Highly Conserved Salt Bridge Stabilizes Rigid Signal Patch at Extracellular Loop Critical for Surface Expression of Acid-sensing Ion Channels*

Background: Plasma membrane expression is vital for the function of ASICs, which act as extracellular proton sensors. Results: Mutations in a conserved salt bridge and its adjacent region impaired cell surface ASIC expression. Conclusion: Surface ASIC expression involves an exposed rigid signal patch at the extracellular loop. Significance: This finding sheds lights on new strategies to prevent excessive neuronal excitability associated with ASIC activation. Acid-sensing ion channels (ASICs) are non-selective cation channels activated by extracellular acidosis associated with many physiological and pathological conditions. A detailed understanding of the mechanisms that govern cell surface expression of ASICs, therefore, is critical for better understanding of the cell signaling under acidosis conditions. In this study, we examined the role of a highly conserved salt bridge residing at the extracellular loop of rat ASIC3 (Asp107-Arg153) and human ASIC1a (Asp107-Arg160) channels. Comprehensive mutagenesis and electrophysiological recordings revealed that the salt bridge is essential for functional expression of ASICs in a pH sensing-independent manner. Surface biotinylation and immunolabeling of an extracellular epitope indicated that mutations, including even minor alterations, at the salt bridge impaired cell surface expression of ASICs. Molecular dynamics simulations, normal mode analysis, and further mutagenesis studies suggested a high stability and structural constrain of the salt bridge, which serves to separate an adjacent structurally rigid signal patch, important for surface expression, from a flexible gating domain. Thus, we provide the first evidence of structural requirement that involves a stabilizing salt bridge and an exposed rigid signal patch at the destined extracellular loop for normal surface expression of ASICs. These findings will allow evaluation of new strategies aimed at preventing excessive excitability and neuronal injury associated with tissue acidosis and ASIC activation.

Acid-sensing ion channels (ASICs) are non-selective cation channels activated by extracellular acidosis associated with many physiological and pathological conditions. A detailed understanding of the mechanisms that govern cell surface expression of ASICs, therefore, is critical for better understanding of the cell signaling under acidosis conditions. In this study, we examined the role of a highly conserved salt bridge residing at the extracellular loop of rat ASIC3 (Asp 107 -Arg 153 ) and human ASIC1a (Asp 107 -Arg 160 ) channels. Comprehensive mutagenesis and electrophysiological recordings revealed that the salt bridge is essential for functional expression of ASICs in a pH sensing-independent manner. Surface biotinylation and immunolabeling of an extracellular epitope indicated that mutations, including even minor alterations, at the salt bridge impaired cell surface expression of ASICs. Molecular dynamics simulations, normal mode analysis, and further mutagenesis studies suggested a high stability and structural constrain of the salt bridge, which serves to separate an adjacent structurally rigid signal patch, important for surface expression, from a flexible gating domain. Thus, we provide the first evidence of structural requirement that involves a stabilizing salt bridge and an exposed rigid signal patch at the destined extracellular loop for normal surface expression of ASICs. These findings will allow evaluation of new strategies aimed at preventing excessive excitability and neuronal injury associated with tissue acidosis and ASIC activation.
We have observed considerable Asp(Glu)-Arg(Lys) interaction pairs in the extracellular loop of the crystal structure of cASIC1, of which functions are not fully explored. Using sequence alignment and alanine-scanning mutagenesis strategies, we systemically examined the roles of these putative salt bridges in ASIC gating. Here, we focus on a conserved salt bridge residing at the finger domain, namely the Asp 108 -Arg 161 interaction pair on cASIC1 and the equivalent pairs, Asp 107 -Arg 160 on human ASIC1a (hASIC1a) and Asp 107 -Arg 153 on rat ASIC3 (rASIC3) channels. The crystal structure revealed that the acidic side chain of aspartate faces toward the basic side chain of arginine with the O-N distance stabilized at ϳ3.0 Å, suggesting a strong electrostatic interaction between these Asp-Arg residues.
Using a diversity of approaches, including comprehensive mutagenesis, electrophysiological recording, charge swapping, cysteine cross-linking, Western blotting, immunocytochemistry, molecular dynamics (MD) simulations, normal mode analysis (NMA), and protein flexibility analysis, we demonstrate that this highly conserved extracellular salt bridge, together with its adjacent residues, which form a rigid signal patch, plays an essential role in normal cell surface expression of ASIC channels.

EXPERIMENTAL PROCEDURES
Solutions, Drugs, Cell Culture, and Transfection-All solutions and drugs were purchased and prepared as described previously (17). All DNA constructs were expressed in Chinese hamster ovary (CHO) cells as described previously (17). Briefly, CHO cells were cultured in F-12 medium at 37°C in a humidified atmosphere of 5% CO 2 and 95% air. Transfections of plasmids were performed using Hilymax (Dojindo Laboratories, Kumamoto, Japan) following the manufacturer's recommendation. Electrophysiological measurements were performed 24 -48 h after transfection, whereas cell surface biotinylation and immunohistochemistry experiments were performed 24 h after transfection.
Site-directed Mutagenesis-As described previously (17), the cDNA of rASIC3 or hASIC1a was subcloned into the pEGFP-C3 vector. Each mutant was generated with the QuikChange mutagenesis kit. Individual mutations were verified by DNA sequence analysis.
Electrophysiology-As described previously (17), ASIC currents were recorded with the whole cell patch-clamp method. In CHO cells, GFP-positive cells were selected for recordings of ASIC currents. Membrane currents were measured using a patch clamp amplifier (Axon 700A, Axon Instruments, Foster City, CA). The membrane potential was held at Ϫ60 mV throughout the experiment under voltage clamp conditions. All experiments were carried out at room temperature (23 Ϯ 2°C). Solutions were applied using a "Y-tube" method throughout the experiments (29).
Cell Surface Biotinylation and Western Blotting Analysis-Surface biotinylation was performed on cultured CHO cells following established protocols (30). Briefly, the cells were washed in chilled PBSϩ/ϩ (containing 1 mM MgCl 2 and 0.1 mM CaCl 2 , pH 8.0) three times and then incubated with 2 mM sulfo-EZ-Link Sulfo-NHS-LC-biotin (Pierce) in the same buffer at 4°C for 30 min. The reaction was terminated by further incubating the cells with 20 mM glycine in PBS. After washing with chilled PBSϩ/ϩ three times, the cells were collected and lysed with radioimmune precipitation assay buffer. Biotinylated proteins were separated from the intracellular protein fraction using agarose resin linked to NeutrAvidin (Pierce) by incubation overnight at 4°C and subsequent centrifugation. The beads were washed five times with ice-cold PBS, and bound proteins were eluted with the boiling SDS sample buffer, whereas a 10% volume of the supernatant (unbound fraction) was diluted with the SDS sample buffer and used as the total protein fraction. Protein samples from the biotinylation assay were analyzed by Western blotting. After boiling at 100°C for 5 min, the samples were separated by SDS-PAGE and transferred to PVDF membrane. The membrane was incubated overnight at 4°C with anti-GFP (1:1000; Roche Applied Science) and anti-ASIC1a (1:500; Santa Cruz Biotechnology, Inc., Santa Cruz, CA), antitransferrin receptor (1:500; Invitrogen), or anti-GAPDH (1:4000; Kang Cheng) antibodies, followed by appropriate HRP-conjugated secondary antibodies and finally visualized using an ECL solution (Pierce) and exposure to x-ray films for 1-3 min.
Insertion of Hemagglutinin (HA) Tag-To visualize the surface expression of ASIC channels, a modified HA epitope was inserted between Glu 296 and Pro 297 of wild-type (WT) or mutant rASIC3. The nucleotide sequence of the HA-tagged region was TACCCATACGATGTTCCAGATTACGCT.
Immunocytochemistry-CHO cells transfected with WT and mutants of EGFP:HA-ASIC3 were used. At 24 h post-transfection, cells were fixed with 4% paraformaldehyde/PBS for 10 min at room temperature. For the permeabilized experiments, cells were treated with 0.1% Triton X-100/PBS for 5 min at room temperature. Then cells were blocked for 1 h with PBS supplemented with 10% FBS. Cells were incubated overnight at 4°C using an anti-HA antibody (Covance, Berkeley, CA) diluted at 1:1000 in 10% FBS/PBS. After rinsing with PBS, the cells were then treated with the secondary antibody conjugated to Alexa 594 (1:2000; Invitrogen) for 1 h at 37°C and washed three times with PBS before the coverslips were mounted onto slides for visualization by confocal fluorescence microscopy.
Homology Modeling-As described previously (17), a homology model of rASIC3 was created based on the cASIC1 structure (Protein Data Bank entry 3HGC, 3.0 Å) using Modeller 9V7 (31). The sequence of rat ASIC3 was retrieved from the UniProt (entry O35240). The alignment of the target sequence with the template was performed mainly using the method reported by Jasti et al. (15). According to the secondary struc-ture information of the template, the sequence alignment was adjusted manually to be more reasonable. The constructed model was checked and validated by the program Procheck (32). pK a values and structures under normal or acidic pH conditions were calculated using PROPKA version 2.0 (33).
Channel Dynamics Simulations-The titrated structures of WT or mutant ASIC3 were used as the starting structures for MD simulations. All simulations used the OPLS-AA force field for proteins and ions and the simple point charge model for water. A large 1-palmitoyl-2-oleoyl-phosphatidylcholine bilayer was constructed. Redundant lipids were subsequently removed from the constructed bilayer to generate a suitable membrane system into which the TM domain of the ASIC3 could be embedded. Counterions were subsequently added to compensate for the net negative charge of the system, and additional salt (150 mM NaCl or 90 mM CaCl 2 ) was added to mimic the microenvironment under which ASIC3 channels would be kept at desensitization or resting state (34,35). All MD simulations were performed by using the program Desmond (36), which uses a particular "neutral territory" method called the midpoint method to efficiently exploit a high degree of computational parallelism. All MD simulations were performed using the default Desmond parameters at constant temperature (310 K) and pressure (1 bar) by using the Berendsen coupling scheme with one temperature group. As described previously (37), NMA was conducted using the Web server developed by Delarue et al. (see the NOMAD-Ref Web site). Root mean square (r.m.s.) fluctuations of each residue from MD trajectories or NMA modes were calculated according to the procedure described previously (37).
Data Analysis-Results are expressed as means Ϯ S.E. Data were analyzed with Clampfit 10.2 (Molecular Devices). Statistical comparisons were made with Student's t test; p Ͻ 0.05 (*) and p Ͻ 0.01 (**) are considered statistically significant.

Highly Conserved Salt Bridge in Extracellular Loop across ASIC Family-
The three-dimensional structures of cASIC1 (15,16) revealed that residues Asp 108 and Arg 161 , which reside in the ␣1 and ␣3 short helices of the extracellular loop, respectively, with their side chains pointing to each other, form a salt bridge (Fig. 1A). Amino acid sequence alignment showed that these two residues are highly conserved across the ASIC family, forming a Asp 107 -Arg 160 pair in hASIC1a and a Asp 107 -Arg 153 pair in rASIC3 (Fig. 1B). Salt bridges are generally seen in flexible structural motifs of ion channels and commonly implicated in channel dynamics (e.g. channel gating, coupling, and conformational transitions) (26,38,39). However, the salt bridge identified here may be less flexible, given its location in, apparently, a rigid ␣-helix (Fig. 1A). Further protein flexibility analysis confirmed the inflexibility of the salt bridge (see below), suggesting an unusual role in ASIC functioning.
Mutations at Salt Bridge and Its Adjacent Residues Decrease Current Density of ASIC3 Channels-To elucidate the functional roles of the salt bridge, we first replaced the two residues with alanine (D107A and R153A) in rASIC3. CHO cells expressing the ASIC3 D107A mutant showed negligible acid-induced currents. Similarly, cells expressing ASIC3 R153A had markedly attenuated acid-induced currents as compared with those expressing WT channels (Fig. 2, A and B). This result suggests a crucial role of the Asp 107 -Arg 153 pair in ASIC3 function. We then expanded mutagenesis to adjacent residues, namely Thr 104 and Gln 149 in rASIC3, which may play a role in stabilizing the conformation of the salt bridge via hydrogen bonding (see Fig. 1A, bottom; equivalent residues of cASIC1, Thr 105 and Glu 157 ). Indeed, alanine substitution of either Thr 104 or Gln 149 also reduced acid-induced currents (Fig. 2, A  and B), consistent with a regulatory role of these residues. Addi- tional mutagenesis with varying charges (D107R, R153D, and R153S) or side chain lengths (D107E, R153K, and T104S) was performed, and functional data confirmed that both steric hindrance and residue charges are important determinants of ASIC3 function because the mutations either fully abolished or tended to decrease acid-induced currents (Fig. 2B). Notably, any change with Asp 107 , even a minor alteration such as the D107E substitution, completely abolished acid-induced currents in CHO cells, supporting the notion that the Asp 107 -Arg 153 pair and their adjacent residues contribute to ASIC3 function.
Impairment of Salt Bridge Does Not Alter pH Sensing-The pH at which a protein is placed is crucial for the stability of salt bridges, largely due to alterations in electrostatic interactions. At pH extremes, two amino acids participating in a salt bridge could lose their ability to interact because one will lose its charge. Thus, the decrease in acid-induced current in the mutated ASIC3 channel could be attributed to impaired pH sensing mediated by the salt bridge and its related conformational transitions (26,38,39). To test this possibility, we compared the apparent proton affinities between WT and the functional mutant channels by measuring pH 50 values (pH required for inducing half-maximal currents). Unexpectedly, in contrast to the marked change of current density, no change or only slight changes were observed for the pH 50 (Fig. 2C). In support of such an observation, we found poor correlation between the current density and the apparent proton affinity (R 2 ϭ 0.35, p Ͼ 0.05; Fig. 2D).
We previously reported that ASIC3 channels could be activated independent of acidosis by either a synthetic small mole-cule (2-guanidine-4-methylquinazoline (GMQ)) or the deprivation of extracellular Ca 2ϩ (17). Taking advantage of these nonproton activation approaches, we further tested the activation of ASIC3 mutants in the absence of pH changes in order to rule out a role of the salt bridge in pH sensing. Indeed, we found that the mutations (D107A, R153A, and T104A) also markedly reduced GMQ-and Ca 2ϩ deprivation-induced ASIC3 activation (Fig. 3, A, B, D, E) without affecting the EC 50 for GMQ (GMQ concentration required for inducing half-maximal currents) in all functional ASIC3 mutants (Fig. 3C). Together, these data suggest that the residues forming the salt bridge do not directly sense protons. Thereby, other mechanisms may play a leading role in channel dysfunction observed in the ASIC3 mutants.
Breaking Salt Bridge Reduces Cell Surface Expression of ASIC3 Channels-Decreased protein expression level or defective trafficking to the cell surface could also account for partial or complete loss of channel function in mutated channels. Indeed, for mutants at Asp 107 -Arg 153 and adjacent Thr 104 and Gln 149 , although the total protein levels were comparable with the WT, cell surface expression levels were significantly decreased (Fig. 4, A, B, and D). As a control, we also tested a known mutant (E79A/E423A) with gating defects (Fig. 2B) unrelated to the salt bridge disruption (17). The expression of the E79A/E423A mutant on the cell surface was comparable with that of WT channels (Fig. 4, C and D). Thus, disruption of the salt bridge reduced ASIC3 current density predominantly through decreasing cell surface expression. This change does not seem to result from altered protein synthesis or degradation because the total ASIC3 protein levels were unchanged (Fig. 4, A-C).
Immunohistochemistry Confirms Functional Role of Salt Bridge in Surface Expression-The reduced surface expression was further verified using immunohistochemistry. For this purpose, we inserted an HA epitope into the extracellular loop of ASIC3 (Fig. 5A). We selected the site of insertion based on alignment of the ASIC family, which revealed an extra segment specific to ASIC3 in the extracellular loop just below the thumb domain ( Fig. 5A; see "Experimental Procedures"). This construct (EGFP:HA-ASIC3) was functional when expressed in CHO cells, with current kinetics and density comparable with the WT channels (Fig. 5B).
To visually measure differences between WT and mutant channels in surface expression, we constructed three additional mutants (HA-ASIC3 D107A , HA-ASIC3 D107E , and HA-ASIC3 R153A ) based on EGFP:HA-ASIC3 (with a GFP tag). Immunolabeling of non-permeabilized cells expressing the HA-tagged WT channel using the anti-HA antibody revealed a pattern of HA labeling (red) around the cell periphery that overlapped with the GFP (green) fluorescence only at the outside margin of cells (orange) (Fig. 6A, top). In permeabilized cells, the GFP and HA fluorescence colocalized perfectly (Fig. 6A,  bottom). In contrast, cells expressing GFP:HA-tagged ASIC3 D107A , ASIC3 D107E , or ASIC3 R153A showed undetectable or very weak background-like fluorescence with HA labeling under non-permeabilized conditions, despite the prominent labeling and colocalization of the HA signal with GFP in per-meabilized cells (Fig. 6, B-D). These results together with the Western blotting assay (Fig. 4) demonstrate an essential role of the salt bridge in supporting cell surface expression of ASIC3 channels.
Equivalent Salt Bridge (Asp 107 -Arg 160 ) Contributes to Surface Expression of ASIC1a-To determine how common among ASIC channels the extracellular salt bridge is in supporting cell surface expression, we analyzed the role of Asp 107 -Arg 160 in hASIC1a (Fig. 1B). As with the mutations of ASIC3, equivalent mutations at ASIC1a (Asp 107 , Arg 160 , Ser 104 , and Glu 156 ) profoundly decreased acid-induced current density of ASIC1a (Fig.  7, A and B). Also similar to the ASIC3 mutants, pH 50 values of the functional ASIC1a mutants did not have an apparent change (Fig. 7C). In addition, surface expression levels of Asp 107 and Arg 160 mutants were significantly reduced although to a less extent compared with the equivalent ASIC3 mutants (Fig.  7, D and E). These results suggest a common mechanism governing ASIC surface expression that requires a conserved salt bridge at the extracellular loop.
Salt Bridge Is Stable under both Acidic and Neutral pH Conditions during MD Simulations-To obtain further insights into the structural characteristics of the salt bridge competent for the surface expression of ASIC3 channels, we created a three-dimensional homology model of ASIC3 (35) (Fig. 8A,  middle), based on the crystal structure of cASIC1, which is ϳ50% identical to rASIC3 at the amino acid level (see "Experimental Procedures"). We then measured pK a values of the salt bridge (pK a ϭ 2.19, 1.82, and 2.07 for Asp 107 of subunit A, B, and C, respectively; pK a ϭ 11.87, 11.31, and 11.80 for Arg 153 of subunit A, B, and C, respectively; see "Experimental Procedures"). These calculated pK a values suggest that the salt bridge should exhibit an excellent stability in response to acidosis. Given that physiological variations in pH are not acidic enough to deprive the negative charges of Asp 107 , these data further support the notion that Asp 107 -Arg 153 may not directly mediate pH sensing (Fig. 2, C and D).
To better predict the stability of the salt bridge, two sets of MD simulations were carried out at either acidic or neutral pH (Fig. 8A). Titrated ASIC3 channels at pH 5.0 and 7.4 were placed in normal (150 mM Na ϩ ) and high Ca 2ϩ (90 mM) solutions, respectively, to mimic the microenvironments under which ASIC3 was kept stable at the desensitized or resting states (34,35) (see "Experimental Procedures"). Both MD simulations and salt bridge interaction analysis indicate that the salt bridge is rigid, with distances between negatively (O Ϫ ) and positively (N ϩ ) charged atoms varying from 3.0 to 3.5 Å during the simulation (Fig. 8, B and  C), suggesting a persistent electrostatic interaction between Asp 107 and Arg 153 . Moreover, although minor fluctuations of hydrogen bonding between Asp 107 and Arg 153 were observed, the total number of hydrogen bonds was limited to 5-6 ( Fig. 8D) in the trimeric channel, further indicating the high stability of the salt bridge. These results, together with the results of pK a calculations, consistently suggest that the Asp 107 -Arg 153 pair restrained by hydrogen bonding and ionic interaction is very stable at either desensitized or resting states.
Rigid Salt Bridge Is Important for Regional Conformational Stability of ASIC Channels-To better understand the functional significance of the high stability of the salt bridge, we performed MD simulations for the ASIC3 D107A mutant, in which the salt bridge is disrupted. r.m.s. fluctuation of C␣ of    adjacent residues was calculated to check the structural fluctuations of the WT and mutant channels. The results revealed that residues adjacent to the salt bridge were more flexible in ASIC3 D107A than in WT channels (Fig. 9A). Moreover, the virtual mutation ASIC3 D107E , which added a methylene at the side chain compared with WT (supplemental Fig. S1A) caused an obvious bending at D107E (supplemental Fig. S1C) and thus broke the original salt bridge. These results implied that the regional conformation of the salt bridge is extremely precise, and the presence of a stable salt bridge may be important for restraining the conformations of adjacent residues that are critical for ASIC3 surface expression.
To test this idea, we designed two strategies (i.e. cysteine cross-linking and charge swap) to mimic and/or restore the restraints conferred by the salt bridge. As expected, the double cysteine substitutions partially restored the ASIC3 surface expression, whereas single substitutions (ASIC3 D107C and ASIC3 R153C ) did not (Fig. 9, B and C). Cysteines can automati- cally connect to each other when their distance is less than 5 Å (27). MD simulations in supplemental Fig. S2A show that the two cysteines are able to form a strong covalent interaction, although the resulting structure does not overlap well with the WT at the finger domain, which suggested that an obvious conformational change still exists (supplemental Fig. S2C). For charge-swapped mutant (ASIC3 D107R/R153D ), the surface abundance showed no significant improvement compared with that of single mutants (ASIC3 D107R and ASIC3 R153D ; Fig. 9, D and E). The result of MD simulations indicates that the side chains of Asp 153 do not face directly toward Arg 107 ; instead, Asp 153 interacts with both Arg 102 (their number of hydrogen bonds is shown in Fig. 9F) and Arg 107 (supplemental Fig. S2B), with a much weaker attraction force and an incorrect orientation compared with that of the salt bridge in the WT channel (supplemental Fig. S2B). Furthermore, the charge-swapped salt bridge exhibited a weaker restraint on the fluctuations of adjacent residues when compared with that of the cysteine-crosslinking mutant during MD simulations (Fig. 9G). Yet, the degree of overlap with the WT for ASIC3 D107C/R153C in the finger domain is better than that for ASIC3 D107R/R153D (supplemental Fig. S2C, blue arrow), although there are still regions where ASIC3 D107C/R153C differs more from the WT than ASIC3 D107R/R153D (supplemental Fig. S2C, yellow arrow). Together, these analyses suggest that the conserved salt bridge located at the extracellular loop is stringently precise, and it is crucial for the normal surface expression of ASIC3 channels.
Salt Bridge Separates Adjacent Rigid Signal Patch from Flexible Finger Domain-The finding that only the stabilized extracellular salt bridge is able to support ASIC surface expression prompted us to look into its detailed structural basis. For this purpose, NMA simulations, a computational approach that can effectively predict the extensive collective dynamics and inherent flexibilities in biological macromolecules (37,40), were carried out on the rASIC3 three-dimensional model. NMA results and r.m.s. fluctuation analysis uncovered an extremely rigid region, which is surrounded by domains, including the finger, thumb, ␤-ball, knuckle, and upper palm domains (Fig. 10, A and  B). This region is formed by Asp 107 , Arg 153 , Arg 99 , Glu 159 , Asp 160 , Asp 216 , Gln 218 , Gln 219 , Glu 220 , Glu 221 , and adjacent residues (Fig. 10B), most of which exhibited low r.m.s. fluctuation values (Fig. 10C). Any single mutation at this rigid region, designated as the "signal patch" (e.g. E159L, D160L, D216A, Q218L, Q219L, E220L, E221L, or R99A) markedly reduced ASIC3 surface abundance (Fig. 11), whereas mutations on identical or adjacent residues in hASIC1a (Q225C, Q226C, I224C, and N96C) also led to significant decreases in cell surface expressions (41). In contrast, the channel gating-related finger domain is the second most flexible domain in the entire extracellular loop, after loop-␤9 (Fig. 10C). Most notably, the Asp 107 -Arg 153 interaction pair is situated in the interface between the flexible finger domain and the rigid signal patch (Fig. 10, B and C).

DISCUSSION
As the key extracellular proton sensors, ASICs are characterized by a large extracellular loop, within which many conserved residues with unknown functions have been identified. High resolution crystal structures of the ASIC1 channel (15,16) caught in the desensitized state informed us of the presence of a salt bridge, lying at the edge of the finger domain of the large extracellular loop between two helices, formed by side chains of two residues that are highly conserved among ASIC family members. We used a range of complementary techniques to demonstrate that the salt bridge and a structurally rigid signal patch adjacent to it are essential for normal cell surface expression of ASIC3 and ASIC1a channels. Along with the resolution of the crystal structures, extensive investigations at the finger domain have demonstrated its role in channel gating. As the most flexible portions, the motion of the thumb and finger domains is vital to channel gating. Shaikh et al. (42) proposed that protons bind to an acidic pocket between the finger and thumb domains, whereas Yang. et al. (37) suggested that the attraction between the thumb and finger is the initial driving force of extracellular loop movement. These previous studies support that the finger domain mainly functions as a gating element. Here we demonstrate the presence of a patch of residues near the finger domain that are structurally rigid in the tertiary structure of ASICs (Fig. 10). This signal patch forms a firm region exposed to the protein surface as revealed in MD and NMA simulations, resisting structural fluctuations potentially exerted by the nearby flexible finger domain, thereby allowing properly folded ASICs to be recognized by the trafficking machinery for export to the cell surface.
It would be speculative, yet interesting, to find out whether this rigid structure is a potential recognition site for binding different auxiliary proteins that regulate ASIC trafficking to the plasma membrane. Indeed, a number of studies have focused on various interacting proteins involved in regulating cell surface expression of ASICs. For example, ASICs contain PDZ (PSD-95, Drosophila Discs-large protein, zonula occludens protein-1) binding motifs at the C termini. The interaction with CIPP (channel-interacting PDZ domain protein) (43), NHERF-1 (Na ϩ /H ϩ exchanger regulatory factor-1) (44), and Lin-7b (45) increased, whereas that with PSD-95 decreased the cell surface expression of ASIC3 (45). The surface expression of