affinity of the NHE inhibitors were located in equivalent positions in a few eukaryotic NHE isoforms, implying that these isoforms share a common binding site (Tables 1S and 2S). Thus, we assumed by extrapolation that the AP inhibitor of EcNhaA binds to an equivalent location on EcNhaA. Accordingly, we designed and isolated seven mutations in residues located in TM2 and TM8 of EcNhaA and examined the sensitivity of their Na+ or Li+/H+ activity to AP inhibition (Table 1 and Fig. 6). The Na+/H+ antiport activity was measured in everted membrane vesicles isolated from EP432 transformed with the plasmids encoding the various mutations. EP432 lacks the chromosome-encoded antiporters (EcNhaA and EcNhaB) and expresses only the EcNhaA variants from a plasmid. Addition of the respiratory substrate, lactate, to these membrane vesicles (downward-facing arrow in Fig. 6) resulted in generation of ΔpH, as monitored by quenching of the fluorescence of acridine orange, a fluorescent probe of ΔpH. Addition of either Na+ or Li+ to the reaction mixture (upward-facing arrow in Fig. 6) initiated the Na+ or Li+/H+ antiport activity, as monitored by dequenching of the fluorescence. EP432 transformed with plasmid pAXH (32Olami Y. Rimon A. Gerchman Y. Rothman A. Padan E. J. Biol. Chem. 1997; 272: 1761-1768Abstract Full Text Full Text PDF PubMed Scopus (94) Google Scholar) or the vector plasmid pBR322 served as positive and negative controls, respectively. To determine the effect of AP on the antiport activity, we added the inhibitor at various concentrations before adding lactate. The half-maximum inhibitory concentration (IC50) of AP was determined as described (5Dibrov P. Rimon A. Dzioba J. Winogrodzki A. Shalitin Y. Padan E. FEBS Lett. 2005; 579: 373-378Crossref PubMed Scopus (19) Google Scholar).TABLE
Referência(s)