From the Institute for Biochemical Pharmacology,
University of Innsbruck, Peter-Mayr Strasse 1, A-6020 Innsbruck,
Austria and the § Department of Membrane Biochemistry & Biophysics, Merck Research Laboratories,
Rahway, New Jersey 07065
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Five novel peptidyl inhibitors of Shaker-type (Kv1) K+ channels have been purified to homogeneity from venom of the scorpion Centruroides limbatus. The complete primary amino acid sequence of the major component, hongotoxin-1 (HgTX1), has been determined and confirmed after expression of the peptide in Escherichia coli. HgTX1 inhibits 125I-margatoxin binding to rat brain membranes as well as depolarization-induced 86Rb+ flux through homotetrameric Kv1.1, Kv1.2, and Kv1.3 channels stably transfected in HEK-293 cells, but it displays much lower affinity for Kv1.6 channels. A HgTX1 double mutant (HgTX1-A19Y/Y37F) was constructed to allow high specific activity iodination of the peptide. HgTX1-A19Y/Y37F and monoiodinated HgTX1-A19Y/Y37F are equally potent in inhibiting 125I-margatoxin binding to rat brain membranes as HgTX1 (IC50 values ~0.3 pM). 125I-HgTX1-A19Y/Y37F binds with subpicomolar affinities to membranes derived from HEK-293 cells expressing homotetrameric Kv1.1, Kv1.2, and Kv1.3 channels and to rat brain membranes (Kd values 0.1-0.25 pM, respectively) but with lower affinity to Kv1.6 channels (Kd 9.6 pM), and it does not interact with either Kv1.4 or Kv1.5 channels. Several subpopulations of native Kv1 subunit oligomers that contribute to the rat brain HgTX1 receptor have been deduced by immunoprecipitation experiments using antibodies specific for Kv1 subunits. HgTX1 represents a novel and useful tool with which to investigate subclasses of voltage-gated K+ channels and Kv1 subunit assembly in different tissues.
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Potassium channels comprise a large family of proteins that control electrical excitability as well as resting membrane potential in many different cell types. The discovery of peptidyl inhibitors in the venoms of different species such as scorpions, snakes, spiders, and sea anemone has played an instrumental role in the development of our current understanding of K+ channels. High affinity and selective peptides have been used to define the physiological role that K+ channels play in tissues of interest. As structural probes, they have guided the identification of the pore-forming region of these proteins. Radiolabeled derivatives of some of the peptides have been produced and used to identify receptor sites in native tissues, to purify these proteins to homogeneity from native tissues, and to determine their subunit composition (1-3). In addition, they have provided a means by which to develop the molecular pharmacology of K+ channels (4, 5).
In this study, we report the identification of five new peptides in the venom of the Central American scorpion Centruroides limbatus. The major peptidyl component, HgTX1,1 has been fully characterized and shares significant sequence homology with MgTX and noxiustoxin. A HgTX1 analog that can be radiolabeled without loss of biological activity has been produced by recombinant techniques and used to define the K+ channel subunit composition of the rat brain HgTX1 receptor. HgTX1, because of its potency and K+ channel pharmacology, represents a novel tool with which to study the distribution of K+ channel subtypes in different tissues.
![]() |
EXPERIMENTAL PROCEDURES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Materials--
Venom from the scorpion C. limbatus
was obtained from Miller International Venoms (Hollywood, FL).
Escherichia coli DH5 was used for plasmid propagation,
and strain BL21(DE3) was used for expression of the fusion protein.
Plasmid PGEMEX-1 was from Promega. 125INa and
86RbCl were purchased from NEN Life Science Products.
Restriction enzymes, Pfu DNA polymerase, T4 DNA ligase,
nucleotide triphosphates, and reagents for polymerase chain reaction
were obtained from Boehringer Mannheim. Glass fiber filters (GF/C) were
from Whatman. Polyethyleneimine and bovine serum albumin (BSA) were
purchased from Sigma. Iscove's modified Dulbecco's medium was from
JHR Biosciences. HEK-293 cells stably transfected with
homotetrameric Kv1 channels were obtained from
Professor Olaf Pongs (Zentrum für Molekulare Neurobiologie, Hamburg, Germany). All other reagents were obtained from
commercial sources and were of the highest purity grade
commercially available.
Peptide Purification-- About 100 mg of crude C. limbatus venom was extracted and purified as described previously with minor modifications (7). Elution from the C18 reversed-phase column was achieved in the presence of a linear gradient of acetonitrile (0-35%, 78 min) at a flow rate of 1 ml/min. Purified peptides were reduced and subjected to Edman degradation as described (7).
Plasmid Construction, Synthesis, and Purification of Recombinant
HgTX1--
The plasmid was created by altering pG9MgTX in
the following manner: Ile2 Val, Asn4
Asp, Gln23
Ile, and Ser24
Arg. These
mutations were introduced by the "gene SOEing" technique using
mutagenic forward and reverse polymerase chain reaction primers (8).
Desired mutations were verified by cDNA sequencing using the
dideoxy chain termination method (9). The fusion protein was expressed
in E. coli and purified as described by Koschak et
al. (10). Composition of the purified material was verified by
electrospray mass spectroscopy and Edman degradation. An identical
procedure was used to generate and purify
HgTX1-A19Y/Y37F.
Iodination of HgTX1-A19Y/Y37F--
A sample of
HgTX1-A19Y/Y37F (38 µg; 9 nmol) dissolved in 100 µl of
100 mM sodium phosphate (pH 6.5) was incubated with 3-5 mCi of 125INa (30-50 µl) in the presence of
IODO-BEAD® (1 bead; Pierce) for 10 min at room
temperature. The bead was removed, and the reaction mixture was applied
to a C18 reversed-phase column (Vydac; 0.45 × 25 cm)
equilibrated with 0.05% trifluoroacetic acid. Elution was achieved in
the presence of a linear gradient of acetonitrile in 0.05%
trifluoroacetic acid (0-35% over 51 min). After lyophilization in the
presence of 0.1% bovine serum albumin, the radioiodinated peptide was
resuspended in 150 mM NaCl, 20 mM Tris-HCl (pH
7.4) and stored in small aliquots at 80 °C. The composition of the
iodinated material was determined by automated Edman degradation and
mass spectroscopy.
Membrane Preparation and Binding Studies-- Rat brain synaptic plasma membrane vesicles were prepared as described previously (5). Membrane vesicles from HEK-293 cells stably transfected with homotetrameric Kv1 channels were prepared as described (11). Binding experiments were carried out essentially as described previously (5). The incubation medium consisted of 20 mM Tris/HCl (pH 7.4), 0.1% bovine serum albumin (rat brain membranes) or 20 mM Tris/HCl (pH 7.4), 5 mM KCl, 0.1% bovine serum albumin (HEK-Kv1 membranes). Nonspecific binding was defined in the presence of 1 nM recombinant HgTX1 or MgTX, and incubation was carried out at 22-25 °C typically for 240 min. Experiments employing low receptor and/or radioligand concentrations (e.g. saturation studies) were allowed to reach equilibrium for >15 h. No detectable decay in receptor activity was observed after this incubation time.
Antibody Production and Immunoprecipitation Studies-- Polyclonal sera were raised against unique regions of Kv1.1, Kv1.2, Kv1.3, Kv1.4, Kv1.5, and Kv1.6 channels (12). Rat brain membranes were incubated with 4-7 pM 125I-HgTX1-A19Y/Y37F at radioligand excess for >15 h at room temperature and solubilized as described previously (5, 13). Immunoprecipitation experiments were carried out in 20 mM Tris-HCl (pH 7.4), 100 mM NaCl, 5 mM KCl, 0.1% digitonin as described previously (5, 12, 13).
Efflux of 86Rb+ from HEK-293/Kv1 Cells-- HEK-293 cells stably transfected with homotetrameric Kv1 channels were plated into 96-well culture plates and maintained in Iscove's modified Dulbecco's medium with L-glutamine and HEPES). Cells were incubated overnight with 86Rb+ (3 µCi/ml). Depolarization-induced 86Rb+ efflux was measured as described previously (11).
Analysis of Data and Protein Determination-- Radioligand binding studies were analyzed as described (5). Protein concentration was determined according to Ref. 14 using bovine serum albumin as a standard.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Purification and Amino Acid Sequence of HgTXs-- Venom of the scorpion C. limbatus contains components that inhibit 125I-MgTX binding to voltage-gated K+ channels in rat brain synaptosomal plasma membrane vesicles. To isolate this inhibitory activity, ~100 mg of crude C. limbatus venom was fractionated on a Mono S HR10/10 cation exchange column. Two active fractions, eluting at ~0.18 M (peak 1) and ~0.20 M NaCl (peak 2), contained all this activity. They were loaded individually on a reversed-phase C18 column. The entire inhibitory activity of Mono S peak 1 was recovered in a single peptide (HgTX1), while peak 2 yielded several active fractions (HgTX2-5). The corresponding peptide sequences are shown in Fig. 1. They define HgTX1 as a 39-amino acid peptide with an overall amino acid sequence homology of 89% to MgTX (15) and 75% to noxiustoxin (16). Although only partial sequences of HgTX2 and HgTX3 were obtained, the data indicate that these two peptides also belong to families of previously identified K+ channel inhibitors. It is interesting that both HgTX2 and HgTX3 have one amino acid residue less between the third and the fourth Cys residue and therefore resemble more the family of ChTX/IbTX/Limbatustoxin/Leiurus toxin 2.
|
Expression and Oxidative Iodination of Recombinant HgTX1-- To verify that the amino acid sequence of HgTX1 corresponds to an active venom component, HgTX1 was expressed in E. coli and purified as described previously (10, 15). Recombinant HgTX1 inhibits 125I-MgTX binding to rat brain membranes with the same potency as native peptide (Ki = 0.47 pM). We failed to produce radioiodinated HgTX1 due to poor base-line separation of native and iodinated peptide. Therefore, we altered two residues in the synthetic HgTX1 gene; Ala was substituted for Tyr at position 19, and a Tyr to Phe conversion was introduced at position 37 (HgTX1-A19Y/Y37F). This approach has been successfully employed to radiolabel IbTX and AgTX1 (13, 17) and takes advantage of the fact that position 19 in these peptides can be modified without loss of biological activity because this residue does not form part of the peptide's interaction surface with K+ channels; only conservative substitutions at position 37 are tolerated. HgTX1-A19Y/Y37F was expressed and purified as described above for HgTX1. Binding of 125I-MgTX to rat brain membranes is inhibited by HgTX1 and HgTX1-A19Y/Y37F with identical Ki values (data not shown), suggesting that substitution of these two residues in HgTX1 does not cause any significant loss in toxin affinity. HgTX1-A19Y/Y37F was reacted with 125INa, and both the mono- and di-iodinated derivatives were well separated from native peptide (data not shown).
Interaction of HgTX1 with Homotetrameric Kv1 Channels-- Functional Effects of HgTX1 and HgTX1-A19Y/Y37F on homotetrameric Kv1 channels were investigated by monitoring depolarization-induced 86Rb+ efflux from HEK-293 cells that have been stably transfected with Kv1.1, Kv1.2, Kv1.3, or Kv1.6. Table I (top) summarizes these results. Both HgTX1 analogs inhibit with high affinities and similar potencies 86Rb+ flux through Kv1.1, Kv1.2, and Kv1.3 channels, whereas they are weaker inhibitors of Kv1.6. This pharmacological profile is identical to that seen with MgTX. It is worth noting that MgTX is indeed a high affinity inhibitor of Kv1.1, although 125I-MgTX does not bind to homotetrameric Kv1.1 channels transiently expressed in COS cells (12). These data suggest that iodination of MgTX at Tyr37 leads to a substantial modification of the peptide's pharmacological properties.
|
|
Interaction of 125I-HgTX1-A19Y/Y37F with Rat Brain Synaptosomal Plasma Membrane Vesicles-- Since the specificity of HgTX1 for homotetrameric K+ channels was established, we next investigated whether the interaction of this ligand with rat brain membranes could be correlated with any given Kv1 channel. 125I-HgTX1-A19Y/Y37F binds to these membranes in a concentration-dependent manner (Fig. 3A) to a single class of binding sites that display a Kd of 0.15 pM and a Bmax of 2.3 pmol/mg of protein (n = 4).
|
Subunit Composition of the Rat Brain HgTX1 Receptor-- To investigate in further detail the molecular components of the HgTX1 receptor, brain membranes were labeled with a saturating concentration of 125I-HgTX1-A19Y/Y37F and solubilized in the presence of digitonin. Solubilized receptors were subjected to immunoprecipitation experiments employing a complete panel of sequence-directed antibodies directed against individual Kv1 channels. The specificity of these antibodies has previously been determined to ensure that they exclusively recognize their corresponding antigens without displaying any cross-reactivity with other Kv1 subunits (5, 11, 12). All antibodies yielded saturable levels of precipitation (Fig. 4A), and, moreover, the presence of the corresponding competing peptide always decreased the level of precipitation by >90% (data not shown). Anti-Kv1.2 antibody precipitated all 125I-HgTX1-A19Y/Y37F receptors, and as a consequence, the combination of anti-Kv1.2 with any other anti-Kv1 antibody (e.g. anti-Kv1.1 or anti-Kv1.4) did not yield any increased level of immunoprecipitation. These results indicate that in virtually all brain HgTX1 receptors, Kv1.2 is an integral component. In the presence of anti-Kv1.1, 75 ± 5% of toxin receptors can be precipitated, whereas anti-Kv1.3 and anti-Kv1.6 only precipitated 17 ± 2% and 14 ± 4% of receptors, respectively. In the presence of saturating concentrations of both anti-Kv1.3 and anti-Kv1.6, additive precipitation levels (29 ± 2%) could be achieved (Fig. 4B), indicating that these two proteins are segregated into distinct HgTX1-sensitive channel complexes consisting of at least Kv1.2/Kv1.3 or Kv1.2/Kv1.6 subunits, respectively. Interestingly, anti-Kv1.4 consistently precipitated 45 ± 4% of receptor-bound 125I-HgTX1-A19Y/Y37F. Since homotetrameric Kv1.4 channels are not labeled by 125I-HgTX1-A19Y/Y37F (see above), and Kv1.2 subunits are always integral parts of the toxin receptor, Kv1.4 should be complexed with Kv1.2. The combination of anti-Kv1.1, anti-Kv1.3, anti-Kv1.4, and anti-Kv1.6 antibodies yielded the same precipitation level observed for anti-Kv1.2 alone, indicating that essentially no homotetrameric Kv1.2 channels are expressed in rat brain, and as a consequence, most Kv1.2 subunits should be assembled with other Kv1 subunits (see below). Only anti-Kv1.5 did not give any measurable level of toxin precipitation, suggesting that this Kv1 subunit does not contribute to the HgTX receptor in rat brain.
|
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
HgTX1 Is a High Affinity Ligand for Voltage-gated K+ Channels-- In this study, we report the purification and characterization of a novel K+ channel-blocking peptide, HgTX1, from the venom of the Central American scorpion C. limbatus. Recombinant HgTX1 inhibits with high affinity 125I-MgTX binding to rat brain membranes but not 125I-IbTX-D19Y/Y36F binding to maxi-K channels, suggesting that HgTX1 is a selective blocker of voltage-dependent K+ channels. The specificity of HgTX1 for members of the Kv1 family indicates that this peptide inhibits with equivalent high potency Kv1.1, Kv1.2, and Kv1.3 channels, but it displays lower affinity for Kv1.6. This pharmacological profile is identical to that of MgTX (Table I). However, it was shown in previous studies that 125I-MgTX does not bind to homotetrameric Kv1.1 channels transiently expressed in COS cells (12). These data suggest that iodination of MgTX at Tyr37 alters the peptide's pharmacological properties. Since Tyr37 is presumed to form part of the peptide's interaction surface with the channel, covalent modification of that residue could lead to alteration in the conformation of the peptide and, as a consequence, to a different pharmacological profile. The bulky hydrophobic iodine at Tyr37 might also prevent an interaction with residues located in the pore of the channel. To circumvent this problem, HgTX1-A19Y/Y37F was prepared and shown to display the same functional characteristics as HgTX1. In contrast to 125I-MgTX, which only binds with high affinity to Kv1.2 and Kv1.3 channels, binding of 125I-HgTX1-A19Y/Y37F occurs with subpicomolar dissociation constants to Kv1.1, Kv1.2, or Kv1.3 channels. However, interaction with Kv1.6 channels is of lower affinity, and the peptide does not bind to either Kv.1.4 or Kv1.5 channels. Thus, iodination of HgTX1-A19Y/Y37F does not alter the pharmacological properties of the peptide and yields a very high affinity ligand selective for a subset of voltage-gated K+ channels.
Subunit Composition of Rat Brain Voltage-gated K+ Channels Labeled by 125I-HgTX1-A19Y/Y37F-- In rat brain membranes, 125I-HgTX1-A19Y/Y37F binds to a single class of binding sites with a dissociation constant identical to that determined for homotetrameric Kv1.1, Kv1.2, and Kv1.3 channels. Thus, the brain receptor is likely to be associated with either one or all of these Kv1 channels. To investigate in further detail the molecular components of the rat brain HgTX1 receptor, we employed antibodies specific for individual Kv1 subunits in immunoprecipitation studies of detergent-solubilized membranes that had been labeled with 125I-HgTX1-A19Y/Y37F. All HgTX1 receptors were found to contain at least one Kv1.2 subunit. A number of putative complexes containing two or three distinct subunits (e.g. Kv1.1·Kv1.2, Kv1.2·Kv1.4, Kv1.1·Kv1.2·Kv1.4, Kv1.1·Kv1.2·Kv1.6, and Kv1.2·Kv1.3·Kv1.4) were unequivocally identified. Neither Kv1.1, Kv1.2, nor Kv1.3 appears to exist as a homooligomer. However, we cannot exclude the possibility that either yet unidentified additional subunits or several copies of one subunit are contained in any of the identified K+ channel complexes. Our findings are in agreement with previously published immunohistochemical distribution studies where Kv1.1·Kv1.2 or Kv1.2·Kv1.4 subunits have been found to be co-localized in specific regions of the rat brain (12, 18-21).
Heterooligomeric association of Kv1 subunits has recently been reported in bovine cerebral cortex (22) by immunoprecipitation experiments and subsequent identification of the channel subunits by Western blotting. However, a quantitative determination of the relative abundance of each individual oligomer was not possible using this approach. Moreover, denaturation of channel oligomers followed by subunit dissociation could have hampered the precise detection of naturally occurring subunit combinations. It is worth mentioning that our precipitation experiments also have the limitation that only complexes to which radiolabeled peptide remains bound at the end of the immunoprecipitation assay can be detected. For instance, homooligomeric Kv1.4 and Kv1.5, channels to which HgTX1 does not bind, would not be detected by this approach. The same is true for putative heterooligomers of Kv1.5 with other HgTX1-sensitive subunits, since it has been shown that Kv1.5 would confer a dominant phenotype to these complexes (23). Even homotetrameric Kv1.6 channels, to which HgTX1 binds with much lower affinity, would be difficult to identify because of significant ligand dissociation during the time course of the study. The development of new ligands that display high affinity interaction with the different Kv1 subunits could therefore be of great interest for defining the total subunit composition of these channels. Regardless of the limitations discussed above, HgTX1 represents a very useful tool for identifying subsets of voltage-gated K+ channels in different tissues. ![]() |
ACKNOWLEDGEMENTS |
---|
We thank Maria Trieb, Emanuel Emberger, and William Schmalhofer for technical contributions and Tracey Klatt for mass spectroscopy studies. We gratefully acknowledge Dr. Robert Koch for experimental support in the early stage of the project and Drs. Jörg Striessnig, Robert Slaughter, and Hartmut Glossmann for continuous discussion.
![]() |
FOOTNOTES |
---|
* The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
A preliminary report of this work has been presented in abstract form (6).
¶ Supported by Austrian Research Foundation Grants S6611-MED and P-11187-MED, Austrian National Bank Foundation Grant 6239, and the European Union BIOMED 2 program Grant BMH4-CT96-2118. To whom correspondence should be addressed. Tel.: 43-512-507-3156; Fax: 43-512-507-2858; E-mail: hans.g.knaus{at}uibk.ac.at.
1 The abbreviations used are: HgTX, hongotoxin; ChTX, charybdotoxin; 125I-ChTX, monoiodotyrosine charybdotoxin; MgTX, margatoxin; 125I-MgTX, monoiodotyrosine margatoxin; IbTX, iberiotoxin; Kv1 channel, voltage-gated K+ channel, Shaker-type; BSA, bovine serum albumin; AgTX, agitoxin.
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|