The Potassium Channel:
A Structural Analysis with Visual Molecular Dynamics
Bryan Q. Spring
Biophysics 490M Final Project
Professor Emad Tajkhorshid
April 22, 2003
Introduction
The functional role of the potassium channel, a phylogentically conserved homotetrameric transmembrane protein, has been elucidated nearly to completion. The focus of this report is to illustrate potassium channel structure using Visual Molecular Dynamics (VMD)1, a state-of-the-art molecular modeling software developed at the University of Illinois Beckman Institute, and to tie these architectural features to their physiological function. Of particular interest are the amino acid and a-helix participants in the selectivity filter, the mechanism of gating, and the coupling of stimulus to potassium cation efflux.
Here we analyze the structure of the voltage-gated potassium channel KcsA from Streptomyces lividans and provide a brief comparison to the structure of the calcium-gated potassium channel MthK from Methanobacterium thermoautotrophicum. Comparison between the KcsA and MthK channels highlights essential structural features common to both voltage-gated and ligand-gated channels. We discuss the pore lining inner helices which may contain a “gating hinge”. All figures in this report were prepared using VMD.
Open (PDB code 1JQ1)2 and closed (PDB code 1BL8)3 structures of the KcsA protein and an open structure of the MthK channel (PDB code 1NQ)4 are available from the Protein Data Bank (PDB). Note that the open structure for KcsA was determined by NMR and contains much less information than the closed structure determined by X-ray crystallography (see figure 8). Figure 1 contrasts the closed KcsA channel and the open MthK channel.
Figure 1 Side view and aligned overhead view of closed KcsA (blue) and open MthK (yellow) structures. Note that for clarity we have truncated the MthK structure, showing only the transmembrane region.
Channel Architecture
Four identical subunits of the KcsA potassium channel tetramerize to form a single ion conduction pore as shown in figure 2. Each subunit of the tetramer consists of six transmembrane a-helices (S1–S6). The structures available from the PDB are actually truncated and include only the pore helix (residues 53-84) and the inner helices (residues 1-52 and 85-119) which form the channel. Seven positively charged residues reside in the S4 domain and two negatively charged residues reside in the S2 domain. These two domains are believed to be the voltage sensors of the K+ channel. The amino and carboxy termini are free within the cytoplasm.
Figure 2 The homotetrameric KcsA channel viewed along an axis perpendicular to the membrane with distinctly colored subunits. The green sphere in the center is a K+ ion.
In figure 3, two subunits of the closed KcsA channel are shown. The upper narrow region formed by the pore loops is the selectivity filter. A wider region, a water-filled cavity, precedes the selectivity filter along the ion conduction pathway. An intracellular gate is located at the bottom of the figure, controlling ion flow into the cavity.
The overall length of the channel is shown in the figure to be roughly 45 angstroms and the pore loops (which compose the selectivity filter) are shown to be about 12 angstroms in length. There are two K+ binding sites within the selectivity filter separated by 7.5 angstroms. (Three K+ ions and one water molecule appear within the selectivity filter. The upper two are separated by about 7.5 angstroms and represent the binding sites within the selectivity filter. The third K+ ion is below the selectivity filter and is thought to be a hydrated cation cloud).3
Figure 3 Two subunits of the closed KcsA channel are shown (yellow and red colored tubes). The pore loops form the selectivity filter, which is preceded by a cavity and intracellular gate formed by the inner transmembrane helices S5 and S6. The green spheres are K+ ions and the red sphere is the oxygen atom of a water molecule.
The potassium channel is responsible for accomplishing both high selectively and rapid K+ ion efflux. These seemingly contrary functions are performed simultaneously by utilizing electrostatic repulsion between two closely positioned K+ ions. In other words, the 7.5 angstrom separation between the binding sites within the selectivity filter allows the K+ ions to push themselves through the filter.
Figure 4 The carbonyl groups lining the selectivity filter are highlighted for the closed KcsA channel. Separations between the oxygen atoms in the selectivity filter are roughly 4 angstroms. The water-filled cavity has a larger width (about 12 angstroms). Transparent objects are the surrounding helices.
Figures 4 and 5 outline the structural features of the selectivity filter and the water-filled cavity. The a-helix dipole moments of the pore helices point their negatively charged ends towards the entrance of the selectivity filter. The protein utilizes the dipole moments to select cations. Figure 4 (above) shows that closely spaced carbonyl groups line the selectivity filter. The oxygen atoms of the carbonyl groups are positioned to exactly reproduce the environment of solvated K+ ions. Thus, the selectivity filter perfectly compensates for the energy penalty of de-solvating K+ ions.3 Furthermore, the selectivity filter geometry allows specific chemical coordination of K+ (1.33 angstrom Pauling radius) in preference to other smaller cations such as Na+ (0.95 angstrom radius). Smaller cations cannot interact with all four subunits of the filter. Consequently, smaller cations are excluded because the do not bind to the selectivity filter strongly enough to overcome for the energetic penalty of desolvation. Larger cations, like Ca2+, are simply too wide to fit into the filter.
The functional purpose of the water-filled cavity is two-fold. Water in the channel is used to stabilize ions in the hydrophobic interior of the membrane before they reach the selectivity filter. Secondly, water molecules in the channel are employed to stabilize the charged groups of the S2 and S4 domains of the transmembrane a-helices, which may prevent efficient K+ entry into the selectivity filter. Figure 5 depicts the location of the K+ binding sites within the selectivity filter. Also, the remainder of the pore surface is shown to be largely hydrophobic. The pore forming amino acids accomplish this by pointing their side groups away from the channel.
An interesting consequence of outward facing side groups along the K+ channel is that these residues are involved in stabilizing the pore loops. In particular, the aromatic residues of the helices surrounding the pore hold the amino acid sequence of the selectivity filter in place. In figure 6, a few of the stabilizing hydrogen bonds within the KcsA potassium channel tetramer are depicted.
Figure 5 A volume slice of the solvent accessible surface of the closed KcsA channel. The blue spheres within the selectivity filter represent K+ ions. The electrostatic potential is colored red in negative areas and blue in positive areas. The green and white regions are indicative of polar and hydrophobic side chains respectively. The large cavity at the center of the channel is shown. / Figure 6 Three hydrogen bonds are shown between side groups surrounding the selectivity filter. Protein side chains are represented as sticks where red and blue represent acidic and basic atoms, respectively.Pore Gating Coupled to Conformational Movements
A complete mapping of the gating mechanism for the voltage-dependent KcsA channel has yet to be completed. The focus of the literature has been to identify possible mechanism for ion channel activation and inactivation. Inactivation is the phenomenon of channel closure – after activation – in spite of maintained stimulus.
Movements of subunits S2 and S4 are thought to result in activation of the voltage-dependent channels. Experimentalists have used elegant techniques in spectroscopy, such as fluorescence resonance energy transfer (FRET), to probe the movements of the positively charged S4 voltage sensor. The data of these experiments seem to suggest that the S4 a-helix undergoes a slight rotation and tilt during membrane depolarization.5,6 Others have also performed experiments to characterize movements of the negatively charged S2 domain.7
Mechanisms for inactivation of voltage-dependent potassium channels have been proposed. The so-called ball and chain model of ion flux inactivation states that an N –terminus of the KcsA channel binds to the pore. The S6 domain bundle crossing can block ion flow by trapping organic molecules that bind to sites within the channel.8
More experimental and computational work is needed to elucidate time-resolved movements of voltage-gated ion channels.
Figure 7 shows the open MthK structure. Upon Ca2+ binding the ligand sites within the protein undergo conformational changes which result in a lateral force applied to the intracellular gate. Note that the upper helices in the figure compose the transmembrane pore. The lower helices make up the ligand binding domain.
Based on structure alignments achieved using VMD we give our own analysis of the possible location of the “gating hinge” in the KcsA potassium channel. The results of structural alignments of the open and closed KcsA channels achieved in VMD are shown in figure 8. Table 1 lists root mean square displacement calculations using the RMSD Calculator in VMD. Residues 86 to 119 were used to compute the alignments. Computing the root mean square displacement between the aligned structures revealed that residue Thr 103, displayed in figure 8 as Van der Waal’s spheres, moved the least amount. This is in contrast to the conclusion of Jiang et al. that a glycine residue serves as the “hinge” between the open and closed states.9
Jiang et al. assigned the “hinge” location based on comparisons between the open MthK and closed KcsA channel structures. Sequence alignments revealed that a glycine residue located below the selectivity filter is conserved amongst MthK and KcsA channels.9 This supports the conclusion, of Jiang et al., that residues Gly 83 in MthK and Gly 99 in KcsA serve as “hinges” between the open and closed states.9
Residue number / Root Mean Square Displacement* (angstroms)86 / 1.81
87 / 2.47
88 / 2.53
90 / 1.76
95 / 1.87
99 / 1.21
100 / 1.23
103 / 0.87
105 / 0.99
110 / 1.02
111 / 2.31
115 / 3.10
119 / 3.97
Figure7 The MthK open structure with each subunit colored distinctly. Ca2+ ions are shown in yellow and are bound to allosteric sites within the protein.
Table 1 Root mean square displacement calculation results from VMD, from the open and closed KcsA structures. Residue Thr 103 has the least value, and therefore must be located near the “gating hinge” *The root mean square displacement between open
(the amino acid in each subunit of the pore and closed structures calculated for specific residues.
helices that remains “fixed” during opening of the
the intracellular gate).
Figure 8 Views of opened (white) and closed (blue) KcsA pore helices. Top: Aligned side view of residues 86-119. Residues Thr 103 have the least root mean square displacement and are represented as Van der Waal’s spheres. The open-pore helices are displaced by about 8 angstroms at the intracellular gate. Left: Comparison of residues 86-119. The distance between opposite helices increases from 13 to 21 angstroms at the intracellular gate in the open conformation. Bottom: Overhead view of aligned open and closed KcsA structures. Note that the open structure contains fewer atoms than the closed structure.
Summary
The potassium channel is constructed of inner helices with a selectivity filter forming a narrow pore region and a water-filled cavity constituting a wide region. The large water-filled cavity and pore helix dipoles accommodate cations that would otherwise face a large energy barrier, preventing their permeation of the low dielectric membrane center. The selectivity filter is lined by carbonyl oxygen atoms, which are constrained to an optimal geometry by the surrounding side groups, such that a dehydrated K+ ion fits with proper coordination but the Na+ is too small. Within the filter, two K+ ions are within close proximity and repel each other. The repulsion overcomes the strong attraction between the ion and the protein allowing rapid conduction of potassium.
Work to map voltage-dependent and ligand-dependent movements of the potassium channel during activation and inactivation has begun. Both species appear to convert protein conformational changes to lateral forces that open the intracellular gate of the channel. However, more data is needed to form a complete picture of gating mechanisms.
Based on available structural data we are able to postulate the location of the gating hinge in the KcsA potassium channel. Our results based solely on root mean square displacement calculations closely correspond to that of the literature. The discrepancy in our result from that of Jiang et al. is due to the fact that root mean square displacement calculations cannot be used as the sole basis for hinge identification. As noted by Jiang et al., the glycine residue below the selectivity filter is conserved among MthK and KcsA potassium channels, suggesting that it is key to the function of the channels.9 Furthermore, glycine is the residue most likely to serve as the hinge because of the conformational mobility conferred by its small side group.
Works Cited
1. Humphrey, W., Dalke, A. and Schulten, K. VMD - Visual Molecular Dynamics. J. Molec. Graphics 14, pp. 33-38 (1996)
2. Liu, Y.-S., Sompornpisut, P. & Perozo, E. Structure of the Kcsa Channel Intracellular Gate in the Open State. Nat.Struct.Biol. 8, 883 (2001)
3. Doyle, D.A., Morais, Cabral J., Pfuetzner, R.A., Kuo, A., Gulbis, J.M., Cohen, S.L., Chait, B.T. & MacKinnon, R. The structure of the potassium channel: molecular basis of K+ conduction and selectivity. Science 280, 69-77 (1998).
4. Jiang, Y., Lee, A., Chen, J., Cadene, M., Chait, B.T. & MacKinnon, R. Crystal structure and mechanism of a calcium-gated potassium channel. Nature 417, 515-22 (2002).
5. Cha, A., Snyder, G.E., Selvin, P.R. & Bezanilla, F. Atomic scale movement of the voltage-sensing region in a potassium channel measured via spectroscopy. Nature 402, 809-13 (1999).
6. Glauner, K.S., Mannuzzu, L.M., Gandhi, C.S. & Isacoff, E.Y. Spectroscopic mapping of voltage sensor movement in the Shaker potassium channel. Nature 402, 813-7 (1999).
7. Milligan, C.J. & Wray, D. Local movement in the S2 region of the voltage-gated potassium channel hKv2.1 studied using cysteine mutagenesis. Biophys J 4, 1852-61 (2000).
8. Yellen, G. The voltage-gated potassium channels and their relatives. Nature 419, 35-42 (2002).
9. Jiang, Y., Lee, A., Chen, J., Cadene, M., Chait, B.T. & MacKinnon, R. The open pore conformation of potassium channels. Nature 417, 523-6 (2002).
2