What can be deduced about the structure of Shaker from available data?

What can be deduced about the structure of Shaker from available data?
of 30
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  WhatcanbededucedaboutthestructureofShakerfromavailabledata? Benoit Roux Weill Medical College of Cornell University, 1300 York Ave, New York, NY 10021, USA  Abstract  . Voltage-gated K + channels are transmembrane proteins that control andregulate the £ow of K + ions across cell membranes in response to changes in membranepotential andareessential forthe propagationof actionpotentialsinthe nervoussystem.Oneofthemoststudiedvoltage-gatedchannelsisShaker.Availableexperimentalresultsclearlyprovidespeci¢cconstraintsonthestructureofthechannel,eventhoughthedirecttranslation of the available information into 3D structures is not trivial. The goal of thisworkistodevelopacomputationalapproachtoconstructandre¢ne3DmodelsofShakerby incorporating and integrating available experimental data. Our approach is based oncomparative modelization and global conformational optimization using energyrestraints extracted from experimental data. 2002 Ion channels     from atomic resolution physiology to functional genomics. Wiley, Chichester (NovartisFoundationSymposium245)p84^108  The activity of voltage-gated ion channels is the basic molecular mechanismunderlying the electrical excitability of nerves and muscles (Hodgkin & Huxley1952). These channels are specialized transmembrane proteins, which control andregulatethe£owofionsacrosscellmembranesbyopeningandclosing(‘gating’)inresponse to changes in membrane potential (Hille 1992). The ¢rst identi¢ed andbest-studied voltage-gated channel is the Shaker K + channel from the fruit£y Drosophilia melanogaster   (Tempel et al 1987); the corresponding voltage-gated K + (K v ) channels inmammals are K v 1.1^K v 1.7 (Jan& Jan 1997).Normallyclosed athyperpolarized resting potentials, Shaker K + channels undergo a conformationaltransitionfromaclosedtoanopenstateatdepolarizationpotentials(Chaetal1999,Glauner et al 1999).Studies have shown that Shaker and all the channels in the K v  family arestructurally and functionally similar. They are formed by four identical orhomologous domains, or subunits (MacKinnon 1991). Analysis of the aminoacid sequence suggests that each subunit contains six putative transmembrane(TM) segments, S1 to S6 (Jan & Jan 1997, Tempel et al 1987). The second (S2) 84  !"# $%&##'()* +,"- ./"-01 2')"(3/0"# 4%5)0"("65 /" +3#1/0"#&( 7'#"-01)* 8"9&,/0) +"3#:&/0"#;5-<")03- =>? ! "#$%&' ()*+,-.', /0 12'3#20 4#56 78, 97&-' :! 1##,';#<02-3=.  !  >#?72.-@ A#%8,7.-#8 (BB(! CD4>E BF)GBFH)IG*FJ  and fourth (S4) segments contain several charged residues, which are a¡ected bychanges in membrane potential and form part of the voltage sensor that controlsthe gating of the channel (Aggarwal & MacKinnon 1996, Bezanilla 2000, Limanetal1991,Logothetisetal1992,Papazianetal1991,Seohetal1996,Yellen1998).Thepartoftheproteinformingtheporeregionresponsible forthe selectivityandconduction of K + ions is located between segments S5 and S6 of a subunit in theregion containing the essential amino acid ‘signature sequence’ TTVGYGDcommon to all K + channels (Heginbotham et al 1992, 1994).Although a growing body of information is available for Shaker and othervoltage-activated channels in the K v  family, the only ion channel for which astructure at atomic resolution is currently available is the KcsA channel from Streptomyces lividans   (Doyle et al 1998). The main features of the crystallographicstructure are shown in Fig. 1. The channel is made of four identical subunitsdisposed symmetrically around a common axis corresponding to the pore (onlytwo are shown in Fig. 1). Although the monomer of KcsA is formed by only twotransmembrane helices, the amino acid sequence is, in fact, very similar to that of segment S5^S6, which is conserved in eukaryotic voltage-gated channels such asShaker (Cortes & Perozo 1997, Doyle et al 1998, Schrempf et al 1995).Furthermore, a combination of structural and functional data with neurotoxinfrom scorpion indicates the extracellular vestibule of KcsA is structurally verysimilar to Shaker (MacKinnon et al 1998).Intheabsenceofadetailedatomicstructure,extensivestudiesusingavarietyof experimental approaches including electrophysiology, site-directed mutagenesis,resonance energy transfer and electron microscopy, have been used to probe thestructure and function of the Shaker K + channel. Undoubtedly, many of theexperimental results obtained so far put very speci¢c constraints on the structureof Shaker, though often indirectly. Nonetheless, the direct translation of all theavailable information into a 3D structure is not straightforward. The purpose of the present work is to develop a computational approach to construct and re¢ne3D models of Shaker by incorporating and integrating all available experimentaldata. The approach that we use is based on comparative modelization and globalconformational optimization using energy restraints extracted from experimentaldata. Given the limited amount of information presently available, we do notexpect to converge towards a unique ‘best’ model of Shaker. Instead, we seek todelineate and clarify, as objectively as possible, the current state of the knowledgeabout Shaker by generating an ensemble of plausible models which are consistentwiththeavailabledata.Ourhopeisthatsuchanensembleof3Dmodelscanplayauseful role in the design of future experiments by indicating the areas of greatestuncertaintyinthestructure,byhelpingtoexaminethespatialrelationshipbetweenfunctionallyimportantresidues,andbyrevealinginconsistenciesbetweendi¡erentexperimental results. THE STRUCTURE OF SHAKER 85  Assumptions The general topology of a Shaker subunit is illustrated schematically in Fig. 2. Inparticular, it is assumed that the segments S1^S4 are in an  a -helical conformationand that the central pore formed by S5^P^S6 is structurally very similar to thecrystallographic structure of the KcsA K + channel. Although these assumptionsare reasonable and currently supported by experimental evidence, some of themmight turn out to be incorrect in the future. Nonetheless, such simpli¢cations arenecessary at this point. 86 ROUX FIG.1. SchematicviewoftheKcsAchannel(onlytwoofthefourmonomersareshown).Theextracellular side is at the top and the intracellular side is at the bottom. The main structuralelements are: the outer helix corresponding to S5 in voltage-gated K + channels (residues byA391 to E418); the P loop formed by the pore helix and the selectivity ¢lter which contains thesignaturesequenceTTVGYGD(residuesI429toT441);andtheinnerhelixcorrespondingtoS6(residues G452 to N482).  HelicalconformationofthetransmembranesegmentsS1^S4 It is assumed that the TM segments S1^S4,S1 (226^247): ARVVAIISVFVILLSIVIFCLES2 (279^300): FFLIETLCIIWFTFELTVRFLAS3 (311^332): VMNVIDIIAIIPYFITLATVVAS4 (358^380): LAILRVIRLVRVFRIFKLSRHSKare in an  a -helical conformation and are roughly perpendicular to the membraneplane, though some tilting of the helix axis is possible.Thishypothesisissupportedbyanumberofobservations.Experimentalstudieshave established that an isolated fragment corresponding to S4 adopts apredominantly  a -helical conformation in methanol and in lipid membranes(Halsall & Dempsey 1999, Haris et al 1994, Mulvey et al 1989). The structure is arandom coil in aqueous solution (Haris et al 1994). Similar studies with TMsegments of the Na + channel have shown that they adopt  a -helical structures indetergent micelles (Doak et al 1996). There are also strong indications that the S1to S4 segments are  a  helical in the channel structure. Ala- (Li-Smerin et al 2000a)and Trp-scanning (Hong & Miller 2000, Monks et al 1999) mutagenesis studiessuggest that S1 and S2 are amphipathic membrane spanning  a  helices that THE STRUCTURE OF SHAKER 87 FIG. 2. Schematic view of one subunit of Shaker with its six transmembrane segments. It isassumed that S1 to S4 are in an  a -helical conformation and that the central pore formedby S5^P^S6 is structurally similar to the crystallographic structure of the KcsA K + channelshown in Fig. 1.  interface directly with the lipid membrane. Helical periodicity of functionalalteration in the voltage-activation curves and gating kinetics were observedthroughout S1 and S2. Trp-tolerant positions in the Shaker K + channel areclustered on approximately half the  a  helix surface, as if the side chains areexposed to the hydrocarbon region of the lipid bilayer (Hong & Miller 2000,Monks et al 1999).Similarly, Ala-scanning mutagenesis in the drk1 K + channel of S1 and S2suggest that these segments are relatively simple amphipathic helices that spanthe full width of the membrane and make extensive contacts with the lipidmembrane (Li-Smerin et al 2000a). The observations for the S3 and S4 segmentsare more complex. In the case of S3, the distribution of Trp-tolerant position isroughly consistent with a helical secondary structure, although the results are notas clear towards the extracellular side (Hong & Miller 2000). Results from Ala-(Li-Smerin et al 2000a) and Lys-scanning (Li-Smerin & Swartz 2001) with thedrk1 K + channel suggest that the S3 segment is entirely helical, but that theN-terminal region interfaces with both lipid and protein, whereas the C-terminalregion interfaces with water. It has been speculated that a conserved proline atposition 322 might induce a kink in the helical segment. Ala scanning of S4reveals helical periodicity in only the C-terminal region (Li-Smerin et al 2000a).However, it seems likely that the absence of helical character in the N-terminalportion, which is exposed to the intracellular side, results from complexities inthe aqueous and protein environment surrounding the segment. StructureofthecentralporeS5^P^S6  It is assumed that the conformation of the central pore of Shaker, formed bythe S5^P^S6 segment, is very similar to the crystallographic structure of KcsA(Doyle et al 1998). This is a very reasonable assumption given the highsequence similarity of the core of Shaker and the KcsA bacterial channel(Fig. 3). For 93 residues, the sequence identity is 31 for a global sequencesimilarity of 49%. This high similarity makes Shaker an excellent candidatefor successful comparative modelling using the KcsA structure as a template(Fiser et al 2000). The structural similarity of Shaker relative to KcsA is alsosupported by experiments: AgTx2 binds to the KcsA channel, demonstratingthat this prokaryotic K + channel has the same pore structure as that of Shaker(MacKinnon et al 1998), and the chemical modi¢cations of cysteines locatedalong S6 by soluble thiol agents are generally consistent with the accessibilityof the corresponding residues in KcsA (Liu et al 1997). Because the KcsA K + channel is in a closed conformation (Roux et al 2000), the crystallographicstructure is probably a better model for the closed state of Shaker.Nonetheless, although it appears to be very reasonable, the assumption that 88 ROUX
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks