| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Department of Biochemistry, University of Cambridge, Cambridge, United Kingdom; Bioinformatique Structurale, Institut Pasteur, Paris, France; and Chemical Computing Group, Montreal, Canada
Correspondence: Address reprint requests to P.-L. Chau, Bioinformatique Structurale, Institut Pasteur, 75724 Paris, France. E-mail: pc104{at}pasteur.fr.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
8 Å long and terminates 20 Å above the membrane. The exit routes for an agonist (5-HT) and an antagonist (granisetron) were similar, with trajectories toward the membrane and outward from the ligand binding site. 5-HT appears to form many hydrogen bonds with residues in the unbinding pathway, and experiments show that mutating these residues significantly affects function. The location of the pathway is also supported by docking studies of granisetron, which show a potential binding site for granisetron on the unbinding route. We propose that leaving the binding pocket along this tunnel places the ligands close to the membrane and prevents their immediate reentry into the binding pocket. We anticipate similar exit pathways for other members of the Cys-loop receptor family. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
-aminobutyric acid type A (GABAA) receptors. These proteins are responsible for fast synaptic transmission, and are the targets of many neuroactive drugs. Similar to other members of the Cys-loop family, the 5-HT3 receptor forms a pentameric arrangement of subunits (Fig. 1). Each subunit contains an extracellular region and a transmembrane region. The transmembrane region consists of four membrane-spanning 
-helices, M1–M4; the M2 segments from each of the five subunits forms the lining of the central ion-conducting pore. A large loop between M3 and M4 constitutes the majority of the intracellular mass of the protein and is responsible for receptor modulation and channel conductance. The extracellular region contains the ligand binding site, and there have been numerous studies to identify the amino acids responsible for receptor-ligand interactions. However, molecular details of the complete receptor structure are still relatively undefined. This is largely because no high-resolution (x-ray crystallographic) structures of Cys-loop receptors have yet been elucidated. Nonetheless, the structure of a protein homologous to the extracellular domain of the nACh receptor, the acetylcholine binding protein (AChBP), has been resolved to 2.1 Å (1
–3). This structure has proved useful for homology models of this domain both in the nACh receptor and for other Cys-loop receptors, and comparison with electron microscope data from the nACh receptor has indicated that movement in this domain may trigger receptor gating (4).
|
). The gorge residues are considered to provide a rapid yet specific entry pathway for acetylcholine. However, the entry and exit route(s) in the Cys-loop receptor family have not yet been elucidated. This information is of considerable importance, as the route of neurotransmitter entry and/or exit has the potential to be a site of action of neuroactive drugs, and mutations in amino acids in the pathway could be responsible for receptor malfunction. Early electron microscope images of the nACh receptor studies showed the presence of "tunnels" between the large extracellular vestibule of the receptor and the synaptic space, suggesting that the neurotransmitter entry and/or exit route(s) may be via the extracellular vestibule (6). However, the current structural data do not support this hypothesis (4). The aim of our work was to determine the exit route for the 5-HT3 receptor, a typical Cys-loop receptor, using molecular dynamics simulations combined with functional data.
Molecular dynamics simulations have been used to explore a variety of ligand-receptor interactions over the last decade (7,8), with "steered" molecular dynamics developed specifically to simulate ligand unbinding from the receptor (9). This method has been applied to various ligand-receptor systems (10–12), for example, to examine unbinding of the streptavidin-biotin complex where the authors identified conformational changes of biotin during the unbinding process and the breaking of hydrogen bonds between the ligand, the receptor, and some surrounding water molecules (13). However, this method requires the predetermination of the unbinding trajectory, which is not always possible. To overcome this problem, the mutual repulsion method was developed to allow the ligand to explore its own unbinding trajectory (14). Here, the distance between the centers of mass of the ligand and the receptor is incrementally increased during the course of the simulation, but the method does not stipulate the exact pathway. This prevents the imposition of a specific trajectory of the unbinding process and allows the ligand to explore the space available and locate its own optimum unbinding trajectory. In this study, we have used this method to locate the unbinding pathway of both an agonist (5-HT) and an antagonist (granisetron) from their binding sites in the 5-HT3 receptor into the extracellular surroundings.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
). We used this structure to create a homology model as previously described (15). Briefly, the three-dimensional model of the 5-HT3 receptor extracellular domain based on the structure of AChBP was built using MODELLER (16). Sequence alignment was obtained using the program FUGUE. The model was energy-minimized in SYBYL using the AMBER force field. The protein model has 5958 atoms, and it is a cylinder of approximate dimensions 62 x 80 Å.
Modified mutual repulsion simulation
Details of this method have been described previously (14). In the mutual repulsion method, the centers of mass of the ligand and the receptor are assigned what can be called "pseudocharges", g, that increase linearly with time. A potential 
(r) is defined:
 | (1) |
The pseudocharges interact with each other, but do not affect the normal electronic partial charges assigned to each atom. They repel or attract each other under rules similar to those for normal electronic partial charges. This method approaches the problem of unbinding as a rare event with a large energy difference between the bound and the unbound states. The (ri) potential artificially reduces this energy difference so that the transition from one state to another is facilitated. In addition, there are two advantages of the method. First, since the force due to the pseudocharges is calculated with respect to the centers of mass, no torque is generated on the molecules. The molecules will be allowed to explore the unbinding path with fewer artificial forces. Second, since the pseudocharge increases slowly, the potential (ri) can be exploited for the calculation of the Helmholtz free energy by the adiabatic switching method.
The position of 5-HT in the 5-HT3 receptor is shown in Fig. 1. The only discernible opening in the extracellular domain is located toward the base of the ligand binding site, and thus the force was directed to push the ligand out of its binding site in this direction (z direction), although in some simulations the direction was varied to ensure this did not bias the data. The protein was centered with its fivefold axis of symmetry coincident on the z axis, with its extracellular side in the negative z direction and its membrane side in the positive z direction. A point directly above the ligand in the negative z direction was then used as the center of repulsion. The force on the protein was artificially reduced to zero to prevent the receptor from spinning.
Simulation details
All simulations were carried out using a modified form of the DL_POLY molecular dynamics simulation package (17) (see also http://www.cse.clrc.ac.uk/msi/software/DL_POLY/), which incorporated the mutual repulsion method. The CHARMm22 potential was used throughout (18). The cut-off for the nonbonded interactions was 10 Å, as used previously (11).
Two series of simulations were carried out. The first was initiated using an agonist (5-HT) docked into the binding site (15), and the second was initiated with an antagonist (granisetron) docked into the same site (19). In each simulation, the structure of the protein-ligand complex was minimized for 20,000 steps using the zero-K minimization method. All nonhydrogen atoms of the ligand and all C atoms of the protein were tethered to their original positions using a harmonic potential. The structure with 5-HT bound was heated to 310 K over 250 ps, followed by an equilibration period of 100 ps. The granisetron-bound structure was heated to 310 K over 100 ps, followed by an equilibrium period of 50 ps.
Two starting conditions were used after equilibration to examine their effect on the final unbinding trajectory. In the first set of conditions, the unbinding simulation continued from equilibration (no velocity rescaling), and in the second set of conditions the velocities of all particles were rescaled to 310 K at the beginning of the data production run (initial velocity rescaling). In all simulations, unbinding forces were increased to achieve an unbinding speed of 20 m/s (0.20 Å/ps). The time step of the simulation was 1 fs. Data were recorded after the initial equilibration step with configuration-dumping every 1 ps. A Nosé-Hoover thermostat was applied with a thermostat constant of 0.1 ps. All simulations were carried out in vacuum. During the unbinding process, the tethering was reduced to the C atoms of every fifth amino acid; the atomic positions were allowed to deviate from the initial position, but with an energy penalty that is a harmonic potential. In half the simulations, the tethering remained unchanged during the whole unbinding process (symmetric tethering). In the remaining simulations, the tethering was lifted from the C atoms of these binding-path amino acids (asymmetric tethering), allowing the system more flexibility.
Hydrogen bond analysis
A hydrogen bond can be described as follows: B-A....H-D, where A is an acceptor atom, D is a donor atom, B is an atom immediately bonded to A, and H is a hydrogen atom. The conditions for a hydrogen bond are met when the following criteria are satisfied: 1), the distance between A and D is <3.5 Å; 2), the distance between A and H is <2.5 Å; 3), the angle subtended by atoms A, H, and D (conventionally known as 
) is between 130° and 180°; and 4), the angle subtended by atoms B, A, and H (conventionally known as 
) is between 90° and 180° (20). Note that for acceptor atoms that are bonded to two atoms (such as the nitrogen on histidines), B is a dummy atom created by taking the mean position of the two atoms bonded to nitrogen.
Unbinding pathway
A computer program was written to construct a surface of the unbinding pathway. The algorithm was inspired by an earlier work of Chau and Dean (21). An axis was extended from one end of the pathway to the other end. At regular intervals of 0.25 Å, rays were drawn radially outward, normal to the axis, at angular intervals of every 10°. The point where the ray hits a van der Waals surface of the protein atom is recorded. Together, the collection of such points defines the unbinding tunnel. Cross-sectional areas of the tunnel can also be readily computed at different points along the tunnel axis.
Cell culture and molecular biology
Human embryonic kidney HEK293 cells were grown on 90-mm tissue culture plates at 37°C and 7% CO2 in a humidified atmosphere. They were cultured in DMEM/F12 (Dulbecco's modified Eagle medium/nutrient mix F12 (1:1)) with GLUTAMAX I containing 10% fetal calf serum, and passaged when confluent. At 70–80% confluency, cells were transfected with mutant or wild-type DNA by electroporation. Mutations were created using the Kunkel method (22) using 5-HT3A(b) subunit DNA (accession AY605711), as described previously (23).
FlexStation assays
These were as previously described (24). Briefly, at 36–48 h posttransfection, cells were washed in flex buffer (115 mM NaCl, 1 mM KCl, 1 mM CaCl2, 1 mM MgCl2, 10 mM glucose, 10 mM HEPES, pH 7.4) and 100 µl of membrane potential dye (Molecular Devices, Wokingham, UK) was added to each well. The plates were incubated at room temperature for 45–60 min before being placed in the FlexStation. Fluorescence was measured every 2 s for a total experimental period of 200 s. At 20 s, 50 µl of either agonist or flex buffer was added to each well.
Softmax Pro (Molecular Devices) or PRISM (Graph Pad, San Diego, CA) was used for data analysis. The percent change in fluorescence, which was calculated as F (peak fluorescence) minus F0 (baseline fluorescence at 20 s) divided by 5-HT Fmax (peak fluorescence at 30 µM 5-HT), was compared across 5-HT concentrations using F = Fmax/(I + EC50/[5-HT]nH), where I is the change in fluorescence, EC50 is the concentration required for the half-maximal response, and nH is the Hill coefficient.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
2.5 Å, which shows that the protein structure was preserved during the course of unbinding in all eight simulations.
|
8 Å in length. The entrance to the tunnel was at the end of the C-loop, 20 Å above the membrane, and the trajectories deviated from each other beyond this point. Nevertheless, they were all broadly similar, indicating that there was no significant effect of either velocity rescaling or the tethering scheme used. To directly compare the different trajectories, amino acids within 5 Å of the ligand in the course of each trajectory were identified (Tables 1 and 2). These data showed that a total of 52 different amino acids lay on the pathways, and 44% of these were within 5 Å of the ligand in all eight unbinding pathways. These can be considered as consensus residues and their positions are shown in Fig. 3 B. The majority of consensus amino acids were located close to the binding pocket, indicating a high degree of conformity in the unbinding trajectories during the early stages of the unbinding process.
|
|
|
Unbinding of 5-HT
The amino acids we previously identified as being within 5 Å of 5-HT in its correct orientation in the binding site are marked in Table 1 (taken from model 4 of Reeves et al. (15)). As expected, these amino acids interacted with the ligand at the early stages of the simulation (white squares) but not at the latter stages (black squares). Three overlapping frames of the position of 5-HT at 0 ps, 48 ps, and 56 ps into the simulation (Fig. 4) show that the unbinding pathway follows the tunnel described above. Forty-three amino acids were identified as being on the 5-HT unbinding pathways, with 74% of these present in all four pathways.
|
). As the ligand left the binding site, it turned slightly so that these interactions were replaced by other hydrogen bonds. Of these, the most persistent were those with E236, which initially acted as a hydrogen-bond acceptor with the H
of 5-HT, then HN2, and finally HN3; Y153, which acts as a hydrogen-bond acceptor first to HN2 and then to HN1 of 5-HT; W183, which acts as an acceptor to HN1, HN2, and then HN3 of 5-HT; K238, which acts as a donor to the O of 5-HT; and E129, which hydrogen-bonds with HN1 and then with HN3. Thus, some of the amino acids contributing to the hydrogen bonds are in the enclosed part of the unbinding pathway, whereas others lie in the groove that extends from this enclosed region toward the membrane. There are no explicit waters in our simulations, but given the size of 5-HT and the volume of the binding pocket, it is unlikely that water would exist in this site.
|
) are indicated in Table 2. Similar to 5-HT unbinding, these residues are seen to interact with the ligand in the early, but not the late, stages of the simulation. The sole exception is residue S206. This is interesting as it is located on the F-loop whose position in the high-resolution structure of AChBP has not yet been finalized. However, it is clear from our previous studies that it plays a role in the binding and/or function of the receptor (19) and our current data support this. Three overlapping frames of the position of granisetron at 0 ps, 10 ps, and 55 ps into the simulation show that the unbinding pathway is similar to that for 5-HT (Fig. 5). In total, 44 residues were located on the granisetron unbinding pathways, with 59% found in all four pathways.
|
1 of N128, between the indazole nitrogen of granisetron and the hydroxyl group of Y153, and the ammonium hydrogen of the azabicyclic ring and the main-chain oxygen of I207.
|
120 Å2 at the location of Y153, constricts at the level of I228, and then widens toward the mouth of the pathway. As it proceeds closer to the membrane, the C-loop ends and the tunnel that was present in the initial stages of the unbinding process becomes a groove. The text and arrows in the center of Fig. 6 show the location of the views in Fig. 7.
|
|
). In these studies, 13 energetically favorable models were created and were found to fall into three main groups based upon the orientation of granisetron in the binding site. Fig. 8 shows the position and orientation of granisetron in the three model types. In model types A (Fig. 8 A) and B (Fig. 8 B), the location of granisetron is broadly similar, with granisetron docked high in the binding site, although the azabicyclic ring is located at different ends of the binding pocket. In model type C (Fig. 8 C), granisetron is located outside the binding pocket in a region significantly closer to the membrane. A comparison of residues in model types A and B (yellow) and model type C (red) is shown in Fig. 8 D (common residues are shown in orange). Amino acids within models A and B were almost all identified as residues that interacted in the "early" part of the granisetron unbinding simulation, whereas amino acid residues close to the ligand in model type C interact with the ligand later in the simulation. This indicates that the potential binding site of granisetron shown in model C is located on the unbinding pathway (Fig. 5).
|
,26–29). As the side chain on alanine cannot form hydrogen bonds, these data are consistent with our hypothesis that these residues may hydrogen-bond to 5-HT. Radioligand binding data are also available for E236A and K238A mutant receptors and show no change in binding affinity compared to wild-type receptors (19). However, our current data show large changes in EC50 in both mutant receptors: E236A mutant receptors did not function at concentrations up to 100 µM 5-HT, and K238A-containing receptors had a 10-fold shift in EC50 (1.36 µM, compared to 0.2 µM for wild-type receptors). As EC50 is a term that encompasses both binding and functional parameters, these data suggest that agonist binding or function is impaired. They therefore support our hypothesis that hydrogen bonding with the agonist is important. Some hydrogen bonding was also indicated between 5-HT and T181, S206, Y234, E129, Q211, S163, and Q83. Examination of receptor function when each of these residues was replaced by alanine suggests that they may also hydrogen-bond to 5-HT: E129A, Q83A, and S206A mutant receptors were nonfunctional up to 100 µM 5-HT, Y234A mutation caused a 70-fold increase in EC50 (23), and T181A resulted in a 10-fold increase (EC50 = 1.95 µM). Q211A and S163A mutant receptors had EC50 values similar to wild-type receptors (0.17 and 0.24 µM, respectively). These data are shown in Fig. 9.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
8 Å long and its entrance is 20 Å above the membrane. These data are supported by functional studies.
The model that we have used to determine the unbinding pathway is the homology model of the 5-HT3 receptor, which is based on the structure of AChBP. AChBP has considerable structural similarity with the extracellular domain of the nACh receptor, and is a good template for the closely related 5-HT3 receptor. However, as our data are based on a model and not on an atomic resolution structure, caution must be applied in interpreting molecular details. Indeed, until a high-resolution structure is available, such studies can only provide qualitative information, although some details of molecular interactions can be obtained from experimental studies. We have previously used such studies to show that the predicted structure of the binding site is reasonably accurate: Mutagenesis has demonstrated that many key residues identified in the binding pocket affect agonist and/or antagonist binding to the receptor. The roles of W183, Y143, Y153, and Y234 have been particularly well studied, and these data have shown that these residues form hydrogen bonds and/or cation-
interactions with ligands (15,19,26,28). Further studies in which all 26 residues proposed to form the binding site were mutated provide yet more support for the model (19).
The 5-HT3 receptor, like all Cys-loop receptors, exists in a number of conformations, including an open, a closed, and one or more desensitized states. It is not yet known how much the structure of the protein varies in these different states, although significant changes in the nACh receptor have been reported (30). AChBP has been proposed to be similar to the open or desensitized state of the nACh receptor, and the location of residues that interact with 5-HT when it is docked into the binding site of the AChBP-based homology model are well supported by experimental evidence, so our positioning of 5-HT in the binding site is probably fairly accurate (1,2,31). Therefore, taking into account the limitations mentioned above, the unbinding pathway that we have identified is probably as accurate as the structure and the molecular simulations allow, although, as the closed state of the receptor has a distinct structure, it is probably not a good representation of the binding pathway. A recent study by Unwin (4) shows that in the resting state of the nACh receptor, the C-loop points away from the center of the receptor, "opening" the binding-site pocket and making it more accessible. Consequently, it is possible that ligands have a number of potential routes into the binding site, and indeed, in the GABAA receptor it has been proposed that the ligand accesses the binding site from the side (32). Maksay et al. (25) alluded to the region we have identified as the entry point of the ligand into the binding site (see Fig. 4 in that article) but the new structural information better supports ligands having a fairly unrestricted access (6). This is entirely reasonable, as entry to the binding site must be rapid to allow opening of the channel within milliseconds of agonist application, whereas ligand unbinding is in the hundreds of milliseconds to seconds time range (33–35). Structural data suggest that in the ligand-bound conformation the C-loop has closed over the ligand, restricting access to the binding site and leaving only a narrow passage through which the ligand can exit (Fig. 3). This exit route would place the ligand some distance from the binding pocket and, we speculate, would prevent immediate reentry of the ligand to the binding site when the receptor reverts to the closed conformation.
There are as yet no high-resolution structural details of the ligand binding pocket of any Cys-loop receptor with antagonist bound. However, in the 5-HT3 receptor we anticipate that this structure would be more similar to the open than to the closed state, as our docking studies of granisetron in the AChBP-like (open state) homology model are well supported by experimental evidence (19). We therefore believe that the starting location for granisetron that we used for the simulations is broadly correct and that the pathway we have identified is qualitatively accurate, although some molecular details may vary. Support for this hypothesis comes from docking studies with granisetron, where a potential binding site for this molecule was located on the unbinding pathway (19). A further binding site for another 5-HT3 antagonist (tetrahydroacridine) has also been located on our proposed pathway (36).
Of the residues thought to participate in the latter part of the unbinding pathway, only S206 has previously been shown to affect antagonist binding affinity when mutated, suggesting a role in binding and/or function of the 5-HT3 receptor (19). It is known from structural studies that this residue is located within the F-loop, which is a flexible region in AChBP. F-loop residues have also been shown to play a role in GABAA and nACh receptor function (37–39) and consequently it is not surprising that mutations in this region also have an effect on 5-HT3 receptor function. Structural details of the F-loop region are poorly resolved in the crystal structures of the AChBP (1,2) and in cryoelectron microscopy images of the nACh receptor (6), suggesting that the homology models may be inaccurate in this region. However, there is increasing evidence from mutagenesis and functional studies that this region undergoes structural changes upon ligand binding.
The tunnel that forms the exit route is considerably shorter than the binding "gorge" described by Sussman et al. (5) in acetylcholinesterase. This might be expected, as in a ligand-gated receptor a neurotransmitter must be able to enter and exit the binding site more rapidly than a substrate or product of an enzymic reaction. Interestingly, both the active site gorge and Cys-loop receptor binding sites are dominated by aromatic amino acids, although there are lower proportions of this type of amino acid in the unbinding pathway, suggesting that hydrophobic interactions have less importance here. However, there are many potential hydrogen bonding partners for 5-HT on the neurotransmitter unbinding pathway. Mutation of those amino acids that dominate the hydrogen-bonding pattern show that they play an important role in receptor function. These data provide support for the pathway and also suggest that these bonds may be more important than aromatic interactions for the movement of the ligand in this protein.
In conclusion, we have defined a potential unbinding pathway for ligands from the 5-HT3 receptor binding pocket using molecular dynamics simulations. The routes of both an agonist and an antagonist were similar, extending from the ligand binding site down toward the membrane. From these and other functional and structural data, we propose the following model. The ligand enters the binding site from one of several possible directions owing to the open nature of the binding pocket in the extended C-loop conformation. Ligand binding pulls the C-loop across the pocket, closing access to the site, and, in the case of an agonist, initiating a series of events that lead to channel opening and subsequent receptor desensitization. The ligand then leaves the binding pocket through the base of the binding site, along the only pathway that remains unobstructed after the movements of the extracellular domain. This places the ligand some distance from the binding site and the protein undergoes a conformational change back into the closed, resting state.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Submitted on June 24, 2005; accepted for publication November 9, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
2. Brejc, K., W. J. van Dijk, R. V. Klaassen, M. Schuurmans, J. van Der Oost, A. B. Smit, and T. K. Sixma. 2001. Crystal structure of an ACh-binding protein reveals the ligand-binding domain of nicotinic receptors. Nature. 411:269–276.[CrossRef][Medline]
3. Bouzat, C., F. Gumilar, G. Spitzmaul, H. L. Wang, D. Rayes, S. B. Hansen, P. Taylor, and S. M. Sine. 2004. Coupling of agonist binding to channel gating in an ACh-binding protein linked to an ion channel. Nature. 430:896–900.[CrossRef][Medline]
4. Unwin, N. 2005. Refined structure of the nicotinic acetylcholine receptor at 4Å resolution. J. Mol. Biol. 346:967–989.[CrossRef][Medline]
5. Sussman, J. L., M. Harel, F. Frolow, C. Oefner, A. Goldman, L. Toker, and I. Silman. 1991. Atomic structure of acetylcholinesterase from Torpedo californica: a prototypic acetylcholine-binding protein. Science. 253:872–879.
6. Miyazawa, A., Y. Fujiyoshi, M. Stowell, and N. Unwin. 1999. Nicotinic acetylcholine receptor at 4.6 Å resolution: transverse tunnels in the channel wall. J. Mol. Biol. 288:765–786.[CrossRef][Medline]
7. Kern, P., R. M. Brunne, and G. Folkers. 1994. Nucleotide-binding properties of adenylate kinase from Escherichia coli—a molecular-dynamics study in aqueous and vacuum environments. J. Comput. Aided Mol. Des. 8:367–388.[CrossRef][Medline]
8. Rognan, D., L. Scapozza, G. Folkers, and A. Daser. 1994. Molecular dynamics simulation of MHC-peptide complexes as a tool for predicting potential T cell epitopes. Biochemistry. 33:11476–11485.[CrossRef][Medline]
9. Leech, J., J. Prins, and J. Herma. 1996. SMD: visual steering of molecular dynamics for protein design. IEEE Comput. Sci. Eng. 3:38–45.
10. Izrailev, S., S. Stepaniants, M. Balsera, Y. Oono, and K. Schulten. 1997. Molecular dynamics study of unbinding of the avidin-biotin complex. Biophys. J. 72:1568–1581.
11. Marrink, S. J., O. Berger, P. Tieleman, and F. Jähnig. 1998. Adhesion forces of lipids in a phospholipid membrane studied by molecular dynamics simulations. Biophys. J. 74:931–943.
12. Kosztin, D., S. Izrailev, and K. Schulten. 1999. Unbinding of retinoic acid from its receptor studied by steered molecular dynamics. Biophys. J. 76:188–197.
13. Grubmüller, H., B. Heymann, and P. Tavan. 1996. Ligand binding: molecular mechanics calculation of the streptavidin-biotin rupture force. Science. 271:997–999.[Abstract]
14. Chau, P.-L. 2001. Process and thermodynamics of ligand-receptor interaction studied using a novel simulation method. Chem. Phys. Lett. 334:343–351.[CrossRef]
15. Reeves, D. C., M. F. Sayed, P. L. Chau, K. L. Price, and S. C. Lummis. 2003. Prediction of 5-HT3 receptor agonist-binding residues using homology modeling. Biophys. J. 84:2338–2344.
16. Sali, A., and T. L. Blundell. 1993. Comparative protein modelling by satisfaction of spatial restraints. J. Mol. Biol. 234:779–815.[CrossRef][Medline]
17. Smith, W., and T. R. Forester. 1996. DL_POLY_2.0: a general-purpose parallel molecular dynamics simulation package. J. Mol. Graph. 14:136–141.[CrossRef][Medline]
18. MacKerell, A. D., D. Bashford, M. Bellott, R. L. Dunbrack, J. D. Evanseck, M. J. Field, S. Fischer, J. Gao, H. Guo, S. Ha, D. Joseph-McCarthy, L. Kuchnir, et al. 1998. All-atom empirical potential for molecular modeling and dynamics studies of proteins. J. Phys. Chem. B. 102:3586–3616.
19. Thompson, A. J., K. L. Price, D. C. Reeves, S. L. Chan, P. L. Chau, and S. C. Lummis. 2005. Locating an antagonist in the 5-HT3 receptor binding site using modeling and radioligand binding. J. Biol. Chem. 280:20476–20482.
20. Chau, P.-L., T. R. Forester, and W. Smith. 1996. Curvature effects in hydrophobic solvation. Mol. Phys. 89:1033–1055.[CrossRef]
21. Chau, P.-L., and P. M. Dean. 1987. Molecular recognition: 3D surface structure comparison by gnomonic projection. J. Mol. Graph. 5:97–100.[CrossRef]
22. Kunkel, T. A. 1985. Rapid and efficient site-specific mutagenesis without phenotypic selection. Proc. Natl. Acad. Sci. USA. 82:488–492.
23. Hargreaves, A. C., M. J. Gunthorpe, C. W. Taylor, and S. C. Lummis. 1996. Direct inhibition of 5-hydroxytryptamine3 receptors by antagonists of L- type Ca2+ channels. Mol. Pharmacol. 50:1284–1294.[Abstract]
24. Price, K. L., and S. C. Lummis. 2005. FlexStation examination of 5-HT3 receptor function using Ca2+ - and membrane potential-sensitive dyes: advantages and potential problems. J. Neurosci. Methods. 149:172–177.[CrossRef][Medline]
25. Maksay, G., Z. Bikadi, and M. Simonyi. 2003. Binding interactions of antagonists with 5-hydroxytryptamine3A receptor models. J. Recept. Signal Transduct. Res. 23:255–270.[CrossRef][Medline]
26. Spier, A. D., and S. C. Lummis. 2000. The role of tryptophan residues in the 5-Hydroxytryptamine(3) receptor ligand binding domain. J. Biol. Chem. 275:5620–5625.
27. Price, K. L., and S. C. Lummis. 2004. The role of tyrosine residues in the extracellular domain of the 5-hydroxytryptamine3 receptor. J. Biol. Chem. 279:23294–23301.
28. Beene, D. L., K. L. Price, H. A. Lester, D. A. Dougherty, and S. C. Lummis. 2004. Tyrosine residues that control binding and gating in the 5-hydroxytryptamine3 receptor revealed by unnatural amino acid mutagenesis. J. Neurosci. 24:9097–9104.
29. Beene, D. L., G. S. Brandt, W. Zhong, N. M. Zacharias, H. A. Lester, and D. A. Dougherty. 2002. Cation- interactions in ligand recognition by serotonergic (5-HT3A) and nicotinic acetylcholine receptors: the anomalous binding properties of nicotine. Biochemistry. 41:10262–10269.[CrossRef][Medline]
30. Unwin, N., A. Miyazawa, J. Li, and Y. Fujiyoshi. 2002. Activation of the nicotinic acetylcholine receptor involves a switch in conformation of the -subunits. J. Mol. Biol. 319:1165–1176.[CrossRef][Medline]
31. Celie, P. H., R. V. Klaassen, S. E. van Rossum-Fikkert, R. van Elk, P. van Nierop, A. B. Smit, and T. K. Sixma. 2005. Crystal structure of AChBP from Bulinus truncatus reveals the conserved structural scaffold and sites of variation in nicotinic acetylcholine receptors. J. Biol. Chem. 280:26457–26466.
32. Berezhnoy, D., Y. Nyfeler, A. Gonthier, H. Schwob, M. Goeldner, and E. Sigel. 2004. On the benzodiazepine binding pocket in GABAA receptors. J. Biol. Chem. 279:3160–3168.
33. Suprenant, A., and J. Crist. 1988. Electrophysiological characterization of functionally distinct 5-hydroxytryptamine receptors on guinea-pig submucous plexus. Neuroscience. 24:283–295.[CrossRef][Medline]
34. Yakel, J. L., and M. B. Jackson. 1988. 5-HT3 receptors mediate rapid responses in cultured hippocampus and a clonal cell line. Neuron. 1:615–621.[CrossRef][Medline]
35. Maricq, A. V., A. S. Peterson, A. J. Brake, R. M. Myers, and D. Julius. 1991. Primary structure and functional expression of the 5HT3 receptor, a serotonin-gated ion channel. Science. 254:432–437.
36. Camilleri, M. 2001. Management of the irritable bowel syndrome. Gastroenterology. 120:652–668.[CrossRef][Medline]
37. Newell, J. G., and C. Czajkowski. 2003. The GABAA receptor 1 subunit Pro174-Asp191 segment is involved in GABA binding and channel gating. J. Biol. Chem. 278:13166–13172.
38. Martin, M., C. Czajkowski, and A. Karlin. 1996. The contributions of aspartyl residues in the acetylcholine receptor and subunits to the binding of agonists and competitive antagonists. J. Biol. Chem. 271:13497–13503.
39. Sedelnikova, A., C. D. Smith, S. O. Zakharkin, D. Davis, D. S. Weiss, and Y. Chang. 2005. Mapping the rho1 GABAC receptor agonist binding pocket. Constructing a complete model. J. Biol. Chem. 280:1535–1542.
This article has been cited by other articles:
 |
|
 |
|
  A. J. Thompson, C. L. Padgett, and S. C. R. Lummis Mutagenesis and Molecular Modeling Reveal the Importance of the 5-HT3 Receptor F-loop J. Biol. Chem., June 16, 2006; 281(24): 16576 - 16582. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
