(Received for publication, March 13, 1997, and in revised form, June 2, 1997)
From the Neuromuscular Unit, Department of Paediatrics and Neonatal
Medicine, Royal Postgraduate Medical School, Hammersmith Hospital,
London W12 0NN, United Kingdom and Novartis Institute for
Medical Sciences, 5, Gower Place,
London WC1E 6BN, United Kingdom
Taicatoxin, isolated from the venom of the Australian taipan snake Oxyuranus scutellatus, has been previously regarded as a specific blocker of high threshold Ca2+ channels in heart. Here we show that taicatoxin (in contrast to a range of other Ca2+ channel blockers) interacts with apamin-sensitive, small conductance, Ca2+-activated potassium channels on both chromaffin cells and in the brain. Taicatoxin displays high affinity recognition of 125I-apamin acceptor-binding sites, present on rat synaptosomal membranes (Ki = 1.45 ± 0.22 nM) and also specifically blocks affinity-labeling of a 33-kDa 125I-apamin-binding polypeptide on rat brain membranes. Taicatoxin (50 nM) completely blocks apamin-sensitive after-hyperpolarizing slow tail K+ currents generated in rat chromaffin cells (mean block 97 ± 3%, n = 12) while only partially reducing total voltage-dependent Ca2+ currents (mean block 12 ± 4%, n = 6). In view of these findings, the use of taicatoxin as a specific ligand for Ca2+ channels should now be reconsidered.
Neurotoxins, found in the venom of a wide variety of poisonous species (snakes, scorpions, spiders, and marine snails) have provided biologists with a formidable armory of molecular probes with which to study the structure and function of ion channels (1). In particular, identification of the wide range of subtypes of potassium channels that are now known to exist, originally owes much to the discovery of neurotoxins with highly selective pharmacological actions (2-5). Three types of potassium channel activated by intracellular Ca2+ (KCa channels)1 can be distinguished on a biophysical basis (2): large conductance, BKCa channels (typically 100-250 pS), intermediate conductance, IKCa channels (typically 20-100 pS), and small conductance, SKCa channels (typically 5-20 pS). Each channel subtype has a distinct and characteristic neurotoxin pharmacology. Many SKCa channels are specifically blocked by apamin, a peptide (~2000 Da) isolated from the venom of the European honey bee, Apis mellifera. Through the use of apamin, SKCa channels have been shown to be present in a wide variety of electrically excitable and non-excitable cells. In neurones, SKCa channels regulate repetitive firing by maintaining a slow after-hyperpolarizing potential following bursts of action potentials (6). In chromaffin cells, SKCa channels have been implicated in the control of adrenaline release (7) and in hepatocytes, they respond to increases in cytosolic [Ca2+] which is in turn specifically regulated by inositol trisphosphate and cAMP (8).
High affinity binding sites for 125I-apamin have been
characterized on plasma membranes prepared from numerous tissues
(9-12). 125I-Apamin binding polypeptides (putative
SKCa channel subunits) have been identified through
cross-linking and photoaffinity labeling strategies (11-16). These
studies indicate that hetero-oligomeric association of high () and
low (
) molecular mass polypeptide subunits may be a general
structural feature of members belonging to this family of
K+ channels. Most recently, members of a new ion channel
gene family have been cloned (17) with functional and structural
properties that implicate them as SKCa channel
subunits.
Apamin is generally recognized as being highly specific for SKCa channels (5, 18), however, it has also been reported to have apparently anomalous effects on Ca2+ channels in heart. Apamin can block slow Ca2+ action potentials in cultured cells originating from the ventricles of 15-day-old chick embryos (19) and also L-type Ca2+ currents of embryonic chick and human fetal heart cells (20). These intriguing observations have prompted an examination of the complementary effects of Ca2+ channel blockers on SKCa channel function. To our surprise, these studies have led us to identify taicatoxin (a previously characterized Ca2+ channel blocker isolated from the venom of the Australian taipan, Oxyuranus scutellatus (21, 22)) as a potent inhibitor of SKCa channels, both in rat brain and chromaffin cells. To the best of our knowledge, taicatoxin is the first SKCa channel blocker to be found in snake venom.
Native apamin was purified from A. mellifera bee venom and radioiodinated as described previously (23). Taicatoxin and other Ca2+ channel blockers were obtained from Alomone Labs. Cardiotoxin was purified from the venom of Naja nigricollis nigricollis and was a generous gift from Dr. A. Menez (Department d'Ingenierie et d'Etudes des Proteines, CEA, Saclay, Gif-sur-Yvette, France). Molecular weight markers were obtained from Pharmacia Biotech Inc. Protease inhibitors, bovine serum albumin (BSA, fraction V, protease free), and hyaluronidase type I-S were obtained from Sigma. Fetal bovine serum was obtained from Life Technologies, Inc. and collagenase type I was obtained from Worthington. Disuccinimidyl suberate was obtained from Pierce Chemical Co. All other chemicals used were reagent grade.
125I-Apamin Binding Assays and Affinity Labeling125I-Apamin binding to rat cerebrocortical synaptic plasma membranes and subsequent analysis of data was performed as described previously (12). The incubation medium (1 ml) consisted of 10 mM KCl, 1 mM EGTA, 25 mM Tris, pH 8.4, containing 0.1% (w/v) BSA. In saturation experiments, aliquots (~40 µg of protein) of purified plasma membranes were incubated with increasing concentrations of 125I-apamin (0.2-150 pM) in the presence or absence of either 0.1 µM native apamin (to determine non-saturable binding) or 2 nM taicatoxin. Following equilibration on ice (1 h) the reaction was quenched by the addition of ice-cold incubation medium and rapid filtration through Whatman GF/B FP-100 filters presoaked (1 h at 4 °C) in 0.5% (v/v) polyethyleneimine. In displacement experiments, aliquots (~100 µg of protein) of plasma membranes were incubated with a single fixed concentration of 125I-apamin (10 pM) in the absence or presence of increasing concentrations of taicatoxin or single fixed concentrations of various ligands (as detailed in Table I). In both saturation and displacement experiments triplicate assays were routinely performed; the standard deviation of the means was typically between 3 and 5%. Affinity labeling of rat synaptic plasma membranes with 125I-apamin using the homobifunctional agent disuccinimidyl suberate was performed as detailed elsewhere (12).
|
Affinity labeled membrane pellets were solubilized by heating (95 °C for 5 min) in sample buffer (4% (w/v) SDS, 10% (v/v) glycerol, 20 mM Tris, pH 6.8) containing protease inhibitors (2 mM EDTA, 2 mM EGTA, 10 µg/ml soybean trypsin inhibitor, 0.2 mM benzamidine, 0.1 mM phenylmethylsulfonyl fluoride, and 25 µg/ml bacitracin) in the presence of 5% (v/v) 2-mercaptoethanol. Aliquots (~150 µg of protein) were analyzed by discontinuous SDS-PAGE using 12% (w/v) acrylamide slab gels. Radioactive bands were identified by exposing the dried gel to x-ray film (Hyperfilm-MP; Amersham Corp.) for 2-5 days using an intensifying screen. Molecular masses were determined by comparison with migration of known standards. Taicatoxin samples were examined by SDS-PAGE using a high-molarity Tris buffer system adopted for analysis of small peptides (24). The resolving gel contained 0.75 M Tris, pH 8.9, and the composition of the running buffer was 0.192 M glycine, 0.05 M Tris, pH 8.9, containing 0.1% (w/v) SDS. Samples were heated (95 °C for 5 min) in sample buffer (as above) containing 2 mM EGTA, 2 mM EDTA and then analyzed by discontinuous gradient pore (15-30% (w/v) acrylamide) SDS-PAGE run under reducing or nonreducing conditions in the presence or absence, respectively, of 5% (v/v) 2-mercaptoethanol. Gels were silver stained for total protein using the method of Morrissey (25).
Cell CultureRat chromaffin cells were prepared by procedures based on the methods of Neely and Lingle (26) and Park (27). In brief, 2-3 rats (~200 g weight) were killed by CO2 asphyxiation, according to Home Office guidelines. The adrenal glands were removed and the medullas isolated by dissection. The medullas were incubated in Ham's F-14 medium containing hyaluronidase (2.4 mg/ml), 0.1% (w/v) collagenase, and 10% (v/v) fetal bovine serum for 45-60 min at 37 °C in a 5% CO2 incubator. The medullas were then washed three times with F-14 medium containing 10% (v/v) fetal bovine serum and the cells dissociated by trituration through a fire polished Pasteur pipette. Cells were plated on polyornithine-coated coverslips and maintained at 37 °C for 3-5 days in a 5% CO2 incubator prior to electrophysiological measurements.
ElectrophysiologyCoverslips with attached chromaffin cells
were placed in a recording chamber and visualized on the stage of a
Nikon Diaphot microscope. The general electrophysiological procedures
were as reported in Bevan and Yeats (28). Membrane currents were
studied with whole cell voltage clamp methods (29) using fire polished patch pipettes. Series resistances were between 3 and 10 M and an
80-85% series resistance compensation was applied. Potassium currents
through small conductance Ca2+-activated channels
(ISK(Ca)) were studied by examination of the tail currents,
evoked when the membrane potential was stepped from an initial holding
potential of
80 mV to more depolarized potentials (to evoke an inward
Ca2+ current), and then subsequently stepped back to
120
mV for 7 s. The external solution contained (mM) 122.5 NaCl, 30 KCl, 10 CaCl2, 1 MgCl2, and 10 HEPES,
adjusted to pH 7.4 with NaOH. The internal solution contained
(mM) 120 potassium aspartate, 20 KCl, 5 MgCl2,
0.1 EGTA, 3 Na2ATP, 0.1 leupeptin, and 20 HEPES, adjusted to pH 7.2 with KOH. The same solutions were also used to study any
effects of taicatoxin on potassium currents through BKCa
channels. Delayed rectifier currents were evoked by 100 ms voltage
steps from
80 mV to more depolarized potentials in
"Ca2+-free" solution. In most experiments drugs were
applied by pressure ejection from a puffer pipette although a U-tube
application system was employed in some cases. Data are presented as
mean ± S.E. of n observations. For Ca2+
current (ICa) experiments the external solution contained
(mM) 150 choline chloride, 10 CaCl2, 1 MgCl2, 10 HEPES, buffered to pH 7.4 with Tris-OH, and the
internal solution contained (mM) 130 CsCl, 1 CaCl2, 10 EGTA, 1 MgCl2 and 10 HEPES, adjusted
to pH 7.2 with CsOH.
A variety of both peptide and non-peptide blockers of
voltage-dependent Ca2+ channels were tested for
their ability to interact with 125I-apamin acceptors
present on rat synaptosomal plasma membranes. Potent and widely used
1,4-dihydropyridine Ca2+ channel antagonists, such as
nifedipine, nitrendipine, nimodipine, nicardipine, and PN 200-110 caused no significant change in 125I-apamin binding when
present at a 100,000-fold molar excess (relative to
125I-apamin) (Table I).
Similarly, Ca2+ channel toxins -agatoxin IVA,
calciseptine,
-conotoxin MVIIC, and
-conotoxin GVIA were without
effect when included in the assay at a 5000-fold molar excess. However,
under identical conditions, the high threshold Ca2+ channel
blocker taicatoxin, a complex oligomeric protein isolated from the
venom of the Australian taipan snake O. scutellatus
scutellatus (21) caused a marked reduction of binding to 11% of
control values (Table I).
Taicatoxin (Mr 52000) has been reported to be an
oligomeric complex of three noncovalently linked polypeptides (21).
Analysis of taicatoxin samples by SDS-PAGE under reducing conditions
followed by silver staining identified only three polypeptides of
~16, 8, and 7 kDa (Fig. 1). Although
the broad nature of the 8-kDa polypeptide band on SDS gels made it
rather difficult to accurately assess the relative abundance of the
three components, our results are consistent with the conclusions of
Possani and colleagues (21) who reported a stoichiometry of 1:1:4 of
the 16-, 8-, and 7-kDa polypeptides, respectively. As no other
polypeptides were observed after prolonged silver staining of SDS gels
(not shown), our results attest to the purity of the toxin sample.
Since the 16-kDa subunit displays phospholipase activity (21), it is
possible that this is responsible for the observed antagonism of
125I-apamin binding. However, this seems unlikely, since
the ability of taicatoxin to inhibit 125I-apamin binding
was unaltered in the presence of either 1 mM EGTA or 2 mM strontium chloride (Sr2+ is a competitor for
Ca2+ binding, which is essential for phospholipase
activity) (data not shown). Nevertheless, since EGTA alone (1 mM) produced no discernible inhibition of
125I-apamin binding, it was routinely included in the
buffer used in all further binding experiments. Another possible
explanation for taicatoxin's block of 125I-apamin binding
could be nonspecific charge effects, owing to the high density of
positive charge on the toxin. To test this possibility, we examined the
ability of a snake venom cardiotoxin to act in a similar manner
(cardiotoxins are extremely basic, surface-active polypeptides that
have a wide range of membrane perturbing activities, thought to be due
to their highly charged character (30)). Cardiotoxin from N. nigricollis nigricollis (31) did not alter 125I-apamin
binding (Table I), suggesting that random charge effects are unlikely
to underlie taicatoxin's inhibitory action.
Further examination of taicatoxin's activity showed that inhibition of
125I-apamin binding was complete over two log units of
taicatoxin concentration (Ki = 1.45 ± 0.22 nM) (Fig. 2A). In
agreement with this finding, saturation experiments (in which
increasing concentrations of 125I-apamin were incubated
with brain membranes in the presence or absence of 2 nM
taicatoxin) demonstrated that the inhibitory effect of taicatoxin was
due to a reduction in the affinity of 125I-apamin for its
acceptor, rather than an alteration of acceptor binding site density.
The Kd determined for 125I-apamin in the
absence and presence of taicatoxin was 5 and 13 pM,
respectively; Bmax varied by <2% (Fig.
2B). Importantly, these data re-affirm our earlier
conclusion that the effects of taicatoxin on 125I-apamin
binding are not due to site depletion through phospholipase activity.
125I-Apamin selectively labels high and low molecular
weight polypeptides associated with SKCa channels on both
brain and liver membranes (16). Such affinity labeling can be blocked
by structurally diverse pharmacological agents known to inhibit
apamin-sensitive SKCa channel K+ currents (12).
Taicatoxin completely abolished the incorporation of
125I-apamin into a 33-kDa polypeptide (putative
SKCa channel -subunit (16)) present on rat brain
membranes (Fig. 3, lane 7).
However, no significant difference in the labeling of this polypeptide was seen in the presence of a large molar excess of other
Ca2+ channel blockers (Fig. 3, lanes 3-6). (The
small decrease in labeling apparent in the presence of
-conotoxin
(lane 3) was not reproducibly observed).
In many cells, an apamin-sensitive Ca2+-activated slow tail
K+ current through SKCa channels
(ISK(Ca)) contributes to the after-hyperpolarization that
is observed following an action potential (6). The effects of
taicatoxin on ISK(Ca) were examined in voltage-clamp
studies using the slow after-hyperpolarizing tail currents found in
cultured rat chromaffin cells. Fig. 4
(parts A-C) demonstrates the currents evoked by stepping the
membrane potential from an initial holding potential of 80 mV to 0 mV
(for 2 s) and then to
120 mV (for 7 s). Under normal
recording conditions, a biphasic outward current was elicited by the
depolarizing voltage step and a long-lasting inward tail current was
seen when the membrane potential was stepped to
120 mV
(e.g. Fig. 4A, trace 1). The current is inward at
120 mV, as this holding potential is negative to the K+
equilibrium (zero current) potential. An application of 2 µM apamin for 1 s during the slow tail current
resulted in rapid and complete block (Fig. 4A, traces 2 and 3), confirming the nature of the current as an
ISK(Ca). Application of 50 nM taicatoxin for
1 s during the tail current had no immediate effect (Fig. 4B, trace 3) but abolished the slow ISK(Ca)
current evoked by subsequent depolarizing-hyperpolarizing voltage steps
(Fig. 4B, trace 4). This low concentration of taicatoxin,
eventually reduced the tail current by 97 ± 3%
(n = 12). In contrast, higher concentrations of
taicatoxin (5 µM) produced an immediate block of the
ISK(Ca) tail current (Fig. 4C, trace 3). The
specificity of taicatoxin for apamin-sensitive currents was tested in 4 cells in which a low dose of apamin (2 nM) was first
applied, followed by combined application of taicatoxin (50 nM) and apamin (2 nM). Application of
taicatoxin had no effect on either the outward currents evoked by
voltage steps from
80 mV to +20 mV or the fast tail currents evoked
when the potential was stepped back to
80 mV. For the outward
current, the ratio (amplitude in the presence of both toxins/amplitude
in the presence of apamin alone) was 1.02 ± 0.06, while for the
tail currents the ratio was 0.97 ± 0.06. These results suggest
that taicatoxin does not block any additional apamin-insensitive components, such as the Ca2+-activated K+
current through BKCa channels.
At the same concentration of taicatoxin (50 nM) that
completely blocked SKCa currents in chromaffin cells, only
a partial block of ICa was achieved in the same cells (mean
block 12 ± 4%, n = 6; Fig.
5A). This is consistent with
the observations of Possani et al. (21) who showed that
taicatoxin blocked calcium currents in heart cells with
IC50 values ranging from 10 to 500 nM,
dependent on holding potential (30 to
80 mV). The relatively small
effect on total ICa observed here is consistent with the
notion that the reduction of ISK(Ca) represents a direct
effect of taicatoxin on SKCa channels. The specificity of
taicatoxin for ISK(Ca) was also tested by examining its
effect on delayed rectifier currents which are present in chromaffin
cells (26). In Ca2+-free media, taicatoxin (50 nM) had no significant effect on the amplitude of the
delayed rectifier K+ currents (Fig. 5B, mean
amplitude 96 ± 3%, n = 5, of control value after
toxin application).
The ability of neurotoxins to selectively recognize different ion channels with high affinity renders them extremely useful probes for determining the distribution, biophysics, pharmacology, and structure of individual channel subtypes. As the number of recognized channel subtypes increases, so too does the demand for novel, highly selective toxins.
Apamin, an octadecapeptide from the venom of the European honeybee
A. mellifera is an established blocker of
SKCa channels in both the central nervous system and in
diverse peripheral tissues (3, 5, 18). The properties of
SKCa channels have been largely deduced through the use of
apamin in both biochemical and physiological studies. Although three
other toxins, scyllatoxin (leiurotoxin I), PO5, and Ts, possessing
apamin-like activity have been characterized from scorpion venoms
(32-36), taicatoxin, purified from the venom of the Australian taipan
O. scutellatus scutellatus (21), is to our
knowledge the first snake venom toxin shown to recognize
apamin-sensitive SKCa channels. This toxin is an oligomeric
complex, consisting of a peptide bearing homology to
-neurotoxins (8 kDa), a neurotoxic phospholipase (16 kDa), and four copies of a serine
protease inhibitor (7 kDa) whose primary structure bears homology to
protease inhibitors from other snake venoms (21). Taicatoxin binds with
high affinity (Ki ~1 nM) to
125I-apamin acceptors present on rat brain membranes and
the inhibition of 125I-apamin binding is both competitive
and phospholipase-independent. Taicatoxin also specifically blocks
affinity labeling of a 33-kDa 125I-apamin binding
polypeptide implicated in the structure of a hetero-oligomeric
SKCa channel. These biochemical findings are supported by
the demonstration that in chromaffin cells, taicatoxin (50 nM) completely blocks the slow ISK(Ca) tail
current. At the same concentration, taicatoxin reduces total
ICa by only ~12%.
Although taicatoxin competes for 125I-apamin acceptors with
high affinity, it is considerably less potent than apamin itself (Kd ~3 pM (12)). One cannot therefore
unequivocally rule out the possibility that activity may be due to the
presence (<0.01% by mass, assuming a Mr of
2000) of an unidentified neurotoxin with affinity analogous to that of
apamin. However, evidence suggesting that this is not the case stems
from the observation that taicatoxin's constituent -neurotoxin
possesses a structural motif consistent with the recognition of
SKCa channels. Apamin has two adjacent arginine residues
(Arg-13 and Arg-14) that are essential for its biological activity
(37-39) and it has been proposed that this motif (more specifically
the approximately 11 Å separation of two positive charges) provides
the basis for specificity of SKCa channel blocking
activity. A similar spatial separation of two positive charges is also
present in dequalinium and tubocurarine (as well as other bisquaternary
neuromuscular blocking agents), and this structural feature is thought
to underlie their ability to inhibit 125I-apamin binding to
liver and brain membranes and to block Ca2+-activated
K+ currents in both hepatocytes and neurones (12, 40, 41). Structure-function studies have also demonstrated that the scorpion toxins PO5 and scyllatoxin possess two functionally critical arginine residues responsible for SKCa channel inhibition. These two
arginine residues are adjacent in PO5 (34) while in the case of
scyllatoxin, they are brought into close proximity through secondary
folding (33, 42). The close apposition of two non-contiguous arginine residues within Ts
is similarly thought to underlie its recognition of 125I-apamin acceptor-binding sites (36). Since the
-neurotoxin of the taicatoxin complex contains two adjacent arginine
residues at its N terminus (21), this suggests a structural basis for the recognition of 125I-apamin acceptor-binding sites. We
have no insights as to a possible structural basis for taicatoxin's
recognition of high threshold Ca2+ channels, although given
the large size and oligomeric composition of taicatoxin, one can
speculate that this may involve a different region of the toxin. The
involvement of a single toxin domain in recognizing two distinct
classes of ion channel would mean that such a toxin motif must exploit
similar regions of like-charge distribution (presumably within the pore
regions) on the two channel types. However, since there is no primary
sequence homology between
-subunits of SKCa channels and
high threshold Ca2+ channels (17, 43), any similarity would
be purely fortuitous. Nevertheless, such a scenario may explain the
anomalous observation of apamin-sensitive Ca2+ channels in
immature heart cells (19, 20). If there is any similarity in
electrostatic topography between SKCa channels and certain
forms of Ca2+ channel, then it is severely restricted; the
overwhelming consensus of data in the literature would indicate that
Ca2+ channels and SKCa channels have distinct
and non-overlapping pharmacology.
In summary, our results indicate that taicatoxin, previously shown to block high threshold Ca2+ channels in heart, interacts with SKCa channels in both chromaffin cells and the brain. The toxin blocks 125I-apamin acceptor sites on rat synaptosomal membranes with high affinity and is an effective inhibitor of ISK(Ca) in rat chromaffin cells. In view of these findings the use of taicatoxin as a specific ligand for voltage-dependent Ca2+ channels should now be reconsidered.
We thank Dr. A. Menez for the generous gift
of cardiotoxin and K. Davidson for photography.