Recreation of neuronal Kv1 channel oligomers by expression in mammalian cells using Semliki Forest virus

A new Kv1.2 channelopathy underlying cerebellar ataxia

A forward genetic screen of mice treated with the mutagen ENU identified a mutant mouse with chronic motor incoordination. This mutant, named Pingu (Pgu), carries a missense mutation, an I402T substitution in the S6 segment of the voltage-gated potassium channel Kcna2. The gene Kcna2 encodes the voltage-gated potassium channel α-subunit Kv1.2, which is abundantly expressed in the large axon terminals of basket cells that make powerful axo-somatic synapses onto Purkinje cells. Patch clamp recordings from cerebellar slices revealed an increased frequency and amplitude of spontaneous GABAergic inhibitory postsynaptic currents and reduced action potential firing frequency in Purkinje cells, suggesting that an increase in GABA release from basket cells is involved in the motor incoordination in Pgu mice. In line with immunochemical analyses showing a significant reduction in the expression of Kv1 channels in the basket cell terminals of Pgu mice, expression of homomeric and heteromeric channels containing the Kv1.2(I402T) α-subunit in cultured CHO cells revealed subtle changes in biophysical properties but a dramatic decrease in the amount of functional Kv1 channels. Pharmacological treatment with acetazolamide or transgenic complementation with wild-type Kcna2 cDNA partially rescued the motor incoordination in Pgu mice. These results suggest that independent of known mutations in Kcna1 encoding Kv1.1, Kcna2 mutations may be important molecular correlates underlying human cerebellar ataxic disease.

The Kv1.2 potassium channel: the position of an N-glycan on the extracellular linkers affects its protein expression and function

Voltage-gated potassium Kv1 channels have three extracellular linkers, the S1-S2, the S3-S4, and the S5-P. The S1-S2 is the only linker that has an N-glycan and it is at a conserved position on this linker on Kv1.1-Kv1.5 and Kv1.7 channels. We hypothesize that an N-glycan is found at only this position due to its effect on folding, trafficking, and/or function of these channels. To investigate this hypothesis, N-glycosylation sites were engineered at different positions on the extracellular linkers of Kv1.2 to determine the effects of N-glycans on channel surface protein expression and function. Our data suggest that for Kv1 channels, (1) placing an N-glycan at non-native positions on the S1-S2 linker decreased cell surface protein expression but the N-glycan still affected function similarly as if it were at its native position, (2) placing a non-native N-glycan on the S3-S4 linker significantly altered function, and (3) placing a non-native N-glycan on the S5-P linker disrupted both trafficking and function. We suggest that Kv1 channels have an N-glycan at a conserved position on only the S1-S2 linker to overcome the constraints for proper folding, trafficking, and function that appear to occur if the N-glycan is moved from this position.

Breaking the bottleneck: Eukaryotic membrane protein expression for high-resolution structural studies

The recombinant expression of eukaryotic membrane proteins has been a major stumbling block in efforts to determine their structures. In the last two years, however, five such proteins have yielded high-resolution X-ray or electron diffraction data, opening the prospect of increased throughput for eukaryotic membrane protein structure determination. Here, we summarize the major expression systems available, and highlight technical advances that should facilitate more systematic screening of expression conditions for this physiologically important class of targets.

Concatemers of brain Kv1 channel alpha subunits that give similar K+ currents yield pharmacologically distinguishable heteromers

At least five subtypes of voltage-gated (Kv1) channels occur in neurons as tetrameric combinations of different alpha subunits. Their involvement in controlling cell excitability and synaptic transmission make them potential targets for neurotherapeutics. As a prerequisite for this, we established herein how the characteristics of hetero-oligomeric K(+) channels can be influenced by alpha subunit composition. Since the three most prevalent Kv1 subunits in brain are Kv1.2, 1.1 and 1.6, new Kv1.6-1.2 and Kv1.1-1.2 concatenated constructs in pIRES-EGFP were stably expressed in HEK cells and the biophysical plus pharmacological properties of their K(+) currents determined relative to those for the requisite homo-tetramers. These heteromers yielded delayed-rectifier type K(+) currents whose activation, deactivation and inactivation parameters are fairly similar although substituting Kv1.1 with Kv1.6 led to a small negative shift in the conductance-voltage relationship, a direction unexpected from the characteristics of the parental homo-tetramers. Changes resulting from swapping Kv1.6 for Kv1.1 in the concatemers were clearly discerned with two pharmacological agents, as measured by inhibition of the K(+) currents and Rb(+) efflux. alphaDendrotoxin and 4-aminopyridine gave a similar blockade of both hetero-tetramers, as expected. Most important for pharmacological dissection of channel subtypes, dendrotoxin(k) and tetraethylammonium readily distinguished the susceptible Kv1.1-1.2 containing oligomers from the resistant Kv1.6-1.2 channels. Moreover, the discriminating ability of dendrotoxin(k) was further confirmed by its far greater ability to displace (125)I-labelled alphadendrotoxin binding to Kv1.1-1.2 than Kv1.6-1.2 channels. Thus, due to the profiles of these two channel subtypes being found to differ, it seems that only multimers corresponding to those present in the nervous system provide meaningful targets for drug development.

The glycosylation state of Kv1.2 potassium channels affects trafficking, gating, and simulated action potentials

We presented evidence previously that decreasing the glycosylation state of the Kv1.1 potassium channel modified its gating by a combined surface potential and a cooperative subunit interaction mechanism and these effects modified simulated action potentials. Here we continued to test the hypothesis that glycosylation affects channel function in a predictable fashion by increasing and decreasing the glycosylation state of Kv1.2 channels. Compared with Kv1.2, increasing the glycosylation state shifted the V(1/2) negatively with a steeper G-V slope, increased activation kinetics with little change in deactivation kinetics or in their voltage-dependence, and decreased the apparent level of C-type inactivation. Decreasing the glycosylation state had essentially the opposite effects and shifted the V(1/2) positively with a shallower G-V slope, decreased activation kinetics (and voltage-dependence), decreased deactivation kinetics, and increased the apparent level of C-type inactivation. Single channel conductance was not affected by the different glycosylation states of Kv1.2 tested here. Hyperpolarized or depolarized shifts in V(1/2) from wild type were apparently due to an increased or decreased level of channel sialylation, respectively. Data and modeling suggested that the changes in activation properties were mostly predictable within and between channels and were consistent with a surface potential mechanism, but those on deactivation properties were not predictable and were more consistent with a conformational mechanism. Moreover the effect on the deactivation process appeared to be channel-type dependent as well as glycosylation-site dependent. The glycosylation state of Kv1.2 also affected action potentials in simulations. In addition, preventing N-glycosylation decreased cell surface Kv1.2 expression levels by approximately 40% primarily by increasing partial endoplasmic reticulum retention and this effect was completely rescued by Kv1.4 subunits, which are glycosylated, but not by cytoplasmic Kvbeta2.1 subunits. The nonglycosylated Kv1.2 protein had a similar protein half-life as the glycosylated protein and appeared to be folded properly. Thus altering the native Kv1.2 glycosylation state affected its trafficking, gating, and simulated action potentials. Differential glycosylation of ion channels could be used by excitable cells to modify cell signaling.

The epilepsy-linked Lgi1 protein assembles into presynaptic Kv1 channels and inhibits inactivation by Kvbeta1

The voltage-gated potassium (Kv) channel subunit Kv1.1 is a major constituent of presynaptic A-type channels that modulate synaptic transmission in CNS neurons. Here, we show that Kv1.1-containing channels are complexed with Lgi1, the functionally unassigned product of the leucine-rich glioma inactivated gene 1 (LGI1), which is causative for an autosomal dominant form of lateral temporal lobe epilepsy (ADLTE). In the hippocampal formation, both Kv1.1 and Lgi1 are coassembled with Kv1.4 and Kvbeta1 in axonal terminals. In A-type channels composed of these subunits, Lgi1 selectively prevents N-type inactivation mediated by the Kvbeta1 subunit. In contrast, defective Lgi1 molecules identified in ADLTE patients fail to exert this effect resulting in channels with rapid inactivation kinetics. The results establish Lgi1 as a novel subunit of Kv1.1-associated protein complexes and suggest that changes in inactivation gating of presynaptic A-type channels may promote epileptic activity.

Over-expression of a mammalian small conductance calcium-activated K+ channel in Pichia pastoris: effects of trafficking signals and subunit fusions

Mammalian SK proteins are Ca2+-activated K+ channels, which show a sub-20 pS conductance. We have expressed the SK2 variant gene in Pichia pastoris and found protein to be produced at considerably higher levels than in brain tissue. The channel was correctly folded as evidenced by its high affinity interaction with apamin, a specific ligand from bee venom. However, the protein was largely unable to reach the plasma membrane, its normal destination, instead remaining in the endoplasmic reticulum. Adding a putative translocation sequence altered the intracellular distribution significantly with enhanced trafficking out of the endoplamic reticulum. Fusion of SK2 with the associated protein calmodulin also altered the channel localisation but in a different manner with channels now found mainly in transit between endoplasmic reticulum and Golgi compartments.

Structural characterisation of neuronal voltage-sensitive K+ channels heterologously expressed in Pichia pastoris

Neuronal voltage-dependent K(+) channels of the delayed rectifier type consist of multiple Kv alpha subunit variants, which assemble as hetero- or homotetramers, together with four Kv beta auxiliary subunits. Direct structural information on these proteins has not been forthcoming due to the difficulty in isolating the native K(+) channels. We have overexpressed the subunit genes in the yeast Pichia pastoris. The Kv1.2 subunit expressed alone is shown to fold into a native conformation as determined by high-affinity binding of 125I-labelled alpha-dendrotoxin, while co-expressed Kv1.2 and Kv beta 2 subunits co-assembled to form native-like oligomers. Sites of post-translational modifications causing apparent heterogeneity on SDS-PAGE were identified by site-directed mutagenesis. Engineering to include affinity tags and scale-up of production by fermentation allowed routine purification of milligram quantities of homo- and heteroligomeric channels. Single-particle electron microscopy of the purified channels was used to generate a 3D volume to 2.1 nm resolution. Protein domains were assigned by fitting crystal structures of related bacterial proteins. Addition of exogenous lipid followed by detergent dialysis produced well-ordered 2D crystals that exhibited mostly p12(1) symmetry. Projection maps of negatively stained crystals show the constituent molecules to be 4-fold symmetric, as expected for the octameric K(+) channel complex.

Heteromeric Kv1 potassium channel expression: amino acid determinants involved in processing and trafficking to the cell surface

Kv1.4 and Kv1.1 potassium channels are expressed in brain as mature glycoproteins that are trans-Golgi glycosylated. When expressed in cell lines these homomers had very different trans-Golgi glycosylation efficiencies and cell surface expression levels with Kv1.4 > Kv1.1 for both parameters (Zhu, J., Watanabe, I., Gomez, B., and Thornhill, W. B. (2001) J. Biol. Chem. 276, 39419-39427). This previous study identified determinants in the outer pore region of Kv1.4 and Kv1.1 that positively and negatively, respectively, affected these events when expressed as homomers. Here we investigated which subunit exhibited positive or negative effects on these processes when expressed as heteromers. Kv1.4/Kv1.1 heteromers, by coexpression or expression as tandem-linked heteromers, were expressed on the cell surface at approximately 20-fold lower levels versus Kv1.4 homomers but they were trans-Golgi glycosylated. The lower Kv1.4/Kv1.1 expression level was not rescued by Kvbeta 2.1 subunits. Thus Kv1.1 inhibited high cell surface expression and partially retained the heteromer in the endoplasmic reticulum, whereas Kv1.4 stimulated trans-Golgi glycosylation. The subunit determinants and cellular events responsible for these differences were investigated. In a Kv1.4/Kv1.1 heteromer, the Kv1.1 pore was a major negative determinant, and it inhibited high cell surface expression because it induced high partial endoplasmic reticulum retention and it decreased protein stability. Other Kv1.1 regions also inhibited high surface expression of heteromers. The Kv1.1 C terminus induced partial Golgi retention and contributed to a decreased protein stability, whereas the Kv1.1 N terminus contributed to only a decreased protein stability. Thus a neuron may regulate its cell surface K+ channel protein levels by different Kv1 subfamily homomeric and heteromeric combinations that affect intracellular retention characteristics and protein stability.

Glycosylation affects rat Kv1.1 potassium channel gating by a combined surface potential and cooperative subunit interaction mechanism

The effect of glycosylation on Kv1.l potassium channel function was investigated in mammalian cells stably transfected with Kv1.l or Kv1.1N207Q. Macroscopic current analysis showed that both channels were expressed but Kv1.1N207Q, which was not glycosylated, displayed functional differences compared with wild-type, including slowed activation kinetics, a positively shifted V 1/2, a shallower slope for the conductance versus voltage relationship, slowed C-type inactivation kinetics, and a reduced extent of and recovery from C-type inactivation. Kv1. 1N207Q activation properties were also less sensitive to divalent cations compared with those of Kv1.l. These effects were largely due to the lack of trans-Golgi added sugars, such as galactose and sialic acid, to the N207 carbohydrate tree. No apparent change in ionic current deactivation kinetics was detected inKv1.1N207Q compared with wild-type. Our data, coupled with modelling, suggested that removal of the N207 carbohydrate tree had two major effects. The first effect slowed the concerted channel transition from the last dosed state to the open state without changing the voltage dependence of its kinetics. This effect contributed to the G-V curve depolarization shift and together with the lower sensitivity to divalent cations suggested that the carbohydrate tree and its negatively charged sialic acids affected the negative surface charge density on the channel's extracellular face that was sensed by the activation gating machinery. The second effect reduced a cooperativity factor that slowed the transition from the open state to the dosed state without changing its voltage dependence. This effect accounted for the shallower G-V slope, and contributed to the depolarized G-V shift, and together with the inactivation changes it suggested that the carbohydrate tree also affected channel conformations. Thus N-glycosylation, and particularly terminal sialylation, affected Kv1.l gating properties both by altering the surface potential sensed by the channel's activation gating machinery and by modifying conformational changes regulating cooperative subunit interactions during activation and inactivation. Differences in glycosylation pattern among closely related channels may contribute to their functional differences and affect their physiological roles.

Kv channel subunits that contribute to voltage-gated K+ current in renal vascular smooth muscle

The rat renal arterial vasculature displays differences in K(+) channel current phenotypes along its length. Small arcuate to cortical radial arteries express a delayed rectifier phenotype, while the predominant Kv current in larger arcuate and interlobar arteries is composed of both transient and sustained components. We sought to determine whether Kvalpha subunits in the rat renal interlobar and arcuate arteries form heterotetramers, which may account for the unique currents, and whether modulatory Kvbeta subunits are present in renal vascular smooth muscle cells. RT-PCR indicated the presence of several different Kvalpha subunit isoform transcripts. Co-immunoprecipitation with immunoblotting and immunohistochemical evidence suggests that a portion of the K(+) current phenotype is a heteromultimer containing delayed-rectifier Kv1.2 and A-type Kv1.4 channel subunits. RT-PCR and immunoblot analyses also demonstrated the presence of both Kvbeta1.2 and Kvbeta1.3 in renal arteries. These results suggest that heteromultimeric formation of Kvalpha subunits and the presence of modulatory Kvbeta subunits are important factors in mediating Kv currents in the renal microvasculature and suggest a potentially critical role for these channel subunits in blood pressure regulation.

Semliki Forest virus vectors for rapid and high-level expression of integral membrane proteins

Semliki Forest virus (SFV) vectors have been applied for the expression of recombinant integral membrane proteins in a wide range of mammalian host cells. More than 50 G protein-coupled receptors (GPCRs), several ion channels and other types of transmembrane or membrane-associated proteins have been expressed at high levels. The establishment of large-scale SFV technology has facilitated the production of large quantities of recombinant receptors, which have then been subjected to drug screening programs and structure-function studies on purified receptors. The recent Membrane Protein Network (MePNet) structural genomics initiative, where 100 GPCRs are overexpressed from SFV vectors, will further provide new methods and technologies for expression, solubilization, purification and crystallization of GPCRs.

Calnexin co-expression and the use of weaker promoters increase the expression of correctly assembled Shaker potassium channel in insect cells

Voltage-gated potassium channels control the membrane potential of excitable cells. To understand their function, knowledge of their structure is essential. However, these channels are scarce in natural sources, and overexpression is necessary to generate material for structural studies. We have compared functional expression of the Drosophila Shaker H4 potassium channel in stable insect cell lines and in baculovirus-infected insect cells, using three different baculovirus promoters. Stable insect cell lines expressed correctly assembled channel, which was glycosylated and found predominantly at, or close to, the cell surface. In comparison, the majority of baculovirus-overexpressed Shaker was intracellular and incorrectly assembled. The proportion of functional Shaker increased, however, if the weaker basic protein promoter was used rather than the stronger p10 or polyhedrin promoters. In addition, co-expression of the molecular chaperone, calnexin, increased the quantity of correctly assembled channel protein, suggesting that calnexin can be used to increase the efficiency of channel expression in insect cells.

Characteristics of brain Kv1 channels tailored to mimic native counterparts by tandem linkage of alpha subunits: implications for K+ channelopathies

Most neuronal Kv1 channels contain Kv1.1, Kv1.2 alpha, and Kvbeta2.1 subunits, yet the influences of their stoichiometries on properties of the (alpha)(4)(beta)(4) variants remain undefined. cDNAs were engineered to contain 0, 1, 2, or 4 copies of Kv1.1 with the requisite number of Kv1.2 and co-expressed in mammalian cells with Kvbeta2.1 to achieve "native-like" hetero-oligomers. The monomeric (Kv1.1 or 1.2), dimeric (Kv1.1-1.2 or 1.2-1.2), and tetrameric (Kv1.1-(1.2)(3)) constructs produced proteins of M(r) approximately 62,000, 120,000, and 240,000, which assembled into (alpha)(4)(beta)(4) complexes. Each alpha cRNA yielded a distinct K(+) current in oocytes, with voltage dependence of activation being shifted negatively as the Kv1.1 content in tetramers was increased. Channels containing 1, 2, or 4 copies of Kv1.1 were blocked by dendrotoxin k (DTX)(k) with similarly high potencies, whereas Kv(1.2)(4) proved nonsusceptible. Accordingly, Kv1.2/beta2.1 expressed in baby hamster kidney cells failed to bind DTX(k); in contrast, oligomers containing only one Kv1.1 subunit in a tetramer exhibited high affinity, with additional copies causing modest increases. Thus, one Kv1.1 subunit largely confers high affinity for DTX(k), whereas channel electrophysiological properties are tailored by the content of Kv1.1 relative to Kv1.2. This notable advance could explain the diversity of symptoms of human episodic ataxia I, which is often accompanied by myokymia, due to mutated Kv1.1 being assembled in different combinations with wild-type and Kv1.2.

Episodic ataxia type-1 mutations in the Kv1.1 potassium channel display distinct folding and intracellular trafficking properties

Episodic ataxia type 1 (EA-1) is a neurological disorder arising from mutations in the Kv1.1 potassium channel alpha-subunit. EA-1 patients exhibit substantial phenotypic variability resulting from at least 14 distinct EA-1 point mutations. We found that EA-1 missense mutations generate mutant Kv1.1 subunits with folding and intracellular trafficking properties indistinguishable from wild-type Kv1.1. However, the single identified EA-1 nonsense mutation exhibits intracellular aggregation and detergent insolubility. This phenotype can be transferred to co-assembled Kv1 alpha- and Kv beta-subunits associated with Kv1.1 in neurons. These results suggest that as in many neurodegenerative disorders, intracellular aggregation of misfolded Kv1.1-containing channels may contribute to the pathophysiology of EA-1.

Determinants involved in Kv1 potassium channel folding in the endoplasmic reticulum, glycosylation in the Golgi, and cell surface expression

Kv1.1 and Kv1.4 potassium channels are expressed as mature glycosylated proteins in brain, whereas they exhibited striking differences in degree of trans-Golgi glycosylation conversion and high cell surface expression when they were transiently expressed as homomers in cell lines. Kv1.4 exhibited a 70% trans-Golgi glycosylation conversion, whereas Kv1.1 showed none, and Kv1.4 exhibited a approximately 20-fold higher cell surface expression level as compared with Kv1.1. Chimeras between Kv1.4 and Kv1.1 and site-directed mutants were constructed to identify amino acid determinants that affected these processes. Truncating the cytoplasmic C terminus of Kv1.4 inhibited its trans-Golgi glycosylation and high cell surface expression (as shown by Li, D., Takimoto, K., and Levitan, E. S. (2000) J. Biol. Chem. 275, 11597-11602), whereas truncating this region on Kv1.1 did not affect either of these events, indicating that its C terminus is not a negative determinant for these processes. Exchanging the C terminus between these channels showed that there are other regions of the protein that exert a positive or negative effect on these processes. Chimeric constructs between Kv1.4 and Kv1.1 identified their outer pore regions as major positive and negative determinants, respectively, for both trans-Golgi glycosylation and cell surface expression. Site-directed mutagenesis identified a number of amino acids in the pore region that are involved in these processes. These data suggest that there are multiple positive and negative determinants on both Kv1.4 and Kv1.1 that affect channel folding, trans-Golgi glycosylation conversion, and cell surface expression.

Twenty years of dendrotoxins

Dendrotoxins are small proteins that were isolated 20 years ago from mamba (Dendroaspis) snake venoms (Harvey, A.L., Karlsson, E., 1980. Dendrotoxin from the venom of the green mamba, Dendroaspis angusticeps: a neurotoxin that enhances acetylcholine release at neuromuscular junctions. Naunyn-Schmiedebergs Arch. Pharmacol. 312, 1-6.). Subsequently, a family of related proteins was found in mamba venoms and shown to be homologous to Kunitz-type serine protease inhibitors, such as aprotinin. The dendrotoxins contain 57-60 amino acid residues cross-linked by three disulphide bridges. The dendrotoxins have little or no anti-protease activity, but they were demonstrated to block particular subtypes of voltage-dependent potassium channels in neurons. Studies with cloned K(+) channels indicate that alpha-dendrotoxin from green mamba Dendroaspis angusticeps blocks Kv1.1, Kv1.2 and Kv1.6 channels in the nanomolar range, whereas toxin K from the black mamba Dendroaspis polylepis preferentially blocks Kv1.1 channels. Structural analogues of dendrotoxins have helped to define the molecular recognition properties of different types of K(+) channels, and radiolabelled dendrotoxins have also been useful in helping to discover toxins from other sources that bind to K(+) channels. Because dendrotoxins are useful markers of subtypes of K(+) channels in vivo, dendrotoxins have become widely used as probes for studying the function of K(+) channels in physiology and pathophysiology.