KcsA-Fab
Created by DeSean Thom
The KcsA-Fab complex (1K4D) is a Streptomyces lividans derived potassium channel complexed to a Fab antibody derived from Mus musculus. The complex has a molecular weight of 59964.71 Da making it a moderately sized protein monomer and has an isoelectric point (pI) of 8.79, which is the pH where it carries no net electric charge. KcsA is a proton-activated, voltage-modulated K+ channel and a key member of the Kv channel superfamily that assists cells in the transport of potassium ions through the plasma membrane. It was complexed to the Fab antibody in order to crystallize it and subsequently resolve its structure.
The KcsA channel functions as a tetramer, therefore the Fab monoclonal antibodies are selected on the basis of being able to recognize the tetrameric form rather than the monomeric form of the channel, but the complex was formed with a stoichiometry of one Fab fragment per channel subunit (1). In addition to functioning as crystallographic chaperones for KcsA, these synthetic antigen-binding fragments (Fabs) assisted in the probing of the mechanistic role of the C-terminal domain during activation gating (2).
The Fab fragments are attached to the K+ channel turrets on the extracellular face of the channel. The K+ channel is suspended so that its detergent micelle was not involved in crystal contacts, while making sure that all of the protein contacts within the crystal were formed between neighboring Fab fragments. The attachment of the Fab fragments to the turrets assure a wide passageway outside of the pore, which leaves ion binding undisturbed by the presence of the Fab fragments. This combined with the packing arrangement accounted for the high quality of the X-ray diffraction.
Potassium channels are tasked with the control of the electric potential across the plasma membrane of cells through selectively and rapidly catalyzing the diffusion of K+ ions down their electrochemical gradient. Through this, the channels have a critical role in setting electrical excitation in nerve and muscle cells while being involved in a wide range of vital physiological processes, including cardiac dysfunction, diabetes, and epilepsy (2). In order to perform the function of ion transport, the channel must remove the K+’s hydration shell in order to accurately select ions on the basis of size and charge. This responsibility is directly conferred from KcsA’s structure, a 160 peptide transmembrane protein that can be divided into 2 domains: the transmembrane domain (approximately 140 residues), and the C-terminal cytoplasmic domain (approximately 40 residues) (2).
The C-terminal domain is formed mainly by a bundle of α helices extending linearly toward the cytoplasm. It is a structure that has proven difficult to crystallize due to the inherent flexibility of this cytoplasmic domain. In KscA this domain also assists in the thermal stabilization of the channel. Removal of this C-terminal domain not only affects thermal stabilization, but also promotes higher basal activity and reduces the efficiency of the mechanisms controlling channel folding and assembly (2).
The tetrameric transmembrane domain has fourfold-symmetry about a central pore, with each subunit containing two α-helices; one facing the central pore and the other facing the lipid bilayer. The inner bundle of helices open on the extracellular side of the membrane, and are closely packed on the intracellular side, resembling an inverted cone (4). The inner helices also participate in subunit-subunit interactions. The outer helices only interact with the helix from their respective subunits; however, and they do so via a coiled-coil heptad repeat and a high number of Gly residues. This high proportion of Gly residues allows for less impedance and tighter interaction due to the lack of Gly side chains.
The inner cavity in KcsA is lined mainly by hydrophobic amino acids from these inner helices (Ile-100, Phe-103) that do not provide strong hydrogen-bonding donor or acceptor groups. Therefore, water in the channel pore is available to interact with the K+ ion without interference from much of the protein surface. This combined with the inner helices ensures that the K+ ion is highly stabilized within the cavity.
The inner and outer helices of these monomers are connected by the P loop, which is a stretch of about 30 residues. The P loops of the four subunits arrange to form the extracellular vestibule and the selectivity filter of the channel, a region whose passage is too narrow to fit a hydrated K+ ion. The selectivity filter contains four ion binding sites to which K+ ions can bind in an essentially dehydrated state, while being surrounded by eight carbonyl oxygen atoms provided by the sequence (75TVGYG79) from each of the four subunits (4). These carbonyls are directed straight outward, forming a ring surrounding the perimeter of the pore entryway. Buried beneath the protein surface, carboxylate-carboxylate pairs formed by the side chains of Glu-71 and Asp-80 provide four negative charges near the entryway. The strong negative charges ensure the attraction of the potassium cation to the electronegative pore entrance.
The positively charged sodium ion is also attracted to this channel, but due to a combination of regulative properties, K+ ions are selected with high fidelity over the significantly smaller Na+ ion by a factor of over 103. The previously mentioned Gly-79 carbonyl oxygen atoms not only assist in the creation of an attractive force for the potassium cation, but these negatively charged carbonyls also aid in the hydration and dehydration of K+ ions at the extracellular entryway (4, 7).
The K+ ion is held in the cavity center by the water structure around the ion, which is formed by weak hydrogen bonding interactions mediated by residues in the cavity wall. There are many other water molecules present, but they are unable to be visualized through crystallography due to their disorder. These water molecules are less ordered than the eight that hydrate the K+ ion due to the highly hydrophobic nature of most of the cavity-lining residues. Electrostatic factors, such as the orientation of the four pore α helices directing their C-terminal ends toward the center of the cavity, create a slight negative dipole, therefore making the center position favorable to the K+ ion. This provides additional stabilization energy for the K+ ion, which is essential for efficient function of the channel. This orientation of the pore dipoles also effectively stabilizes monovalent ions at the center of the membrane but not divalent ions. This is important to the selectivity of the structure. Since the filter is located at the extracellular end of the pore, cations other than K+ can probably enter the pore from the intracellular side and penetrate two-thirds of the way across the membrane. This can result in the potential blocking of the pore at the selectivity filter. The selectivity of the cavity for monovalent ions over divalent ions such as Mg2+ that are present in the intracellular environment counteracts the potential blocking of the channel. The much higher concentration of K+ present in the cytoplasm compared with that of Na+ will ensure that the monovalent ion in the cavity will be predominantly K+ (4).
The selectivity filter is stabilized by a network of hydrogen bonds to the amide nitrogen atoms that point away from the pore and into the protein core, including a short hydrogen bond between the carboxylic group of Glu-71 and that of Asp-80, and the bond between a buried water molecule and the amide nitrogen of Gly-79 (1). Under physiological conditions, the selectivity filter normally contains two K+ ions, resulting in mutual repulsion between the ions. The mutual repulsion subsequently balances the attractive force between the K+ ion and its binding site. Therefore the driving force for the movement of the ions through the channel is the coordinated movement of the two ions, which will occupy either the 1 and 3 or 2 and 4 positions in the filter, separated by a single water molecule (4).
Interestingly, the previously described structure varies significantly in conformation and function depending on the concentration of the K+ ions. Specifically, the residues at sites 3 and 2, Val-76 and Gly-77 respectively, adopt alternate conformations in the presence and absence of K+ ions. In the absence of K+, the carbonyl of Val-76 points away from the central pore and the α carbon of Gly-77 points inwards, which blocks the pore. This blockage represents the nonconductive state and demonstrates that only in the presence of high K+ ion concentrations does the selectivity filter adopt the structure in which it can conduct K+ ions with high efficiency (4, 5, 6).
This is enforced by the observation that the K+ concentration inside the cell is greater than 100 mM, whereas in the exterior the K+ concentration is usually less than 5 mM. The K+ channel gate which opens and closes the pore is located between the selectivity filter and the intracellular solution, and when this gate is open, the filter is exposed to a high K+ concentration from inside the cell, and when it is closed the filter is exposed to the low extracellular K+ concentration (1). This characteristic of the channel allows it to adequately respond to varying concentrations of K+ and not remain constitutively open (or closed).
Given these large conformational differences between the high and low K+ structures, it is reasonable that their rate of interconversion occurs more on the timescale of gating (which takes milliseconds) than that of ion conduction (which only takes nanoseconds). This demonstrates a drastic discrepancy on the order of 106. Through these observations, it has been determined that the K+ channel only begins to conduct ions once the filter snaps into a main chain conformation similar to that of the high-K+ structure. It is also noted that in the high- K+ ion concentrations in which conduction occurs, the structure of the filter is significantly less flexible. This variation in structure asserts the sophistication within this structure, and offers a physical explanation for the phenomena of “permeant ion effects on gating” (1). It may also explain apparent gating transitions that occur by mechanisms other than opening and closing the activation gate.
Intracellular pH also plays a role in the gating of the KcsA channel, with it assuming the closed conformation at a neutral or basic pH and in the open conformation at acidic pH. It has been proposed that helix bending and rearrangement deep in the membrane opens K+ channels due to there being a strong effect on a conserved Gly residue (7).
In its role as a membrane transport channel, KcsA is surrounded by the lipid molecules of the plasma membrane, and while the importance of the lipid bilayer and the role of the lipid-protein interactions in ion channel structure and function is not well understood, it has been determined that KcsA binds a negatively charged lipid molecule (3).
Trp residues are often found at the ends of transmembrane α helices and have been hypothesized to serve as anchors for these helices in the lipid bilayer. Trp residues in KcsA form girdles at the two faces of the lipid bilayer with the rings of the Trp residues being oriented almost parallel to the surface of the membrane. On the extracellular side of the membrane, Tyr residues also form a girdle above the band formed by the Trp residues. Above and below these girdles formed by the aromatic residues on the two sides of the membrane are girdles of charged residues, Arg-52, Arg-64, Arg-89 and Glu-51 on the extracellular side and Arg-27, Arg-117, Arg-121, Glu-118 and Glu-120 on the intracellular side. These residues are presumed to provide the charge required for interaction with the lipid head group region of the bilayer (4).
In the crystallization of the Fab-KscA complex, two partial lipid molecules were also identified. One of these is modeled as nonan-1-ol and the other as diacylglycerol with a C14 and C9 chain. The diacylglycerol is presumed to be a distorted phosphatidylglycerol molecule with its diacylglycerol moiety bound between two KcsA monomers into a groove on the surface near Trp-87. Considering that the diacylglycerol is in fact a phosphatidylglycerol, its anionic head-group probably interacts with Arg-64 and Arg-89 located in the girdle of charged residues above Trp-87. The single chain modeled as nonan-1-ol is located in a groove between the transmembrane helices of a single monomer (those that respectively face the channel cavity and the lipid bilayer), and could possibly correspond to the alkyl chain of the detergent dodecylmaltoside (4).
Lipid headgroup structure also plays a role in the proper function of KcsA. Studies show that a functional channel could only be obtained in the presence of anionic phospholipid, however specificity regarding the anionic lipid is not important. This could therefore be a phosphatidylglycerol, phosphatidylserine, or a cardiolipin. The presence of this anionic lipid cofactor bound between the transmembrane α helices could be essential in the gating process, because movement of these helices is required for the opening and closing of the gate mechanism (4).
Lipids also contribute to the folding process of KcsA, and it has been demonstrated that lipids are required for the in vitro refolding of the KcsA tetramer from the unfolded monomeric state. Unlike the mediation of the ion conduction however, this process is not dependent on negatively charged lipids. Through the degradation of the KcsA tetramer into its monomeric form, refolding only by the insertion of the monomers into lipid vesicles, and subsequent checking for the tetrameric form, it was determined that lipids are required in the refolding reaction of KcsA (3).
Utilizing the Dali/Blast search engines in order to find proteins structurally similar to the Fab-KcsA complex was virtually impossible, mainly because the Fab antibody is so prevalent (along with antibodies in general). Therefore the Fab antibody was compared to a different antibody as an alternative. The light chain of the green-fluorescent antibody 11G10 in complex with its hapten (2G2R) shows considerable structural and functional similarity to the Fab antibody heavy chain. It is also derived from M. musculus, has a Z-score of 25.5 (via DALI) and an E-value of 7e-111 (via protein Blast), and has an identical length of 219 peptides in its primary structure (5, 6, 10). A Z-score above 2 indicates that the proteins have a similar tertiary structure and folding, and an E-value lower than .05 demonstrates significant similarity in the primary structure of the proteins with a very high homology and low gap number. Considering that these are both antibodies, this was expected since they share a very similar function in the binding of other proteins. Upon investigation of images of their secondary and tertiary structure, they contain a high proportion of beta sheets, with these residing in the anti-parallel configuration. Also, both antibodies were generated in order to assist in the characterization of other entities through binding and review of the complex, Fab through crystallization and GFP through fluorescence (8, 9, 10).