Pentameric assembly of the Kv2.1 tetramerization domain
The Kv family of voltage-gated potassium channels regulate neuronal excitability. The biophysical characteristics of Kv channels can be matched to the needs of different neurons by forming homotetrameric or heterotetrameric channels within one of four subfamilies. The cytoplasmic tetramerization (T1) domain plays a major role in dictating the compatibility of different Kv subunits. The only Kv subfamily lacking a representative structure of the T1 domain is the Kv2 family. Here, X-ray crystallography was used to solve the structure of the human Kv2.1 T1 domain. The structure is similar to those of other T1 domains, but surprisingly formed a pentamer instead of a tetramer. In solution the Kv2.1 T1 domain also formed a pentamer, as determined by inline SEC–MALS–SAXS and negative-stain electron microscopy. The Kv2.1 T1–T1 interface involves electrostatic interactions, including a salt bridge formed by the negative charges in a previously described CDD motif, and inter-subunit coordination of zinc. It is shown that zinc binding is important for stability. In conclusion, the Kv2.1 T1 domain behaves differently from the other Kv T1 domains, which may reflect the versatility of Kv2.1, which can assemble with the regulatory KvS subunits and scaffold ER–plasma membrane contacts.
Voltage-gated potassium channels (Kvs) are essential for regulating membrane potential, propagating action potentials and controlling potassium homeostasis in a diverse array of neuronal and non-neuronal tissues (Allen et al., 2020; Wulff et al., 2009). There are 27 Kv genes in the human genome, which are classified into five subfamilies with distinct expression patterns and biophysical properties: Kv1 (Shaker), Kv2 (Shab), Kv3 (Shaw), Kv4 (Shal) and the regulatory KvS (Silent). Functional diversity is achieved by mixing and matching Kv monomers to form heterotetrameric channels in addition to homotetrameric channels (Isacoff et al., 1990). Only Kv subunits within the same subfamily can interact. The KvS proteins are an exception to this: they do not form homomeric channels, but instead must assemble with Kv2 subunits (Ottschytsch et al., 2002; Bocksteins, 2016). The major constraint on inter-subfamily assembly is the cytoplasmic tetramerization (T1) domain (Li et al., 1992; Shen & Pfaffinger, 1995; Xu et al., 1995; Shen et al., 1993). How the T1 domain confers selective assembly is poorly understood.
The T1 domain is one of four structural classes in the BTB (or POZ) superfamily of protein–protein interaction domains (Stogios et al., 2005). The core BTB fold comprises ∼95 amino acids and is comprised of five α-helices organized into two sets of hairpins that are capped at one end by three β-strands. Deletions, insertions or extensions in the core BTB fold characterize different structural classes, with the T1 class from Kv proteins being the most similar to the core BTB fold. BTB folds are most often found as homodimers or, in the case of T1 domains, tetramers. One interesting exception is the BTB fold from KCTD proteins, which may form monomers, dimers, tetramers, pentamers or even hexamers in vitro, independent of their stoichiometry in native protein complexes (Pinkas et al., 2017; Sereikaite et al., 2019).
While there are no representative T1 structures from either the Kv2 or KvS subfamilies, structural analysis of T1 domains from Kv1, Kv3 and Kv4 proteins have provided key insights into how the selective assembly of tetramers occurs. Some subfamily-specific individual differences in charged or hydrophobic amino acids facing the tetramerization interface have been identified as important for assembly (Bixby et al., 1999; Shen & Pfaffinger, 1995; Kreusch et al., 1998; Nanao et al., 2003; Stogios et al., 2005). Another critical factor in selective T1-domain assembly is inter-subunit coordination of zinc. An HX5CX20CC motif is conserved in Kv2, Kv3 and Kv4 proteins but is absent from Kv1 proteins (Bixby et al., 1999). In Kv4.2 and Kv3.1 T1 domains Zn2+ binding has been shown to be essential for monomers to assemble into tetramers (Jahng et al., 2002; Strang et al., 2003; Nanao et al., 2003). The role of Zn2+ binding has not been tested for Kv2 or KvS T1 domains, but it is logical to assume that it plays a similar role in their assembly. Thus, zinc-dependent assembly distinguishes the Kv1 T1 domain from the Kv2, Kv3 and Kv4 T1 domains. A feature that distinguishes Kv2 and KvS T1 domains from Kv3 and Kv4 T1 domains is the presence of a CDD motif (Bocksteins et al., 2009). The authors who identified the CDD motif used the structure of Kv4.2 and SWISS-MODEL to generate a homology model of Kv2.1 and determined that the critical aspartates in the CDD motif are not at the T1 homotetrameric interface but are closer to the surface of the domain (Bocksteins et al., 2009). This raises the possibility that interaction between the surface of Kv T1 domains and other regions of the channel could add to the subfamily-selective assembly dictated by T1-domain tetramerization. Supporting this idea is the observation that the N- and C-termini of Kv2.1 can interact (Ju et al., 2003; Mohapatra et al., 2008; Bocksteins et al., 2009). Solving the structures of Kv2 and KvS T1 domains will be necessary steps towards the goal of completing a comprehensive comparative analysis of how T1 domains confer selective assembly.
In this study, we solved the crystal structure of the human Kv2.1 T1 domain (Kv2.1 T1) and investigated the role of Zn2+ in the stability of the protein. To our surprise, Kv2.1 T1 was pentameric both in the crystal and in solution. Zn2+, along with a salt bridge formed by the aspartates in the CDD motif, provides stability. The different assembly state of the isolated Kv2.1 T1 domain compared with the isolated T1 domains from Kv1, Kv3 and Kv4 proteins indicates that multiple interactions between Kv2.1 subunits are required in vivo to ensure proper assembly into tetramers.
Human Kv2.1 residues 29–147, corresponding to the T1 domain, were cloned using NdeI and BamHI sites in a modified pET-28a vector (Novagen) to express a fusion protein with an N-terminal hexahistidine (6×His) affinity tag followed by a Tobacco etch virus (TEV) protease cleavage site (ENLYFQG). Transformed Escherichia coli strain BL21 (DE3) cells were grown in LB medium supplemented with 50 µg ml−1 kanamycin at 37°C to an optical density (OD600) of 0.7, and 1 mM isopropyl β-D-1-thiogalactopyranoside was added to induce expression for 15 h at 18°C. The cells were harvested by centrifugation and stored at −80°C. The harvested cells were resuspended in lysis buffer (20 mM Tris–HCl pH 8.0, 150 mM NaCl, 5% glycerol, 0.01% Triton X-100, 20 mM imidazole, 5 mM β-mercaptoethanol) and supplemented with DNase I and protease-inhibitor cocktail (Roche). The cells were disrupted using an Emulsiflex C3 (Avestin) to release the protein. The His-tagged fusion protein was purified using an Ni–NTA column (Qiagen) and the elution fraction was incubated with TEV protease for 15–18 h at 4°C to cleave the N-terminal His tag. The cleaved protein was further purified using a HiLoad 16/600 Superdex 200 pg gel-filtration column (GE Healthcare) in 20 mM Tris–HCl pH 8.0, 150 mM NaCl, 5% glycerol, 0.01% Triton X-100, 1 mM DTT. The purified protein was concentrated using Amicon Ultra centrifugal filter devices (Millipore, 10 kDa cutoff). The protein samples for dynamic light-scattering (DLS) and circular-dichroism (CD) measurements were dialyzed into phosphate buffer (20 mM Na2HPO4 pH 7.4, 150 mM NaCl, 5% glycerol, 1 mM DTT) at 4°C overnight before use.
Kv2.1 T1 was crystallized using the hanging-drop vapor-diffusion method by mixing 8 mg ml−1 Kv2.1 T1 with the reservoir solution in a 1:1 ratio. Initial crystals were obtained at 291 K in 0.2 M magnesium chloride, 15% PEG 400, 0.1 M sodium HEPES pH 7.5. The crystals were flash-cooled in liquid nitrogen using 30% PEG 400 as a cryoprotectant (see PDB entry 7spd). Crystals of Kv2.1 T1 prepared by supplementing the size-exclusion buffer with 50 µM zinc sulfate were obtained in 0.2 M magnesium chloride hexahydrate, 0.1 M Tris pH 8.5, 3.4 M 1,6-hexanediol and were flash-cooled in liquid nitrogen without additional cryoprotection (see PDB entry 7re5). X-ray diffraction data were collected at 100 K on beamline 4.2.2 at the Advanced Light Source (ALS), Berkeley, California, USA.
Diffraction data were indexed, integrated and scaled using XDS (Kabsch, 2010). Initial phase estimates for Kv2.1 T1 were obtained by molecular replacement with Phaser in Phenix (Bunkóczi et al., 2013) using the structure of Kv3.1 T1 (PDB entry 3kvt; Bixby et al., 1999) as a search model. Refinement was performed using phenix.refine (Afonine et al., 2012), followed iteratively by manual building using Coot (Emsley et al., 2010). The final refined structure includes a Kv2.1 T1 pentamer with residues 29–133 resolved. Structures were visualized in PyMOL (version 2.5; Schrödinger). The buried surface area was determined using the PISA server (Krissinel & Henrick, 2007). The structural biology applications used in this project were compiled and configured by SBGrid (Morin et al., 2013).
2.4. Size-exclusion chromatography coupled with multi-angle light scattering and small-angle X-ray scattering (SEC–MALS–SAXS)
SEC–MALS–SAXS data sets were collected on BioCAT beamline 18-ID-D at the Advanced Photon Source, Argonne, Illinois, USA. The samples were centrifuged for 5 min at 13 000 rev min−1 to remove any potential aggregates before injection. 250 µl aliquots containing 1.5 mg ml−1 Kv2.1 T1 were loaded at a flow rate of 0.5 ml min−1 onto a 24 ml Superdex 200 Increase 10/300 column on an Agilent 1300 chromatography system. Following elution from the column, the samples were analyzed inline by the UV absorbance detector of the Agilent 1300 chromatography system followed by DAWN HELEOS II light-scattering and OptiLab T-rEX refractive-index detectors in series. An accurate protein molecular weight was determined using the ASTRA software (Wyatt Technology). The elution trajectory was redirected into the SAXS sample-flow cell. Scattering data were collected every 1 s using 0.5 s exposure on a PILATUS3 1M pixel detector (Dectris) covering a q range of 0.0045 < q < 0.35 Å−1 (q = 4π/λsinθ, where λ is the wavelength and 2θ is the scattering angle). The BioXTAS RAW software was used to collect the SAXS data (Hopkins et al., 2017).
Following data reduction and buffer subtraction, the SAXS data were further analyzed using the BioXTAS RAW software. The forward scattering intensity I(0) and the radius of gyration (Rg) were calculated from the Guinier fit. The normalized Kratky plot, the pair distance distribution plot P(r) and the Porod volume were calculated using GNOM embedded in BioXTAS RAW. Low-resolution ab initio bead-based models of proteins were constructed from the experimental data using GASBOR (Svergun, 1999). The calculation of theoretical scattering curves for the crystal structure was performed by CRYSOL (Manalastas-Cantos et al., 2021), which also determines the discrepancy (χ2 value) between the simulated and experimental scattering curves.
Far-UV CD spectra of Kv2.1 T1 were acquired using a J-815 spectrometer (JASCO) connected to a Peltier temperature controller. All experiments were performed in duplicate. For each measurement, 300 µl sample consisting of 20 µM Kv2.1 T1 was added to a 1 mm path-length quartz cuvette (Hellma Analytics). The CD ellipticities (θ) were measured from 200 to 260 nm with a 1 nm increment, a scanning speed of 50 nm min−1 and a data-integration time of 2 s with a standard sensitivity. Buffers were used for baseline measurements. The final ellipticities were recorded as an average of four baseline-corrected scans. The ellipticities (θ) were used to calculate the mean residue ellipticity using the formula [θ] = θ/cnl, where c is the concentration of the protein in moles, n is the number of residues and l is the path length of the cuvette. To examine the effect of Zn2+ on the stability of Kv2.1 T1, 1 mM EDTA was added to Kv2.1 T1 samples and incubated at 4°C for 12 h, and the far-UV CD spectra were then measured in the same way as for the samples in the absence of EDTA. To detect the thermal unfolding, denaturation experiments were carried out using an automated 1°C incremental temperature ramp in the interval 20–95°C, with a 30 s equilibration time at each measurement step. The thermal unfolding profiles of the samples were characterized using the mean residue ellipticity minimum at 222 nm (θ222) to determine Tm by fitting the Boltzmann sigmoid equation using Prism (GraphPad).
2.7. Dynamic light scattering (DLS)
Thermal ramp stability measurements were made by DLS in a plastic cuvette with 1 mm path length at various temperatures ranging from 10 to 80°C, with a ramp rate of 1°C min−1. Before measurements, the sample was incubated at 10°C for 3 min. The DLS measurements were averaged from five acquisitions of 1 s each. To examine the effect of Zn2+ on the stability of Kv2.1 T1, 1 mM EDTA was added to Kv2.1 T1 and incubated at 4°C for 12 h and the size was then measured by DLS and compared with that of samples without adding any EDTA. The DLS data were analyzed using Wyatt software and were fitted by the linear intersection method to determine the Tonset values.
For negative staining, 3 µl protein solution at 0.01 mg ml−1 was added to a glow-discharged carbon-coated copper grid (Electron Microscopy Science). The grids were stained in 2%(w/v) uranyl acetate (twice for 3 s each, followed by a third staining for 25 s). The grids were blotted with filter paper (Whatman) to absorb residual solution between each step.
The grids were imaged using a Hitachi HT7800 electron microscope equipped with a tungsten filament and operated at 120 kV. Images were collected at a magnification of 100 000×, resulting in a pixel size of 1.93 A on the specimen. Only top views of particles were manually picked using cryoSPARC to determine the oligomeric state of Kv2.1 T1 (Punjani et al., 2017). The manually picked particles were classified using 2D reference-free classification and new templates were created for automatic picking to select more particles. Two rounds of 2D reference-free classification were calculated to exclude `bad' particles and were used for data analysis.
The T1-domain sequences from Kv2.1 and Kv8.2 (in a 3:1 ratio) were submitted to the ColabFold notebook to generate the heterotetramer using AlphaFold2-Multimer (Mirdita et al., 2022). The models were inspected and the top-ranked model was used for the addition of zinc. Potential zinc-binding sites in the predicted complex structure were identified using the method recently reported by Wehrspan et al. (2022). Briefly, this method seeks to superimpose a user-specified list of ligands at all possible plausible locations within a protein structure and retains all positions where a ligand can be placed free of steric clashes. The method has previously been applied to entire structural proteomes predicted using AlphaFold2 (Tunyasuvunakool et al., 2021) and has identified thousands of potential zinc-binding sites. Here, the method was applied using the same parameters as used previously. Three zinc-binding sites were identified, all of which had r.m.s.d.s lower than 0.3 Å, indicating high-quality fits, and with all coordinating side chains having pLDDT scores exceeding 88, indicating high-confidence predictions from AlphaFold2 (Jumper et al., 2021). The code is available at https://github.com/Elcock-Lab/Metalloproteome
3.1. The crystal structure of the Kv2.1 tetramerization (T1) domain
The T1 domain of voltage-gated potassium channel Kv2.1 (human Kv2.1 T1; residues 29–147) was expressed in E. coli BL21 (DE3) cells and purified by Ni–NTA chromatography followed by TEV cleavage to remove the N-terminal His tag and finally Superdex 200 chromatography. Crystals of Kv2.1 T1 belonged to space group P41212. Initial phase estimates were obtained by molecular replacement using a Kv3.1 T1-domain monomer (PDB entry 3kvt) as a search model, and the resulting Kv2.1 T1 structure was refined at a resolution of 2.5 Å (PDB entry 7re5). Electron densities for residues 29–133 were identified; the missing C-terminal residues (134–147) may be disordered. Surprisingly, five protein molecules were present in the asymmetric unit of the Kv2.1 T1 crystals, forming a pentameric ring, as shown from the C-terminal (top) and side views (Figs. 1a and 1b). Diffraction data and refinement statistics are presented in Table 1. A metal ion was identified in each of the individual T1 domains, coordinated by a conserved zinc-binding HX5CX20CC motif. The Kv2.1 T1 pentameric rings were stacked as dimers in the crystal and a closer examination of the interface between rings revealed a second metal ion coordinated by residues introduced by the N-terminal NdeI cloning site. Representative electron densities and verification of the metal as Zn2+ by X-ray fluorescence spectroscopy are shown in Supplementary Fig. S1. We also obtained a lower resolution (2.7 Å) structure in the alternative space group C2221, which was also a pentamer but with partial zinc occupancy (for details, see PDB entry 7spd). The core BTB fold of the Kv2.1 T1 monomers is similar to the monomers of the pentameric T1-like domain from KCTD5 (Fig. 1c) and the tetrameric Kv4.2 T1 domain (Fig. 1d). The five Kv2.1 T1 subunits are rotated from the central axis of the ring by 70.5°, 71.3°, 71.5°, 73.2° and 73.5°, which is more open than the arrangement in KCTD5 (69.5°, 69.5°, 72.8°, 73.4° and 74.7°). Overall, the pentameric ring of Kv2.1 T1 subunits is less compact than that for KCTD5.
The surface of the interface between adjacent Kv2.1 T1 subunits is characterized by complementary charges (Fig. 2a). Electrostatics make a particular contribution to the N-terminal portion of the T1 domain, with a positively charged bulge fitting into a negatively charged concave area on the adjacent subunit. The residues constituting the negatively charged bulge are highly conserved across the T1 family. However, the residues constituting the matching positive patch are less conserved and thus may contribute to the selectivity of T1-domain assembly (Fig. 2b).
Two motifs that have previously been implicated in selective higher-order assembly of T1 domains are the Zn2+ binding and CDD motifs (Fig. 3a; Jahng et al., 2002; Strang et al., 2003; Nanao et al., 2003; Bocksteins et al., 2009; Bixby et al., 1999). The HX5CX20CC Zn2+-binding motif is found in all Kv2, Kv3 and Kv4 proteins, as well as in four of the ten KvS proteins (Kv8.2, Kv6.1, Kv6.2 and Kv6.3) that assemble with Kv2 proteins. This motif, occupied by Zn2+, is at the C-terminal end of the T1 domain, which in the context of the full-length protein would be the end closest to the transmembrane domains. In Kv2.1 it is formed by His105, Cys132 and Cys133, with the third cysteine, Cys111, provided by the adjacent subunit (Fig. 3b, inset 1). The CDD motif is found in a loop between α3 and β3 at the other end of the subunit interface from zinc (Fig. 3b, inset 2). Asp74 and Asp75 of this motif are in the interface and form a salt bridge with Arg31 and Arg32. The arginines are only conserved in Kv2.1 and Kv2.2. Together, these two motifs are well positioned to provide stability to higher-ordered assemblies of Kv2.1 T1 domains. Functional Kv channels are tetramers, and we asked whether these two motifs could still be involved in stabilizing either homotetrameric or heterotetrameric assemblies of Kv2.1 T1 by examining the interface of models generated using YASARA or AlphaFold2. It is likely that Zn2+ binding increases the stability of all forms, while the CDD motif should make Kv2.1 homodimers or homotetramers more compact and potentially more stable (Supplementary Fig. S2).
We used multi-angle light scattering and small-angle X-ray scattering coupled with size-exclusion chromatography (SEC–MALS–SAXS) as an independent method to determine the stoichiometry of Kv2.1 T1 complexes in solution. The molecular weight of the Kv2.1 T1 monomer is 14.6 kDa, so the theoretical mass of a tetramer is 58.4 kDa, while that of a pentamer is 73 kDa. The absolute molecular mass of Kv2.1 T1 measured by MALS was 71 ± 1 kDa (Fig. 4a). The MALS data are most consistent with a pentameric assembly of Kv2.1 T1. From the SAXS data, it was determined that the Kv2.1 T1 protein was monodisperse (linear in a Guinier plot) and globular (with a Gaussian peak in a Kratky plot) (Supplementary Fig. S3). The experimental scattering curve of Kv2.1 T1 was fitted to a tetrameric homology model of Kv2.1 T1 generated using YASARA (based on Kv3.1 T1; PDB entry 3kvt), but the fit was very poor, with χ2 = 27. Fitting the experimental scattering curve to the pentameric crystal structure resulted in a good fit with χ2 = 1.7 (Fig. 4b). To aid in visualization, the calculated surface envelope for Kv2.1 T1 was superimposed with cartoons for the tetrameric homology model or pentameric crystal structure (Fig. 4c). The pentameric structure best fits the SAXS data. We also used negative-stain electron microscopy (EM) to investigate the oligomeric state of Kv2.1 T1. Top views of ring-shaped particles for Kv2.1 T1 could be seen in the raw electron micrographs, and after 2D class averaging the rings could be more clearly seen to be pentamers (Fig. 5). In conclusion, analysis of Kv2.1 T1 in solution by SEC–MALS–SAXS complemented with direct visualization by EM is consistent with the pentameric assembly revealed by the crystal structure.
The role of Zn2+ in Kv3 and Kv4 T1 domains has been determined to be critical for the transition from monomers to tetramers as well as for protein stability (Jahng et al., 2002; Strang et al., 2003). To investigate the effect of bound Zn2+ on the assembly and stability of Kv2.1 T1, we used chelation by excess EDTA (1 mM) and assessed the effect that this had on the protein using circular-dichroism (CD) and dynamic light-scattering (DLS) thermal unfolding experiments. The CD spectra calculated for Kv1.1, Kv3.1 and Kv4.1 T1 domains determined using PDB2CD (Mavridis & Janes, 2017) was very similar to the experimental spectra obtained for Kv2.1 T1. A strong α-helical signal for Kv2.1 T1 persisted at 10°C after the addition of EDTA, implying that Kv2.1 T1 remains in a native fold in this temperature range independent of bound Zn2+ (Fig. 6a).
The thermal unfolding experiments were monitored by CD at 222 nm at 1°C intervals from 20 to 95°C (Fig. 6b). The fraction of unfolded protein (fU) as a function of temperature exhibited two transitions. The α-helical signal began to diminish at 40°C and unfolding increased until ∼60°C, and at ∼80°C the amount of unfolded protein again increased. In the presence of EDTA the curve was shifted to the left and the differences between the two phases were less striking. To calculate the melting temperature (Tm, defined as fU = 0.5), the fraction of unfolded protein (fU) as a function of temperature from 10 to 70 °C (to encompass the first transition) was fitted by a sigmoidal curve. The Tm of Kv2.1 T1 was 50 ± 1°C and decreased to 38 ± 1°C in the presence of EDTA. Our results imply that Kv2.1 T1 undergoes a large conformational change at lower temperature in the absence of bound Zn2+. Using DLS as a complementary approach, we monitored the hydrodynamic radius as a function of temperature (10–80°C). The onset transition temperature (or the threshold where protein unfolding begins; Tonset) of Kv2.1 T1, defined by a linear transition plot, was 42.5 ± 0.5°C. Addition of EDTA reduced the Tonset to 29.5 ± 0.5°C (Fig. 6c). These data support the CD experiments and we conclude that Zn2+ binding increases the stability of the protein.
Here, we report the crystal structure of the T1 domain of Kv2.1. This T1 domain is very similar to previously described T1 domains from other Kv subfamilies, except that it forms a pentamer. All of the previously reported structures of isolated Kv T1 domains are tetramers, which is consistent with the tetrameric assembly of full-length Kv subunits into functional channels. Our X-ray structural data are supported by solution analysis of the Kv2.1 T1 domain using SEC–MALS–SAXS and EM imaging.
In the larger family of BTB folds (of which the T1 domain is one class), there is a variety of stoichiometries from monomers to hexamers (Stogios et al., 2005; Sereikaite et al., 2019). Illustrative to Kv2.1 T1 are the T1-like domains from the Cullin-dependent ubiquitin E3-ligases, where the stoichiometry is particularly diverse. In one study, the T1-like domains of SHKBP1, KCTD13, KCTD16 and KCTD17 crystallized as a monomer, a tetramer, an open pentamer and a closed pentamer, respectively, even though most of these T1-like domains are pentamers in the context of the full-length native protein. In solution, the T1-like domains of SHKBP1 and KCTD13 were predominantly mixtures of monomers and dimers, but mixing with the binding partner Cullin3 induced pentamers (Pinkas et al., 2017).
We assume that the pentameric state that we observed for Kv2.1 T1 represents either an intermediate state that is possible as Kv2.1 monomers assemble into functional, tetrameric channels or a simple epiphenomenon of working with the T1 domain out of the context of the full-length channel. We propose that for full-length native Kv2.1 the T1-domain assembly state is restrained to a tetramer by interactions with either other parts of the channel or accessory subunits. In support of this idea, interactions between the cytoplasmic N-termini and C-termini of Kv2 have been observed (Ju et al., 2003; Mohapatra et al., 2008; Bocksteins et al., 2009). Intriguingly, a recent structural study discovered that the orientation of the T1 domain in Kv3 channels differs from that in structures of the Kv1.2–2.1 paddle chimera. The most C-terminal α-helix of the Kv3.1a T1 domain is rotated towards and contacts the linker between the voltage-sensor and pore domains, which may stabilize the pore and contribute to the unique kinetic properties of Kv3 channels (Chi et al., 2021). It remains to be determined whether a similar arrangement occurs in Kv2 channels.
In terms of accessory subunits, the Kvβ subunit is best characterized for its interaction with Kv1 (Shaker) channels, and several KChIP proteins serve a similar role to Kvβ for Kv4 channels (Pongs et al., 1999; Bähring, 2018). Both Kvβ and KChIP proteins make contacts with the T1 domain of the cognate Kv protein and in the case of KChIP can even rescue mutants of Kv4 T1 domains that interfere with tetramerization (Kunjilwar et al., 2004; Liang et al., 2010). For Kv2.1 the best characterized interaction proteins are the electrically silent (KvS) α subunits (Bocksteins, 2016) and those mediating the plasma membrane–ER contact sites organized by phosphorylated Kv2.1 in a scaffolding role independent of ion conduction (Deutsch et al., 2012; Weigel et al., 2012; Fox et al., 2013, 2015; Johnson et al., 2018; Antonucci et al., 2001; Misonou et al., 2004; Kirmiz, Palacio et al., 2018; Kirmiz, Vierra et al., 2018). The recent acceleration in solving structures of ion channels, with and without accessory proteins, by cryo-EM makes it reasonable to expect that a structure of full-length Kv2.1 will be determined and provide information about the interactions that influence the T1 domain.
The interface between Kv2.1 T1 subunits in the pentameric ring involves two major features: electrostatics and inter-subunit coordination of zinc. At the N-terminal end of the domain there is a large highly conserved patch of surface negative charges that fits into a concavity of positive charges on the adjacent subunit. This is also where the CDD motif is found, a motif that is recognized for its selective conservation in Kv2 and KvS proteins (Bocksteins et al., 2009). Charge-reversal mutants of the aspartates in the CDD motif interfered with the assembly of Kv2.1. The original homology modeling of Kv2.1 T1 positioned the CDD residues away from the T1 interface, so that the assembly problem was thought to have been due to disruption of an interaction with the C-terminus (Mohapatra et al., 2008). However, in our structure the aspartates forming the CDD motif participate in a salt bridge with two arginines at the N-terminus of T1 which contributes to stabilizing the T1–T1 interface. Caution must be employed because this salt bridge may be disrupted in the physiologically relevant tetrameric form of the Kv2.1 T1–T1 interface, although our modeling predicts that this is only true for heterotetrameric assemblies of Kv2.1 with KvS T1 domains.
Inter-subunit coordination of zinc by Kv3 and Kv4 T1 domains has been shown to be critical for the stability of the protein as well as for the assembly of tetramers. Kv4.2 T1 is a particularly interesting case because chelation with EDTA to remove zinc has been shown to convert tetramers to monomers, which could reform into tetramers upon the replacement of zinc (Jahng et al., 2002). For Kv2.1 T1 we obtained two structures, one with full zinc occupancy and one with partial zinc occupancy; in both structures the T1 domain was a pentamer. This indicates that zinc binding alone may not be sufficient to dictate the stoichiometry of the Kv2.1 T1 domain. Using CD and DLS, we were able to show that removing zinc by EDTA chelation reduced the stability of the Kv2.1 T1 protein; this latter observation is consistent with the role of zinc binding in stabilizing Kv3 and Kv4 T1 domains.
Our observations of the pentameric Kv2.1 T1–T1 interface have implications for understanding why only T1 domains from Kv2.1 or Kv2.2 can interact with T1 domains from the ten-member KvS family. Four of the KvS T1 domains (Kv8.2, Kv6.1, Kv6.2 and Kv6.3) appear to have lost the ability to bind zinc since the histidine in the HX5CX20CC Zn2+ coordination motif is not conserved in these proteins. This alteration would not prevent these T1 domains from providing the cysteine needed to coordinate the Zn2+ ion bound in an adjacent Kv2.1 T1 domain, but it would result in only partial zinc occupancy of the tetrameric channels. Since partial zinc occupancy did not prevent the crystallization of a Kv2.1 homopentamer, it is possible that partial zinc occupancy is not as deleterious to Kv2.1-containing channels as it is predicted to be for Kv3- or Kv4-containing channels. This could explain why at least four KvS T1 domains do not assemble with Kv3 or Kv4 T1 domains. A feature that impacts all ten KvS T1 domains is the lack of a pair of arginines to form the salt bridge with the aspartates in the CDD motif. If this salt bridge is important in the tetramer as well as the pentamer that we see here, it would imply that the Kv2 and KvS proteins would be most stable in a 3:1 stoichiometry. Both 2:2 and 3:1 stoichiometries have been observed for Kv2 and KvS proteins in transfected cells, but the state that predominates in native channels is unknown (Pisupati et al., 2018, 2020; Möller et al., 2020; Kerschensteiner et al., 2005).
In conclusion, we provide the first structural analysis of the human Kv2.1 T1 domain. Since structures of Kv1, Kv3 and Kv4 T1 domains have been solved, we started this work with the expectation of providing a representative T1 domain for the missing channel-forming Kv subfamily. We were consequently surprised to find that the Kv2.1 T1 domain behaves differently from the others in the stoichiometry that we observed. Future work will need to determine how the Kv2.1 T1 domain is arranged in the context of the full-length channel. While this manuscript was under review, an abstract was published indicating that a cryo-EM structure of full-length Kv2.1 will soon be available for just such an analysis (Fernández-Mariño et al., 2022).
The authors would like to thank Srinivas Chakravarthy and Jesse Hopkins at BioCAT beamline 18-ID-D at APS for SEC–MALS–SAXS data collection, Jay Nix at beamline 4.2.2 at ALS for X-ray diffraction and fluorescence data collection and Lokesh Gakhar for helpful discussions. Adrian H. Elcock and Zachary J. Wehrspan provided helpful assistance with the addition of zinc to the AlphaFold2 model of the Kv2.1–Kv8.2 T1-domain heterotetramer using their Metalloproteome code. We acknowledge the use of resources at Carver College of Medicine's Protein and Crystallography Facility and Central Microscopy Research Facility at the University of Iowa. This research used the resources of the Advanced Photon Source, a US Department of Energy (DOE) Office of Science User Facility operated for the DOE Office of Science by Argonne National Laboratory under Contract No. DE-AC02-06CH11357 and supported in part by a grant from the National Institute of General Medical Sciences of the National Institutes of Health (P30 GM138395 to the Illinois Institute of Technology). This research also used the resources of beamline 4.2.2 of the Advanced Light Source, a US DOE Office of Science User Facility under Contract No. DE-AC02-05CH11231, supported in part by the ALS-ENABLE program funded by the NIH, National Institute of General Medical Sciences (P30 GM124169-01).
The following funding is acknowledged: National Institute of General Medical Sciences (grant No. P30 GM138395).
Afonine, P. V., Grosse-Kunstleve, R. W., Echols, N., Headd, J. J., Moriarty, N. W., Mustyakimov, M., Terwilliger, T. C., Urzhumtsev, A., Zwart, P. H. & Adams, P. D. (2012). Acta Cryst. D68, 352–367. Web of Science CrossRef CAS IUCr Journals Google Scholar
Allen, N. M., Weckhuysen, S., Gorman, K., King, M. D. & Lerche, H. (2020). Eur. J. Paediatr. Neurol. 24, 105–116. PubMed Google Scholar
Antonucci, D. E., Lim, S. T., Vassanelli, S. & Trimmer, J. S. (2001). Neuroscience, 108, 69–81. PubMed CAS Google Scholar
Bähring, R. (2018). Channels, 12, 187–200. PubMed Google Scholar
Bixby, K. A., Nanao, M. H., Shen, N. V., Kreusch, A., Bellamy, H., Pfaffinger, P. J. & Choe, S. (1999). Nat. Struct. Mol. Biol. 6, 38–43. CAS Google Scholar
Bocksteins, E. (2016). J. Gen. Physiol. 147, 105–125. CAS PubMed Google Scholar
Bocksteins, E., Labro, A. J., Mayeur, E., Bruyns, T., Timmermans, J. P., Adriaensen, D. & Snyders, D. J. (2009). J. Biol. Chem. 284, 31625–31634. PubMed CAS Google Scholar
Bunkóczi, G., Echols, N., McCoy, A. J., Oeffner, R. D., Adams, P. D. & Read, R. J. (2013). Acta Cryst. D69, 2276–2286. Web of Science CrossRef IUCr Journals Google Scholar
Chi, G., Liang, Q., Sridhar, A., Sader, K., Radjainia, M., Qian, P., Castro-Hartmann, P., Venkaya, S., Singh, N. K., McKinley, G., Fernandez-Cid, A., Mukhopadhyay, S., Burgess-Brown, N., Delemotte, L., Covarrubias, M. & Duerr, K. (2021). Research Square, https://doi.org/10.21203/rs.3.rs-828003/v1. Google Scholar
Deutsch, E., Weigel, A. V., Akin, E. J., Fox, P., Hansen, G., Haberkorn, C. J., Loftus, R., Krapf, D. & Tamkun, M. M. (2012). Mol. Biol. Cell, 23, 2917–2929. CAS PubMed Google Scholar
Emsley, P., Lohkamp, B., Scott, W. G. & Cowtan, K. (2010). Acta Cryst. D66, 486–501. Web of Science CrossRef CAS IUCr Journals Google Scholar
Fernández-Mariño, A. I., Tan, X. F., Bae, C., Stix, R., Chang, T. H., Jiang, J. S., Faraldo-Gómez, J. D. & Swartz, K. (2022). Biophys. J. 121, 295a. Google Scholar
Fox, P. D., Haberkorn, C. J., Akin, E. J., Seel, P. J., Krapf, D. & Tamkun, M. M. (2015). J. Cell Sci. 128, 2096–2105. CAS PubMed Google Scholar
Fox, P. D., Haberkorn, C. J., Weigel, A. V., Higgins, J. L., Akin, E. J., Kennedy, M. J., Krapf, D. & Tamkun, M. M. (2013). Mol. Biol. Cell, 24, 2703–2713. CAS PubMed Google Scholar
Hopkins, J. B., Gillilan, R. E. & Skou, S. (2017). J. Appl. Cryst. 50, 1545–1553. Web of Science CrossRef CAS IUCr Journals Google Scholar
Isacoff, E. Y., Jan, Y. N. & Jan, L. Y. (1990). Nature, 345, 530–534. CAS PubMed Google Scholar
Jahng, A. W., Strang, C., Kaiser, D., Pollard, T., Pfaffinger, P. & Choe, S. (2002). J. Biol. Chem. 277, 47885–47890. PubMed CAS Google Scholar
Johnson, B., Leek, A. N., Solé, L., Maverick, E. E., Levine, T. P. & Tamkun, M. M. (2018). Proc. Natl Acad. Sci. USA, 115, E7331–E7340. CAS PubMed Google Scholar
Ju, M., Stevens, L., Leadbitter, E. & Wray, D. (2003). J. Biol. Chem. 278, 12769–12778. PubMed CAS Google Scholar
Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., Tunyasuvunakool, K., Bates, R., Žídek, A., Potapenko, A., Bridgland, A., Meyer, C., Kohl, S. A. A., Ballard, A. J., Cowie, A., Romera-Paredes, B., Nikolov, S., Jain, R., Adler, J., Back, T., Petersen, S., Reiman, D., Clancy, E., Zielinski, M., Steinegger, M., Pacholska, M., Berghammer, T., Bodenstein, S., Silver, D., Vinyals, O., Senior, A. W., Kavukcuoglu, K., Kohli, P. & Hassabis, D. (2021). Nature, 596, 583–589. Web of Science CrossRef CAS PubMed Google Scholar
Kabsch, W. (2010). Acta Cryst. D66, 125–132. Web of Science CrossRef CAS IUCr Journals Google Scholar
Kerschensteiner, D., Soto, F. & Stocker, M. (2005). Proc. Natl Acad. Sci. USA, 102, 6160–6165. PubMed CAS Google Scholar
Kirmiz, M., Palacio, S., Thapa, P., King, A. N., Sack, J. T. & Trimmer, J. S. (2018). Mol. Biol. Cell, 29, 2410–2432. CAS PubMed Google Scholar
Kirmiz, M., Vierra, N. C., Palacio, S. & Trimmer, J. S. (2018). J. Neurosci. 38, 7562–7584. CAS PubMed Google Scholar
Kreusch, A., Pfaffinger, P. J., Stevens, C. F. & Choe, S. (1998). Nature, 392, 945–948. CAS PubMed Google Scholar
Krissinel, E. & Henrick, K. (2007). J. Mol. Biol. 372, 774–797. Web of Science CrossRef PubMed CAS Google Scholar
Kunjilwar, K., Strang, C., DeRubeis, D. & Pfaffinger, P. J. (2004). J. Biol. Chem. 279, 54542–54551. PubMed CAS Google Scholar
Li, M., Jan, Y. N. & Jan, L. Y. (1992). Science, 257, 1225–1230. PubMed CAS Google Scholar
Liang, P., Chen, H., Cui, Y., Lei, L. & Wang, K. (2010). Biophys. J. 98, 2867–2876. CAS PubMed Google Scholar
Manalastas-Cantos, K., Konarev, P. V., Hajizadeh, N. R., Kikhney, A. G., Petoukhov, M. V., Molodenskiy, D. S., Panjkovich, A., Mertens, H. D. T., Gruzinov, A., Borges, C., Jeffries, C. M., Svergun, D. I. & Franke, D. (2021). J. Appl. Cryst. 54, 343–355. Web of Science CrossRef CAS IUCr Journals Google Scholar
Mavridis, L. & Janes, R. W. (2017). Bioinformatics, 33, 56–63. Web of Science CrossRef CAS PubMed Google Scholar
Mirdita, M., Schütze, K., Moriwaki, Y., Heo, L., Ovchinnikov, S. & Steinegger, M. (2022). bioRxiv, 2021.08.15.456425. Google Scholar
Misonou, H., Mohapatra, D. P., Park, E. W., Leung, V., Zhen, D., Misonou, K., Anderson, A. E. & Trimmer, J. S. (2004). Nat. Neurosci. 7, 711–718. CrossRef PubMed CAS Google Scholar
Mohapatra, D. P., Siino, D. F. & Trimmer, J. S. (2008). J. Neurosci. 28, 4982–4994. CrossRef PubMed CAS Google Scholar
Möller, L., Regnier, G., Labro, A. J., Blunck, R. & Snyders, D. J. (2020). Proc. Natl Acad. Sci. USA, 117, 9365–9376. PubMed Google Scholar
Morin, A., Eisenbraun, B., Key, J., Sanschagrin, P. C., Timony, M. A., Ottaviano, M. & Sliz, P. (2013). eLife, 2, e01456. Web of Science CrossRef PubMed Google Scholar
Nanao, M. H., Zhou, W., Pfaffinger, P. J. & Choe, S. (2003). Proc. Natl Acad. Sci. USA, 100, 8670–8675. CrossRef PubMed CAS Google Scholar
Ottschytsch, N., Raes, A., Van Hoorick, D. & Snyders, D. J. (2002). Proc. Natl Acad. Sci. USA, 99, 7986–7991. CrossRef PubMed CAS Google Scholar
Pinkas, D. M., Sanvitale, C. E., Bufton, J. C., Sorrell, F. J., Solcan, N., Chalk, R., Doutch, J. & Bullock, A. N. (2017). Biochem. J. 474, 3747–3761. CrossRef CAS PubMed Google Scholar
Pisupati, A., Mickolajczyk, K. J., Hancock, W. O. & Jegla, T. (2020). Proc. Natl Acad. Sci. USA, 117, 29288–29289. CrossRef CAS PubMed Google Scholar
Pisupati, A., Mickolajczyk, K. J., Horton, W., van Rossum, D. B., Anishkin, A., Chintapalli, S. V., Li, X., Chu-Luo, J., Busey, G., Hancock, W. O. & Jegla, T. (2018). J. Gen. Physiol. 150, 1702–1721. CrossRef CAS PubMed Google Scholar
Pongs, O., Leicher, T., Berger, M., Roeper, J., Bähring, R., Wray, D., Giese, K. P., Silva, A. J. & Storm, J. F. (1999). Ann. N. Y. Acad. Sci. 868, 344–355. CrossRef PubMed CAS Google Scholar
Punjani, A., Rubinstein, J. L., Fleet, D. J. & Brubaker, M. A. (2017). Nat. Methods, 14, 290–296. Web of Science CrossRef CAS PubMed Google Scholar
Sereikaite, V., Fritzius, T., Kasaragod, V. B., Bader, N., Maric, H. M., Schindelin, H., Bettler, B. & Strømgaard, K. (2019). J. Med. Chem. 62, 8819–8830. CrossRef CAS PubMed Google Scholar
Shen, N. V., Chen, X., Boyer, M. M. & Pfaffinger, P. J. (1993). Neuron, 11, 67–76. CAS PubMed Google Scholar
Shen, N. V. & Pfaffinger, P. J. (1995). Neuron, 14, 625–633. CrossRef CAS PubMed Google Scholar
Stogios, P. J., Downs, G. S., Jauhal, J. J., Nandra, S. K. & Privé, G. G. (2005). Genome Biol. 6, R82. Web of Science CrossRef PubMed Google Scholar
Strang, C., Kunjilwar, K., DeRubeis, D., Peterson, D. & Pfaffinger, P. J. (2003). J. Biol. Chem. 278, 31361–31371. CrossRef PubMed CAS Google Scholar
Svergun, D. I. (1999). Biophys. J. 76, 2879–2886. Web of Science CrossRef PubMed CAS Google Scholar
Tunyasuvunakool, K., Adler, J., Wu, Z., Green, T., Zielinski, M., Žídek, A., Bridgland, A., Cowie, A., Meyer, C., Laydon, A., Velankar, S., Kleywegt, G. J., Bateman, A., Evans, R., Pritzel, A., Figurnov, M., Ronneberger, O., Bates, R., Kohl, S. A. A., Potapenko, A., Ballard, A. J., Romera-Paredes, B., Nikolov, S., Jain, R., Clancy, E., Reiman, D., Petersen, S., Senior, A. W., Kavukcuoglu, K., Birney, E., Kohli, P., Jumper, J. & Hassabis, D. (2021). Nature, 596, 590–596. Web of Science CrossRef CAS PubMed Google Scholar
Wehrspan, Z. J., McDonnell, R. T. & Elcock, A. H. (2022). J. Mol. Biol. 434, 167377. CrossRef PubMed Google Scholar
Weigel, A. V., Fox, P. D., Akin, E. J., Ecklund, K. H., Tamkun, M. M. & Krapf, D. (2012). Biophys. J. 103, 1727–1734. CrossRef CAS PubMed Google Scholar
Wulff, H., Castle, N. A. & Pardo, L. A. (2009). Nat. Rev. Drug Discov. 8, 982–1001. CrossRef PubMed CAS Google Scholar
Xu, J., Yu, W., Jan, Y. N., Jan, L. Y. & Li, M. (1995). J. Biol. Chem. 270, 24761–24768. CrossRef CAS PubMed Google Scholar
This is an open-access article distributed under the terms of the Creative Commons Attribution (CC-BY) Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original authors and source are cited.