• voltage-gated;
  • K+ channels;
  • α and β subunits;
  • hetero-oligomers;
  • dendrotoxins.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgement
  7. References

K+ channels from the Kv1 subfamily contain four α-subunits and the combinations (from Kv1.1–1.6) determine susceptibility to dendrotoxin (DTX) homologues. The subunit composition of certain subtypes in rat brain was investigated using DTXk which only interacts with Kv1.1-containing channels and αDTX (and its closely related homologue DTXi) that binds preferentially to Kv1.2-possessing homo- or hetero-oligomers. Covalent attachment of [125I]DTXk bound to channels in synaptic membranes unveiled subunits of Mr = 78 000 and 96 000. Immunoprecipitation of these solubilized and dissociated cross-linked proteins with IgG specific for each of the α-subunits identified Kv1.1, 1.2 and 1.4; this led to assemblies of Kv1.1/1.2 and 1.1/1.4 being established. Kv1.2-enriched channels, purified from rat brain by chromatography on immobilized DTXi, contained Kv1.1, 1.2 and 1.6 confirming one of the above-noted pairs and indicating an additional Kv1.1-containing oligomer (Kv1.1/1.2/1.6); the notable lack of Kv1.4 excludes a Kv1.1/1.2/1.4 combination. On the other hand, channels with Kv1.1 as a constituent, isolated using DTXk, possessed Kv1.4 in addition to those found in the DTXi-purified oligomers; this provides convergent support for the occurrence of the three combinations established above but adds a possible fourth (Kv1.1/1.4/1.6), though this was not confirmed. Moreover, sequential purification on DTXi and DTXk resins yielded channels containing only Kv1.1/1.2 but with an apparent predominance of Kv1.1, reaffirming the latter multimer.




voltage-gated K+ channels


K+ channels isolated on DTXi and DTXk resin, respectively.

Voltage-gated K+ channels (Kv) act as key determinants of membrane excitability, regulating the duration of action potentials, firing frequencies and interspike intervals [1]. Kv channels in the brain exhibit a high degree of diversity in their physiological and pharmacological properties due to the expression of several subfamilies (Kv1–4) of these membrane proteins (reviewed in [2–6]). The Shaker-related group (Kv1) is the most intensively studied; its six major members (Kv1.1–1.6) have been well characterized individually at the structural and functional levels. Different combinations of these α-subunit isoforms can coassemble into hetero-multimeric channels, known to contain four α- and four β-subunits [7,8]. Although substantial progress has been made in molecular and electrophysiological studies of cloned Kv1 subunits, little is known about how the biophysical characteristics of native K+ currents relate to the subunit compositions of the requisite channel oligomers. This is due to the considerable overlap in functional properties between the cloned Kv members [6] and the unknown subunit stoichiometry of channels at different neuronal locations [5]. Several human disorders such as epilepsy, cerebral anoxia and long QT syndrome are associated with K+ channels [9]. In particular, Kv1.1 has been found to be involved with seizures and episodic ataxia [10–12], and the dominant negative effect of the mutated Kv1.1-subunit [12] highlights the functional importance of Kv1.1-containing channels. Because of such medical relevance, it is of particular importance to establish the macromolecular properties of these native K+ channels.

The approaches taken to date towards understanding the neurobiological functions of distinct Kv1 channel subunits (or combinations) have included mapping the expression pattern of subunit mRNAs [13], immunocytochemical localization of the various subunits [14–20] and labelling the channels in the nervous system with radio-iodinated dendrotoxin (DTX) homologues [21–24]. These studies showed that different neuronal cell types can express distinct patterns of Kv1 channel subunits. Although some subunits were found to be co-localized, knowledge of the combinations in oligomers is limited due to a lack of direct biochemical evidence for their coassociation. Recently, sequential immunoprecipitation with specific anti-Kv1 antibodies [25] identified conclusively one fully defined tetramer Kv1.2/1.3/1.4/1.6, and several other possible multimeric channels in synaptic membranes from bovine cerebral cortex. Also, specific K+-channel-binding toxins have helped to establish the heterogeneity of these channels. Use of αDTX and DTXi (reviewed in [5,26–28]), which block Kv1.2 with higher potency than Kv1.1 or 1.6 while being ineffective towards Kv1.3 and 1.4 [29,30], identified a family of hetero-tetrameric K+ channels in bovine cerebral cortex which all appear to contain Kv1.2, ≈ 50% possess Kv1.1, with Kv1.4 and 1.6 being minor components [31]. [125I]-labelled margotoxin, which binds to Kv1.2 and 1.3 with KD = 0.08 pm, was used to establish the subunit composition of K+ channels in rat cerebellum and gave similar results except for a somewhat higher content of Kv1.1 in Kv1.2-containing channels [32].

Herein, a complementary strategy exploiting the specificity of DTXk for the Kv1.1-subunit (all other Kv1.X members being virtually insensitive) [33,34] was employed to identify other partners with which it is coassembled in the α-subunit tetramers found in rat whole brain. [125I]DTXk was covalently attached to its bound channels in rat synaptic membranes, followed by solubilization and immunoprecipitation of the dissociated labelled subunits; additionally, affinity chromatography on immobilized DTXk and DTXi resins allowed purification of K+ channel subtypes. These procedures unveiled two types of Kv1.1-containing channels: Kv1.1/1.4, and Kv1.1/1.2 coassembled with or without Kv1.6.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgement
  7. References

Cross-linking of [125I]DTXk to K+ channels in synaptic membranes and immuno-precipitation of the dissociated individual α-subunits

Covalent attachment of [125I]DTXk bound to K+ channels in membranes from rat whole brain was performed by the method of Black and Dolly [35]. An aliquot of synaptic membranes (200 µg) was incubated in 250 µL with 3 nm[125I]DTXk (prepared as outlined below) in 0.25 m triethanolamine HCl, pH 8.5 containing 90 mm NaCl and 5 mm KCl, in the presence and absence of specified concentrations of unlabelled DTXk or αDTX. After incubation for 1 h at room temperature, dimethylsuberimidate was added to a final concentration of 1 mg·mL−1 and the incubation continued for 1 h. To terminate the cross-linking, the mixtures were centrifuged at 10 000 g for 2 min and the pelleted membranes washed three times with 0.5 m Tris/HCl, pH 6.8. Finally, the samples were dissolved in SDS extraction buffer (2% SDS in 0.25 m Tris/HCl, pH 6.8 supplemented with 5% (v/v) 2-mercaptoethanol [36].

For immunoprecipitation, the solubilized cross-linked samples were diluted 20-fold with NaCl/Pi (10 mm Na2HPO4, pH 7.4, 150 mm NaCl) containing 5 mm EDTA, 5 mm EGTA and 1% Triton X-100, followed by centrifugation at 100 000 g for 45 min. The resultant concentration of 0.1% SDS is known not to interfere with antibody activity [17], and the final concentration (0.25%) of mercaptoethanol used prevents aggregation of the dissociated subunits. IgG reactive with Kv1.1, 1.2, 1.3 or 1.4, whose specificities have been established [25], were incubated at 10 µg·mL−1 with each of the solubilized samples (1 mL) overnight at 4 °C, followed by reaction with 50 µL protein A-, for anti-Kv1.1, 1.3 and 1.4, or protein G-, for anti-Kv1.2, agarose at 4 °C for 4 h. The resin was sedimented, washed three times with NaCl/Pi supplemented with 0.5 m NaCl, once with H2O, and then dissolved in SDS sample buffer. The samples were subjected to SDS/PAGE using 10% gels [25]; after Coomassie staining, the gels were destained and dried prior to exposure to X-ray film.

Purification of K+ channels on immobilized DTXk and DTXi

Synaptic membranes, isolated from rat brain by gradient centrifugation [37], were solubilized [25,38] (10 mg·mL−1) with 4% (v/v) Thesit in 62.5 mm imidazole, pH 7.4 containing 250 mm KCl, 2.5 mm EDTA, 25 µg·mL−1 soybean trypsin inhibitor, 50 µg·mL−1 bacitracin, 0.25 mm benzamidine and 0.5 mm phenylmethanesulfonyl fluoride (extraction buffer). The solubilized extract was obtained by centrifugation at 30 000 g for 20 min and 100 000 g for 45 min at 4 °C, after a 2.5-fold dilution of the mixture with water.

DTXk and DTXi affinity resins were prepared by the method of Parcej and Dolly [39], at a concentration of 0.3 mg·mL−1 resin. Each toxin coupled Sepharose resin (15 mL) was poured into 50-mL columns, equilibrated with 50 mL of 2.5-fold diluted extraction buffer and the solubilized membrane extract (400 mL) was loaded at a rate of 20 mL·h−1. After washing the columns for 1–2 h with buffer A (25 mm imidazole, pH 8.1, 100 mm KCl, 1 mm EDTA, 0.5 mm benzamidine, 0.2 mm phenylmethanesulfonyl fluoride) supplemented with 0.1% (w/v) Tween 80 (buffer B) at 300 mL·h−1, the bound K+ channels were selectively eluted with 120 mL buffer B containing 10 mm dithiothreitol at a rate of 60 mL·h−1, directly onto a 2-mL DEAE Sepharose column which had been equilibrated with buffer B. The DEAE column was then washed with 30 mL buffer B at 60 mL·h−1, followed by 15 mL buffer C [buffer A and 0.5% (w/v) octylglucoside]. The bound K+ channels were eluted with 15 mL buffer D (buffer C and 0.3 m KCl) at a rate of 20 mL·h−1 and, after adding 0.1% (v/v) Tween 80 to the eluate, dialysed overnight at 4° against buffer A. For further fractionation on another toxin resin, the samples were diluted fivefold with buffer B supplemented with 1% (w/v) BSA, and processed exactly as above. After concentration using Centricon 100 microconcentrator (Amicon), the samples were subjected to SDS/PAGE and silver stained or blotted as detailed elsewhere [25] using 1–3 µg·mL−1 of mouse or rabbit purified IgG reactive with Kv1.1, 1.2, 1.3, 1.4 or 1.6, or Kv β2.1-subunits [25]. Immunoreactivity was detected using anti-species IgG coupled with horseradish peroxidase, in combination with the ECL system (Amersham).

Binding of [125I]labelled DTXk and αDTX to the solubilized and purified K+ channels

A rapid gel filtration assay [38] was employed for measuring the binding of [125I]DTXk or [125I]αDTX [40] to the two solubilized and purified K+ channel preparations. An aliquot of each (50 µL) was incubated for 1 h at room temperature with 3 nm[125I]labelled toxin (final concentration) and various amounts of unlabelled toxin in a buffer containing 50 mm imidazole, pH 7.4, 90 mm NaCl and 5 mm KCl in a total volume of 250 µL. A sample (200 µL) of each incubation mixture was then loaded onto a 2-mL Sephadex G-100 column which had been equilibrated with 2 mL of buffer containing 25 mm imidazole, pH 7.4, 50 mm NaCl, 10 mm KCl, and centrifuged for 2 min at 100 g. The void volume containing the radioactive toxin bound to K+ channels was collected and quantified in a γ-counter. Data were analyzed by using the grafit (Erithacus Software Ltd) and ligand[41] programs. Assays were performed in triplicate and repeated at least three times; average values (± SD) are presented.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgement
  7. References

Cross-linking of [125I]DTXk to K+ channels in the rat synaptic membranes

As αDTX and charybdotoxin bind to the extended pore region of K+ channels formed by four α-subunits [42–44] in a bimolecular fashion, i.e. only one toxin binds to a channel oligomer [8], it can be assumed that the bound toxin molecule would become covalently attached to any of the constituent subunits (see Discussion). Cross-linking by dimethylsuberimidate of [125I]DTXk bound to K+ channels in rat brain membranes, followed by SDS/PAGE/autoradiography, gave two recognizable bands with relative molecular masses of 85 000 and 103 000 (Fig. 1), together with additional material at the top of the gel that arose from intersubunit cross-linking, as observed previously using [125I]αDTX [37]. Most of the labelling resulted from saturable binding of [125I]DTXk because the band intensities were diminished by an excess of unlabelled DTXk; in contrast, the cross-linking of other weaker bands remained unchanged, indicative of these arising from nonsaturable interaction (Fig. 1). Use of higher dimethylsuberimidate concentrations yielded an increase in the density of the two bands specified above but without any significant change in their relative intensities (data not shown), establishing that the larger component did not arise from the covalent attachment of multiple toxin molecules to one subunit. After deducting 7000 for the size of the single bound toxin, the relative molecular masses of these two proteins are 78 000 and 96 000, respectively. It is notable that labelling of both proteins was equally attenuated by DTXk (Fig. 1), in a concentration range approximating to that required for inhibition of [125I]DTXk binding. In contrast, αDTX inhibited cross-linking of the pair of subunits with different potencies (Fig. 1); the faster migrating band was blocked with relatively low amounts of αDTX, whilst the highest concentration tested caused only partial antagonism of labelling of the larger component. These findings indicate the existence of at least two subpopulations of Kv1.1-containing K+ channels that exhibit dissimilar affinities for αDTX.


Figure 1. Covalent attachment of[125I]DTXkto its binding proteins in rat synaptic membranes: inhibition by DTXkand αDTX. Cross-linking of [125I]DTXk-K+ channel complexes with dimethylsuberimidate, in the absence or presence of nonradioactive DTXk or αDTX was carried out, as detailed in Materials and methods. SDS/PAGE in 10% gels, followed by autoradiography, showed two radioactive bands (arrows) with molecular masses of 103 000 and 85 000. DTXk blocked labelling of both of these with equal affinity whereas αDTX proved less effective in preventing cross-linking of the upper band despite displaying similar potency to DTXk in blocking the labelling of the faster migrating subunit. The radioactive diffuse band at the top of the gels is due to intersubunit cross-linking as well as some nonspecifically labelled material.

Immunoprecipitation of K+ channel α-subunits labelled by [125I]DTXk

To identify the subunits to which [125I]DTXk had become attached, the complexes were solubilized from the membrane and dissociated into their constituents; after dilution of the detergent, the extract was incubated with antibodies specific for each Kv1 family member. The lack of precipitation by anti-Kv1.3 (Fig. 2) indicates a low abundance of Kv1.3 subunit in rat brain and/or its absence from the channels labelled by [125I]DTXk. As the toxin only binds to Kv1.1, the presence of this radiolabelled subunit (Mr ≈ 85 000) in the anti-Kv1.1 antibody precipitate was expected; the detection of a larger band corresponding to the Kv1.4 α-subunit (Fig. 2) arose from the known cross-reactivity of this IgG preparation with Kv1.4 [25], whilst the broad band seen near the top of the gel represents intersubunit cross-linked material. Anti-Kv1.4 precipitated a band of relative molecular mass 103 000, whereas anti-Kv1.2 sedimented a subunit with a molecular mass of 85 000 (Fig. 2); after allowing for the contribution of the bound toxin, the sizes of these bands correspond to Kv1.4 and 1.2 subunits seen in mammalian brain ([31] and derived above). These findings revealed two possible combinations of α-subunits, Kv1.1/1.2 and 1.1/1.4 (a third assembly, Kv1.1/1.2/1.4 was later excluded), with the presence of Kv1.1 being essential for binding of DTXk; also, the existence of a Kv1.1 homo-tetramer was not excluded. Similar analysis of membranes whose labelling had been inhibited incompletely by 300 nmαDTX gave a different, but simpler, pattern of subunits; strikingly, no Kv1.2 was precipitated by anti-Kv1.2 antibody but Kv1.1 and 1.4 were pelleted by their respective IgGs (data not shown). This provides evidence for the existence of a subpopulation of Kv1.1-containing channels possessing Kv1.4 but apparently lacking Kv1.2; αDTX displays low affinity for this subtype, being unable to completely block [125I]DTXk binding.


Figure 2. Immuno-identification of α-subunits cross-linked to[125I]DTXkin intact oligomeric K+channels from rat brain. Cross-linking of [125I]DTXk to synaptic membranes, was performed as outlined in Fig. 1; after solubilization and dissociation into individual subunits, the resultant sample was incubated in the absence (–) or presence of antibodies specific for Kv1.1, 1.2, 1.3 or 1.4, prior to subjecting the pellets to SDS/PAGE/autoradiography. Anti-Kv1.2 and -Kv1.4 IgGs each precipitated a single band (arrows), whereas the coprecipitation of Kv1.4 with Kv1.1 (same size as Kv1.2) resulted from the known cross-reactivity of anti-Kv1.1 IgG towards the Kv1.4-subunit. Note that the large molecular mass material at the top of the gel (compare with Fig. 1) was not dissociated under the conditions used.

Characterization of DTXk and αDTX binding to solubilized synaptic membranes

Although conditions had been established for measurement of [125I]αDTX binding to a detergent extract of synaptic membranes [38,39], these did not prove successful for assaying the interaction of [125I]DTXk with K+ channels. This necessitated the inclusion in the standard buffer (50 mm imidazole/HCl, pH 7.4) of 90 mm NaCl and 5 mm KCl during the incubation of the extract with [125I]DTXk, and addition of 50 mm NaCl/10 mm KCl to the buffer for equilibrating the sizing column used to separate the channel–toxin complex. With this optimization (detailed in Materials and methods), the binding of [125I]DTXk observed was displaceable by DTXk(Fig. 3A), yielding a biphasic curve with a Ki value of 0.3 nm (± 0.07; n = 3) for the higher-affinity site. Although it proved difficult to accurately measure the Ki for the low-affinity inhibition of [125I]DTXk binding to the extract, the estimate (Fig. 3, legend) is comparable to that reported for synaptic membranes [45]. Under these conditions, a single set of high-affinity sites was seen for [125I]αDTX with a Ki of 0.61 nm (± 0.05; n = 3) (Fig. 3B) very similar to that reported previously [38], indicating that these conditions did not alter αDTX binding. An extended competition curve was observed for DTXk antagonism of [125I]αDTX binding; this was best fitted by a two-site model, giving Ki values of 0.57 nm (± 0.14) and 140 nm (± 49) (n = 3). Clearly DTXk, unlike αDTX, is capable of binding discriminatively to subpopulations of K+ channels presumably due to their different contents of Kv1.1.


Figure 3. DTXkcompetes with[125I]DTXkor[125I]αDTX for binding to solubilized K+channels. DTXk (•) showed biphasic inhibition of both [125I]DTXk (a) and [125I]αDTX (b) binding. The Ki values of DTXk derived from competition for [125I]DTXk binding were 0.3 nm (± 0.07) and 199 nm (± 115), while 0.57 nm (± 0.14) and 140 nm (± 49) were obtained for its antagonism of [125I]αDTX binding. A Ki value of 0.61 nm (± 0.05) was calculated for αDTX (□) inhibiting [125I]αDTX binding; note the less extended curve than those for DTXk. The data presented (± SD; n = 3) represents the fractional binding of radioiodinated toxin in the absence (B0) and presence of competitor (B).

Fractionation of K+ channel oligomers from rat brain by affinity chromatography

DTXi has been used to purify a family of K+ channels that all contain Kv1.2 [31], presumably due to its higher affinity for this subunit than Kv1.1 or 1.6 (1.3 plus 1.4 are insensitive [29,30]). As DTXk exclusively binds to the Kv1.1 subunit, utilization of DTXk affinity resin should isolate channels that contain Kv1.1 and partners. The oligomers purified on DTXk (PKK) and DTXi (PKI) resins revealed two protein bands with relative molecular masses of 78 000 and 39 000 (Fig. 4), corresponding in size to the major α-subunits (Kv1.1 and 1.2 which comigrate [25,31]) and β2 polypeptide of K+ channels, respectively; an absence of other protein bands established the purity of these preparations and the utility of the isolation procedure. The larger band seen for the PKI sample represented mainly Kv1.2 whereas in the PKK preparation the predominant constituent was Kv1.1 (see below and Fig. 4 legend). Western blotting of the PKK sample identified Kv1.1, 1.2, 1.4 and 1.6 (Fig. 4), all of which showed the expected molecular mass values [25,31]; note that the latter two subunits were not readily detectable by protein staining presumably due to their lower content, as observed previously [46]. A notable lack of staining with anti-Kv1.3 established that Kv1.3 subunit is absent from these Kv1.1-containing channels (data not shown). The above identification of Kv1.6 extends the subunit combinations, derived initially from toxin cross-linking experiments, by the nominal addition to the pair of deduced oligomers (1.1/1.2 and 1.1/1.4) to include: Kv1.1/1.2/1.6 and 1.1/1.4/1.6; due to lack of evidence, other possible oligomers (Kv1.1/1.6, 1.1/1.2/1.4, 1.1/1.2/1.4/1.6 or Kv1.1 homo-oligomers) are not considered further. PKI contained Kv1.1, 1.2 and 1.6 subunits but not Kv1.4 (Fig. 4), leading to four possible combinations of Kv1.2 with 1.1 or 1.6, or both, and Kv1.2 homo-oligomers but only those containing Kv1.1 are considered in detail because of the focus on these subtypes. Estimation of the relative contents of α-subunits (by the semiquantitative method described in Fig. 4 legend) showed that Kv1.2 was the prominent protein in PKI with very much smaller quantities of Kv1.1, and the reverse was true for PKK; it should be noted that the determination of the stoichiometries of the subunits in different oligomers is a difficult task that remains to be accomplished. Both PKI and PKK preparations were additionally found to contain the β2.1-subunit (Fig. 4). In attempts to further fractionate these subtypes, PKI channels were loaded onto a DTXk column whilst DTXi-resin was used for chromatographic separation of the PKK preparation. PKI channels were indeed fractionated on the DTXk resin, with the resultant eluant (PKK-I) being found to contain Kv1.1 and 1.2 subunits (Fig. 4) but no detectable Kv1.6 (or Kv1.4 as expected from above). However, chromatography of the PKK channels on the DTXi resin proved unsuccessful (not shown), possibly due to the low proportion of channel oligomers that contain Kv1.2 subunit, as was reflected by the minimal inhibition by αDTX of [125I]DTXk binding to the PKK channels (shown below). Thus, two Kv1.2/1.1 oligomers, distinguished by the presence or absence of Kv1.6, were isolated on each of the toxin resins; this corroborating evidence established conclusively their existence in rat brain.


Figure 4. K+channel subtypes fractionated from rat brain by affinity chromatography on DTXiand DTXkresins. K+ channels were purified from rat brain on immobilized DTXk (▴) or DTXi (▪) and termed PKK and PKI, respectively. Note the α-subunit band stained weakly with silver, as described previously [7], and contains Kv1.1 and 1.2 which comigrated. Other less prevalent subunits (e.g. Kv1.4 and 1.6), detected by sensitive Western blotting, could not be visualized following protein staining, apparently due to their lower content, as reported elsewhere [46]. Western blotting established that Kv1.1, 1.2 and 1.6 subunits occur in both the PKK and PKI preparations, but only the PKK channels (▴) contained Kv1.4. The PKI channels fractionated further on DTXk affinity resin (•) showed Kv1.1 and 1.2 but no detectable Kv1.4 or 1.6 subunits. β2.1 (Mr ≈ 39 000) was found in PKK and PKI. The blots shown were developed for different times to reveal subunits present in low amounts; however, when they were processed for the same period, Kv1.2 showed up as the major constituent of the PKI whereas Kv1.1 predominated in the PKK. All the α-subunit bands are aligned for ease of presentation; relative molecular mass values calculated from the mobilities of standard proteins (not shown) are 78 000, 95 000 and 58 000 for Kv1.1 and 1.2, 1.4 and 1.6, in agreement with published data [25,31].

Binding of [125I]DTXk to the purified channel subtypes

[125I]DTXk binding to the channels eluted from immobilized DTXk and DTXi, representing Kv1.1- and Kv1.2-enriched channels, were measured in competition experiments (Fig. 5). It appears that DTXk binds equally well to PKK (Fig. 5A) and PKI (Fig. 5B) channels; in each case, the extended curves obtained are best fitted by a two-site model with calculated IC50 values of 2.3 ± 1.5 and 30 ± 16 nm, and 2.7 ± 2 and 23 ± 12 nm (n = 3), respectively. The comparable affinity of DTXk for the PKI and PKK samples is attributable to its specific binding to channels that contain the Kv1.1 subunit, while the two affinities seen with each are likely to be due to the different copy numbers of Kv1.1 in some of these channel oligomers. In contrast to DTXk, αDTX gave only partial inhibition of [125I]DTXk binding to both PKK and PKI channels at the maximal concentration (1 µm) that was practical to use, reflecting relatively weak interactions; this reduced affinity was more pronounced for PKK (Fig. 5).


Figure 5. DTXk and αDTX antagonize [125I]DTXk binding to the K+ channels purified on immobilized DTXk or DTXi. The curves obtained for competition by DTXk (•) of [125I]DTXk binding to both purified preparations [PKK(a), PKI(b)] were best fitted by a two-site model which gave IC50 values 2.3 ± 1.5 nm and 30 ± 16 nm for PKK, and 2.7 ± 2 nm and 23 ± 12 nm. The relatively weak and incomplete inhibition observed for αDTX precluded determination of IC50 values but a higher affinity for PKI than PKK was apparent.

In summary, these various experimental approaches identified conclusively two types of Kv1.1-containing channels in rat brain, typified by the presence of Kv1.2 (type 1) or Kv1.4 (type 2); also, evidence was provided for type 1 having at least two members distinguished by whether or not Kv1.6 is a constituent.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgement
  7. References

Kv1.1 and 1.2 are the most abundant subunits [31], with their mRNAs being about of equal abundance in the adult nervous system [13]; hence, elucidation of the subunit combinations of K+ channels possessing one or both of these subunits would cover most Kv1 members. For this purpose, we exploited the exclusive interaction of DTXk with oligomers that possess Kv1.1, and preferential binding of DTXi to K+ channels containing the Kv1.2 subunit. It must be emphasized that only channels in which Kv1.1 is a constituent were analysed here and, thus, the occurrence of others (e.g. homo-oligomers of Kv1.2 or 1.4) are not discussed.

Type 1 K+ channels were first defined (Kv1.1/1.2) by cross-linking of [125I]DTXk. Analysis of the material purified on DTXk resin suggests they possibly occur with and without associated Kv1.6 . Chromatography on the DTXi gel confirmed the presence of Kv1.6 and the absence of Kv1.4 in the type 1 channels; furthermore, sequential fractionation on DTXi and DTXk resins established conclusively that a proportion of the DTXi-purified preparation contained only Kv1.1/1.2, and so a fraction could also possess Kv1.6. Importantly, these subunit combinations (Kv1.1/1.2, 1.1/1.2/1.6) correspond to those identified by sequential precipitation with antibodies specific for each subunit in bovine brain [25] and in rat cerebellum [32]. From the results presented herein, it seems possible to isolate at least some of the oligomeric subtypes in active form, an advance that will allow their further characterization. The striking convergence of these findings from the various experimental strategies employed consolidate such an assignment of subunit combinations for these two members of type 1 channels. Additionally, the lower efficacy observed for αDTX-antagonizing [125I]DTXk binding to PKK than PKI preparations could arise from the former showing a lower content of Kv1.2 and proportionally higher amounts of Kv1.1 and/or 1.6. This hypothetical explanation for the composition of type 1 channels is favoured because contributions of other known α-subunits to the tetramers being excluded for the following reasons: (a) an absence of Kv1.3 from the Kv1.1-containing oligomers (Fig. 2); (b) Kv1.5 is known to be lacking from brain [13]; and (c) Kv1.4 resides only in type 2 channels.

Published results on the mRNA levels and expression patterns of Kv1.1 and 1.2 proteins in brain regions support the existence of the type 1 channels. For example, the highest level of Kv1.1 and 1.2 mRNAs was seen in caudal regions of the brain, a pattern not so apparent in the case of the other Kv1 mRNAs [13]. Accordingly, distinct though extensively overlapping expression of Kv1.1 and 1.2 proteins has been observed in brain sections [16,18–20]. Therefore, the type 1 subtype identified herein, in which Kv1.1 and 1.2 are coassembled, provides a basis for their overlapping distribution. In the case of the basket cell terminal plexus that surrounds the initial segment of Purkinje cells, a high concentration of Kv1.1 and 1.2 [14–16,18–20] was found to be precisely colocalized, as also noted in the dentate gyrus of the hippocampal formation. Type 1 channels also explains the significant densities of binding sites observed for [125I]αDTX or [125I]δDTX (a close homologue of DTXk) [21] in the synapse-rich molecular layer of hippocampus and cerebellum. The combination of different copies of Kv1.1 or 1.2 in this K+ channel type, as proposed above, would create oligomers with subtly different functional properties, including susceptibilities to the toxins; in the accompanying paper [47], mutants of DTXk were used to establish the basis of the toxin’s selective interaction with Kv1.1-containing channels.

Type 2 channels were recognized because a fraction of the cross-linked [125I]DTXk, not inhibitable by 300 nmαDTX, represented labelled Kv1.4 and 1.1. Consistently, these two subunits were detected in the channels isolated by affinity chromatography on DTXk, together with Kv1.2 and 1.6. As Kv1.2 was found not to be coassociated with Kv1.4 (described above), this provides additional indirect evidence for the coexistence of Kv1.4/1.1. An absence of Kv1.2 from the type 2 channels accords with the fact that a high molecular-mass band corresponding to Kv1.4 not being labelled when channels in rat cortex synaptic membranes possessing Kv1.2 were covalently attached to radio-iodinated αDTX or DTXi[48]. This, however, is in contrast with the results obtained for bovine brain membranes, using sequential immunoprecipitation, in which Kv1.4 was found to be associated with Kv1.2; thus, two other Kv1.1-containing combinations (Kv1.1/1.2/1.4 and Kv1.1/1.2/1.4/1.6) were postulated [25], though direct evidence for their existence was lacking. In any case, if a discrepancy exists it may be due to: (a) different tissue sources, instead of bovine cortex rat whole brain being used herein – several brain regions in the rat are known to show immunostaining for Kv1.1 and 1.4 only, but not Kv1.2 [16]; (b) different experimental techniques – immunoprecipitation is based on the specificity and avidity of antibodies reactive with individual Kv1 subunits in the oligomer whilst affinity chromatography depends on the toxins’ selectivities and affinities for subtypes of whole channels. Also, the latter requires a relatively large quantity of channels; therefore, sparse species might not necessarily be purified by the toxin resins. In this regard, it is notable that, with the exclusion of Kv1.2 from the oligomers postulated for bovine cortex, the resultant combinations (Kv1.1/1.4 and Kv1.1/1.4/1.6) correspond to the two members of type 2 channels described here. Notwithstanding the fact that the channels examined do not represent all subtypes in whole brain, our failure to detect Kv1.1/1.2/1.4 and Kv1.1/1.2/1.4/1.6 in either of the purified preparations is indicative of their low abundance or weak affinity for DTXk and DTXi.

The case presented here for the occurrence in rat brain of Kv1.4/1.1 channels is strengthened considerably by immuno-cytochemical evidence for their colocalization in certain brain regions [16]. For example, double staining of Kv1.4 and 1.1 has been observed in the middle sector of the molecular layer of the hippocampus. In the CA3 subfield, there was prominent immuno-reactivity for both in a narrow zone immediately above the striatum pyramidale; furthermore, the observation of a pronounced zone of Kv1.1 staining matched by a similar but less intense Kv1.4 labelling may have arisen from Kv1.4/1.1 protomers with a larger number of copies of Kv1.1. It is important to note that the presence of Kv1.4 in these type 2 channels could contribute N-type inactivation via its inactivation ball [30]; this may create fast transient A-type K+ currents in contrast to the delayed-rectified properties predicted for the type 1 variety. Finally, no direct experimental data could be obtained to ascertain if Kv1.6 assembles with Kv1.4/1.1, a possibility created by this subunit being found with the latter in the PKK preparation. Nevertheless, the existence of this second member of type 2 channels gains support from the pattern of immuno-reactivity reported for Kv1.4/1.1/1.6 in the neurophil of the molecular layer of the cerebellar cortex [16]. Finally, evidence could not be gained to confirm or exclude the existence of Kv1.1/1.6 or Kv1.1 homo-oligomers, though the latter was not detected in bovine synaptic membranes [25]. With respect to the β subunits, Kvβ2.1 is the most abundant isoform [25,49] and was found in both PKI and PKK channel preparations (Fig. 4); therefore, these Kv1.1-containing channels can coassemble with β 2.1-subunit.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgement
  7. References

This work was financed by a Medical Research Council grant (to J. O. D)


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgement
  7. References
  • 1
    Hille, B. (1992) Ionic Channels of Excitable Membranes. Sinauer Assoc, Sunderland, MA.
  • 2
    Jan, L.Y. & Jan, Y.N. (1992) Structural elements involved in specific K+ channel functions. Ann. Rev. Physiol. 54, 535555.
  • 3
    Jan, L.Y. & Jan, Y.N. (1997) Cloned potassium channels from eukaryotes and prokaryotes. Ann. Rev. Neurosci. 20, 91123.
  • 4
    Pongs, O. (1992) Structural basis of voltage-gated K+ channel pharmacology. Trends Pharmacol. Sci. 13, 359365.
  • 5
    Dolly, J.O. & Parcej, D.N. (1996) Molecular properties of voltage-gated K+ channels. J. Bioenerget. Biomembranes 28, 231253.
  • 6
    Chandy, K.G. & Gutman, G.A. (1995) Voltage-gated potassium channel genes. In Handbook of Receptors and Channels: Ligand and Voltage-Gated Ion Channels (North, A., ed.,), pp. 171. CRC Press, Boca-Raton, FL.
  • 7
    Parcej, D.N., Scott, V.E.S. & Dolly, J.O. (1992) Oligomeric properties of alpha-dendrotoxin-sensitive potassium ion channels purified from bovine brain. Biochemistry 31, 1108411088.
  • 8
    MacKinnon, R. (1991) Determination of the subunit stoichiometry of a voltage-activated potassium channel. Nature 350, 232235.
  • 9
    Goldstein, S.A.N. & Colatsky, T.J. (1996) Ion channels: too complex for rational drug design? Neuron 16, 913919.
  • 10
    Browne, D.L., Gancher, S.T., Nutt, J.G., Brunt, E.R., Smith, E.A., Kramer, P. & Litt, M. (1994) Episodic ataxia/myokymia syndrome is associated with point mutations in the human potassium channel gene, KCNA1. Nat. Genet. 8, 136140.
  • 11
    Browne, D.L., Brunt, E.R., Griggs, R.C., Nutt, J.G., Gancher, S.T., Smith, E.A. & Litt, M. (1995) Identification of two new KCNA1 mutations in episodic ataxia/myokymia families. Hum. Mol. Genet. 4, 16711672.
  • 12
    Smart, S.L., Lopantsev, V., Zhang, C.L., Robbins, C.A., Wang, H., Chiu, S.Y., Schwartzkroin, P.A., Messing, A. & Tempel, B.L. (1998) Deletion of the Kv1.1 potassium channel causes epilepsy in mice. Neuron 20, 809819.
  • 13
    Beckh, S. & Pongs, O. (1990) Members of the RCK potassium channel family are differentially expressed in the rat nervous system. EMBO J. 9, 777782.
  • 14
    McNamara, N.M.C., Muniz, Z.M., Wilkin, G.P. & Dolly, J.O. (1993) Prominent location of a K+ channel containing the α subunit Kv1.2 in the basket cell nerve terminals of rat cerebellum. Neuroscience 57, 10391045.
  • 15
    McNamara, N.M.C., Averill, S., Wilkin, G.P., Dolly, J.O. & Priestley, J.V. (1996) Ultrastructural localization of a voltage-gated K+ channel alpha subunit (Kv1.2) in the rat cerebellum. Eur. J. Neurosci. 8, 688699.
  • 16
    Rhodes, K.J., Strassle, B.W., Monaghan, M.M., Bekelearcuri, Z., Matos, M.F. & Trimmer, J.S. (1997) Association and colocalization of the Kv beta 1 and Kv beta 2 beta-subunits with Kv1 alpha-subunits in mammalian brain K+ channel complexes. J. Neurosci. 17, 82468258.
  • 17
    Sheng, M., Liao, Y.J., Jan, Y.N. & Jan, L.Y. (1993) Presynaptic A-current based on heteromultimeric K+ channels detected in vivo. Nature 365, 7275.
  • 18
    Veh, R.W., Lichtinghagen, R., Sewing, S., Wunder, F., Grumbach, I.M. & Pongs, O. (1995) Immunohistochemical localization of 5 members of the Kv1 channel subunits – contrasting subcellular locations and neuron-specific co-localizations in rat brain. Eur. J. Neurosci. 7, 21892205.
  • 19
    Wang, H., Kunkel, D.D., Martin, T.M., Schwartzkroin, P.A. & Tempel, B.L. (1993) Heteromultimeric K+ channels in terminal and juxtaparanodal regions of neurons. Nature 365, 7579.
  • 20
    Wang, H., Kunkel, D.D., Schwartzkroin, P.A. & Tempel, B.L. (1994) Localization of Kv1.1 and Kv1.2, two K-channel proteins, to synaptic terminals, somata, and dendrites in the mouse-brain. J. Neurosci. 14, 45884599.
  • 21
    Awan, K. & Dolly, J.O. (1991) K+ channel sub-types in rat brain: characteristic locations revealed using β-bungarotoxin, α- and δ-dendrotoxins. Neuroscience 40, 2939.
  • 22
    Mourre, C., Bidard, J.N. & Lazdunski, M. (1988) High-affinity receptors for the bee venom MCD peptide – quantitative autoradiographic localization at different stages of brain-development and relationship with MCD neurotoxicity. Brain Res. 446, 106112.
  • 23
    Pelchen-Matthews, A. & Dolly, J.O. (1988) Distribution of acceptors for β-bungarotoxin in the central nervous system of the rat. Brain Res. 441, 127138.
  • 24
    Pelchen-Matthews, A. & Dolly, J.O. (1989) Distribution in the rat central nervous system of acceptor sub-types for dendrotoxin, a K+ channel probe. Neuroscience 29, 347361.
  • 25
    Shamotienko, O.G., Parcej, D.N. & Dolly, J.O. (1997) Subunit combinations defined for K+ channel Kv1 subtypes in synaptic membranes from bovine brain. Biochemistry 36, 81958201.
  • 26
    Dolly, J.O., Muniz, Z.M., Parcel, D.N., Hall, A.C., Scott, V.E.S., Awan, K.A. & Owen, D.G. (1994) Subtypes of fast-activating, voltage-gated K+ channels in the nervous system: location and molecular properties unveiled with α-dendrotoxin and homologues. In Neurotoxins and Neurobiology (Tipton, K.F. & Dajas, F., eds), pp. 103122. Ellis Horwood Ltd, Chichester.
  • 27
    Garcia, M.L., Hanner, M., Knaus, H.G., Koch, R., Schmalhofer, W., Slaughter, R.S. & Kaczorowski, G.J. (1997) Pharmacology of potassium channels. Adv. Pharmacol. 39, 425471.
  • 28
    Harvey, A.L. (1997) Recent studies on dendrotoxins and potassium ion channels. Gen. Pharmacol. 28, 712.
  • 29
    Grissmer, S., Nguyen, A.N., Aiyar, J., Hanson, D.C., Mather, R.J., Gutman, G.A., Karmilowicz, M.J., Auperin, D.D. & Chandy, K.G. (1994) Pharmacological characterization of 5 cloned voltage-gated K+ channels, types Kv1.1, Kv1.2, Kv1.3, Kv1.5, and Kv3.1, stably expressed in mammalian-cell lines. Mol. Pharmacol. 45, 12271234.
  • 30
    Stuhmer, W., Ruppersberg, J.P., Schroter, K.H., Sakmann, B., Stocker, M., Giese, K.P., Perschke, A., Baumann, A. & Pongs, O. (1989) Molecular basis of functional diversity of voltage-gated potassium channels in mammalian brain. EMBO J. 8, 32353244.
  • 31
    Scott, V.E.S., Muniz, Z.M., Sewing, S., Lichtinghagen, R., Parcej, D.N., Pongs, O. & Dolly, J.O. (1994) Antibodies specific for distinct Kv subunits unveil a heterooligomeric basis for subtypes of α-dendrotoxin-sensitive K+ channels in bovine brain. Biochemistry 33, 16171623.
  • 32
    Koch, R.O., Wanner, S.G., Koschak, A., Hanner, M., Schwarzer, C., Kaczorowski, G.J., Slaughter, R.S., Garcia, M.L. & Knaus, H.G. (1997) Complex subunit assembly of neuronal voltage-gated K+ channels – basis for high–affinity toxin interactions and pharmacology. J. Biol. Chem. 272, 2757727581.
  • 33
    Hopkins, W.F., Miller, J.L. & Miljanich, G.P. (1996) Voltage-gated potassium channel inhibitors. Curr. Pharmaceut. Design 2, 389396.
  • 34
    Robertson, B., Owen, D., Stow, J., Butler, C. & Newland, C. (1996) Novel effects of dendrotoxin homologues on subtypes of mammalian Kv1 potassium channels expressed in Xenopus oocytes. FEBS Lett. 383, 2630.
  • 35
    Black, A.R. & Dolly, J.O. (1986) Two acceptor sub-types for dendrotoxin in chick synaptic membranes distinguishable by β-bungarotoxin. Eur. J. Biochem. 156, 609617.
  • 36
    Connolly, C.N., Krishek, B.J., McDonald, B.J., Smart, T.G. & Moss, S.J. (1996) Assembly and cell surface expression of heteromeric and homomeric gamma-aminobutyric acid type A receptors. J. Biol. Chem. 271, 8996.
  • 37
    Black, A.R., Breeze, A.L., Othman, I.B. & Dolly, J.O. (1986) Involvement of neuronal acceptors for dendrotoxin in its convulsive action in rat brain. Biochem. J. 237, 397404.
  • 38
    Black, A.R., Donegan, C.M., Denny, B.J. & Dolly, J.O. (1988) Solubilisation and physical characterization of acceptors for dendrotoxin and β-bungarotoxin from synaptic membranes of rat brain. Biochemistry 27, 68146820.
  • 39
    Parcej, D.N. & Dolly, J.O. (1989) Dendrotoxin acceptor from bovine synaptic plasma membranes: binding properties, purification and subunit composition of a putative constituent of certain voltage-activated K+ channels. Biochem. J. 257, 899903.
  • 40
    Dolly, J.O. (1992) Polypeptide neurotoxins as probes for certain voltage-dependent K+ channels. In Receptor-Ligand Interactions: a Practical Approach (Hulme, E.C., ed.), pp. 3761. IRL Press, Oxford.
  • 41
    Munson, P.J. & Rodbard, D. (1980) LIGAND: a versatile computerized approach for characterization of ligand-binding systems. Anal. Biochem. 107, 220239.
  • 42
    Hurst, R.S., Busch, A.E., Kavanaugh, M.P., Osborne, P.B., North, R.A. & Adelman, J.P. (1991) Identification of amino acid residues involved in dendrotoxin block of rat voltage-dependent potassium channels. Mol. Pharmacol. 40, 572576.
  • 43
    MacKinnon, R. & Miller, C. (1989) Mutant potassium channels with altered binding of charybdotoxin, a pore-blocking peptide inhibitor. Science 245, 13821385.
  • 44
    Stocker, M., Pongs, O., Hoth, M., Heinemann, S.H., Stuhmer, W., Schroter, K.H. & Ruppersberg, J.P. (1991) Swapping of functional domains in voltage-gated K+ channels. Proc. Roy. Soc. Lond. B 245, 101107.
  • 45
    Smith, L.A., Reid, P.F., Wang, F.C., Parcej, D.N., Schmidt, J.J., Olson, M.A. & Dolly, J.O. (1997) Site-directed mutagenesis of dendrotoxin K reveals amino acids critical for its interaction with neuronal K+ channels. Biochemistry 36, 76907696.
  • 46
    Scott, V.E.S., Rettig, J., Parcej, D.N., Keen, J.N., Findlay, J.B.C., Pongs, O. & Dolly, J.O. (1994) Primary structure of a beta-subunit of alpha-dendrotoxin-sensitive K+ channels from bovine brain. Proc. Natl Acad. Sci. USA 91, 16371641.
  • 47
    Wang, F.C., Bell, N., Reid, P., Smith, L.A., McIntosh, P., Robertson, B. & Dolly, J.O. (1999) Identification of residues in dendrotoxin K responsible for its discrimination between neuronal K+ channels containing Kv1.1 and 1.2 subunits. Eur. J. Biochem. 263 222, –224.
  • 48
    Muniz, Z.M., Diniz, C.R. & Dolly, J.O. (1990) Characterization of binding sites for δ-dendrotoxin in guinea-pig synaptosomes: relationship to acceptors for the K+ channel probe, α-dendrotoxin. J. Neurochem. 54, 343346.
  • 49
    Rhodes, K.J., Keilbaugh, S.A., Barrezueta, N.X., Lopez, K.L. & Trimmer, J.S. (1995) Association and colocalization of K+ channel α subunit and β subunit polypeptides in rat-brain. J. Neurosci. 15, 53605371.