点击显示 收起
Department of Cell Biology and Physiology, Washington University School of Medicine, St. Louis, Missouri (G.L., C.G.N.)
CellGate Inc., Redwood City, California (L.J.M.)
Abstract
Polyamines cause inward rectification of Kir K+ channels by blocking deep within the channel pore. We investigated structural constraints of polyamine block of strongly rectifying mutant KATP channels (Kir6.2[L164C,N160D,C166S] + SUR1). We studied three groups of polyamine analogs: 1) conformationally restricted linear tetra-amines with a cycloalkyl or alkene group between the second and third amines (CGC-11047, CGC-11093, CGC-11099, and CGC-11098), 2) conformationally restricted linear deca-amines with a cycloalkyl or alkene group between the fifth and sixth amines (CGC-11150, CGC-11179, and CGC-11241), and 3) cyclic tetra-amines (CGC-11174, CGC-11197, CGC-11199, and CGC-11254). All linear analogs cause a voltage-dependent block similar to that of spermine, but slightly weaker (at 1 e, V1/2 for spermine block = -10 ± 1 mV, Z = 2.9 ± 0.1, n = 19; V1/2 for analogs varies from polyamine -7 to +10 mV, Z = 2.6-3.9). These data indicate tolerance for conformational restriction and an upper limit to the voltage dependence of the blocking process. There was no voltage-dependent block by the cyclic compounds; instead, they induce irreversible rundown of the current. Structural models of Kir channels suggest that a narrow entry at the top of the cytoplasmic pore may exclude cyclic analogs from the inner cavity, thereby explaining the structure-activity relationship that we observe.
Inward rectification of Kir channels is distinct from the voltage dependence of Kv channels in that conductance is not a function of the absolute membrane potential. Instead, conductance is related to the K+ reversal potential and is conferred by diffusible intracellular compounds, in particular the polyamines spermine and spermidine (and to a lesser extent putrescine and Mg2+ (Nichols and Lopatin, 1997).
There is much evidence consistent with the concept that rectification is conferred by positively charged ions binding to or near negatively charged residues in the channel, but details are lacking and there is no consensus as to where exactly these ions bind (Kubo and Murata, 2001; Xie et al., 2002, 2003; Guo and Lu, 2003; Guo et al., 2003; Kurata et al., 2004). Mutagenesis has indicated residues involved in the blocking process, and recent structural studies demonstrate the nature of the permeation path of Kir channels, within which the polyamines must bind (Nishida and MacKinnon, 2002; Kuo et al., 2003; Pegan et al., 2005). Further information may be obtained by examining the blocking abilities of a large collection of polyamine analogs that have been developed primarily in efforts to find anticancer agents (Seiler et al., 1998). We demonstrated previously that only the polyamines conferred strong rectification, whereas other bulkier, dipolar or nonlinear molecules (e.g., GABA, creatinine, lysine) fail to block these channels (Lopatin et al., 1994). These results indicate that a molecule must possess both the correct structure and charge density or distribution to confer strong rectification. Systematic examination of monoaminoalkanes and diaminoalkanes with alkyl chains composed of 2 to 12 methylene groups further indicates that changes in molecular length and charge have profound effects on channel block, with longer chain lengths increasing blocking affinity (Pearson and Nichols, 1998; Guo and Lu, 2000). Thus, the increased blocking affinity of spermine and spermidine, compared with diamines, is attributable not solely to increased blocker charge but also to increasing size or hydrophobic interactions.
The determination of high resolution structures of bacterial K channels (Doyle et al., 1998; Zhou et al., 2001a,b; Jiang et al., 2002a,b) provides a template for comparative modeling of Kir channel pores. Such models predict that polyamines may experience no major energetic barriers in the pore and may actually enter the selectivity filter in causing strong inward rectification, with longer polyamines binding progressively deeper (Dibb et al., 2003; Rose and Nichols, 2003). Experimental evidence for the ability of negative charges throughout the inner cavity, including the entrance to the selectivity filter (Kurata et al., 2004), to confer voltage-dependent block provides further consistency with this picture of inward rectification.
High-resolution structural determination also shows that the Kir cytoplasmic domain includes a 30- extension of the channel pore into the cytoplasm (Nishida and MacKinnon, 2002; Kuo et al., 2003; Pegan et al., 2005), but the consequences of this extension for the mechanism of rectification are unclear. In the present study, we have examined the blocking ability of further series of more complex polyamine analogs in three structural groups. The data indicate additional important structural features of the blocking moiety and argue for a role of the cytoplasmic pore in controlling polyamine-induced rectification.
Materials and Methods
Expression of KATP Channels in COSm6 Cells. COSm6 cells were transfected with pCMV6b-Kir6.2 (with mutations as described), pECE-SUR1, and pGreenLantern (Invitrogen, Carlsbad, CA), as described previously (Loussouarn et al., 2000). Point mutations were prepared by overlap extension at the junctions of relevant residues by sequential polymerase chain reaction as described previously (Loussouarn et al., 2000). The `control' Kir6.2 construct had a deletion of 36 amino acids from the C-terminal end, as well as N160D and C166S mutations (Loussouarn et al., 2000). An additional mutation (L164C) in M2 renders the channels very insensitive to ATP and excised patch currents extremely stable (Loussouarn et al., 2001). Patch-clamp experiments were made at room temperature, in a chamber that allowed the solution bathing the exposed surface of the isolated patch to be changed rapidly. Data were normally filtered at 0.5 to 2 Hz; signals were digitized at 3 to 5 kHz and stored directly on computer hard drive using Clampex software (Axon Instruments, Union City, CA). The standard bath (intracellular) and pipette (extracellular) solution used in these experiments had the following composition: 140 mM KCl, 10 mM K-HEPES, and 1 mM K-EGTA, pH 7.3.
Data Analysis. Off-line analysis was performed using Fetchan, pSTAT (Axon Instruments), and Microsoft Excel (Microsoft, Redmond, WA). Wherever possible, data are presented as mean ± S.E.M. Microsoft Solver was used to fit data by least-square algorithm.
Chemicals. CGC- (previously SL) compounds were synthesized as described previously (Valasinas et al., 2001, 2003; Frydman et al., 2004). TB34, which was synthesized by the method of Igarashi et al. (1997), was kindly provided by Kazuei Igarashi (Chiba University, Chiba, Japan). Spermine was purchased from Fluka AG (Buchs, Switzerland).
Results
We examined the block of a strong inward rectifier channel generated by coexpression of Kir6.2[N160D, L164C, C166S, 36] (Kir6.2-rec) with SUR1 (Loussouarn et al., 2001) by internally applied CGC- compounds (Fig. 1), with the objective of comparing their blocking potency and voltage dependence with the prototypical spermine.
Alkyl Extended Linear Analogs Produce Spermine-Like Block. We first examined the blocking abilities of a series of conformationally restricted tetra-amine analogs with bis-ethyl extensions of the primary amines (Fig. 1, group 1). These compounds are flexible linear analogs (22- extended length), but minimum width of the molecule is >2.2 , even at the central cyloalkyl groups. Two voltage protocols (Fig. 2, A and B) were used to determine both the steady-block and the blocking and unblocking kinetics. As is apparent from the recordings in Fig. 2, the tetra-amine class 1 analogs, exemplified by CGC-11047 (Fig. 2, A and B) all caused voltage-dependent rectification of Kir6.2-rec with a voltage-dependence and potency that was similar to spermine (Fig. 3A). Using a single Boltzmann function to fit Grel-V curves, the estimated voltage-dependence of spermine block (z 2.9) is slightly lower than has previously been reported (Phillips and Nichols, 2003), which is probably the result of the use of a lower concentration of spermine (1 e) than may be typically examined. Except for CGC-11047, all showed a slightly reduced potency compared with spermine that was accounted for by 2-fold slowing of the on rate and increase of the off rate. An immediate conclusion from these experiments is that a free primary amine at the end of the blocking molecule is not a requirement for steep inward rectification (see Discussion). Likewise, a steric restriction in the center of the molecule does not significantly impede block. The relatively enhanced potency of CGC-11047 compared with the other group 1 compounds may be related to the spermine-like propyl groups between the primary and secondary amines.
Very Long Deca-Amines Indicate a Limiting Steepness to Rectification. Previous studies of rectification induced by diamine analogs indicate increasing voltage-dependence of block, as alkyl chain length is increased from 2 to 10 (i.e., up to the length of spermine) (Pearson and Nichols, 1998; Guo and Lu, 2003; Guo et al., 2003). The availability of a series of extended deca-amines (Fig. 1, class 2) affords the opportunity to test whether much longer, or more charged, polyamines can cause much steeper rectification. The extended length of these molecules is 56 , but their minimal width is >2.2 . Figure 2, A and B, shows recordings from one example (CGC-11150) of these conformationally restricted deca-amines. Again, the blocking characteristics are similar to those of spermine, with a similar apparent gating charge but with faster on and off rates (see Discussion) underlying a slightly less potent block (Fig. 3). The similarity of apparent gating charge for CGC-11150, CGC-11179, and CGC-11241, to spermine, even though these compounds have 10 positive charges instead of 4, is striking and implies an upper limit to the charge movement associated with block is reached in these compounds (see Discussion). Equally striking is the considerable difference in the off rates of two compounds (>10-fold; Fig. 3), which differ only in the number of carbon atoms spacing the amines (three versus four) and in the cis (CGC-11179) versus trans (CGC-11150) arrangement of the penta-amine extensions. It is interesting that CGC-11241, which has the same structure as CGC-11150 but contains a cyclopropyl group in place of the cis-ethenyl group, has almost identical kinetics and potency to CGC-11150 (Fig. 3).
Cyclic Analogs Do Not Block the Channel. In sharp contrast to the group 1 and 2 analogs, CGC-11174, CGC-11197, and CGC-11199 induce no voltage-dependent block of the channel at a concentration of 1 e (Fig. 4), suggesting that they do not enter the channel pore. At this concentration, however, they do provoke a rundown of the channel [such as polylysine (Shyng and Nichols, 1998)], possibly by screening the phospholipid charges that keep the channel open. An action of these compounds on the membrane lipids is suggested by a separate series of experiments, in which we applied each of the CGC- compounds to Xenopus laevis oocytes. At a concentration of 100 e, none of the group 1 or 2 molecules had any noticeable effect, but each of the group 3 molecules caused a breakdown of the membrane and death of the oocytes within 2 h (data not shown).
Each of the above group 3 compounds contains a carbonyl group, which could be the cause of the lack of rectification induction. However, another related compound CGC-11254 (Fig. 1), lacking the carbonyl group, also failed to induce rectification at 1 e (Fig. 5A) and caused rapid breakdown of the membrane patch at higher concentrations.
These results suggest that the cyclized polyamine analogs, with a minimal diameter across the cyclic group of 5.5 , may not access the necessary site for induction of rectification, perhaps because they do not enter the inner cavity. Another interesting molecule in this regard is TB34, a derivative of spermidine, with a benzyl group attached to each nitrogen. This molecule causes potent and steeply voltage-dependent block of NMDA receptor channels (Igarashi et al., 1997). However, with a minimal diameter of 2.8 across the benzyl ring, it also failed to induce any rectification of Kir6.2-rec channels (Fig. 5B).
Discussion
Similarity of Rectification Induced by Extended Polyamines. Strong inward rectification of Kir channels is conferred by the naturally occurring polyamines (spermine, spermidine, and putrescine), which block the channel from the intracellular side of the membrane (Nichols and Lopatin, 1997). Molecular modeling, based on the crystal structures of KcsA and MthK (Doyle et al., 1998; Zhou et al., 2001; Jiang et al., 2002), has led us to propose that the voltage dependence of polyamine block arises primarily from entry of the compounds into the ion selectivity filter, thereby displacing charge-carrying K+ ions from the filter, to the outside of the membrane (Rose and Nichols, 2003). This view has been challenged, however. At first, Nishida and MacKinnon (2002) suggested that polyamines should bind in the extended cytoplasmic pore of the channel and that the voltage dependence of block results from the obligate movement of a column of K+ ions that extends from this region, through the inner cavity and selectivity filter. This view has been modified and further argued, in light of energetic coupling experiments, by Lu and colleagues (Guo and Lu, 2003; Guo et al., 2003).
The present experiments add more insight to the potential location and structural basis of polyamine block. Previous studies have indicated a requirement for a linear polyamine structure for steep voltage dependence of block (Lopatin et al., 1995; Pearson and Nichols, 1998). An untested possibility is that a primary `head' amine must enter the blocking site first. This is clearly not required, because tetra-alkylamines with bis-ethyl extensions (group 1) can cause rectification that is as steep, and almost as potent, as spermine (Fig. 3). The voltage dependence of on and off rates for each of these compounds is remarkably consistent and similar to that of spermine (Fig. 3B). Further patterns emerge in the details: 1) the central restriction in group 1 molecules, whether cycloalkyl or an unsaturated double bond, seems not to limit the blocking ability. This may be expected if only the first alkyl segment (i.e., between N1 and N2) is required to enter the very restricted space of the selectivity filter, with the remainder of the molecule accommodated in the much wider inner vestibule. 2) There is a systematic consequence of the first alkyl chain being a propyl (C3) or butyl (C4) group; the C3 compounds (CGC-11047, spermine) are systematically more potent (approximately 5- to 10-fold slower off rate) than C4 compounds (CGC-11093, CGC-11098, and CGC-11099). The same difference in potency is observed for C3 (CGC-11179) versus C4 (CGC-11150 and CGC-11241) deca-amines. It is tempting to speculate that this difference results from the more optimal N-N spacing with C3 linkers for accommodation in the K+ binding sites in the filter (Fig. 6).
A Limiting Voltage Dependence of Rectification. Previous studies with diaminoalkanes indicate that the charge movement associated with block increases essentially monotonically as chain length increases, up to at least Z = 4 or 5 for C12 or spermine (Pearson and Nichols, 1998; Guo and Lu, 2003; Guo et al., 2003; Rose and Nichols, 2003; Kurata et al., 2004). The possibility that longer diamines might move yet more charge is not readily testable, because solubility becomes a problem, and there may be significant partition into the membrane, making experiment and interpretation difficult. However, the deca-amines studied here (group 2), with C4 or C3 alkyl spacers are readily soluble, and allow some examination of these questions. The blocking potency and voltage-dependence are strikingly similar to those of the tetra-alkylamines.
However, both on and off rates are speeded up 4- to 10-fold compared with the tetra-amines. The extreme length of deca-amines makes it necessary that even if the head group can reach the selectivity filter, the tail will still be in the cytoplasmic pore (Nishida and MacKinnon, 2002). There is clearly a role for this region in controlling on and off rates of polyamines from the deep binding sites (Yang et al., 1995; Kubo and Murata, 2001; Xie et al., 2002; Kurata et al., 2004), and the possibility exists that prepositioning of the long deca-amine in the cytoplasmic pore is responsible for speeding both on and off rates. In this regard, it is important to note that in Kir2 family members, residues within the cytoplasmic pore [Glu224 and Glu299 (Yang et al., 1995), and the recently identified Asp255 and Asp259 (Pegan et al., 2005)] seem to control the potency of polyamine block, a result that may be interpreted to imply a role of these residues in forming the PA binding site (Guo et al., 2003). However, it is also conceivable that these residues instead form a `predocking' site for polyamines (Xie et al., 2003; Kurata et al., 2004), thereby controlling the apparent affinity of a deeper site. Differential interactions of polyamines and CGC- analogs with cytoplasmic cavity residues could affect the estimated affinities of the deeper blocking site.
Sizing the Channel Entry. Figure 6 shows the structure of KirBac1.1 (Kuo et al., 2003), a model for eukaryotic Kir channels, together with space-filling models of representative polyamine analogs. All of the structures are quite flexible, with the exception of the cycolalkyl ring segments (in group 1 and 2), which are rigid planes, and the cyclic (group 3) compounds. The cycloalkyl rings cause a `kink' of 10 to 20° in what is otherwise the straightest conformations (as can be seen for CGC-11098 versus spermine in Fig. 6), but they do not limit the overall flexibility of the molecule; all of the group 1 and 2 compounds are very flexible and can adopt very twisted as well as relatively straight conformations. The cyclized group 3 compounds maintain a planar cyclized segment, when analyzed dynamically, that is 9.5 long by 5.5 at the widest. All group 1 and 2 compounds have a maximum diameter of less than 2.2 , even at the cyclobutyl group of CGC-11098. TB34 is also a flexible structure, except for the benzyl rings which have a minimum diameter of 2.8 (C-C) across the ring.
The cytoplasmic extension of the Kir pore that is generated by the cytoplasmic domain (Nishida and MacKinnon, 2002; Kuo et al., 2003; Pegan et al., 2005) meets the inner cavity at what may be the region of the channel gate (Phillips and Nichols, 2003; Phillips et al., 2003). Relatively large, symmetric quaternary ammonium ions, at least up to tetrapentylammonium, can block Kir channels (Oliver et al., 1998; Guo and Lu, 2001) and in doing so must at least enter the cytoplasmic pore and block either at the top of the cytoplasmic pore (Guo and Lu, 2001) or in the inner cavity (Zhou et al., 2001b). Because 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide-propyl triethylammonium can modify cysteine residues inside the inner cavity, it seems most likely to us that these molecules can cross into the inner cavity; presumably, however, there is a limit to the size of molecules that can gain entry. This seems to be exceeded by the cyclized group 3 compounds. None of these cause any voltage-dependent block, despite their similar chemical composition compared with the linear tetra-alkylamines (group 1) and considerably smaller size compared with the group 2 deca-amines Figs. (4 and 5). The inability of TB34 to induce block, although it is a potent and steeply voltage-dependent blocker of NMDA receptor channels (Igarashi et al., 1997), suggests that the lower limit may be reached at the benzyl group, with a minimum diameter of 2.8 .
Acknowledgements
We are grateful to Decha Enkvetchakul for performing oocyte experiments, to Thierry Rose for insights regarding computer modeling of Kir channel structures and to Kazuei Igarashi for the gift of TB34.
1 Current address: INSERM U-533, l'Institut du Thorax, Nantes, France.
References
Dibb KM, Rose T, Makary SY, Claydon TW, Enkvetchakul D, Leach R, Nichols CG, and Boyett MR (2003) Molecular basis of ion selectivity, block and rectification of the inward rectifier Kir3.1/Kir3.4 K+ channel. J Biol Chem 278: 49537-49548.
Doyle DA, Morais Cabral J, Pfuetzner RA, Kuo A, Gulbis JM, Cohen SL, Chait BT, and MacKinnon R (1998) The structure of the potassium channel: molecular basis of K+ conduction and selectivity. Science (Wash DC) 280: 69-77.
Frydman B, Bhattacharya S, Sarkar A, Drandarov K, Chesnov S, Guggisberg A, Popaj K, Sergeyev S, Yurdakul A, Hesse M, et al. (2004) Macrocyclic polyamines deplete cellular ATP levels and inhibit cell growth in human prostate cancer cells. J Med Chem 47: 1051-1059.
Guo D and Lu Z (2000) Mechanism of IRK1 channel block by intracellular polyamines. J Gen Physiol 115: 799-814.
Guo D and Lu Z (2001) Kinetics of inward-rectifier K+ channel block by quaternary alkylammonium ions. Dimension and properties of the inner pore. J Gen Physiol 117: 395-406.
Guo D and Lu Z (2003) Interaction mechanisms between polyamines and IRK1 inward rectifier K+ channels. J Gen Physiol 122: 485-500.
Guo D, Ramu Y, Klem AM, and Lu Z (2003) Mechanism of rectification in inward-rectifier K+ channels. J Gen Physiol 121: 261-276.
Igarashi K, Shirahata A, Pahk AJ, Kashiwagi K, and Williams K (1997) Benzyl-polyamines: novel, potent N-methyl-D-aspartate receptor antagonists. J Pharmacol Exp Ther 283: 533-540.
Jiang Y, Lee A, Chen J, Cadene M, Chait BT, and MacKinnon R (2002a) Crystal structure and mechanism of a calcium-gated potassium channel. Nature (Lond) 417: 515-522.
Jiang Y, Lee A, Chen J, Cadene M, Chait BT, and MacKinnon R (2002b) The open pore conformation of potassium channels. Nature (Lond) 417: 523-526.
Kubo Y and Murata Y (2001) Control of rectification and permeation by two distinct sites after the second transmembrane region in Kir2.1 K+ channel. J Physiol 531: 645-660.
Kuo A, Gulbis JM, Antcliff JF, Rahman T, Lowe ED, Zimmer J, Cuthbertson J, Ashcroft FM, Ezaki T, and Doyle DA (2003) Crystal structure of the potassium channel KirBac1.1 in the closed state. Science (Wash DC) 300: 1922-1926.
Kurata HT, Phillips LR, Rose T, Loussouarn G, Herlitze S, Fritzenschaft H, Enkvetchakul D, Nichols CG, and Baukrowitz T (2004) Molecular basis of inward rectification: polyamine interaction sites located by combined channel and ligand mutagenesis. J Gen Physiol 124: 541-554.
Lopatin AN, Makhina EN, and Nichols CG (1994) Potassium channel block by cytoplasmic polyamines as the mechanism of intrinsic rectification. Nature (Lond) 372: 366-369.
Lopatin AN, Makhina EN, and Nichols CG (1995) The mechanism of inward rectification of potassium channels—long-pore plugging by cytoplasmic polyamines. J Gen Physiol 106: 923-955.
Loussouarn G, Makhina EN, Rose T, and Nichols CG (2000) Structure and dynamics of the pore of inwardly rectifying KATP channels. J Biol Chem 275: 1137-1144.
Loussouarn G, Phillips LR, Masia R, Rose T, and Nichols CG (2001) Flexibility of the Kir6.2 inward rectifier K+ channel pore. Proc Natl Acad Sci USA 98: 4227-4232.
Nichols CG and Lopatin AN (1997) Inward rectifier potassium channels. Annu Rev Physiology 59: 171-191.
Nishida M and MacKinnon R (2002) Structural basis of inward rectification. Cytoplasmic pore of the G protein-gated inward rectifier GIRK1 at 1.8 resolution. Cell 111: 957-965.
Oliver D, Hahn H, Antz C, Ruppersberg JP, and Fakler B (1998) Interaction of permeant and blocking ions in cloned inward-rectifier K+ channels. Biophys J 74: 2318-2326.
Pearson WL and Nichols CG (1998) Block of the Kir2.1 channel pore by alkylamine analogues of endogenous polyamines. J Gen Physiol 112: 351-363.
Pegan S, Arrabit C, Zhou W, Kwiatkowski W, Collins A, Slesinger PA, and Choe S (2005) Cytoplasmic domain structures of Kir2.1 and Kir3.1 show sites for modulating gating and rectification. Nat Neurosci 8: 279-287.
Phillips LR, Enkvetchakul D, and Nichols CG (2003) Gating dependence of inner pore access in inward rectifier K+ channels. Neuron 37: 953-962.
Phillips LR and Nichols CG (2003) Ligand-induced closure of inward rectifier Kir6.2 channels traps spermine in the pore. J Gen Physiol 122: 795-804.
Rose T and Nichols CG (2003) Molecular modeling of Kir channel-polyamine interactions. Biophys J 84: 80a.
Seiler N, Atanassov CL, and Raul F (1998) Polyamine metabolism as target for cancer chemoprevention (review). Int J Oncol 13: 993-1006.
Shyng SL and Nichols CG (1998) Membrane phospholipid control of nucleotide sensitivity of KATP channels. Science (Wash DC) 282: 1138-1141.
Valasinas A, Reddy VK, Blokhin AV, Basu HS, Bhattacharya S, Sarkar A, Marton LJ, and Frydman B (2003) Long-chain polyamines (oligoamines) exhibit strong cytotoxicities against human prostate cancer cells. Bioorg Med Chem 11: 4121-4131.
Valasinas A, Sarkar A, Reddy VK, Marton LJ, Basu HS, and Frydman B (2001) Conformationally restricted analogues of 1N,14N-bisethylhomospermine (BE-4-4-4): synthesis and growth inhibitory effects on human prostate cancer cells. J Med Chem 44: 390-403.
Xie LH, John SA, and Weiss JN (2002) Spermine block of the strong inward rectifier potassium channel Kir2.1: dual roles of surface charge screening and pore block. J Gen Physiol 120: 53-66.
Xie LH, John SA, and Weiss JN (2003) Inward rectification by polyamines in mouse Kir2.1 channels: synergy between blocking components. J Physiol 550: 67-82.
Yang J, Jan YN, and Jan LY (1995) Control of rectification and permeation by residues in two distinct domains in an inward rectifier K+ channel. Neuron 14: 1047-1054.
Zhou M, Morais-Cabral JH, Mann S, and MacKinnon R (2001a) Potassium channel receptor site for the inactivation gate and quaternary amine inhibitors. Nature (Lond) 411: 657-661.
Zhou Y, Morais-Cabral JH, Kaufman A, and MacKinnon R (2001b) Chemistry of ion coordination and hydration revealed by a K+ channel-Fab complex at 2.0 resolution. Nature (Lond) 414: 43-48.