Effects of Conus striatus venom on voltage-gated K channels
in frog sympathetic ganglia neurons
BRETT E. ETCHEBARNE
Experimental Neurobiology
June 15, 2001
William F. Gilly, Advisor
Acknowledgements
1 would like to thank Professor Gilly for his commitment to all of his students in
the course and drive to emphasize the importance of proper experimentation and data
analysis in the pursuit of knowledge. His influence throughout this experience will have a
lasting impression throughout my lifetime in all future work.
Dr. Joseph Schulz was a constant source of guidance and consultation throughout
this quarter. He made this project possible, and allowed me to see the light at the end of
the tunnel during the tough times.
Mat Brock was extremely helpful in providing crucial technical support in the
face of many ominous problems.
Josh Rosenthal is a funny guy. Go Sixers.
Table of Contents:
Acknowledgements.
Table of Contents
List of Figures..........................................................4
Text....c...................................................
Literature Review........................................
Abstract.........................
Introduction...............4
Materials and Methods................................................................................10
Results...................................................................................................10
Discussion.Z0
References4
DiagramsZ0
.....29
Figures
Figure Legends0
List of figures
Diagram 1: Cone snails
Figure 1: Basic venom effect. Externally applied Conus venom causes time-dependent
block of frog sympathetic ganglia voltage-gated Kchannels
Figure 2: Conus venom does not alter the K conductance versus voltage relationship
Figure 3: Venom block of Ix occurs in a voltage-independent manner
Figure 4: Concentration dependence of steady-state Ig block
Figure 5: Time course of lg block at two relative venom concentrations
Figure 6: S-Nitrosodithiothreietol does not modify sympathetic ganglia Ix
Figure 7: Sympathetic ganglia neuron Ig is TEA sensitive
Review of Conus venom effects on voltage-gated
sodium and potassium channels
Venomous gastropods of genus Conus
One of the most dangerous and unlikely predators in the ocean is presented to us
in the form of a snail. Cone snails are marine gastropods (Class Gastropoda, Order
Neogastropoda, Superfamily Toxoglossa, Family Conidae) that have found their
ecological niche in the coral reef habitat and proliferated greatly in the past 55 million
years, growing to a size of about 500 different species as possibly the largest single
marine genus.
An unequivocal reason for the adaptive success and radiation of Conus is present
in its employment of neurotoxins as a method of prey capture, escape from and defense
against predators, and deterrent of competitors. These toxins are small, structurally
constrained peptides of 12-30 amino acid residues in length held together by several
internal disulfide bonds. Conotoxins have evolved to select targets comprised of specific
voltage- and ligand-gated ion channels and G protein-linked receptors and act to disrupt
normal neuronal communication (Olivera, 1997). Currently, researchers have identified
10 major classes of Conus peptides, each targeting a different aspect of the neuronal
communication process including K, Na', Ca", and ACh receptors (Mclntosh et al.,
1999
Studies of the molecular genetics of toxin genes in Conus species has shown that
gene duplication and selection for diversity are resulting in functionally variable
conotoxins which are further linked to the evolutionary spread and success of the genus
(Duda and Palumbi, 1998). Furthermore, it has been postulated that in the more complex
venom batteries of Conus species there may be over 100 different peptides present, each
of which may be unique to that species as a means of prey capture (Olivera et al., 1995).
An adept system of conotoxin delivery into both prey and possible enemies is
based on a highly sophisticated venom production-apparatus and delivery system. The
cone snail employs chemical and tactile sensory systems in its extensible proboscis to
locate and track targets, and then "harpoons" it's victim using a specialized barbed radula
tooth that is tethered to its proboscis. The radula tooth at the distal end of a venom duct
acts to spear and hypodermically envenom the prey (Kohn, 1956). The proximal end of
the duct emerges from a muscular bulb-like structure.
To be a successful predator, each of the 500 different species of cone snail has
developed an arsenal of conotoxin peptides geared specifically toward incapacitation of
their specific prey items. For example, fish-eating (piscivorous) cones use a battery of
peptides evoking “excitotoxic shock" in their victim through the combination of a ô¬
conotoxin that increases sodium-channel conductance while a complementary k-
conotoxin blocks potassium-channel conductance. This typically produces a strong
depolarization in poisoned nerve processes leading to repetitive firing and disruption of
coordinated locomotion by the fish. Meanwhile, a different group of toxins in the same
venom cocktail acts synergistically on different target cell types to inhibit neuromuscular
transmission and muscular contraction, thereby paralyzing the fish. These are the o¬
conotoxin PIVA, which targets the nicotinic acetylcholine receptor, and u-conotoxin
PIIIA, which blocks skeletal muscle sodium channels. Together, this combination of
peptides very quickly renders the fish helpless and enables the slow-moving snail to catch
its faster prey (Terlau et al., 1996).
Conotoxins are remarkably specific to the prey type for each Conus species.
Studies carried out by Endean and Rudkin (1965) showed that crude venom samples from
several Conus species that were fish eaters had no effect on mollusks or worms, whereas
toxins from mollusk-eating species did not affect fish. This is consistent with the idea that
particular Conus species selected for a match between prey-type and toxin complement
during their evolutionary history. In some cases, however, toxins from species that
selectively prey on other gastropod mollusks are very potent in mammalian systems, such
as those from Conus geographus, C. magus and C. textile. Specificity of such peptides
occurs not because the cone snails have selected for a conotoxin that is directed against
mammalian systems, but because a particular feature of the channel or receptor target is
highly conserved in both mammals and mollusks (Olivera et al., 1991).
Conus striatus
Conus striatus is a relatively large cone snail found in the Indo-Pacific. It is one
of the most widely distributed cones, and was the first ever described as specializing in
fish hunting (Kohn, 1956).
Peptide batteries in C. striatus and C. magus have been shown to differ from other
Indo-Pacific fish hunters in their composition. All of the Indo-Pacific Conus described
make peptides, which target the block of neuromuscular transmission. C. purpurascens
produces both a noncompetitive nicotinic antagonist (aA-conotoxin) and a competitive
nicotinic antagonist (y-conotoxin). C. ermineus contains a competitive nicotinic agonist
only. The peptides of C. striatus and C. magus, however, are o-conotoxins which target
calcium channels and are not present in the other fish hunters. Additionally, the KA¬
conotoxin SIVA has been described in C. striatus which causes spastic paralysis in fish,
appears to block voltage-gated Shaker K channels, and causes repetitive action
potentials following a single stimulus in the frog neuromuscular preparation (Craig et al.,
1998).
Conus toxins directed against voltage-gated K and Na channels
My research at the present time is limited to the effects of Conus striatus venom
on voltage-gated Na' and voltage-gated K channels. For this reason I have focused my
literature review on those articles pertaining to the three main categories of Conus venom
components that are known to affect either of these two types of channels. These include
the k-conotoxins, which selectively inhibit K channels, and u-conotoxins and ô¬
conotoxins, which selectively affect Na channels.
k-Conotoxins
k-Conotoxins have been classified as those peptides isolated from Conus species,
which produce an effect on voltage-gated K channels. Thus far, only two k-conotoxins
have been identified. The first, k-conotoxin PVIIA, isolated from Conus purpurascens,
has affects on Shaker H4 channels. In voltage-clamp experiments, the toxin produces a
reversible reduction in peak K currents (Ig) with dose dependence, consistent with a 1:1
binding stoichiometry for the toxin and receptor site (Shon et al., 1997).
In the Shaker channel, PVIIA is known to interact directly with the external TEA
binding site in the external vestibule region, and in this regard is analogous to the
charybdotoxin (CTX) family of scorpion toxins (Terlau et al. 1999). However, the
mutation F425G in this channel that acts to increase CTX-binding affinity by three orders
of magnitude makes the channel insensitive to k-PVIIA. Thus, the two toxins, which
have been determined to be structurally similar in protein conformation and folding, do
not appear to share the identical amino acid residue at the position that mates with the
F425 binding site. Since additional amino acid residues on the K channel are also
important for binding of CTX (Miller, 1995), some of these may overlap with those
important for K-PVIIA binding. Other residues are likely to be unique to each toxin.
Thus, both toxin types are potentially very useful in studying and mapping the structure
of the K channel outer vestibule from a diverse group of K channel targets.
KA-Conotoxin SIVA, isolated from C. striatus, is a peptide that has also been
shown to affect K channels, but is structurally different from the k-conotoxin PVIIA
taken from C. purpurascens (Craig et al., 1998). This class of k-conotoxins is an
antagonist of Shaker K channels as well, but the block of Ix in this case is only slowly
reversible. Using a frog nerve-muscle preparation, the peptide produced repetitive action
potentials. Thus, the block of voltage-gated K channels by KA-SIVA may be involved in
this neuroexcitatory activity.
u-Conotoxins
u-Conotoxins are those Conus peptides that target sodium channels and impair
propagation of action potentials. In voltage-clamp experiments, the u-conotoxins
produce a decrease in voltage-gated Na’ current (Iya). u-Conotoxin GIIIA was first toxin
of this class isolated (from C. geographus) and is selective for vertebrate skeletal muscle.
One distinguishing characteristic for u-GIIIA is that it competes with classical Na
channel blockers tetrodotoxin (TTX) and saxitoxin (STX) and produces similar effects
(Fainzilber et al., 1995). However, mutagenesis studies have shown that each of the three
toxins have distinct binding sites. It has been postulated that a specific arginine residue
(RI3) of the toxin is critical for activity of u-GIIIA and that the guanidinium side chain
of R13 is analogous to the critical guanidino groups of TTX and STX (Dudley et al.,
1995). Complete block of Na' current with u-conotoxin GIIIA is possible at nanomolar
concentrations. u-GIIIA has a structural rigidity that makes it a good candidate for
probing pairwise interplay between toxin and channel amino acids. Interactions between
specific channel and toxin residues have allowed limited mapping of the outer vestibule
architecture of the voltage-gated Na" channel. Such an approach is valuable, because
ascertaining the structure of large membrane proteins like ion channels has been hindered
by problems with application of procedures like NMR, and x-ray crystallography
(Chahine et al., 1998). lon channel structure and function will be better understood in the
future as a result of such studies. This understanding of the topography of the channel
should also facilitate drug design.
Another variety of u-conotoxin has been isolated from C. purpurascens, u-PIIIA,
that shows different target specificity than that of u-GIIIA. Arginine-14 is a key residue
for u-PIIIA, with substitution by alanine abolishing Na" channel-blocking activity. u¬
PIIIA irreversibly blocks amphibian muscle Na' channels but not those in motor nerves,
making it a useful tool in synaptic electrophysiology. Type II Na channels from the
mammalian CNS expressed in Xenopus oocytes, which are TTX-sensitive, were
reversibly blocked by u-PIIIA, whereas they were unaffected by u-GIIIA. Additionally,
u-Conotoxin PIIIA appears to target a wider variety of mammalian voltage-gated Na
channel subtypes in the mammalian CNS than does u-GIIIA.
TTX/STX-sensitive Na channels could be divided into three categories based on
the distinguishing features of u-PIIIA and u-GIIIA,: 1) Those channels sensitive to both
u-PIIIA and u-GIIIA, for example the skeletal muscle subtype in both frog and
mammalian systems. 2) Those sensitive to TTX and u-PIIIA but more resistant to u¬
GIIIA, such as rat brain Type II Na channels. 3) Channels resistant to both u-PIIIA and
u-GIIIA, with most of the total CNS Na channels falling into this category (Shon et al.,
1998). The varieties of u-conotoxins presently known and those yet to be discovered will
be useful in determining the roles of Na channel subtypes in neurons and circuits when a
variety of molecular forms of voltage-gated Na' channels are present.
Another, divergent group of Na channel-directed peptides is the O-glycosylated
uO-conotoxins. These Conus peptides, isolated from C. marmoreus and C. pennaceus,
produce a biological effect similar to that of the previously described u-conotoxins, yet
they are structurally distinct and show different disulfide-bond arrangements. Two
peptides from C. marmoreus, MrVIA and MrVIB have been characterized which block
Na channels in Aplysia neurons. Classical sodium channel blockers TTX and STX have
not been found to be efficient in Aplysia and other marine gastropods. The u-conotoxin
GIIIB, when injected intraperitoneally in rodents acts as a powerful paralytic agent, yet if
injected intracranially no detectable symptoms are observed. In contrast, uO-conotoxin
MrVIA is inactive peripherally in rodents, but induces ataxia andor reversible coma
within minutes when injected intracranially. This indicates that MrVIA is potent in the
mammalian central nervous system, whereas GIIIB targets the skeletal muscle subtype
with high affinity yet does not affect central sodium channels (Mclntosh et al., 1995).
Two peptides from C. pennaceus, PnIVA and PnIVB, show the same disulfide¬
framework as MrVIA and MrVIB, but have different amino acid compositions. These
also block Iya when applied to Aplysia and Lymnaea neurons (Fainzilber et al., 1995). It
thus appears that these toxins would also fall into the uO-conotoxin family. Each of these
peptides and those yet to be found will also help to bring about subtype-specific ligands
for further research of Na channels.
8-Conotoxins
8-Conotoxins are a category of conopeptides, which also affect voltage-gated Na"
but produce a different response by the targeted channels. These toxins bind to Na
channels and act to disrupt the inactivation process. 8-conotoxin PVIA, isolated from C.
purpurascens, caused hyperactivity in fish followed by spasms, but without death or
paralysis. This toxin has been found to complement the action of other conotoxins in
contributing to the in vivo effect on the fish upon injection by the snail (Terlau et al.,
1996).
Effects of Conus striatus venom on voltage-gated K channels
in frog sympathetic ganglia neurons
Abstract
Cone snails are predatory marine gastropods of the genus Conus which
proliferated to over 500 species world wide largely through the use of novel peptide
toxins (conotoxins) that are injected into prey items through a harpooning apparatus.
Conotoxins have evolved for maximal effect on prey species hunted by each cone snail
species (Duda and Palumbi 1999). All cone snail venoms examined thus far contain
peptides that act with great specificity on a variety of channels including K, Na', Ca,
and ACh receptors (Olivera et al. 1999).
Conus striatus is a fish-hunting snail indigenous to the Indo-Pacific region.
Venom samples were "milked" from the snails and were further diluted during
experiments. Voltage-dependent K' currents were recorded from frog sympathetic
ganglia (SG) neurons using the whole-cell patch-clamp technique. When applied to the
bathing solution, the venom was found to induce a reduction in Ig during a voltage step.
This reduction was evident at venom dilution levels equivalent approximately eight times
higher than those shown to elicit repetitive action potentials in frog nerve-muscle
preparations (Schulz and Gilly, unpublished results). It is thus possible that the block of
voltage-gated K' channels in frog SG neurons by the venom component may be involved
in this neuroexcitatory activity.
Further studies have shown that Ig in these cells was found to be insensitive to
externally applied S-nitrosodithiothreitol (SNDTT), a compound that shows a high
degree of specificity for block of the Kvl subfamily (Brock, Mathes, and Gilly, 1997).
Additionally, the venom-sensitive current was found to overlap with that blocked by
externally applied tetraethylammonium (TEA), which has been shown to selectively
inhibit delayed rectifier (DR) Ix in frog SG neurons (Spruce et al., 1987 and Klemic, et
al., 1998).
Together, these results suggest the existence of a previously undescribed blocker
of DR Ix in C. striatus venom which functions in a manner most similar to snake-derived
dendrotoxins in frog peripheral neurons (Anderson and Harvey, 1985) and is not
targeting Kvl-type channels.
Introduction
Peptidic components of the venoms of predatory Conus species have been used in
previous studies to characterize external regions of ion channels because of the great
affinity and specificity they show for binding sites. Many of these conopeptides target
voltage-gated calcium and sodium channels, as well as the acetylcholine receptor. In this
study, voltage-gated K channels were affected by some component of the raw venom of
Conus striatus in a way that has not been previously described. An active component in
this venom induces a reduction in steady-state Ig in frog SG very similar to that seen with
externally applied tetraethylammonium (TEA) (Spruce et al., 1987), as well as by snake¬
venom derived dendrotoxins in the peripheral nervous system (Anderson and Harvey,
1985).
Previous work showed that application of a 100 nM concentration of peptide
purified from C. striatus venom induced repetitive activity in frog neuromuscular prep
and repetitive action potentials when applied to frog sympathetic ganglion neurons (Craig
et al., 1998). In the same study, this peptide was also presented as a blocker of Shaker K-
channels at a concentration of 2.5 M, which led investigators to propose K channel
block as a means of modifying action potentials in the peripheral nervous system. It was
hypothesized that K channel block in the sympathetic ganglia could be the mechanism
by which spontaneous activity in the neuromuscular prep was being produced.
Other studies focusing on K currents in frog SG have also explored block of Ik
and its relationship to action potential firing activity. Up to five distinct voltage-activated
K channels have been identified in frog SG cells. Together, all five can act to produce
hyperpolarization and repolarize the action potential. Each, however, has a conductance
of a different size, speed and voltage threshold and act together to create cell electric
properties (Adams and Galvan, 1986). Studies using 3,4-diaminopyridine (DAP) (Goh,
Kelly, and Pennefather, 1989) as a specific blocker of DR Ig in SG neurons have pointed
to a mechanism in which channel block of this type leads to attenuation of action
potential spikes in a train of repetitive firing. Block of DR Ix by TEA leads to spike
broadening in the SG, as repolarization is inhibited (Spruce et al., 1987). The block of Ix
by these compounds alone, however, does not seem to be sufficient to initiate repetitive
firing in the SG neuron, as models have shown that IM (current with muscarine
antagonist) and IAHP (K current after hyperpolarization) are more likely responsible for
this mechanism because of their greater activation at more-negative membrane potentials
(Adams et al., 1986).
Dendrotoxins are small peptides isolated from the venom of the snake mamba
genus, Dendroaspis. Originally, it was found that the venom of the green mamba,
Dendroaspis augusticeps, contained a toxin that increased quantal release of
acetylcholine at the neuromuscular junction, leading to excitotoxic paralysis. Additional
dendrotoxins have been shown to selectively block K channels, including cloned
members of the Kvl subfamily, with high affinity. These have made them useful in
studies of K channel structure and physiology (Harvey, 2000). The selective block of
DR channels in the frog peripheral nervous system by dendrotoxins is very similar to that
seen with C. striatus venom and frog SG Ix reduction. Building on this, experimental
evidence in this paper shows that Ix in SG neurons tested were insensitive to SNDTT, a
compound that selectively blocks the Kvl subfamily of K channels (Brock, Mathes and
Gilly, 1997). This suggests that those channels affected in the SG neurons are not of the
Kvl type.
These results point to the existence of a previously undescribed blocker of frog
SG K currents that acts specifically on TEA-sensitive DR channels.
Methods
Preparation of venom
C. striatus venom was obtained by the "milking" procedure. This technique was
performed by attracting the snail's attention with a fishtail fragment attached to a latex
membrane covering a small Eppendorf tube. Following the snail strike and injection of
venom through the fish tail and membrane, the tube was pulled from the animal. The
contents were then centrifuged and stored at -70'C along with a pool of venom collected
from the same snail within a similar post-captivity timeframe. This venom stock was
thawed, brought to final concentrations, mixed, and added to the reservoir during data
acquisition.
Cells
Neurons were isolated from caudal paravertebral sympathetic ganglia of adult
frogs (Rana pipiens). Immediately following decapitation and pithing, the ganglia were
removed and placed in a Ringer solution of the following composition (mM): 100 NaCl.
2.5 KCl, 4 MgCh, 10 Glucose, 10 HEPES. To this solution Collagenase A (3 mg/ml),
and Protease Type XIV (1-1.5 mg/ml) were added and the cells treated for 45 minutes.
Neurons were then teased onto small glass cover slips that had been treated with
protamine. These cells were then stored in 50:50 2.5 mM KCl, 120 mM NaCl, 2 mM
Cach, 5 HEPES, pH 7.2 frog Ringer/L15 culture medium at 4 ’C for up to one week.
Cells appeared most healthy 2 days following plating.
Electrophysiological Recording
Conventional whole-cell patch clamp experiments were performed with the frog
sympathetic ganglia neurons. External frog solution contained 5 mM KCl, 30 mM NaCl,
115 TMA-Cl, 1 mM CaCh, 4 mM MgCh, 10 HEPES with a pH of 7.2 (adjusted with
NaOH). Currents were recorded at room temperature using a voltage clamp amplifier and
a low-pass filter set at 15 KHz. Electrode resistance with 7052 glass was 1.0 to 3.5 MQ
when filled with an internal solution containing either (40 mM KF, 35 mM KCl, 40 mM
KAsp, 2 mM EGTA), or (0 mM F, 40 mM K-Glu', 40 mM K-Asp, 35 mM KCl, 1 mM
NazEDTA, 1.3 mM CaCh, 1 mM BAPTA, 2 mM Tris-ATP), both at a pH of 7.3. Results
did not appear to be influenced by which solution was used.
All currents were compensated for linear capacitive and ionic currents using a
standard P/4 subtraction protocol. Voltage commands were generated using software and
a direct memory access (DMA) interface developed by D. Matteson, University of
Maryland. Data acquisition and analysis were performed with the same program. Holding
potential was -80 mV throughout all experiments. Recordings were made at room
temperature (20-22°C).
Application of venom
The setup for frog sympathetic ganglia recording included an Air Cadet
vacuum/pressure station (Cole-Parmer Instrument) to allow exchange of the external
solution. A volume of 0.5 to 1.5 ml of the experimental solution was perfused into the
chamber with this system while aspirating the overflow. Recording chamber volume was
200 ml.
Results
Outward Ik is blocked by a C. striatus venom component
Control Ig (Fig. 1) was evoked in response to a 25-ms depolarization to +20 mV
and represents maximal activation of K conductance. In the frog sympathetic ganglia
neurons tested, K channel activation was generally at its peak at around 25 ms at each
voltage tested.
C. striatus venom did not act to alter the conductance versus voltage (G-V)
relationship for K currents. Conductance was measured and normalized for both control
and venom sensitive K+ currents in different cells. When conductance of venom-sensitive
K channels was following the subtraction of all Ix unaffected by venom from the control
Ik levels, no detectable difference was seen between the control and +venom curves at
either venom concentration tested (Fig. 2 A,B). Similarly, venom block of Ig occurred in
a voltage-independent manner, as the I-V plot of Icontol versus Lyenon over a series of
voltages gave a similar value for voltage at 50% conductance of around-5 mV (Fig. 3A).
Block induced by venom introduction occurred in a concentration-dependent
manner in which steady-state Ig was blocked by about 80% and 60% by 1:500 (Fig. 4A)
and 1:1000 (Fig. 4B) dilution venom, respectively. The full effect of block of Ig occurs
slowly at both concentrations tested for all neurons. Time to 50% of maximal block was
achieved, on average, approximately twice as fast with the higher concentration 1:500
dilution venom at 150-200 sec. for 1:500 to around 320 sec. For 1:1000 (Fig. 5 A,B).
Frog sympathetic ganglia Ik is insensitive to S-Nitrosodithiothreietol (SNDTT)
It was found in this study that control Ix and Ig in the presence of 10 mM SNDTT
do not show differences in either current amplitude or kinetics (Fig. 5B, 6). SNDTT is a
selective blocker of the Kvl family of K+ channels (Brock, Mathes and Gilly, 1997).
When 1:1000 dilution C. striatus venom was applied to the same cell that did not show a
SNDTT effect, Ix was reduced to a level of 45% of control over 16 minutes. The
convincing lack of effect of SNDTT in a cell that showed reduction of Ix following
venom application indicates that Kvl channels are not being targeted in these SG
neurons.
Sympathetic ganglia K currents targeted by venom component are TEA sensitive
When 10 mM TEA was applied to neurons of similar size and with similar K
current kinetics to venom-sensitive neurons from previous studies (Fig 7A), a decrease in
steady-state current to approximately 20% of control was seen (Fig. 7B). This TEA¬
sensitive current, which has been previously described as delayed rectifier current in SG
neurons (Klemic, Durand and Jones, 1998), would seem to correlate very closely with
Conus venom-sensitive Ig. In the Klemic study TEA-sensitive channels showed 50%
activation between -10 mV and 0 mV. Venom sensitive cells also activated within a
similar range (Fig. 2AB, 3B). These results point to delayed rectifier Ig as the probable
target of the Conus venom component.
Discussion
This study demonstrates that a component of the venom of Conus striatus blocks
Ix in frog sympathetic ganglia neurons. The mechanism of this block seems to involve
the direct interaction of the venom component with K channels in a manner that is very
similar to the block of delayed rectifier channels in frog SG by externally applied
tetraethylammonium ions as described by Goh et al. (1989). Similarly, this block, which
seems to target the non-inactivating, delayed rectifier-type K currents in SG cells has
been observed with a class of snake venoms, the dendrotoxins (Anderson and Harvey,
1985).
An increased understanding of the various K channel types seen in frog SG
neurons has emerged from the experiments performed in this study. In one cell the
application of 1:500 dilution venom showed no effect of K channel block (Fig. 5A). This
cell, designated with “60 pF cell", covered approximately twice as much surface area as
all other cells tested. The range for venom sensitive cells was from 20-35 pF. These
results suggest that different, venom-insensitive K channels exist in very large SG
neurons. The general insensitivity of the large, 60 pF cell to the venom component
suggests that at least two types of non-inactivating K channel types are present in the
ganglia, and can be distinguished based on cell size. This also fits into the categorical
designation of at least two "B" and one “C" cell types seen in the frog SG (Dodd and
Horn, 1983). Furthermore, the results involving K current insensitivity to externally
applied 10 mM SNDTT suggest that the Kvl family subtype is not present, at least in any
quantifiable amount, in these small, venom-sensitive SG neurons. This result is
interesting, as it has been shown that dendrotoxins have been found that are specific both
for Kyl channels and peripheral nervous system K currents in rat dorsal root ganglia
(Stansfelf and Feltz, 1988). Further studies should proceed with examining the sensitivity
of various Kyl channel clones to the venom component, to draw more conclusions about
the relation between this active component and dendrotoxins.
An apparent correlation between TEA- and venom-sensitive K currents was also
presented in this study. This is interesting in its implication that the venom component is
targeting delayed rectifier-type channels in the frog SG. Future studies would concentrate
on eliminating the IAHP, IM, and Ca“-activated channels from recordings using
pharmacological blockers specific for each. This would help to narrow down the probable
channel type that this venom component could be targeting.
Originally, the frog SG neurons were chosen as an assay to study C. striatus
venom sensitivity in order to correlate channel block with the spontaneous firing effect
evoked by application of 100 nM concentrations of Conus peptide to the frog nerve-
muscle prep (Craig et al., 1998). It has been reported that simply blocking delayed
rectifier Ig in the SG neurons is not enough to induce spontaneous action potential
generation (Goh, Kelly and Pennefather, 1989). Öther K channels that are activated
closer to membrane resting potential (IM and IAHp) must be modified to induce this effect
(Adams and Galvan, 1986). However, the possibility still exists that this K blocker effect
could be important for inducing such an effect. Further studies will have to focus on
eliminating these other K currents from the recordings using their known
pharmacological blockers (Block and Jones, 1996). While the milked venom was active
at a 1:1000 dilution, a dilution of-1:8000 has been shown to induce spontaneous firing in
the frog muscle preparation. This eight-fold difference is not too large to rule out an
overlapping effect for K channel block and spontaneous firing. Again, however, the long
delay that was observed that seems to be necessary for block of Ig is not consistent with a
rapidly-immobilizing excitotoxic shock effect produced by a C. striatus strike on its prey.
Similar to the effect of Conus venom, it was reported that dendrotoxins shown to block
SG Ig were also not easily washed out and took time to show an effect (Harvey and
Anderson, 1985). It is possible that this Conus toxic effect is working in secondary
manner to other active peptides also present in the venom. This multiple-target venom
attack would then be much like the secondary effect of dendrotoxins on Kchannels that
works in addition to the effect of inducing an increase in synaptic quantal release in the
neuromuscular junction (Harvey and Anderson, 1985).
Preliminary results studying the effects of C. striatus venom on Na- channels in
frog SG and dorsal root ganglia neurons indicated that a venom dilution of 1:1000
induced a -10 mV shift in channel activation. This shift in Na' channel conductance
could possibly give rise to the spontaneous action potential firing effect seen in the frog
neuromuscular prep, as channels would be shifted closer to their action potential firing
threshold. The Na" shifting effect and K channel block could work in conjunction to
produce the spontaneous firing.
With further isolation and purification of the active component of this C. striatus
venom, a greater understanding of the nature of the Conus attack will emerge. In
addition, pharmacological block of Ix by TEA and dendrotoxins have been used to mar
the pore regions as well as dissect the electrophysiological properties of various K
channel types over the years and have led to a greater understanding of the diversity seen
in this group of ion channels. With time, it is possible that this Conus venom K channel
blocking component could emerge as a means by which this complicated area of research
can further be developed.
References
Adams, P. and M. Galvan. 1986. Voltage-dependent currents of vertebrate neurons and
their role in membrane excitability. Adv. Neurol. 44:137-170.
Anderson, A. and A. Harvey. 1985. Electrophysiological actions of facilitatory toxins
from mamba venoms on mammalian neuromuscular junctions. Br. J. Pharmac.
86: 588P.
Armstrong, C.M. 1971. J. Gen. Physiol. 58: 413-437.
Block, B. and S. Jones. 1996. lon permeation and block of M-type and delayed rectifier
potassium channels. J. Gen. Physiol. 107: 473-488.
Brock, M.W., C. Mathes and W.F. Gilly. 1997. SNDTT: A novel open-channel blocker
of Kvl channels. Biophys. J. 78: A94.
Chahine, M., J. Sirois, P. Marcotte, L.-Q. Chen, and R.G. Kallen. 1998. Extrapore
residues of the SS-S6 loop of Domain 2 of the voltage-gated skeletal muscle
sodium channel (rSKMI) contribute to the u-conotoxin GIIIA binding site.
Biophys. J. 75:236-246.
Chen, C., and Okayama. 1987. High-efficiency transformation of mammalian cells by
plasmid DNA. Mol. Cell. Biol. 7:2745-2752
Choi, L.L., R.W. Aldrich, and G. Yellen. 1991. Tetraethylammonium blockade
distinguishes two inactivation mechanisms in voltage-gated K channels. Proc.
Natl. Acad. Sci. USA. 88:5092-5095.
Craig, A.G., G. Zafaralla, L.J. Cruz, A.D. Santos, D.R. Hillyard, J. Dykert, J.E. Rivier,
W.R. Gray, J. Imperial, R.G. DelaCruz, A. Sporning, H. Terlau, P.J. West, D.
Yoshikami, and B.M. Olivera. 1998. An O-Glycosylated neuroexcitatory Conus
peptide. Biochemistry. 37:16019-16025.
DeCoursey, T.E. 1995. Mechanism of K+ channel block by verapamil and related
compounds in rat alveolar epithelial cells. J. Gen. Physiol. 106:745-779.
Dodd, J. and J. Horn. 1983. A reclassification of B and C neurones in the ninth and tenth
paravertebral sympathetic ganglia of the bullfrog. J. Physiol. 344: 255-269.
Duda, T.F. Jr., and S.R. Palumbi. 1999. Molecular genetics of ecological diversification:
Duplication and rapid evolution of toxin genes of the venomous gastropod Conus.
Proc. Natl. Acad. Sci. USA. 96:6820-6823.
Dudley, S.C. Jr., H. Todt, G. Lipkind, and H.A. Fozzard. 1995. A u-Conotoxin¬
insensitive Na channel mutant: Possible localization of a binding site at the outer
vestibule. Biophys. J. 69:1657-1665
Fainzilber, M., T. Nakamura, A. Gaathon, J.C. Lodder, K.S. Kits, A.L. Burlingame, and
E. Zlotkin. 1995. A new cysteine framework in sodium channel blocking
conotoxins. Biochemistry. 34:8649-8656.
Harvey, A. 2001. Twenty years of dendrotoxins. Toxicon. 39: 15-26.
Harvey, A. and A. Anderson. 1985. Dendrotoxins: Snake toxins that block potassium
channels and facilitate neurotransmitter release. Pharmac. Ther. 31: 33-55.
Goh, J., M. Kelly and P. Pennefather. 1989. Electrophysiological function of the delayed
rectifier (Ig) in bullfrog sympathetic ganglion neurones. Pflugers Arch. 413: 482-
486.
Klemic, K., D. Durand and S. Jones. 1998. Activation kinetics of the delayed rectifier
potassium current of bullfrog sympathetic neurons. J. Amer. Physiol. Soc. 98:
2345-2357.
Mclntosh, J.M., A. Hasson, M.E. Spira, W.R. Gray, W. Li, M. Marsh, D.R. Hillyard, and
B.M. Olivera. 1995. A new family of conotoxins that blocks voltage-gated
sodium channels. J. Biol. Chem. 270:16796-16802.
Mclntosh, J.M., B.M. Olivera, and L.J. Cruz. 1999. Conus peptides as probes for ion
channels. Methods Enzymol. 294:605-624.
Miller, C. 1995. The charybdotoxin family of K-channel blocking peptides. Neuron.
15:5-10.
Olivera, B.M., D.R. Hillyard, M. Marsh, and D. Yoshikami. 1995. Combinatorial peptide
libraries in drug design: lessons from venomous cone snails. Tibtech. 13:422-426.
Olivera, B.M. 1997. Conus venom peptides, receptor and ion channel targets, and drug
design: 50 million years of neuropharmacology. Molecular Biol. Cell. 8:2101-
2109.
Olivera, B.M. 1999. Conus venom peptides: correlating chemistry and behavior. J. Comp.
Physiol. A. 185:353-359.
Shon, K.-J., B.M. Olivera, M. Watkins, R.B. Jacobsen, W.R. Gray, C.Z. Floresca, L.J.
Cruz, D.R. Hillyard, A. Brink, H. Terlau, and D. Yoshikami. 1998. u-conotoxin
PIIIA, a new peptide for discriminating among tetrodotoxin-sensitive Na channel
subtypes. J. Neurosci. 18:4473-4481.
Shon, K.-J., M. Stocker, H. Terlau, W. Stuhmer, R. Jacobsen, C. Walker, M. Grilley, M.
Watkins, D.R. Hillyard, W.R. Gray, and B.M. Olivera. 1998. k-Conotoxin PVIIA
is a peptide inhibiting the Shaker K channel. J. Biol. Chem. 273:33-38.
Stansfeld, C. and A. Feltz. 1988. Dendrotoxin-sensitive K+ channels in dorsal root
ganglion cells. Neurosci. Lett. 93: 49-55.
Terlau, H., K.-J. Shon, M. Grilley, M. Stocker, W. Stuhmer, and B.M. Olivera. 1996.
Strategy for rapid immobilization of prey by a fish-hunting marine snail. Nature.
381:148-151.
Terlau, H., A. Boccaccio, B.M. Olivera, and F. Conti. 1999. The block of Shaker K
channels by k-conotoxin PVIIA is state dependent. J. Gen. Physiol. 114:125-140.
Diagram 1: Cone snails
500 species
X 50 peptides
harpoon
0.5 mm
eyesiphon
radule sac
venom bulb
and duct
proboscis with
harpoon in place
pharynx
foot
Figure 1: Basic venom effect. Externally applied Conus venom causes
time-dependent block of frog sympathetic ganglia Kchannels
Contol
+ venom




20 m
80—
k
25 ms
Figure 2: Conus venom does not alter the Kconductance
versus voltage relationship
A
G/Gmax

-60 -40
G/Gmax
-40
1:1000 venom conc.
1.2
0.8
0.4
0
-20
20
Voltage (mV)
1:500 venom conc.
1.2
0.8
0.6
045
282

0
-20
Voltage (mV)
20


Control
Venom-sensitive kk
40
60


Control
Venom sensitive lk
40
60
-50
Figure 3: Venom block of Ig occurs in a voltage-independent manner
Current vs. voltage relationship before and after venom application
12
10
Current (nA)
Control
8
+ 1:1000 venom




-30
-10
10
30
Voltage (mV)
Relative conductance in same cell
1.2
G/Gmax

0.8
Control
venom sensitive lk
04
0.2
g
40
-60 -40
-20
20
60
Voltage (mV)
Figure 4: Concentration dependence of steady-state Ig block
Control
A
Block
measured at
12:05 after
application
+ 1:500 venom
1 nA

F
w
20

-80 mV
25 ms
Conto
Block
B
measured
16:46 after
+ 1:1000 venom
application

2 nA
—

20
80 mV
25 ms
Figure 5: Time course of Ig block at two relative venom concentrations
Relative current vs. time 1:500 dilution
A

4
□ 60 pF cell
0.8
1:500 venom 41
Relative lk
A 1:500 venom 42
0.6
0.4

0.2

-1000 -750 -500 -250
250 500 750 1000 1250 1500
Time (sec)
Relative current vs. time 1:1000 dilution
Sa

1 +10 mM SNDTT
1:1000 venom 41
Relative lk
A 1:1000 venom 42
0.6
1:1000 venom f3
0.4

0.2
80 400 200
0 200 400 600 800 1000 1200
Time (sec)
Figure 6: S-Nitrosodithiothreietol does not modify sympathetic ganglia Ig
Control


L


+ 10 mM SNDTT


4 nA
+ 1:1000 venom
—
—

20 mV
-80 -
+
25 ms
Figure 7: Sympathetic ganglia neuron Ig is TEA sensitive

Control
+ 1:1000 venom

2 nA

m
w
20 mV
+
-80 —
25 ms

Control
2 nA
+ 10 mM TEA


80
—.
25 ms
Figure Legends
Figure 1: Basic venom effect. Externally applied Conus venom causes time-dependent
block of frog sympathetic ganglia K channels. Ig evoked by 25 ms
depolarizations to +20 mV in the absence (control) and presence of venom.
Figure 2: Conus venom does not alter the K conductance versus voltage relationship.
A) Conductance (G) is measured and normalized to Gmax for all traces within a
voltage series at control and following 1:1000 venom application using the
equation G= I/(V-Vg) where Vx is the reversal potential for Ix measured by tail
currents for each individual cell. G.venon is calculated following subtraction of all
Ix unaffected by venom from the control Ig levels so that only venom sensitive G
is measured.
B) Same set of calculations performed for a cell at 1:500 venom concentration.
Variation is seen between the two cells in voltage at 50% conductance, but not
within each cell at Gcontrol VS. G4venom¬
Figure 3: Venom block of Ix occurs in a voltage-independent manner.
A) The current versus voltage relationship is measured between Icontol and Lavenom
traces. Block of Ig appears to be independent of voltage, as the voltages for 50%-
activation are about equal (approximately -5 mV).
B) Relative conductance measured in the same cell. Same set of calculations as
performed in Fig. 2 A,B. Conductance in this cell was nearly identical with and
without venom present.
Figure 4: Concentration dependence of steady-state Ix block.
A) Ix at +20 mV in the presence of 1:500 dilution of Conus venom. 80% of steady
state Ig blocked at 12:05 minutes after venom application.
B) Different cell Ig at +20 mV in the presence of 1:1000 dilution of Conus
venom. 44% of steady-state Ix was blocked 16:46 minutes after venom
application.
Figure 5: Time course of Ig block at two relative venom concentrations.
A) Steady-state Ig at different times was measured relative to that at venom
application (time 0) at a constant voltage for three different cells at a venom
concentration of 1:500. Block of Ix levels off at a constant level of about 20% of
control. 50% of maximal block occurs at around 150-200 seconds.
B) Same procedure followed with three cells exposed to 1:1000 venom
concentration. One cell (gray squares) was monitored in the presence of 10 mM
SNDIT. Block of Ix levels off at a constant level of about 40% of control. 50% of
maximal block occurs at about 320 seconds.
Figure 6: S-Nitrosodithiothreietol does not modify sympathetic ganglia Ik.
Ig at control and in the presence of 10 mM SNDTT do not show differences in
current amplitude or kinetics. Ix is blocked by 55% in the presence of 1:1000
venom concentration.
Figure 7: Sympathetic ganglia neuron Ix is TEA sensitive.
A) Approximately 55% of control IK is blocked in the presence of 1:1000 dilution
venom.
B) 10 mM externally applied TEA removes 80% of control IK.