Structure-function relationships of the major neurotoxin from the sea anemone Stichodactyla helianthus with a new sodium...


Material Information

Structure-function relationships of the major neurotoxin from the sea anemone Stichodactyla helianthus with a new sodium channel receptor site
Physical Description:
xi, 188 leaves : ill. ; 29 cm.
Pennington, Michael William, 1962-
Publication Date:


Subjects / Keywords:
Cnidarian Venoms   ( mesh )
Neurotoxins   ( mesh )
Sodium Channels   ( mesh )
Biochemistry and Molecular Biology thesis Ph.D   ( mesh )
Dissertations, Academic -- Biochemistry and Molecular Biology -- UF   ( mesh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph.D.)--University of Florida, 1988.
Statement of Responsibility:
by Michael William Pennington.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001071588
oclc - 20298848
notis - AFF6023
System ID:

This item is only available as the following downloads:

Full Text







This dissertation is dedicated
to my mother and father, to Kim Lassiter
and to the memory of Dr. Carolyn Bourne.


I would like to express my gratitude to my advisor, Dr. Ben

M. Dunn, and Dr. William R. Kem as well as the rest of the

members of my committee for their advice and assistance in the

preparation of this dissertation. In addition, I am indebted to

Dr. Jan Pohl, whose valuable insight into the characterization

of peptides as well as helpful suggestions, gave me a role model

to follow. Special thanks to all those who learned to work

around my rather bizzare schedule and still remained my friends.

Finally, I would like to thank the following people for their

help and/or support: Benne Parten, Jeff Weidner, Mark Carter,

Rohit Cariappa, Mike Campa, Dave Marriott, Susie Pennington-

Hammersley, John Hamilothoris, Kevin Krajniak, Tom Nutter, and

Dr. Peter Anderson.




ABSTRACT . . . x



Historical Elucidation of the Sodium Channel. ... 1
Identification of Toxin Binding Sites . 7
Tissue Localization of the Sodium Channel 20
Molecular Characterization and Cloning of the
Sodium Channel . . 22
Alpha-Scorpion Toxins . . 30
Sea Anemone Neurotoxins . . 43
Unanswered Questions. . .. 59


Introduction. . . .. 60
Experimental Procedures . . 63
Results . . 72
Discussion. . . .. 91


Introduction . . 98
Experimental Procedures . 100
Results. . . .. 108
Discussion . . 119


Introduction. . . .. 126
Experimental Procedures . . 130
Results . . 140
Discussion. . . .. 154


APPENDIX. . . .. 166
















[13C] NMR

Androctonus austrialis Hector

acetic acid

Anthopleura fuscovirdis

Anemonia sulcata

Asn analog of ShN at position 11

Asn analog of ShN at position 6

Asn analog of ShN at position 7

adenosine triphosphate

Anthopleura xanthogrammical

maximum theoretical binding sites

Bolecera tuediae

bovine serum albumin

carbon thirteen nuclear magnetic



octadecyl silica

circular dichroism


Condylactus gigantea

counts per minute

Centruroides sculpturatus Ewing

Centruroides suffusus suffusus






Condyl III




















1,1,1 trichloro-2,2 bis


degrees Celsius

distilled water


effective dose for 50% effect

ethylenediaminetetra-acetic acid

antigen binding domain of digested


femptomoles (10-15 moles)

Gin analog to ShN I at position 8

glutathione, reduced form

glutathione, oxidized form


hydrochloric acid


ethanesulfonic acid

anhydrous Hydrogen Floride

Heteractis macrodactylis

Heteractis paumotensis

high pressure liquid chromatography

insect toxin isolated from Buthus


displacement of 50% of labeled




















Na channel



dissociation constant


lethal dose for 50% of the animals

Lieurus quinquestriatus


moles per liter

microgram (10-6 g)

microliter (10-6 1)

micromolar (10-6 moles/l)

micromole (10-6 mole)

milligram (10-3 g)



sodium channel

radioactive isotope of sodium

N-acetyl Lys analog of ShN I at

position 4

nanomolar (10-9 moles/1)

nanomole (10-- mole)

nuclear magnetic resonance


phosphate buffered saline

-log [H+]

dissociation constant of a proton

picomole (10-12 moles)

proton nuclear magnetic resonance









[lH] NMR










Tris HCl

Tris OAc









Stichodactyla gigantea

Stichodactyla helianthus


trifluoroacetic acid

Tityus serulatus toxin r

Tityus serulatus toxin





tritiated derivative of


tritiated deriative of saxitoxin

tritiated derivative of tetrodotoxin


two-dimensional nuclear magnetic


analog incorporating Tyr at first

position of ShN I

void volume

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy




December, 1988

Chairman: B.M. Dunn, Ph.D.
Major Department: Biochemistry and Molecular Biology

We have determined that ShN I, a 48-residue type 2 sea

anemone toxin, delays the inactivation of the Na channel in

lobster olfactory somas. The effects of this toxin are similar

to those observed for the a-scorpion toxins and type 1 sea

anemone toxins, yet ShN I is structurally distinct from the type

1 toxins. Using a polyclonal antibody prepared against ShN I,

no crossreactivity was observed to exist between the type 1 and

type 2 anemone toxins.

The receptor for ShN I was identified in vesicle

preparations of neuronal tissues from both crustaceans and

mammals; however, the KD value for the former is more than 1,000

fold lower than for the later. The binding of [1251]-ShN I to

this receptor was determined to be unaffected by Anemonia

sulcata II, depolarization of the membrane, or veratridine.

ShN I was unable to displace [1251]-Androctonus austrialis

Hector II, whereas unlabeled AaH II and As II displaced the

labeled scorpion toxin from rat brain synaptosomes. This is the

first characterization of a new Na channel receptor site which

specifically binds type 2 anemone toxins.

To study the interactions that specific amino acid residues

of ShN I have with this receptor, we developed a strategy using

solid phase peptide synthesis. Prior to the synthesis of

analogs to ShN I, we assembled the native ShN I sequence and

reoxidized the three intramolecular disulfide bonds. Chemical,

physical, and pharmacological characterization of the purified

synthetic ShN I showed it to be indistinguishable from the

natural toxin.

Monosubstituted analogs of ShN I were synthesized to probe

the high density of charged residues localized in the N-

terminal region of the molecule. Following oxidation of the

disulfide bonds, purification and characterization, these

analogs were tested in vivo on fiddler crabs and in vitro on

vesicles prepared from crab walking leg nerves. Both of these

studies implicated the tri-anionic region of Asp6-Asp7-Glu8 as

essential for activity. Substitutions at the positions of Lys4

and Asp11 resulted in analogs with markedly reduced activity and

increased K0.5 values. Substitution of tyrosine for the N-

terminal residue had almost no effect on either of these values.



Historical Elucidation of the Sodium Channel

Action potentials are the rapidly propagating electrical

messages that travel along axons of the nervous system and over

the surface of some muscle and glandular cells. In axons, they

are characterized as being short-lived signals, which travel at

a constant velocity and maintain a constant amplitude (Hille,


The axon can be considered as a cylinder of axoplasm

surrounded by a continuous surface membrane. The membrane is

essential in that it serves as a barrier for the development of

an ionic gradient. Bernstein (1902, 1912) was the first to

develop a hypothesis considering the membrane to be central in

the propagation of an action potential. The membrane potential

is defined as the intracellular potential minus the

extracellular charge. Measurement of the membrane potential can

be made with a special glass microelectrode filled with a stock

solution of 3 M KC1. This microelectrode can be carefully

passed through the cell membrane of a neuron to measure the

potential (Hille, 1984).

Cole and Curtis (1939) were the first to record the changes

in the electrical conductivity of the membrane from the squid



(Loligo forbesi) giant axon. They observed that the membrane

undergoes a large change in conductance which occurs in the same

time scale as the electrical change. This experiment supported

the theory of an increase in ionic permeability but did not

determine the ions involved.

Hodgkin and Huxley (1939, 1945) and Curtis and Cole (1940,

1942) were the first to measure the full action potential of an

axon using an intracellular recording microelectrode. Shown in

Figure 1:1, an action potential develops following an electrical

stimulus to the neuron. The action potential develops

approximately 0.5 msec after the stimulus at which time the

membrane is observed to depolarize. As the action potential

propagates down the length of the axon, the cell then

repolarizes. This response is characteristic of all

electrically excitable tissues and it is always an all-or-none

response (Hille, 1984).

An unexpected result observed in these classical experiments

was that the action potential overshot zero and reversed in sign

to positive membrane potential values (Hodgkin and Huxley, 1939,

1945; Cole and Curtis, 1940, 1942). This positive overshoot was

not understood until Hodgkin and Katz tested their sodium

permeability theory (Hodgkin and Katz, 1949). This theory

incorporated previously reported results that the membrane

potential at resting values in squid giant axon was primarily

0 1 2
50 -



0 1 2


closed open inactivated closed

0 1 2
time (ms)

Figure 1:1: Action potential recording following a brief pulse
of current that partially depolarizes the membrane. In the
middle graph, the response of the Na channel opening and
subsequent inactivation are shown in the solid line. The broken
line shows the membrane response in the absence Na channels.
The bottom illustration shows the sequence of resting, open and
closed (inactivated) states following a stimulus. Taken from
Alberts et al., 1983).


due to selective potassium permeability (Goldman, 1943).

Following a stimulus, the positive overshoot of an action

potential resulted from the inward passage of Na+ into the axon

through the membrane, causing a change in the resting membrane

potential to values near that of the equilibrium potential of

Na+. Decreasing the Na+ concentration in the extracellular

medium with nonpermeable salts such as choline chloride or other

small molecules such as glucose decreased the amplitude of the

action potential (Hodgkin and Katz, 1949). These results

determined Na+ to be the ion responsible for the permeability

changes observed.

Following these experiments, radioactive 22Na+ was used to

investigate the movement of Na+ across the membrane in squid

giant axon. Keynes and Lewis (1951), using axoplasm activation

analysis, determined that a single impulse resulted in the net

movement of 20,000 Na+ ions through 1 pm2. This value was shown

to be within experimental error of the number of ions necessary

to result in the 120 mV voltage change associated with an action

potential (Huxley, 1964).

The voltage clamp technique was developed at this time in

several labs (Marmont, 1949; Cole, 1949; and Hodgkin et al.,

1949, 1952). This technique maintains a constant potential

across the membrane with a feedback amplifier, allowing changes

in the ionic currents to be measured. Using this technique,

Hodgkin and Huxley (1952) were able to show that the action

potential in squid giant axon results from the time and voltage-

dependent increases in the axonal membrane permeability to Na+

and K+. These experiments showed that during a maintained

depolarization, the Na+ permeability initially increases for a

few msec, then returns to resting values. The K+ permeability

lags behind that of Na+ and is rising to the maximum value as

the Na+ permeability returns to resting values. As the resting

membrane potential is reachieved, the K+ permeability decreases

to resting values. The Hodgkin and Huxley experiments (1952)

showed that these changes can be separated into two voltage-

dependent processes: activation, which controls the rate and

voltage dependence of the Na+ permeability increase following a

depolarization, and inactivation, which controls the rate and

voltage-dependence of the subsequent return of the Na+

permeability to the resting level during a maintained

depolarization. This gives rise to the simplistic model system

which consists of three states or groups of states: resting,

activated (open), and inactivated (closed). The resting and the

inactivated states are both nonconducting processes which differ

from each other in that the inactivated state is refractory to

further stimulus until repolarization of the membrane.

Pharmacological studies using TTX, a natural product isolated

from puffer fish of the order Tetraodontiformes (Halstead,

1978), resolved the question concerning the actual mechanism by

which ions pass through the membrane. In voltage clamp

experiments, Narahashi et al. (1964) showed that TTX blocks the

Na current without affecting the K current on lobster giant


axon. The toxin exerted these effects at nanomolar

concentrations. Similar studies with STX, a natural product

isolated from dinoflagellates of the genus Gonyaulax responsible

for the "red tide" (Taylor and Selinger, 1979), showed identical

effects when tested on the electroplax of Electrophorus

electricus (Nakamura et al., 1965), lobster giant axon

(Narahashi et al., 1967), and frog node of Ranvier (Hille, 1967,

1968a, 1968b). The discovery of a specific K+ current blocking

compound, tetraethylammonium, by Tasahi and Hagiwara, (1957)

provided further evidence for the existence of separate

transporters (channels) for these two ions.

The size of the pore responsible for the passage of Na+ was

estimated by Hille (1971). In this study, the ability of

several nonmethylated organic cations such as ammonium,

guanidinium, hydroxylammonium, and hydrazinium to pass through

the axonal membrane was measured. The results of these

experiments determined a minimum pore size of 3 A by 5 A.

The pore mechanism for ionic transport was supported by

determination of the capacity of the channel. Using the voltage

clamp technique on squid giant axon to measure Na current and a

[3H]-TTX derivative to estimate Na channel density, a

conductance of 2.5-8.6 pS was measured (Levinson and Meves,

1975; Almers and Levinson, 1975). Other methods including patch

clamping, a technique which measures the ionic currents through

a small isolated patch of membrane (Sigworth and Neher, 1980;

Nagy et al., 1983), and Fourier transform analysis of Na current


fluctuations (Conti et al., 1975, 1976) determined values

between 4.1-18 pS. These values would result in ion transport

rates in excess of 107 ions/sec, which exceed those measured for

small antibiotic ion carriers such as valinomycin and gramicidin

with measured rates of 104 ions/sec (Stark et al., 1971).

Selectivity for Na+ ions is another important property of the

Na channel. Of the rare earth ions, Na+ is the most permeable,

with K+ being only 8% as permeable as Na+. The other rare earth

ions, Rb+ and Cs+, are even less permeable than K+ (Chandler and

Meeves, 1965; Moore et al., 1966). The selectivity of the Na

channel was postulated to reside inside the pore of the channel

as an ion selectivity filter which would block the passage of

molecules with dimensions greater than 3 A by 5 A (Hille, 1971,


Identification of Toxin Binding Sites

Following the electrophysiological dissection of the action

potential, the discovery of a Na+ specific channel was not fully

accepted until the discovery that TTX specifically blocks Na

current reported by Narahashi et al., (1964). The

pharmacological and biochemical characterization of the Na

channel was initiated with the compounds TTX and STX, and

greatly advanced through the discovery of other toxins which

bind with high affinity and specificity to the Na channel.

These toxins which act on excitable tissues have been given the

name neurotoxins. The number of neurotoxin binding sites on the


Na channel continues to increase with the discovery of new toxic

compounds. The neurotoxins have been classified according to

the effects which they exert on the Na channel.

The first class of toxins consist of the guanidinium toxins

STX and TTX as well as the p-conotoxins, which are a series of

polypeptides isolated from the marine snail Conus geographus

(Sato et al., 1983; Cruz et al., 1985). The structures for the

guanidinium toxins are shown in Figure 1:2, and the sequences of

the p-conotoxins are shown in Figure 1:3. The guanidinium

toxins act in very low concentrations (nM) at a small number of

discrete sites on the membrane. Initial attempts at

quantitating the number of sites was based upon bioassay of the

bathing medium following blockage of the action potential in

lobster giant axon. The results of the experiments suggested an

upper limit of 13 sites/pm2 of axonal membrane (Moore et al.,

1967). Second generation experiments to determine the number of

binding sites used radiolabeled derivatives of both TTX and STX.

Using tritium exchange, exchangeable protons were replaced with

tritium to yield [3H]-TTX and [3H]-STX derivatives with specific

activities of 10 Ci/mmol (Ritchie et al., 1976). Binding

studies with these radiolabeled derivatives determined saturable

binding components with KD values from 1 to 10 nM for all the

tissues studied (Ritchie et al., 1976; Ritchie and Rogart, 1977

a,b; Catterall and Morrow, 1978; Strichartz et al., 1979). More

recently, studies on Na channel in cardiac tissue have shown at

least two types of Na channel: one with high affinity for STX






Figure 1:2: Molecular structures of the guanidinium toxins,
Tetrodotoxin (TTX) and Saxitoxin (STX). Taken from Hille, 1984.



Figure 1:3. Amino acid sequences of the p-conotoxins. Taken
from Cruz et al. (1985).


(Renaud et al., 1986), and a second type which has a

significantly higher KD value for STX and TTX (1 pM) (Catterall

and Coppersmith, 1981a, 1981b).

The p-conotoxins have been shown to have no effect on either

neuronal (both rat and lobster) or cardiac (rat) forms of the Na

channel. Competition assays have shown that p-conotoxins

displace [3H]-STX from rat skeletal muscle membranes and

electroplax membranes. The KD values for the p-conotoxins GIIIA

and GIIIB were determined to be 50 nM for electroplax membranes,

KD values of 25 and 140 nM, respectively, were determined for

rat skeletal muscle membranes (Moczydlowski et al., 1986). The

p-conotoxin GII was determined to have a K0.5 values of 35 nM on

rat T-tubular membranes and 60 nM on rat skeletal muscle

homogenates (Ohizumi et al., 1986). These results indicate a

tissue specific distribution of different Na channel types.

The binding site for the guanidinium toxins has been proposed

to reside at or near the ion selectivity filter (Kao and

Nishiyama, 1965; Henderson et al., 1974; Hille, 1975). The ion

selectivity filter had been shown to contain an essential

carboxyl group with a pKa value of 5.4. This acidic group must

be ionized for channel function (Woodhull, 1973). Similarly,

TTX and STX binding are completely inhibited by protonation of

an acidic group with a pKa value of 5.4 (Henderson et al., 1973;

Henderson et al., 1974; Balerna et al., 1975). Experiments

where this carboxyl group was modified through carbodiimide

mediated amidation (Shranger and Porfera, 1973) or alkylated


with trialkyloxonium salts (Reed and Raftery, 1976; Baker and

Rubinson, 1975, 1976) determined that the modified Na channel

was unable to bind TTX. Furthermore, following treatment with

trimethyloxonium, the Na channel is irreversibly modified;

however, the ability to generate action potentials is not

abolished. It was decreased to 40% of the original value

(Sigworth and Spalding, 1980). This implies that the reactive

carboxyl group is close to the pore, but it does not determine

ionic selectivity. Following modification, an unmodified

carboxyl group still exists within the pore which can be

protonated with acid solutions to block Na channel function.

Therefore, at least two carboxyl groups have been implicated in

the ion selectivity filter. When modification reactions were

performed in the presence of saturating amounts of TTX,

alkylation of these carboxyl groups was blocked (Spalding,


The guanidinium toxins are believed to bind to the Na channel

receptor through the interaction of the charged guanidinium

group with a carboxyl on the extracellular face of the

transmembrane pore (Kao and Nishiyama, 1965; Hille, 1968a),

which may possibly be part of the ion selectivity filter (Hille,

1971). These molecules function by blocking the Na current

directly without altering the properties of activation or

inactivation (Narahashi et al., 1964; Nakamura et al., 1965;

Hille, 1968). Binding of these compounds is not affected by

voltage (Almers and Levinson, 1975; Catterall et al., 1979) or


by other toxins which bind and alter different properties of the

channel (for review see Catterall, 1980). In tissues with high

affinity TTX and STX binding sites, no interactions between

binding and gating have been found (Catterall, 1980; Ulbricht,


The p-conotoxin family is a group of very basic 22 residue

peptides containing three disulfide bonds. The high content of

Arg and Lys residues suggests the interaction with a carboxylate

group or groups on the Na channel. The effects induced on the

Na channel resemble those of an STX blocked channel

(Moczydlowski et al., 1986). The selectivity exhibited by these

toxins for muscle or electroplax sodium channels suggests that

they share a common or overlapping receptor with STX in these

tissues not found in neuronal tissues (Ohizumi et al., 1986;

Moczydlowski et al., 1986).

Neurotoxins of the second type are lipophilic compounds

isolated from several plants, a tropical frog or synthesized in

the laboratory. These compounds include BTX, veratridine,

aconitine, grayanotoxin, insecticidal pyrethrins, DDT and

allethrin (Figure 1.5) (for review see Catterall, 1980; Hille,

1984). Studies with these compounds on several different

tissues leads to hyperexcitability and depolarization of the

excitable membrane (Schmidt, 1960; Peper and Trautwein, 1967;

Deguchi and Sakai, 1967; Ulbricht, 1969; Albuquerque et al.,

1971; Herzog et al., 1974). All of these compounds have been

found to shift the membrane potential of activation of the Na




C-CH C-0 CH3
CH/ 17

Allethrin I

Figure 1:4: Lipophilic alkaloid toxins acting at the site II
receptor. Taken from Hille (1984).






Grayanotoxin I



, I


channel to more negative values, block inactivation, and

reduce ion selectivity. The net result of these effects is the

persistent activation of the Na channel. The guanidinium toxins

TTX and STX have been shown to block the effects of these

compounds noncompetitively (Peper and Trautwein, 1967; Ulbricht,

1969; Albuquerque et al., 1971; Narahashi et al., 1971; Ohata et

al., 1973; Schmidt and Schmitt, 1974; Catterall, 1975; Seyama

and Narahashi, 1981; Vijvenberg et al., 1982).

Interestingly, the agonist effects of these toxins are all

different. As shown by Catterall (1975, 1977a), BTX is a full

agonist of the channel in neuroblastoma cells activating

approximately 95% of the cells. The other compounds

(grayanotoxin, veratridine and aconitine) activated

approximately 51%, 8%, and 2%, respectively. Competition

displacement experiments have shown that all of these compounds

interact competitively with a single class of binding sites in

neuroblastoma cells (Catterall, 1977a) and rat brain

synaptosomes (Ray et al., 1978). Studies with several different

local anesthetics such as dibucaine and diphenhydramine have

shown that [3H]-BTX-benzoate, a chemically modified derivative

of BTX, is competitively displaced from the site II receptor by

these compounds (Creveling et al., 1983). This binding site is

believed to reside within the membrane spanning region of the Na

channel which would exploit the hydrophobic properties that all

of these toxins possess (Catterall, 1980).


An allosteric model for the function of these lipophilic

neurotoxins has been proposed by Catterall (1977b). This model

utilizes the ideas introduced by Monod, Wyman and Changeux

(1965) for allosteric interactions in enzymes. The model is

based on the assumption that all of these toxins bind with

higher affinity to an activated state of the Na channel and

cause a shift in the voltage-dependent equilibrium between the

active and inactive states. These toxins bind with high

affinity to an active state and translate the energy of binding

into shifts in the activation properties of the channel. This

causes the channels to remain active (open) at the resting

membrane potential (Catterall, 1977b).

The third class of toxins consists of the cc-scorpion toxins

and the sea anemone neurotoxins. Detailed description of each

of these two types of polypeptide toxins will be presented later

in the text. Each of these toxin types have been shown to bind

to the Na channel and delay the inactivation process

(Koppenhofer and Schmidt, 1968; Narahashi et al., 1969).

Studies on neuroblastoma cells have shown that both types of

toxins stimulate the flux of 22Na+ in the presence of the

alkaloid toxin veratridine (Catterall 1975; Catterall and

Beress, 1978; Jacques et al., 1978). Studies in vitro using

radiolabeled derivatives of both of types anemone and c-scorpion

toxins have shown that anemone toxins displace oc-scorpion

toxins, but the reverse is not true (Ray et al., 1978; Vincent

et al., 1980). Ray et al. (1978) have shown that a positive


heterotrophic cooperativity exists between the site II

lipophilic compounds and the oc-scorpion toxins. This

interaction results in stimulated transport of Na+, lower KD

values, and an increase in specific binding.

Pharmacological characterization of several neurotoxins

isolated from the venoms of Centruroides suffusus suffusus,

Centruroides sculpturatus Ewing and Tityus serrulatus resulted

in the characterization of a new toxin binding site (Jover et

al., 1980; Barhanin et al., 1982; Couraud et al., 1982). These

have been classified as the fl-scorpion toxins (sequences for

several of the P-scorpion toxins are shown in Figure 1:8).

These toxins have no effect on the inactivation process like the

cx-scorpion toxins. The pharmacological effects are manifested

in altering the activation of the Na channel. Voltage clamp

studies on frog myelinated nerve with C. sculpturatus Ewing

venom induced repetitive firing due to the appearance of an

abnormal Na current upon repolarization of the nerve (Cahalan,

1975). Purified toxins, C. suffusus suffusus II (Couraud et

al., 1982) and C. sculpturatus Ewing IVa (Wang and Strichartz,

1983), produced similar effects. Kinetic analysis of the Na

current after the addition of both oc-scorpion and fl-scorpion

toxins, showed that the same Na channels were modified

simultaneously by both toxins (Wang and Strichartz, 1983).

Binding experiments on rat brain synaptosomes with a

radiolabeled derivative of TiTXr, [1251]-TiTxr, have shown that

Css II displaces [125I]-TiTxr with a K0.5 of 900 pM. The KD

value for [1251]-TiTxr on this same preparation was determined

to be 4 pM. Binding of the P-scorpion toxins has been shown not

to be affected by any of the other types of toxins. Depolar-

ization of the membranes, an effect which abolishes c-scorpion

toxin binding, has no effect on P-scorpion toxins (Barhanin et

al., 1982). Similar results have been obtained using an [1251]-

Css II derivative on electroplax membranes (Wheeler et al.,

1982); [1251]-TiTxT has been shown to bind to cardiac Na channel

(KD value of 15 pM) (Lombet and Lazdunski, 1984) as well as to

skeletal muscle membranes with a KD value of 10 pM. However,

this same derivative failed to bind to T-tubule membranes

(Barhanin et al., 1984). Thus, TiTxr has the highest affinity

for the neuronal, cardiac and surface skeletal forms of the Na

channel of any toxin isolated at this time.

The fifth class of neurotoxins which have been

pharmacologically characterized includes ciguatoxin and

brevetoxin. Ciguatoxin has been isolated from the

Gambierdiscus toxicus, the dinoflagellate that infects the reef

fish Gymnothorax javanicus (Chanteau et al., 1976). Structural

determination has not been reported for this toxin at this time.

Brevetoxins are isolated from the dinoflagellate Ptychodiscus

brevis (Baden et al., 1979). The structure of brevetoxin B is

shown in Figure 1:5. Ciguatoxin (Bidard et al., 1984) and

brevetoxin A (Catterall and Gainer, 1985) both bind to the Na

channel in neuroblastoma cells and cause repetitive firing as

well as stimulation of 22Na+ flux in the presence of the

Figure 1:5: Molecular structure of Brevetoxin B.
Nakanishi (1985).

Taken from


alkaloid site II or the site III sea anemone and c-scorpion

polypeptide toxins. The effects of this stimulated transport

are blocked noncompetitively by TTX. Brevetoxin causes a three

fold increase in the specific binding of [3H]-BTX-B to rat brain

synaptosomes at a concentration of 100 ng/ml (Catterall and

Gainer, 1985). Ciguatoxin has no effect on the binding of

radiolabeled f-scorpion toxin [125I]-TiTxr, a-scorpion toxin

[1251]-AaH II or sea anemone toxin [1251]-As II (Bidard et al.,

1984). The synergistic effects associated with the increase in

transport have identified this as the newest class of Na channel

specific toxins.

Tissue Localization of the Sodium Channel

The voltage-dependent Na channel has been identified in

several different types of electrically excitable tissues. The

Na channel must have evolved before the separation of the

vertebrate and invertebrate species (Hille, 1984). Thus, a

physiologically similar Na channel has been observed in

mollusks, arthropods, annelids, and vertebrates (Hodgkin and

Huxley, 1952; Julian et al., 1962; Frankenhauser, 1963; Goldman

and Schauf, 1973; Stampfli and Hille, 1976; Chiu et al., 1979).

A major difference between vertebrate and invertebrate nervous

tissue is the myelination of many vertebrate axons. Myelination

serves as an insulator increasing the cabling properties of the

nerve. This, in turn, causes the Na channels to be clustered at

the nodes of Ranvier, axonal hillocks and synaptic terminals


(Waxman and Ritchie, 1985). This clustering of Na channels

results in the characteristic saltatory nerve impulse seen in

myelinated nerves. The density of the Na channels in the nodes

has been estimated to exceed 103 channels/pm2 of membrane

(Neumcke and Stampfli, 1982). In contrast, the internodal

density is less than 25 channels/pm2 membrane (Ritchie and

Rogart, 1977a, 1977b) and synaptic terminals are approximately

25 channels/pm2 membrane (Ray et al., 1978).

Unmyelinated nerves are considered to have a uniform

distribution of Na channels, which causes a continuous mode of

impulse conduction. The densities in vertebrate unmyelinated

nerves (rabbit cervical vagus nerve) and invertebrate nerves are

very similar approximately 110 and 90 channels/pm2 of membrane,

respectively (Ritchie et al., 1976).

Other electrically excitable tissues in which the voltage-

dependent Na channel has been characterized include: eel

electroplax (Agnew et al., 1978), vertebrate skeletal muscle

(Barchi et al., 1979), rat T-tubular membranes (Kraner et al.,

1985) and cardiac tissue (Lombet and Lazdunski, 1984). Although

the channel densities in these tissues have not been determined

directly, binding of radiolabeled guanidinium toxins have

provided the following estimates: 560 fmol [3H]-STX/mg protein

in cardiac homogenates (Lombet et al., 1981), approximately 10

pmol [3H]-TTX/mg protein in sarcolemma homogenates (Barchi et

al., 1979), 228 pmol [3H]-STX/mg protein in eel electric organ

homogenates, and 6-10 pmol [3H]-STX/mg protein in T-tubule

membranes (Barchi, 1983). These values are approximately 1/3,

3, 125 and 3 times the [3H]-STX binding in rat brain

synaptosomes (Ray et al., 1978). The abundance of Na channels

in these membranes has been exploited to yield the purified


Molecular Characterization and Cloning of the Sodium Channel

The Na channel TTX receptor was initially isolated from the

electroplax organ of the eel E. electricus (Agnew et al., 1978).

Improvements in the purification were made by using a monoclonal

antibody affinity purification scheme. Using this methodology,

(Nakayama et al., 1982) the TTX receptor from electroplax was

purified to greater than 90%. Binding of [3H]-STX was observed

to decrease through the purification scheme. Agnew and Raftery

(1979) found that, following detergent solubilization, addition

of phospholipids through the remainder of the purification steps

preserved [3H]-STX binding. The purified TTX receptor was

observed to be a glycosylated single polypeptide component (Mr =

260,000) with no smaller subunits (Agnew et al., 1978).

More recently, using modern cloning technology Noda et al.

(1984) cloned cDNA sequences complementary to the mRNA coding

for the Na channel from electroplax. Analysis of the amino acid

sequence coded by the cDNA clone has determined the Na channel

to consist of 1,820 residues. Sequence analysis revealed four

repeated homology domains of approximately 300 residues which

are flanked by regions of non-homologous residues. The degree


of homology between the repeats is approximately 50%. This high

degree of homology supports the hypothesis that the repeat units

all arose from a common ancestral gene which would allow them to

adopt a similar secondary structure.

Secondary structure analysis of the Na channel sequence

allowed Noda et al. (1984) to propose a model. Each repeat

subunit was suggested to consist of six oc-helical segments, four

of these spanning the membrane while two were proposed to be

intracellularly located. The possibility of positioning all six

helices in the membrane was also suggested; however, the model

they proposed was constructed with only four transmembrane

helices per repeat. The four repeats were arranged into a

square array where the transmembrane pore is formed by the walls

of one of the membrane spanning cc-helices per repeat (Figure

1:6). Between repeat subunits II and III, a 200 residue stretch

is found which contains four equally spaced clusters of

negatively charged residues. Noda et al. (1984) proposed that

these negatively charged clusters interact with positively

charged residues present in each repeat subunit to form the

activation-inactivation gate which is sensitive to a gating


Identification of the Na channel in rat brain was first

accomplished using a radiolabeled photoactivatable derivative of

scorpion toxin. Covalent attachment of the [1251]-scorpion

toxin to the Na channel resulted in the labeling of two

components with molecular weights of 250,000 and 32,000

Figure 1:6: Molecular model of the eel electroplax Na channel.
Taken from Noda et al. (1984).


(Beneski and Catterall, 1980). Labeling of these components was

completely blocked when the rat brain synaptosomes were

depolarized or incubated with excess unlabeled scorpion toxin.

The purification of the Na channel from rat brain was similar

to that established for electroplax. An absolute requirement

for phospholipid during the purification was also observed. The

purification resulted in a 1380-fold enrichment of [3H]-STX

binding over the starting homogenate. The specific activity was

determined to be 2910 pmol [3H]-STX/mg protein or 0.9 mol [3H]-

STX per mol of Na channel (Hartshorne and Catterall, 1981,


The isolated channel was composed of a large 260,000 MW

component and two smaller components of 39,000 and 37,000 MW

(Hartshorne and Catterall, 1984). These have been classified as

the =, fl and f2 subunits, respectively. All three of these

subunits are heavily glycosylated. The carbohydrate content by

weight constitutes 20% for the a subunit and 36% for each of the

two small subunits (Messner and Catterall, 1985). The rat brain

Na channel has been proposed to have of a subunit stoichiometry

of 'l(1)1(#2)1 (Hartshorne and Catterall, 1984). The P2

subunit has been determined to be disulfide bonded to the large

c subunit (Hartshorne et al., 1982), whereas the fl subunit is

noncovalently associated with the o subunit (Hartshorne and

Catterall, 1981). Reconstitution of the solubilized Na channel

(Tamkun and Catterall, 1981) and purified channel (Tamkun et

al., 1984) into lipid vesicles has shown that this protein does


mediate neurotoxin stimulated 22Na+ flux; however, a-scorpion

toxin binding properties are slightly different. Thus, both rat

brain and eel Na channels are composed of a large 260,000 Da

component, but the rat brain channel also has two smaller

subunits of 39,000 Da and 37,000 Da associated with the large


Cloning of two different rat brain Na channel mRNAs was

recently reported by Noda et al. (1986). Nucleotide sequence

analysis of the cDNA clones for these two messages revealed the

primary structures for rat brain Na channels I and II. The

amino acid sequences of each these clones coded for two large o

subunit proteins of 2009 (I) and 2005 (II) amino acids.

Furthermore, the sequences of these two show approximately 82%

homology to each other and 62% homology to the eel electroplax


Rat brain Na channels I and II each possess four homologous

repeat subunits of 300 amino acid residues. These regions in

the rat brain and eel electroplax are highly conserved whereas

the cytoplasmic domains show much lower conservation. Secondary

structural analysis of the rat brain channel has been used by

Noda et al. (1986) to construct a model. The model positions

the six o-helical sections (S1-S6) of each of the four repeat

domains as membrane spanning segments (Figure 1:7). This is

contrary to the electroplax model in which they proposed only

four transmembrane o-helices per repeat (Noda et al., 1984).

The transmembrane pore in the rat brain Na channel is

Figure 1:7: Molecular model of rat brain Na channel.
Noda et al. (1986).

Taken from


suggested to be formed by the walls of the S2 helix (Figure 1:7,

stippled helix) of each of the repeat subunits which are

positioned in a square array. Interestingly, the domain of

clusters of charged residues observed between repeats II and III

of electroplax Na channel are not as prevalent. This newer

model of the transmembrane topology of the channel presented by

Noda et al. (1986) is based upon theoretical consideration of

the measured gating currents which suggest the movement of four

to six charged residues across the membrane to open the channel

(Hodgkin and Huxley, 1952). This suggests the intramembrane

location of many dipoles which only move small distances.

The functional significance of the small subunits present in

the rat brain and rat skeletal muscle Na channel is not known.

The cloned rat brain mRNA coded for the o subunit of two related

Na channels (Noda et al., 1986). Microinjection of cDNA

selected mRNA of the rat brain x subunit into Xenopus laevis

oocytes induced the synthesis of voltage sensitive, TTX

sensitive Na channels (Goldin et al., 1986). High salt

treatment or incubation at 37 *C selectively removes the fl

subunit from purified rat brain Na channels. This has been

shown to abolish [3H]-STX binding as well as to eliminate

stimulated 22Na+ flux in reconstituted vesicles (Messner et al.,

1986). More recently, Tejedor et al. (1988) have shown that

detergent solubilized rat brain Na channel, following high salt

treatment to dissociate the 61 subunit, reacted with

carbodiimides in the absence of added nucleophiles to form


intramolecular isopeptide bonds in the a-subunit. This

derivative was shown to have a 3- to 4-fold higher affinity for

[3H]-STX. Evidently, the isopeptide bonds are stabilizing the

conformation which binds [3H]-STX. This suggests that P1 is

involved in maintenance of the conformation of the Na channel as

opposed to directly binding [3H]-STX. Selective removal of the

P2 subunit by mild reduction of the purified Na channel has no

effect on either of the binding of [3H]-STX or [1251]-Lqq V or

22Na+ flux (Messner et al., 1986). The P2 subunit is believed

to be involved in the intracellular transport and/or membrane

insertion of the ap1i complex (Schmidt et al., 1985; Schmidt and

Catterall, 1986).

Cloning technology has also been exploited to isolate a gene

in the fruit fly Drosophila melanogaster which codes for a Na

channel (Salkoff et al., 1987). The gene codes for a protein

which possesses the identical patterning of four repeat subunits

of 300 amino acid residues. These repeat subunits contain the

same patterning of six a-helical segments that the other cloned

channels possess. Homology between the channel sequences was

55% and 51% when compared to rat and eel and sequences,

respectively. Sequences that distinguish each of the repeat

subunits are absolutely conserved between fly and vertebrate

proteins. This supports the theory that a common ancestral gene

underwent two rounds of gene duplication to give the modern oc

subunit structure (Hille, 1984; Salkoff et al., 1987).

Alpha-Scorpion Toxins

The venom delivered by the sting of scorpion is a fairly

complex mixture of components including: mucopolysaccharides,

phospholipases, hyaluronidase, protease inhibitors, histamine

releasers, serotonin, and neurotoxic polypeptides (Couraud and

Jover, 1987). Only the Buthidae family of scorpions produce

these neurotoxic secretions. This family is further divided

into four sub-families by geographical and morphological

considerations of which the Buthinae, Centrurinae and Tityinae

are the most dangerous (Bucherl, 1971). From these three sub-

families, two classes of neurotoxic polypeptides have been

identified and classified according to the properties of the Na

channel which they modulate. The c-scorpion toxins delay the

inactivation, and the f-scorpion toxins affect the activation of

the Na channel (Catterall, 1980).

Both classes of these scorpion polypeptides are similar in

that they are basic proteins, containing four intramolecular

disulfide bonds, with molecular weights of approximately 7,000.

In Figure 1:8, the sequences of fifteen different scorpion

toxins are shown and classified according to structural homology

(Rochat et al., 1979). The positioning of the cysteine residues

is absolutely conserved among all of the scorpion toxins

sequenced to date. Thus, the positioning of the disulfide bonds

is likely to be same in these toxins.

1 10 20 30 40 50 60 70






Figure 1:8: Amino acid sequences of scorpion toxins. Taken
from Kopeyan et al. (1985).


Although a small number of toxins have been isolated which

selectively paralyze insects (Zlotkin et al., 1971; Zlotkin et

al., 1979; Lester et al., 1982; Zlotkin et al., 1985) or

crustacean species (Zlotkin et al., 1975), the pharmacological

effects of the majority of the scorpion toxins appear to be

directed towards vertebrate species (Jover and Couraud, 1987).

Among the best studied of the vertebrate toxins are the a-toxins

AaH II, Lqq V and Tityustoxin and f-toxins Css II and TiTxr.

Electrophysiological, pharmacological and structural studies

with these toxins have provided much of the data for

characterization of the site III and IV receptors on vertebrate

Na channels (Catterall, 1980).

Molecular Structure of Scorpion Toxins. The first structure

determined by x-ray crystallographic analysis for the a-scorpion

toxin variant III from Centruroides sculpturatus Ewing at 3.0 A

resolution was reported by Fontecilla-Camps et al., (1980). The

structure was refined to a resolution of 1.8 A in a later report

(Almassy et al., 1983). A schematic representation of the a-

carbon backbone of this toxin is shown in Figure 1:9. The

tertiary structure consists of several loops protruding from a

dense core, that contains three of the four disulfide bonds. An

a-helical segment of two and one half turns from residues 23-32

and a short three-strand stretch of antiparallel f3-sheet are the

other prominent structural features of this toxin.

Figure 1:9: X-Ray crystal structure of CsE III. Taken from
Fontecilla-Camps et al. (1980).


The molecular structure of this CsE III toxin when compared

to the structures of the snake toxins erabutoxin (Low et al.,

1976; Kimball et al., 1979), a-cobratoxin (Walkinshaw et al.,

1980) and ac-bungarotoxin (Agard and Stroud, 1982) show minor

similarities. These snake toxins act post-synaptically by

binding to the acetylcholine receptor. Both the scorpion and

snake toxins are basic proteins containing between 60-75 amino

acid residues and either four or five disulfide bonds. They

exhibit the same protruding loop-type structure, which

originates from a disulfide rich core. However, the snake

toxins contain three extended hairpin loops that originate from

the core structure, where as the scorpion toxins only have one

major loop. Also the o-helical segment and antiparallel f-sheet

region are absent from the snake toxins (Almassy et al., 1983).

The molecular structure of CsE III has a relatively high

density of aromatic and hydrophobic amino acid residues located

on the front surface running from the top to the bottom of the

structure (Almassy et al., 1983). This hydrophobic patch

contains several conserved amino acid residues present in

scorpion toxins. These conserved residues occur at positions 1-

4, 47, 52, and 53 in the amino acid sequences of the scorpion

toxins. Most of these conserved residues form a hydrophobic

patch which nearly spans the width of the molecule (Fontecilla-

Camps et al., 1981).

Chemical modification studies on another a-scorpion toxin AaH

II suggest that this front hydrophobic surface may be involved


in the interactions by which the biological effects of scorpion

toxins are mediated. When a single highly reactive lysine

residue is alkylated, AaH II was inactivated (Sampieri and

Habersetzer-Rochat, 1978). This lysyl residue is in close

proximity to a surface oriented tyrosine residue in the

hydrophobic patch of CsE III (Almassy et al., 1983). Reduction

and methylation of a single disulfide bridge markedly reduced

activity (Habersetzer-Rochat and Sampieri, 1976). Modification

of the acidic residues in AaH II resulted in inactivation of the

toxin (Sampieri and Habersetzer-Rochat, 1976). Two of the

reacted acidic groups in AaH II are located on this front

surface patch of CsE III (Almassy et al., 1983). When any of

these chemical modifications on AaH II is taken into context

with the elucidated structure of CsE III, the front surface

region appears to be essential.

Nuclear Magnetic Resonance Structural Properties. Solution

structural properties determined by NMR (Krishna et al., 1983)

show strong similarities to the crystal structure of CsE III

(Fontecilla-Camps et al., 1980; Almassy et al., 1983). At 22

C, the molecule assumes a very well defined folded structure

which undergoes reversible thermal denaturation at 80 *C.

Studies on the solution structure of another cc-scorpion toxin,

Buthus eupeus IX, are in agreement with those reported for CsE

III (Pastikov et al., 1986). Perhaps the most interesting data

obtained by NMR concerns the 35-residue insect toxin, 15A,



Figure 1:10: Comparison of molecular structure for two different
types of scorpion toxins. A, Insect toxin, 15A, from Buthus
eupeus. B, Toxin variant III from Centruroides sculpturatus



isolated from the scorpion Buthus eupeus (Arseniev et al.,

1984). This toxin is presumed to act post-synaptically on the

insect glutamate receptor (Grishin et al., 1982). The primary

structure for 15A shows no similarity to the larger scorpion

toxins, but the spatial structures show many similarities

(Figure 1:10). The two and one half turn c-helical segment, the

three strands of antiparallel 6-sheet as well as the same

spatial packing of atoms makes these two toxins structurally

very similar. These structural elements have been proposed to

be responsible for the biological activity of the longer

scorpion toxins (Fontecilla-Camps et al., 1982). Therefore, the

structural similarities of long and short scorpion toxins are

not directly connected with the mechanism of action, and may

actually be related to a common ancestral gene (Arseniev et al.,


Immunological Properties of Scorpion Toxins. Scorpion toxins

have been classified into five classes based upon structural

homology (Rochat et al., 1979). These five classes consist of

four x-toxin groups and one #-toxin group. Furthermore,

immunocrossreactivity studies have confirmed this structural

classification scheme for the four o-toxins (Delori et al.,

1981). This data was also corroborated using radioimmunoassay

techniques as well (Tessier et al., 1978; El Ayeb et al.,

1983a). Despite the structural and antigenic differences, the


a-toxins all bind to the Na channel at the same receptor site

(Rochat et al., 1979; Catterall, 1980).

Probing the molecular topology of the AaH II class of c-

toxins with antibodies has identified four antigenic domains of

the toxin. Initially, El Ayeb et al. (1983b) identified these

four domains based upon the binding stoichiometry of four Fab

fragments per toxin molecule. Predictions for exposed regions

by the method of Hopp and Woods (1981) determined four

hydrophilic regions present on AaH II (El Ayeb et al., 1983b).

Analysis of thirteen other a-toxins showed four similar

hydrophilic regions, suggesting that these domains occur in

homologous regions present in the topography of all of the

toxins despite variations in the amino acid sequences. These

epitopes consist of: (a) residues 23 through 32 which contains

the a-helical segment in CsE III (b) residues 30 through 41 (c)

residues 50 through 59 which contains a P-turn in CsE III (d)

and a discontinuous epitope consisting of residues 5 through 14

disulfide bonded to residues 60 through 64 (El Ayeb et al.,


Peptides representing each of these four antigenic sites have

been synthesized and these have been used to prepare

monospecific antibodies to each of these determinants (Granier

et al., 1984; El Ayeb et al., 1984; Bahraoui et al., 1986).

Studies with monospecific antibodies prepared to each one of

these sites have shown that each of these epitopes are

recognized on native AaH II. However, the monospecific antibody


which recognizes the a-helical segment failed to neutralize

toxicity in mice. This same antibody was also determined to

retain binding under conditions where AaH II is bound to the Na

channel receptor (Bahraoui et al., 1987). Other antibodies

directed against the discontinuous epitope containing the

disulfide bond or the a-turn region no longer bind to the toxin

under the same conditions (El Ayeb et al., 1986). These results

suggest that the c-helical segment is not directly involved in

the receptor-toxin complex, whereas the a-turn region and the

disulfide linked region are involved with the receptor binding


Alpha-Scorpion Toxin Receptor. The a-scorpion toxins affect

the mammalian neuromuscular system by interacting with the

voltage-sensitive Na channel. This interaction results in the

delayed inactivation of the Na channel (Koppenhoer and Schmidt,

1968). The result of delayed inactivation is hyperexcitability

and a massive release of neurotransmitters (Romey et al., 1976).

Detection of the a-scorpion toxin receptor in a number of

tissues was facilitated through the use of [1251]-labeled

derivatives of Lqq V or AaH II (Catterall, 1977; Rochat et al.,

1977). Binding analyses with these derivatives identified the

a-scorpion receptor in neuroblastoma cells (Catterall, 1976;

Bernard et al., 1977), rat brain synaptic particles (Ray et al.,

1978; Jover et al., 1978), rat skeletal muscle membranes


(Catterall, 1979) and chick cardiac muscle (Couraud et al.,


Identification of the a-scorpion toxin receptor utilized a

photoactivatable derivative of [1251]-Lqq V. This derivative

upon photolysis was covalently attached to the Na channel of rat

brain synaptosomes. Following polyacrylamide gel

electrophoresis and autoradiography, two components with

molecular weights of 250,000 (a subunit) and 32,000 (fl subunit)

were identified (Beneski and Catterall, 1980). A later study,

using the same photoactivatable [1251]-Lqq V derivative, showed

the a-scorpion binding site to be preferentially located on a

34,000 MW p1 subunit (approximately 85% of the label) and only

slight labeling of a large 300,000 MW (x~2 complex) component

was observed. In this same report, the large subunit upon

reduction shifted to a molecular weight 272,000 (a subunit)

without affecting the 34,000 MW (Pl subunit) band. No label was

observed to appear on the released 30,000 MW (02 subunit)

component (Jover et al., 1988). Thus, the a-scorpion receptor

appears to be associated with oai complex possibly near the

interface of these two noncovalently associated subunits.

Solubilization of the rat brain Na channel destroys the

binding of the a-scorpion toxins (Catterall et al., 1979). When

the solubilized Na channel is reconstituted into phospholipid

vesicles, binding of a-scorpion toxins is recovered. The

reconstituted Na channel possesses the three major toxin binding

sites (Tamkun and Catterall, 1981a, 1981b), and it facilitates


Na+ transport in the presence of veratridine (Talvenheimo et

al., 1982).

With the successes of purification and reconstitution of the

Na channel, conditions were established which either the 1il or

P2 subunits were selectively removed from the complex. These

experiments determined that removal of the P2 subunit, leaving

the cpl complex, did not affect c-scorpion toxin binding, [3H]-

STX binding or Na+ flux in the presence of veratridine.

However, selective removal of the P1 subunit abolished

cc-scorpion toxin binding as well as [3H]-STX binding. This ap2

complex was unable to mediate veratridine stimulated Na+ flux

(Messner et al., 1986). Thus, =cl complex appears to display

all of the functional characteristics of the Na channel.

The binding of a-scorpion toxins to the Na channel receptor

in rat brain synaptosomes (Ray et al., 1978; Jover et al., 1978)

and neuroblastoma cells (Catterall, 1976; Bernard et al., 1977)

have shown KD values of approximately 1 nM. The binding of the

a-scorpion toxins are greatly affected by the membrane

potential. Upon depolarization of neuroblastoma cells

(Catterall, 1977) and synaptic particles (Ray et al., 1978), the

KO.5 values increased significantly and the binding of [1251]-

Lqq V was inhibited. The a-scorpion toxin binding site showed

no competition with any other toxins acting on the Na channel,

except for the type 1 sea anemone toxins (Ray et al., 1978;

Jover et al., 1980; Schweitz et al., 1985).


A final characteristic of the a-scorpion toxin receptor is

the synergistic interaction this site experiences with the site

II lipophilic alkaloid toxins such as BTX and veratridine

(Catterall, 1977a; Ray et al., 1978). These two binding sites

exhibit positive heterotrophic cooperativity which results in

markedly increased binding of [1251]-Lqq V (4-fold enhancement

in the presence of BTX) (Ray et al., 1978). An allosteric model

has been presented by Catterall (1977b) to describe this

behavior. This model assumes two distinct states for the Na

channel: active and inactive. The alkaloid toxins bind

preferentially to the active state, scorpion toxins enhance

activation by alkaloid toxins by lowering the equilibrium

constant for the transition between the two states.

Angelides and Brown (1984) have mapped the distance between

the alkaloid site and the c-scorpion site using a fluorescent

derivative of BTX. In these studies, the effect of binding Lqq

V caused a 20 nm red shift in the fluorescence emission maxima

of the BTX derivative, indicating the BTX derivative was

experiencing a more hydrophilic environment. Fluorescence

resonance energy transfer measurements showed that the distance

separating the two receptors was approximately 37 A. Also,

using the same measurement techniques, the TTX receptor was 35 A

from the c-scorpion receptor. Upon binding of BTX, this

distance increased to 42 A indicating a conformational change

associated with the binding of BTX (Angelides and Nutter, 1984).


Sea Anemone Neurotoxins

The phylum coelenterata is composed of animals which exhibit

one of the simplest levels of organization in the Metazoa. All

members of this phylum possess stinging nematocysts or cnidae

which are used to paralyze prey to facilitate its capture. The

nematocyst consists of a coiled hollow thread, contained within

a chemically or mechanically triggered capsule. Upon

stimulation, the nematocyst discharges, and the thread uncoils

delivering toxins through the hollow thread (Rathmayer, 1979).

Richet (1903) was the first to investigate the toxic

components of the sea anemone Anemonia sulcata. Beress et al.

(1975) succeeded in isolating two polypeptides from this

anemone. Purification of toxins from several other anemones

followed the successful establishment of a purification scheme.

The primary structures for several of these toxins were

determined in a number of different labs (Fig. 1.11). As shown

in Figure 1:11, the sequences of eleven toxins are aligned such

that the positioning of the six cysteine residues is conserved.

These cysteine residues form three intramolecular disulfide

bonds in all of these toxins. Only the disulfide pairs of As II

have been chemically determined. These are between Cys 4 and

Cys 46, Cys6 and Cys36, and Cys29 and Cys47 (Wunderer, 1978).

It is believed that the same disulfide pairs are present in the

other toxins due to the conservation of the cysteine residue

positions in the sequences. Other conserved residues include

positions 6 and 8, Gly9, Prol0, Argl3, Glyl9, Trp30, and Lys47.


S0 0 0 a 04 9 9 9 w
M M M M M M M ps o
010 01 1 1 ~ ta ta ta taG
U0 U C. C. U C. .
4C. C. U C. U U u S u C o
!5 ru u u u ru a
000000 0
o 0 H > H H
E44 H E-4 E4 4 04 N H
04 4 4 0 0 0 U E-4 0 E-4 0
0 0 U C! t 0 0 > >4 >4
a a a a a >4 > >4 >4 a

z z z E4 4g
inM M M M M M M < E- < < *5
C.) U % 0 C U ) C C C C

N a,
S2 S H 2 2 > > > > 0
3 0 0 0 0 : : M: 33 a,

4 4 4 4:)A4 E- > 4
n 0

HU 4U O 40 04 2 2 2
UO C.) C. 0 C.) C. C. C. U C.)

0 C 0 0 0 0 A > a > >
400 0>

0>40 0 A
H H H H i 3 C) C) Q a,
> H E H H H > > > >
o 0 u u a u X g

NO 0 0 0 000 0 >4 E4 4 E4
(O M M W M W O) O U U U

U) Cl U) U U) U U) 0* 0 0 0
^4 E' U E-i E E4 E4 ^ ca~ ^ al
B~ 2B 2 2B 2 2 SB 04 E4< 04 04 44*

> > 14> >> 04 o ao
2 U) 2 U) U) U) r > > > H 0,
0440044004 >4 C )x 2 C) 0

Mco o o n n oo M M '

0 0 0 0 0 0 04 0 0 04 0
1-44 4 4 9- U )0
C) ) C) C) C C) C 0 0 00

> > > > > a E;(OS .<
reU U C C U 0 0 C < U C< C)

H ~H H
Lf .1 M .. s. M1 != I U C. U U

S044 04 04 U 04 U 0 U U


Also shown in Figure 1:11, is the classification of these

sequences into two distinct classes based upon sequence

homology. Among the members of each group, there exists

approximately 60% homology, and between the two groups there

exists approximately 30% homology. Immunochemical studies

support this classification system (Schweitz et al., 1985;

Chapter 2, this dissertation). Furthermore, the members of the

type 1 class all belong to the Actiniidae family and the type 2

class all belong to the Stichodactylidae family (Pennington et

al., 1987).

A third class of smaller sea anemone toxins containing 31

amino acids have been isolated from two different anemones.

Toxin III from Anemonia sulcata (Martinez et al., 1977) and PaTX

from Parasicyonis actinostoloides (Nishida et al., 1985) are the

only two examples of these small anemone toxins. Interestingly,

they are active only on crustaceans (Martinez et al., 1977;

Fujita and Warshina, 1980). Furthermore, these small anemone

toxins have four disulfide bonds as opposed to three in the

longer toxins.

Molecular Structures of Long Sea Anemone Toxins. Although

the crystal structure for an anemone toxin has not been

reported, a considerable amount of information about the

solution structures of several anemone toxins has been reported.

Initially, Prescott et al. (1976) used laser Raman spectroscopy

to investigate the structural properties of As II. They


reported that the characteristic absorption peaks for a-helices

and P-structure were not observed, and concluded that the

peptide backbone is predominantly disordered. Application of

laser Raman spectroscopy, CD, fluorescence emission, NMR

rotational correlation time, and Chou-Fasman rules on Ax I

suggested that this toxin was roughly spherical with numerous f-

turns, some #-sheet structure and possibly some c-helix

structure. The aromatic residues Trp23 and Tyr25 appear to be

exposed to the solvent, whereas one or more of the other Trp

residues appear to be buried (Ishizaki et al., 1979). Norton

and Norton (1979) using 13C-NMR, determined that Tyr25 of Ax I

is exposed to the aqueous environment. Furthermore, many

resonances that were assigned were very similar to those of

model small peptides suggesting that Ax I lacks extensive

structure. One particularly low pKa of 2.0 was determined for

one of the Asp side-chain carboxylates. Normal pKa values for

this carboxylate in small peptides is approximately 3.9 (Keim et

al., 1973).

Natural abundance 13C-NMR studies on As II determined that

the conformation of As II is very similar to Ax I. Approx-

imately half of the aromatic resonances and nearly all of the

carboxylate resonances were identical between the two toxins. A

similar low pKa for an Asp carboxylate sidechain was observed

(Norton et al., 1980). This low pKa is proposed to result from

the formation of a salt bridge with the E-amino group of Lys37

Tr L 2 'CHZ

Trp OH 0

Tr p


Figure 1:12: Proposed model of the molecular structure of Ax I.
Taken from Nabiullin et al. (1982).

9 T-
A "\" 0 B
o-o o 0
/ / / \

.-N./...Oc -c N- 0 0 1 H 0 H
e0 ... c s." N 'NNY'A
SC-0 / N 'T / C- a 0 A 0 H 0
/ \ .\\ "-p 4 14

LIa- a -CC4 CN4 0* 0 0 HM H HN
.... ...... / 1 1 o\. I'
-C ,,,9.* 0 H H'

MOH, 1 \\" ///, 0

*-/04 M 0 H M 0 M H

Figure 1:13: Antiparallel f3-sheet structures in Type 1 Sea
Anemone Toxins. A, As II (Gooley et al., 1986). B, As Ia
(Widmer et al., 1987).

(Norton and Norton, 1979; Norton et al., 1980). Using 1H NMR to

study the structure of As I, the same general structural

properties as those of As II and Ax I were observed. The

structure appears to be generally open without a hydrophobic

core. An abnormally low pKa for an Asp carboxylate sidechain

was observed. Interestingly, the As I sequence has a Lys

substitution for the Asp7 present in the As II and Ax I

sequences. Thus, the salt bridge most likely involves the Asp9

in all three of the toxins (Gooley et al., 1984a).

A preliminary model of Ax I was reported by Nabiullin et al.

(1982). This model (shown in Figure 1:12) consists of a short

N-terminal oc-helix, a two strand antiparallel f-sheet and

several f-turns. More recently, 2D-NMR techniques have been

applied to both Ax I and As I (Norton and Gooley, 1986; Widmer

et al., 1988). In both cases, the major secondary structural

element identified was four strands of antiparallel 8-sheet as

shown in Figure 1:13. Also identified was a type II reverse

turn between residues 30 through 33 in Ax I (Gooley and Norton,

1986) and between residues 28 through 31 in As Ia (Widmer et

al., 1988). (The previously reported sequence of As I contained

an ambiguity at position 3. Further purification resolved the

ambiguity to the presence of two isotoxins which contained only

a single amino acid sequence difference at position 3). No c-

helical region was detected in either of Ax I or As Ia.

Aromatic resonances have identified a hydrophobic cluster on the

surface of Ax I, As I and As II. The aromatic residues


determined to be involved in this are Trp23 and Tyr25 (Gooley et

al., 1986). Conformational heterogeneity in the As II and Ax I

molecules has also been observed. The conformational change has

been proposed to occur due to the cis-trans isomerization of the

Gly40-Pro41 peptide bond. The As I molecule was not observed to

display this behavior (Gooley et al. 1984b). These NMR studies

have revised the original model presented by Nabiullin et al.

(1982), in that no c-helical region is formed, the 6-sheet

structure is much more extensive, and the positioning of the p-

turns are slightly different.

These 2D-NMR techniques have recently been applied to type II

anemone toxins. Wemmer et al. (1986), using 2D-NMR, revised the

sequence originally reported by Schweitz et al. (1985) for Hp

II. In this report, they identified two sequences of 6-sheet

most likely joined in a discontinuous distorted sheet, connected

by fl-turns and extended loops without any o-helical regions.

The secondary structural elements of ShN I have been reported by

Fogh et al. (in press) The structural elements are much the same

as for Hp II, a large four strand antiparallel 9-sheet, #-turns

and no o-helical structure. The Trp30 residue appears to be

oriented to the aqueous environment and at least one of the Tyr

residues of ShN I is accessible to flavin dye (Norton et al., in


Immunochemical Studies. As previously mentioned, the

differentiation of long sea anemone toxins into at least two


classes is based upon structural as well as immunochemical

methods. Schweitz et al. (1985) were able to develop an anti-

Hp II antibody which showed no reactivity with any type 1

anemone toxin tested. We have also developed a radioimmunoassay

with an anti-ShN I antibody, which we also determined it to have

no immunocrossreactivity with any of the type 1 toxins tested

(Chapter II, this dissertation).

Immunochemistry has also allowed the surface structure of the

type 1 toxins to be probed. El Ayeb et al. (1986) isolated

monoregion-specific antibodies to As I and As II based upon the

stoichiometry of Fab binding to As II. The region recognized

was determined to contain Asp7, Asp9 and Gln47 of As II.

Furthermore, this site remained accessible to the antibody when

As II was bound to the rat brain synaptosomal Na channel

receptor. These studies suggest that this region is not

directly involved in the receptor toxin complex.

Chemical Modification Studies. Chemical modification

experiments to determine which residues may be involved in

binding to the receptor surface have yielded some information;

however, many conflicting reports have caused confusion in

identifying these residues. The first modifications reported

were the incorporation of radioactive iodine, [125I], into

histidine residues of As I and As II (Hucho, 1978; Habermann and

Beress, 1979). In both cases, radiolabeling the polypeptide did

not result in a significant loss of activity. The [125I]-As II


derivative was used to characterize the receptor in rat brain

synaptosomes (Vincent et al., 1980). Determination of the site

of iodination was later reported to be His37 (Barhanin et al.,

1981). This derivative was determined to possess greater than

80% of the toxicity of the native toxin. Carbethoxylation

modification of histidine residues decreased the toxicity of As

II on crabs and mice by factors of 5 and 13, respectively.

Reversal of this modification resulted in complete recovery of

the native toxin toxicity (Barhanin et al., 1981).

The role the conserved Argl4 residue has been investigated

using reagents which specifically modify the guanidinium group.

Vincent et al. (1980) and Barhanin et al. (1981) determined that

modification with 1,2 cyclohexanedione abolished the binding and

toxic properties of As II. Other groups have presented

conflicting data where Arg modification did not destroy the

activity (Kolkenbrock et al., 1983). Newcomb et al. (1980)

using phenylglyoxal reported that Arg modification did not

abolish activity of Ax I. The absolute conservation of this

residue among all anemone toxins implies its importance in

either structural maintenance or in the binding domain.

The role of amino groups has been probed with reagents which

retain a charge or eliminate the charge associated with either

the c-amino group or the e-amino group of lysine residues. The

results indicated that conservation of charge by guanidylation

of the Lys residues did not affect activity. However,

acetylation or treatment with fluorescamine resulted in


derivatives with activities which were reduced by factors of 8

and 14, respectively, on crabs and mice (Baharnin et al., 1981).

Specific modification of the oc-amino group or the e-amino group

of Lys35 of As II, through reduction of the Schiff's base that

forms when reacted with pyridoxal phosphate, was used to

introduce a negative charge at these residues. This

modification reduced the activity by a factor of three

(Stengelin et al., 1981). Thus, the basic groups of As II are

important in preserving the toxic properties of the molecule.

Chemical modification of the carboxyl groups of As II by

carbodiimide-mediated amidation yielded derivatives which were

modified at all three, at two or at one of the carboxyl groups.

Only the derivative that was modified at all three carboxyl

groups was studied in detail (Barhanin et al., 1981). They

reported that this derivative was devoid of the biological

properties of the native toxin, however, this derivative bound

to rat brain synaptosomes with a similar affinity to that of the

native toxin. They describe this derivative as a competitive

antagonist of the native toxin. Although the mono- and di-

substituted derivatives failed to kill crabs or mice at

concentrations up to 10 times the LD50 of the native toxin,

both of these derivatives were reported to bind to rat brain

synaptosomes with affinities identical to the native toxin.

Gruen and Norton (1985) reported that carboxyl modification on

Ax I resulted in modification of both Asp7 and Asp9. The

conformation of this derivative was greatly altered and the

biological activity was lost.

Chemical modification of only one type 2 toxin has been

reported (Kozlovskaya et al., 1982). They reported that

modification of arginine residues abolished the activity of Hm

I. Modification of the indole ring of the tryptophan residue

with 2-hydroxy-5-nitrobenzyl bromide reduced the activity of Hm

I by a factor of 2. Modification of the lysine residues with

2,4 pentanedione resulted in a derivative with only 10% activity

of the native toxin. Thus, the basic amino acid residues appear

to be essential for activity and the tryptophan residue does not

appear to be as essential.

Pharmacological Properties. Investigation of the

pharmacological properties of anemone toxins were initiated with

the purification of a neurotoxic protein (CTX) from Condylactis

gigantea and the subsequent electrophysiological

characterization. Narahashi et al. (1969) applied CTX to squid

giant axon and observed the prolongation of the action potential

due to the delayed inactivation of Na channel. These properties

resembled those of a-scorpion toxins applied to the same

preparation (Koppenhofer and Schmidt, 1968). With the success

of isolating pure neurotoxins from several other anemones,

determination of the toxic properties on invertebrates and

vertebrates was studied. Beress et al. (1975) observed a wide

range of toxicities for the isotoxins isolated from Anemonia


sulcata. The LD50 values for the toxins isolated from several

different anemones are shown in Table 1:1. One of the more

interesting aspects of these toxins is the wide range of

toxicities that these toxins possess for crabs and mice,

especially when two toxins in the same class, (Hm III and ShN

I), represent the two extremes. These two toxins have nearly

80% sequence identity (Fig 1:11), yet the toxicities that they

display on crabs and mice are inversely related. Assuming that

the species selectivity of these two toxins results from the

differences in the amino acid sequence differences between the

two toxin molecules, it may be possible to determine which of

these residues confers mammalian versus crustacean activity.

The binding properties of anemone toxins were first

investigated in experiments where [125I]-Lqq V was competitively

displaced by As II from neuroblastoma cells with an apparent KD

of 90 nM (Catterall and Beress, 1978). Furthermore, they showed

that the binding of As II decreased upon depolarization of the

cells and that As II stimulated the flux of 22Na+ in the

presence of veratridine. Investigation of the binding to rat

brain synaptosomes showed that As II displaced [1251]-Lqq V with

a KD of 400 nM (Ray et al., 1978). Use of the radiolabeled

[125I]-As II derivative on rat brain synaptosomes determined

that unlabeled As II displaces the labeled As II with a KD of

240 nM (Vincent et al., 1980). In this same report, they showed

that As II competitively displaces [125I]-AaH II with a KD of

Table 1:1. Pharmacological Properties of Purified Sea Anemone

Toxicity Rat Brain
Toxin Crab (pg/kg) Mice Synaptosome
LD50a LD50b KD (nM)

ShN I 3 >15,000 31,000
As I 8 4,000 7,000
As II 8 100 150
As III 14 >18,000 >10,000
As V 20 19 50
Hp III 26 53 300
Sg I 28 >2,000 >10,000
Hp II 40 4,200 >100,000
Ax I (Anthopl. A) 44 66 120
Ax II (Anthopl. B) 160 8 35
Hp IV 230 40 10,000
Hm III 820 2 ---

alnjection into intrahemocoelic space.

blntraperitoneal injection.

(Taken from Kem, 1988).


200 nM but that AaH II does not displace [1251]-As II.

Calculation of the As II receptor density was reported to be 10

times that of the AaH II receptor. The authors suggest the

possibility of different Na channel classes of which AaH II

recognizes only a single type.

The identification of a second class of anemone toxins has

recently been reported (Schweitz et al., 1985; Zykova et al.,

1986; Metrione et al., 1987; Kem et al., submitted). Binding

properties for rat brain synaptosomes showed that Hp toxins were

unable to displace [125I]-As V. All Hp toxins except Hp II were

able to competitively displace [1251]-AaH II. The Hp toxins did

not affect the binding of TiTxr (Schweitz et al., 1985). It is

our goal to further characterize the binding properties of this

class of anemone toxins.

Cardiac Properties. As described earlier in this chapter,

cardiac tissue possesses a Na channel which exhibits similar

toxin binding properties as neuronal Na channels.

Interestingly, cardiac tissue appears to contain Na channel

which is relatively insensitive to STX (Catterall and

Coppersmith, 1981a, 1981b). Cultured cardiac cells displaying

this same property towards STX have different properties for

anemone toxins and cc-scorpion toxins. In neuronal tissue the KD

for Lqq V and As II are 2 and 200 nM, respectively, whereas in

cultured cardiac cells, the KO.5 values are 120 and 20 nM,

respectively, (Catterall and Coppersmith, 1981a, 1981b). As II


was also observed to exhibit biphasic behavior compared to the

hyperbolic results obtained with Lqq V.

Studies where As II has been applied to guinea pig atria show

that it causes long prolongation of the action potential, and

that it acts as a positive inotroph (Ravens, 1979). Ax I,

another type 1 anemone toxin, has been shown to possess very

strong positive inotroph properties (Shibata et al., 1976,

1978). Ax II, an isotoxin of Ax I, has been determined to have

a 12.5-fold higher cardiac stimulant activity than Ax I or As II

(Reiner et al., 1985). Type 2 anemone toxins (Hp) showed only a

moderate increase in contractility relative to that of As II, Ax

I or Ax II (Renaud et al., 1986). Thus, the type 1 anemone

toxins appear to possess much stronger cardiac binding

properties than the type 2 anemone toxins. This is consistent

with the existence of at least two separate anemone toxin


Ax I has been determined to have a 200 fold greater

inotrophic activity than digoxin (on a molar basis) and a higher

therapeutic index in vivo (Scriabine et al., 1979). The

mechanism of action of Ax I has been determined to involve a

delay in the inactivation of the cardiac Na channel, causing a

prolongation of the Na+ current (Kodama et al., 1981) without

affecting neuromuscular function (Scriabine et al., 1979).

Thus, the potential exists to use these type 1 anemone toxins as

templates from which drugs designed for the treatment of


congestive heart failure may be fashioned (Shibata and Norton,


Unanswered Questions

Although the sea anemone toxins have been utilized in the

intense study of the Na channel in recent years, there are still

a number of unanswered questions relating to toxin structure,

the species selectivity, receptor characterization and binding

properties. In this dissertation, the binding properties of

ShN I to the Na channel receptor in crustacean and mammalian in

vitro systems are investigated through use of a radiolabeled

derivative of ShN I. A program was initiated to chemically

synthesize, refold, and purify the native 48-residue

polypeptide. Furthermore, elucidation of several residues which

are essential for toxicity as well as receptor binding were

determined through the preparation of six monosubstituted

synthetic analogs directed at the N-terminal domain of ShN I.

Chemical and physical methods have been used to study the

refolded synthetic products. The information obtained from

these studies is expected to give the first detailed

characterization of the type 2 anemone toxin receptor of the Na





The Na channel is the transmembrane, glycoprotein responsible

for the selective passage of Na+ ions through the lipid bilayer

in electrically excitable tissue. A variety of pharmacological

agents have been utilized to investigate the mechanism of this

ion channel. The number of different binding sites continues to

increase, but at this time at least five different classes have

been identified (for review see Catterall, 1986). These five

classes are as follows: (i) tetrodotoxin (TTX) and saxitoxin

(STX) block the entry of Na+ ions through the channel by

preventing the opening of the channel (Narahashi, 1974; Ritchie

and Rogart, 1977; Ritchie, 1980); (ii) the lipophilic alkaloids

[ e.g., batrachotoxin (BTX) and veratridine as well as

pyrethrins bind to an open form of the channel and stabilize it

in this conformation (Ulbricht, 1969; Albuquerque and Daly,

1976; Narahashi, 1976; Jacques et al., 1980); (iii) the a-

scorpion and sea anemone polypeptide toxins delay the

inactivation of the channel (Romey et al., 1975, 1976; Bergman

et al. 1976), and have also been shown to have positive

heterotropic cooperativity with the site II lipophilic toxins



(Ray et al. 1978; Catterall and Tamkun, 1981); (iv) the 6-

scorpion toxins such as those isolated from Tityus serratulus

alter the activation properties of the channel (Vijvenberg et

al., 1984; Barhanin et al., 1984); (v) ciguatoxin and brevetoxin

bind to a specific class of sodium channels (Bidard et al.,

1984; Huang et al., 1984).

Recently, a novel class of anemone toxins have been isolated

from the Stichodactylid sea anemone family (Schweitz et al.,

1985; Zykova et al., 1986; Kem et al. 1986). These toxins

differ significantly from the Actiniid toxins in amino acid

sequence homology. Members of each class of anemone toxins

possess greater than 60% sequence identity. However, only about

30% identity exists between the two classes. The major

contribution to the interclass sequence identity is the

positioning of cysteine residues. As a result of the cysteine

residue positioning, it is believed that all anemone toxins

possess the same disulfide pairings although Anemonia sulcata II

(As II) is the only toxin for which the disulfide pairs have

been reported (Wunderer, 1978).

Probing the two classes of sea anemone toxins for

immunocrossreactivity has shown that there is no cross

reactivity between polyclonal antibodies against Heteractis

paumotensis (formerly Radianthus paumotensis) III, (Hp III) and

As II, As V, Ax I, or Ax II (Schweitz et al. 1985).

In binding studies using rat brain synaptosomes, Hp II was

unable to displace [125I]-As V, [125I]-AaH II (Androctonus


austrialis Hector II) or [125I]-TiTxr. However, Hp III and

other Heteractis paumotensis toxins were successful in

displacing [125I]-AaH II but not the [125I]-As V or [125I]-TiTxr

(Schweitz et al., 1985).

Our group recently isolated and sequenced another

Stichodactylid toxin from the sea anemone Stichodactyla

helianthus (Kem et al., submitted). The pharmacological

properties of this toxin are among the most interesting of any

sea anemone toxin isolated to date. This toxin has the greatest

range of selective toxicity reported for sea anemone toxins.

The LD50 for crustaceans is 5000 fold lower than the LD50 for

mammals (Kem et al., submitted).

In this report, we examine the electrophysiological effects

of ShN I on lobster olfactory somas. Secondly, we describe the

preparation of a mono-iodinated derivative of Stichodactyla

helianthus neurotoxin I ([125I]-ShN I). The [125I]-ShN I

derivative was then used to characterize the pharmacological

properties of ShN I on rat brain synaptosomes and Blue Crab

walking leg axolemma vesicles. Both of these preparations have

been well characterized, and have been shown to be enriched in

the sodium channel protein. Use of these two systems may help

to identify the existing differences between them. Finally,

using a polyclonal antibody prepared against ShN I, we have

examined the immunocrossreactivity between type 1 and type 2

anemone toxins.


Experimental Procedures

Materials. Sea anemone toxin I from Stichodactyla helianthus

(ShN I) was purified as described previously (Kem et al.,

submitted). [3H]-STX (specific activity 10 Ci/mmol) was kindly

provided by Dr. Gary Strichartz (Department of Anesthesia,

Harvard Medical School, Boston Massachusetts). Androctonus

austrialis Hector II was a kind gift from Dr. Herve Rochat

(Laboratoire de Biochimie, Marseille, France). Anemonia sulcata

II and Bolecera tuediae II were kindly provided by Dr. Lazlo

Beress (Institut fur Meereskunde an der Universitat Kiel, Kiel,

West Germany). Nal25I was purchased from New England Nuclear

(Boston, Massachusetts) and had a specific activity of 2200

Ci/mmol. Di-0-C5 (3,3'dipentyloxacarbocyanine iodide) was

obtained from Molecular Probes, Inc. (Junction City, Oregon).

Staphylococcus Protein A was purchased from Boehringer Mannheim

(Indianapolis, Indiana). Affi-gel 15 was purchased from Bio-Rad

Inc., (Richmond, California). Antisera to ShN I was obtained

from New Zealand White Rabbits. TTX was obtained from

Calbiochem (La Jolla, California). Trypsin, collagenase (Type

V) and veratridine were obtained from Sigma Chemical Co., (St.

Louis, Missouri). All other reagents were the highest

commercial grade available.

Biological Assay. Intrahemocoelic injection of 3-5 g fiddler

crabs (Uca pugilator) was performed with purified ShN I diluted

at a constant interval with normal saline containing 0.1 mg/ml


BSA (142 mM NaCl, 2 mM CaCl2, 40 mM KC1, 9 mM Dextrose, and 10

mM Tris HC1, pH 7.4). Paralytic response was determined 15 min

following the injection by placing the animals on their backs

and measuring their ability to "right" themselves in a 2 min

interval. Intracerebroventricular injection of 24-31 g white

mice was performed with a constant dose interval of ShN I. At

low doses (1-2 ymol/kg), a slight tremor could be detected by

holding the animal upside down by its tail. At higher doses,

the injection resulted in paralysis and eventual death of the


Preparation of lobster olfactory somas. Antennules excised

from the Florida spiny lobster, Panulirus argus, were perfused

with cold saline to remove the hemolymph, and cut into 1 cm

sections which were split longitudinally. These sections were

bisected again longitudinally. The half that bore the axon

bundles was discarded and the half that bore the sensillia

contained the receptor cells. The clumps of receptor cells were

dislodged from the connective tissue with a gentle stream of

saline from a pipette. Isolated receptor cells were obtained

from these clumps, by dissociating them with the enzymes

collagenase followed by trypsin. Cells were treated with 100

IU/ml of collagenase in saline for 90 min, with gentle

agitation, followed by 30 min in 0.4-0.6 mg/ml trypsin in Ca2+-

free saline. The cells were then rinsed several times in normal

saline and transferred to the recording bath. All recordings


were done in normal saline (140 mM NaCI, 2mM CaC12, 5.4 mM KC1,

9mM dextrose and 10 mM Hepes, pH 7.4)

Intracellular lobster olfactory soma recording. Recordings

on the isolated receptor cells was according to the method of

Anderson and Ache (1985). Briefly, preparations were examined

at 200 X using a fixed stage microscope (Aus Jena) equipped with

Modulation Contrast optics. Intracellular recordings were

obtained with the whole-cell, patch clamp technique (Hamill et

al., 1981). High impedance (>1 GO) seals were obtained using

unpolished patch pipettes pulled from borosilicate glass and

filled with a high K+/low Ca2+ solution containing: 140 mM KC1,

1 mM CaC12, 11 mM EGTA, 10 mM HEPES, and 696 mM glucose. The pH

of the solution was adjusted to 7.0 with 5N KOH. Final K+

concentration was 210 mM. With this solution, electrodes had

impedances of 3-5 MO. Recorded signals were amplified with a

Dagan Instruments 8900 Patch Clamp Amplifier equipped with a 1

GO feedback resistor. The bandwidth of the recording system was

10 kHz. Signals were displayed on a Nicolet 2090 Digital

Oscilloscope and stored on floppy disks. Hard copies were

obtained on a Houston Instruments 100 X-Y Plotter.

Preparation of Axolemma vesicles. Axolemma membranes were

obtained from the Blue Crab (Callinectes sapidus), since this

species was locally available and the walking leg nerves could

be readily removed by breaking the leg joints and pulling the

nerves out. The method used was based on that described for

spiny lobster (Panularus argus) axonal vesicles (Barnola et

al., 1973) with minor modifications. In a typical preparation,

the 8 walking legs and chelae neurons were dissected from 25

adult Blue Crabs. The neurons (wet weight 12 g) were minced

with a Polytron rotary stainless steel knife mincer in 0.33 M

sucrose, 2 mM MgSO04, 10 mM Tris HC1, pH 7.4, at 4 *C (8 ml

buffer per g wet weight tissue). This mixture was homogenized

with Teflon-glass (10 strokes) at 500 rpm. The homogenate was

centrifuged for 10 min at 3000 x g at 4 *C. The supernatant was

retained and the pellet was rehomogenized in 40 ml 0.33 M

sucrose, 2mM MgS04, 10 mM Tris HC1, pH 7.4, 4 *C, with a Teflon

glass homogenizer (10 strokes). The rehomogenized pellet was

centrifuged at 3000 x g for 10 min at 4 *C. The two

supernatants were combined and the pellet was discarded. The

supernatant was then centrifuged at 17,000 x g for 80 min at 4

C. At this point, the supernatant was discarded and the pellet

was recovered and homogenized (4 strokes) in 20 ml of 0.11 M

sucrose, 2 mM MgS04, 5 mM Tris HC1, pH 7.4 at 4 C. This

suspension was carefully layered on top of cellulose nitrate

tubes containing discontinuous sucrose gradients consisting of

10 mM Tris HC1, pH 7.4 buffered sucrose solutions of 1.2 M, 1.0

M, 0.8 M, 0.6 M and 0.4 M (6 ml each). These tubes were

centrifuged at 100,000 x g in a SW-28 rotor for 8 h at 4 *C.

The dense white band at the 1.0-1.2 M sucrose interface was

carefully removed and diluted dropwise to 0.33 M sucrose with


ice cold dH20. This mixture was centrifuged at 40,000 x g in a

SW-28 rotor for 45 min at 4 C. The pellets were recovered and

rehomogenized (2 strokes by hand) in 20 ml of the homogenization

buffer and stored at -78 C. The 1.0-1.2 M sucrose fraction was

used in all experiments. Characterization of this fraction

included assay of [3H]-STX binding to calculate receptor

density, measuring the ability to retain a membrane potential as

monitored by the fluorescence dye method using di-0-C5

(Blaustein and Goldring, 1975). Protein concentration was

determined by the bicinchonic acid (BCA) assay using bovine

serum albumin as a standard (Redinbaugh and Turley, 1986).

Enrichment of the axolemma fraction for the marker enzyme Na+ -

K+ ATPase was determined by the method of Kilberg and

Christensen (1979) where the released phosphate was quantitated

spectrophotometrically (Fiske and Subbarow, 1925).

Preparation of Rat Brain Synaptosomes. Synaptosomes were

prepared by a modification of the method of Gray and Whittaker

(Gray and Whittaker, 1962). The brains were removed from 3 male

Sprague-Dawley rats and homogenized in ice-cold 0.32 M sucrose

containing 10 mM Tris HC1, pH 7.4 (10 ml/g tissue) with 10

strokes of teflon glass homogenizer (500 rpm). The resulting

homogenate was sedimented at 1000 X g for 10 min at 4 *C. The

supernatant was saved and the pellet was resuspended in 10 ml of

0.32 M sucrose and resedimented at 1000 X g for 10 min at 4 *C.

The resulting supernatant was pooled with the first and


sedimented at 17,000 X g for 1 h at 4 C. The supernatant was

discarded and the pellet was resuspended in 8 ml of 0.32 M

sucrose containing 10 mM Tris HC1, pH 7.4. This suspension was

layered on to a stepwise gradient consisting of 6 ml of 1.2,

1.0, 0.8, 0.6, and 0.4 sucrose solutions buffered with 10 mM

Tris HC1, pH 7.4, and sedimented at 100,000 X g for 105 min at 4

*C. The dense white band isolated at the 1.0-1.2 interface was

used in all binding studies due to the presence of a high

density of sealed vesicles (Michaelson and Whittaker, 1963).

Characterization of this preparation for Na+, K+ ATPase

enrichment, [3H]-STX binding, and ability to maintain a membrane

potential was performed as described earlier in the text for

axolemma vesicles.

Electron microscopy. Axolemma vesicles and rat brain

synaptosomes sucrose gradient fractions were monitored for the

presence of sealed structures by negative stain electron

microscopy. A solution of vesicles in 0.32 M sucrose buffered

with 10 mM Tris HC1, pH 7.5 was allowed to adhere to a 400 mesh

formvar carbon coated grid for 30 sec, blotted to remove excess

solution, and then stained with a 2% solution of uranyl acetate

for 30 sec. Excess stain was adsorbed with filter paper and air

dried prior to examination in a Joel 100 CX electron microscope

operated at 80 kV. Sealed vesicles and fragments were observed

at a magnification of 165,000 X.

Radiolabeling Androctonus austrialis hector II with [125I1.

Androctonus austrialis Hector II (5 pg) was labeled by the

lactoperoxidase-H202 and purified by gel filtration on a

combination Sephadex G-15 and G-50 column as previously

described (Rochat et al., 1977). The specific activity of the

[125I]-AaH II derivative was determined to be 460 Ci/mmol.

Radiolabeling ShN I with [12511. The chloramine-T method of

iodination was used to incorporate [1251] into the Tyr residues

(Hunter and Greenwood, 1962). ShN I (0.2 mg) was dissolved in

0.25 ml of ice cold 20 mM Tris HC1, pH 8.64. To this solution

1.0 nmol of Nal carrier was added followed by a 10 pl aliquot

of Na [125I] (1 mCi). The iodination reaction was initiated by

the addition of 10 pl of a 10 mM chloramine-T solution. Two

additional 10 pl aliquots of the chloramine-T solution were

added at 3 min intervals. The reaction was maintained at 0 C

for 30 min, at which point the entire reaction mixture was

placed on a Sephadex G-25 column equilibrated in 50 mM NaP04, pH

7.5. The labeled protein eluted in the void volume of this

column. The labeled protein was separated from unlabeled

protein by reverse-phase HPLC using on a C18 column, with a

gradient of 10% 60 % acetonitrile into 0.1% TFA in 40 min at

2.1 ml/min. The unlabeled toxin eluted at 21 min into the

gradient and the labeled material eluted at 22.3 min into the



Binding Experiments. [1251]-ShN I and [3H]-STX binding to

synaptosomes and axolemma vesicles was measured by rapid

filtration assay with glass fiber filters (Whatman GF/B). The

standard binding medium consisted of 140 mM choline chloride, 50

mM Hepes adjusted to pH 7.4 with Tris base, 5.5 mM glucose, 0.8

mM MgS04, 5.4 mM KC1, and 1 mg/ml of bovine serum albumin.

Samples were mixed and incubated for 30 min at 4 C for [3H]-

STX and 37 C for [1251]-ShN I. Following 30 min incubation,

the samples were diluted with 3.0 ml of ice cold wash buffer

(163 mM choline chloride, 5 mM Hepes (adjusted to pH 7.4 with

Tris base), 1.8 mM CaC12, 0.8 mM MgS04, and 0.1 % bovine serum

albumin) and immediately collected on glass fiber filters and

washed with 3 X 3.0 ml of ice cold binding medium. The bound

radioactivity was then determined by liquid scintillation

counting for tritium or gamma emission for [1251]. Counting

efficiency for iodine and tritium was determined by the method

of internal standards. Non-specific binding was measured in the

presence of excess ligand (150 nM TTX or 10 pM ShN for axolemma

vesicles and 250 AM ShN I for rat brain synaptosomes) and

subtracted from the results. Binding measurements at each

ligand concentration were done at least in duplicate, and the

error measured for each data point was less than 2%.

Purification of Mono-specific Anti-ShN Antibodies. A

Staphylococcal Protein A column was prepared by reacting 10 mg

of Protein A with 3.0 ml of preswollen Affi-gel 15 for 4 h in


100 mM NaCO3, 150 mM NaCl, pH 8.2 at 4 *C.. The remaining

unreacted imido esters on the resin were treated with a 1.0 M

solution of ethanolamine (Kodak Eastman, Kingsport, Tennessee)

overnight at 4 C. The binding capacity of the column was

determined by monitoring the absorbance at 280 nm of the

coupling buffer containing unreacted Protein A and determined to

be greater than 3.2 mg Protein A per ml of resin. After being

filtered through a Millipore (0.45 pm) antisera to ShN I was

then passed over the column, and the column was washed

extensively with PBS (25 mM NaPO4, 140 mM NaCI, pH 7.8) until

the A280 returned to 0. The IgG fraction was then eluted from

the column with a buffer consisting of 0.58 % acetic acid and

140 mM NaCI. The eluted IgG fraction was immediately

neutralized with a solution 1.5 M Tris base and 140 mM NaCI to

pH 7.8.

Radioimmunoassay for Crossreactivity. Radioimmunoassay was by

the method of Chandler et al. (1984). A 96 well microtiter

plate was coated with a 25 pl of 0.1 mg/ml solution of Protein A

dissolved in PBS. The microtiter plate was incubated for 1 h at

room temperature and then washed with 3 X 100 p1 of PBS. The

IgG fraction was diluted with PBS to 1/2000 of serum volume and

25 1l was plated out in each well. The plate was then incubated

for 1.5 h at room temperature and subsequently washed with 5 X

100 pj of RIA buffer (25 mM NaPO4, 140 mM NaCi, 0.05% Tween 20,

and 0.3% BSA at pH 7.8). To each well was added 25 pl of a


1/100 dilution of [1251]-ShN I (approximately 4000 cpm), and 25

pl of the appropriate dilution of unlabeled toxin sample.

Following an incubation of 1.5 h at room temperature, each well

was washed with 5 X 100 pl of RIA buffer. The wells were then

punched out and read directly on a Beckman 5500 gamma counter.

Each toxin dilution was run in triplicate and the data

represents an average of the values obtained where the standard

error measured for each point was less than 3%.


Intracellular Recording. Preliminary electrophysiological

studies using whole cell patch clamping of lobster (Panularus

argus) olfactory somas indicate that ShN I delays the

inactivation of the Na+ current (Figure 2:1). In control

experiments, application of a stimulus resulted in an initial

Na+ current which rapidly returned to resting values (Figure

2:1A). Following application of ShN I, the Na+ current was

observed to be prolonged with a slow return to resting values

(Figure 2:1B). Application of TTX to the external saline

solution abolished the effects observed with ShN I. TTX

immediately blocked the inward Na+ currents (data not shown).

These results are consistent with those observed with other

polypeptides which bind to the Na channel and delay inactivation

(for review see Catterall, 1980)




mV o


TIME (msec)

Figure 2:1: Electrophysiological effects of ShN I. Whole cell
patch clamping was used to monitor the effects of ShN I. (A)
Control recording following a 2 mV stimulus. (B) Recording of
same cell following application 100 il of 1 pM ShN I solution.
Final toxin concentration in the bath was 100 nM.




E -. o- 60_ 4x105 :

S0.1 .
< -.. -35- 3x105

ii 10o 2x105


0 5 10 15 20 25 30 35 40
TIME (min)

Figure 2:2: Purification of [1251]-ShN I. ShN I (300 ig) was
iodinated by the chloramine-T method at 0 C. Following the
iodination reaction, the reaction mixture was immediately
filtered on a Sephadex G-25 column (0.75 X 21 cm) equilibrated
with 50 mM NaP04, pH 7.5. The void volume peak was collected
and rechromatographed on reverse-phase HPLC using a C18 column.
The gradient consisted of 10-60% acetonitrile into 0.1% TFA in
40 min at 2.0 ml/min. The chromatography was followed by
automatic recording of absorbance at 254 nm (--) and by
measurement of radioactivity of a 10 1l aliquot removed from
each collected fraction monitored in a gamma counter (----).


Preparation of Radiolabeled Derivative. ShN I was iodinated by

the chloramine-T method (Hunter and Greenwood, 1962). Specific

activities varied between 50 and 800 Ci/mmol depending upon the

inclusion of cold 1271 into the reaction mixture. Purification

of the iodinated toxin involved an initial desalting step on a

Sephadex G-25 column followed by reverse-phase HPLC. This final

purification step on reverse-phase HPLC was necessary to isolate

the [1251]-ShN I from the unreacted ShN I. The separation data

is shown in Figure 2:2. Manipulation of the gradient of

acetonitrile increased the retention time of the derivatized

material allowing it to be easily separated from the native

toxin. The protein concentration was then determined through

measurement of the intrinsic absorption at 280 nm due to the

presence of two Tyr and one Trp residue (A280 of 1% solution

13.92). Following determination of the protein concentration

the iodinated toxin was lyophilized and redissolved in the

standard binding buffer which contained img/ml BSA (see

"Experimental Methods").

Pharmacological properties of the [1251]-ShN I derivative

determined the LD50 on fiddler crab to be approximately 3.2

pg/kg (native toxin LD50 = 2.92 pg/kg). Thus, the iodinated

derivative possessed greater than 90% of the activity of the

native toxin.

Characterization of Crab Axolemma Vesicles and Rat Brain

Synaptosomes. Blue Crab walking leg axolemma vesicles and rat


brain synaptosomes were prepared as described previously. The

purified axolemma and synaptosome preparation were then tested

with [3H]-STX for the presence of voltage sensitive sodium

channels. The binding capacity for [3H]-STX was approximately

14 pmol/mg for axolemma vesicles and 1.8 pmol/mg for the

synaptosomes. The ability of the axolemma vesicles as well as

the synaptosomes to maintain a membrane potential was determined

with the voltage sensitive dye di-O-C5, (Blaustein and Goldring,

1975). The fluorescence properties of the dye could be modified

by changing the K+ concentration or by adding veratridine in

Na+ buffer. Lysis of the vesicles with a detergent such as

Triton X-100 caused the return of fluorescence back to baseline

values. Enrichment for the marker enzyme Na+,K+ ATPase was

approximately 12 times that of the starting homogenate for

axolemma vesicles and 6 fold enrichment for synaptosomes.

Vesicle preparations were also monitored by electron microscopy

(Figure 2:3). Sealed vesicles (Figure 2:3A) were easily

distinguished from the unsealed fragments (Figure 2:3B) found

in fractions taken from the lower density bands of the

discontinuous sucrose gradient.

1I25I]-ShN I Binding to Axolemma Vesicles. Initial

experiments were designed to detect saturable binding of [1251]-

ShN I to sites with a KD of 1-1000 nM. In these experiments,

the vesicles were incubated in the presence of increasing

concentrations of [125I]-ShN I (Figure 2:4, left O).


Figure 2:3: Electron Micrographs of Sealed Axolemma and
Synaptosome Vesicles. A, axolemma vesicles isolated from
discontinuous sucrose gradient centrifugation. Magnification x
125,000. B, unsealed axolemma vesicles isolated from the 0.6 -
0.8 M sucrose interface.


-3 1------------ 50------------
E I50
E 5
0 40--
z 3- 0
'c X 2
I U-
2 20

1 10--

0 0
0 10 20 30 40 50 60 0 1 2 3 4 5 f

Concentration of [1251]-ShN I (nM) [1251]-ShN I bound (pmol/mg)

Figure 2:4: Scatchard Analysis of ShN I binding to axolemma
vesicles. Left, axolemma vesicles were incubated with
increasing concentrations of [1251]-ShN I as indicated on the
abscissa in standard binding media ( 0 ) or in standard binding
media containing 10 AM ShN I (saturation concentration)( ).
Bound [1251J-ShN I was then measured as described under
"Experimental Procedures". Right, specific binding, calculated
as the difference between binding in the presence and absence of
10 pM ShN I, is plotted as a Scatchard Plot.

-% 110
6 100 0 0
o 90-
0 80
o 0
-. 50
z 40 0
c, 30
c 20-
-0 I .I ... I
1E-1 1 10 100
Concentration of Toxin (nM)

Figure 2:5: Competition between [12511-ShN I and ShN I or As II
for binding to axolemma vesicles. [125I]-ShN I (0.5 nM) was
incubated with axolemma vesicles (0.5 mg of protein per ml) and
increasing concentrations of ShN I ( 0 ) or As II ( 0 ) in the
standard binding media. Following 30 min incubation at 37 *C,
the radioactivity bound was then determined as described under
"Experimental Methods".


Nonspecific [1251]-ShN I binding was measured in the presence of

an excess of unlabeled ShN I (10 pM) (Figure 2:4, left *).

Saturable binding, which is defined as the difference between

specific and nonspecific binding curves in Figure 2:4, is

presented in the form of a Scatchard plot in Figure 2:4 (right).

The linearity of the Scatchard analysis (Scatchard, 1949)

indicates a single class of receptors with KD of 14 nM and a

binding capacity of 5.8 pmol/mg.

125II-ShN I versus ShN I Binding to Axolemma Vesicles.

Axolemma vesicles were incubated with 0.5 nM [1251]-ShN I and

increasing concentrations of unlabeled ShN I from 1 to 500 nM.

Assuming that there are saturable binding sites for ShN I, these

sites should be increasingly occupied by unlabeled ShN I,

resulting in decreased [1251]-ShN I binding. The results of

these experiments (Figure 2:5) determined that unlabeled ShN I

inhibits greater than 85% of [125I]-ShN I binding with 50%

maximal inhibition at 25 nM. These results suggest a saturable

ShN I receptor exists in the axolemma vesicles with a K0.5 of

approximately 25 nM. In these experiments, nonsaturable binding

was determined in the presence of 10 pM ShN I and was subtracted

from the results.

Recently, long anemone toxins have been divided into two

classes according to sequence homology (Schweitz et al., 1985;

Kem, 1988). As II is classically used as the representative

type 1 long anemone toxin in competition binding experiments

o 110--i

o 90- O 0
80- o
"u 70- 0
I 60 0

S 50

10-- 0
S--- I

10 100 1000 1E4 1E5 1E6
Concentration of ShN I (nM)

Figure 2:6: Competition between [1251]-ShN I and ShN I for
binding to rat brain synaptosomes. [125I]-ShN I (10 nM) was
incubated with rat brain synaptosomes (1 mg protein per ml) and
increasing concentrations of ShN I. Following 30 min incubation
at 37 *C, the radioactivity bound was then determined as
described under "Experimental Methods".

-^ 2000-
E 1800-
o 1600-
'.- 1400-
- 1200-
v, 1000-
'=- 800-
e 600-
J 200-
0 0
M 0



0 20


0 40 80 120 160

Concentration [1251]-ShN I (uM)

0 200 400 600 800 10001200
[1251]-ShN I bound (fmol/mg)

Figure 2:7: Scatchard analysis of [1251]-ShN I binding to rat
brain synaptosomes. Left, synaptosomes were incubated with
increasing concentrations of [125I]-ShN I as indicated on the
abscissa in standard binding media ( 0 ) or in standard binding
media containing 250 pM ShN I ( 0 ). Bound [1251]-ShN I was
then determined as described under "Experimental Methods".
Right, specific binding was then calculated as the difference
between binding with and without 250 jiM ShN I, is plotted in the
form of a Scatchard plot.

- O


0 0

o (
0 @ ***


with scorpion toxins and other sea anemone toxins (Catterall and

Beress, 1978; Vincent et al., 1980). Experiments were carried

out, as described above for unlabeled ShN I, to determine the

ability of As II (Figure 2:5, 0 ) to displace [125I]-ShN I on

the axolemma vesicles. As II was unable to displace the [1251]-

ShN I in these experiments over a concentration range from 10 nM

to 500 nM. Axolemma vesicles were selected for these studies

because of the specificity which ShN I exhibited in binding to

these versus rat brain synaptosomes. Similar results for the

competition of Heteractis paumotensis toxins with [1251-As V on

rat brain synaptosomes have been reported (Schweitz et. al.

1985). However, ShN I exhibits a much lower toxicity to mammals

than the Hp toxins. Thus, these competition experiments were

not attempted on rat brain synaptosomes. Our results along with

those of Schweitz et al. (1985) suggest the possibility of a

separate anemone type 2 toxin receptor on both axolemma vesicles

and synaptosomes.

Binding of ShN I to Rat Brain Synaptosomes. Following the

same logic as used for the axolemma vesicles, a class of

receptors was suspected to exist with a substantially higher KD

value due to the much lower LD50 that ShN I exhibited on mice.

Thus, synaptosomes were incubated with 10 nm [1251]-ShN I and

increasing concentrations of unlabeled ShN I from 50 to 500,000

nM. Assuming the presence of a saturable receptor for the ShN I

toxin, these sites should be increasingly occupied by the


unlabeled ShN I resulting in the same decrease of [1251]-ShN I

binding observed in the axolemma vesicles. The results of these

experiments (Figure 2:6, 0) show that unlabeled ShN I displaces

greater than 90% of the [1251]-ShN I with 50% maximal inhibition

observed at 26,000 nM, suggesting the presence of saturable

receptor with a KD of approximately 26,000 nM.

The presence of a saturable receptor was also determined

through experiments where the concentration of [125IJ-ShN I was

varied. In these experiments, total ShN I binding is measured

in incubation mixtures containing increasing concentrations of

[125IJ-ShN I (Figure 2:7, left 0). Nonsaturable [1251J-ShN I

binding was determined in the presence of excess unlabeled ShN I

(250 AM) (Figure 2:7, left 0 ). The difference of these two

curves represents the saturable binding component and is

presented in the form of a Scatchard Plot (Figure 2:7, right).

The Scatchard analysis indicates the presence of a single class

of receptor sites for [1251]-ShN I with a KD = 31 AM and a Bmax

of 1.08 pmol/mg of protein.

Characterization of ShN I Receptor. Investigation of the

synaptosome toxin receptor was also probed using the c-scorpion

toxin, [125I]-AaH II. In these experiments, synaptosomes were

incubated with 0.02 nM [1251]-AaH II and increasing

concentrations of AaH II (Figure 2:8, 0), As II (Figure 2:8, *),

or ShN I (Figure 2:8, A ). The unlabeled AaH II displaced the

[1251]-AaH II with 50% maximal inhibition observed at 450 pM.





100 1000
Toxin (nM)

Figure 2:8: Competition between [1251]-AaH II and ShN I or
As II binding to rat brain synaptosomes. [1251]-AaH II (0.02
nM) was incubated with rat brain synaptosomes (1 mg protein per
ml) and increasing concentrations of AaH II ( 0 ), ShN I (A )
or As II ( 0 ) in standard binding media. Following 30 min
incubation at 37 *C, the radioactivity bound was determined as
described under "Experimental Methods".

1E-2 1E-1 1


0 O

0 O

--4----- i i G1 a m '
:: O











o 000

0 0

""O O -, .. ..



10.000 100.000 1000.000

Concentration of Verotridine (MuM)

Figure 2:9: Enhancement of [125I]-ShN I binding by
veratridine. Axolemma vesicles (0.5 mg protein per ml) were
incubated with 0.5 nM [1251]-ShN I, in the presence of
increasing concentrations of veratridine in standard binding
media. Following 30 min incubation at 37 C, the bound
radioactivity was determined as described under "Experimental



Similarly, As II displaced the [1251]-AaH II with 50% maximal

inhibition at 300 nM; however, ShN I was unable to displace any

of the [1251]-AaH II from its receptor. Competition assays with

the [1251]-AaH II on axolemma vesicles were not attempted due to

the low toxicity which this toxin has on crustaceans. These

results suggested that the ShN I was binding to a separate

receptor population in the synaptosomes.

Classical site III polypeptides have been shown to interact

cooperatively with site II lipophilic toxins such as BTX and

veratridine (Catterall, 1977; Ray et al., 1978). Alpha-scorpion

toxins enhance the activation of the Na+ channels by these

compounds through an allosteric mechanism (Catterall, 1977b).

Thus, the binding of scorpion toxin is enhanced in the presence

of these compounds in both synaptosomes and neuroblastoma cells

(Catterall, 1977a; Ray et al., 1978). In order to determine if

the binding of [1251]-ShN I is affected by these site II

compounds, experiments were carried out exclusively in choline

substituted buffers in order to eliminate any effect of membrane

potential dependent binding. The increase in binding of [1251]-

ShN I was measured with increasing concentrations of veratridine

(Figure 2:9). No increase in binding of [125I]-ShN I was

detected up to a concentration of 1000 jM veratridine in either

of the synaptosomal or axolemmal preparations. These results

indicate that there is little or no heterotrophic cooperativity

between veratridine and ShN I.


Site III binding polypeptides have all shown some type of

membrane potential dependent binding (Catterall et al., 1976;

Catterall, 1977; Ray et al., 1978). However, scorpion toxin

[e.g., Lqq V] binding is affected much more severely than

anemone toxins such as As II (Catterall and Beress, 1978; Ray et

al., 1978; Vincent et al., 1980). In order to test the effects

of membrane potential on binding, incubations were carried out

in buffers containing 130 mM KCI which completely depolarizes

the synaptosomes (Blaustein and Goldring, 1975), or 130 mM NaCIl

with gramicidin D, a Na+ ionophore, which completely depolarizes

the synaptosomes (Blaustein and Goldring, 1975). These results

are presented in Table 2:1. Neither the incubation in 130 mM

KC1 nor the Na+/gramicidin experiment showed any significant

reduction in the binding of [1251]-ShN I to either synaptosomes

or axolemma vesicles. Osmotic lysis of the vesicle preparations

is expected to result in depolarization (Blaustein and Goldring,

1975). Lysis of either preparation in dH20 at 0 *C prior

addition of [1251]-ShN I also resulted in no apparent reduction

of binding of [1251]-ShN I (Table 2:1). Therefore, we conclude

that this particular receptor class is not dependent on the

presence of a membrane potential.

Immunocrossreactivity of anemone toxins. The Protein A purified

anti-ShN I antibody was diluted to 1/2000 the serum volume. The

antibody was immunoreactive with the labeled [1251]-ShN I

derivative as shown in Figure 2:10 ( 0). The standardization

Table 2:1. Effects of depolarization on [1251]-ShN I binding

Conditions % [125I]-ShN I

Experiment I

A. Axolemma
130 mM NaCl 89
130 mM KC1 94
130 mM NaCI, gramicidin D (10 pg/ml) 92

B. Synaptosomes
130 mM NaCIl 99
130 mM KCI 104
130 mM NaCl, gramicidin D (10 pg/ml) 103

Experiment II
A. Axolemma normal 100
Lysed 89

B. Synaptosomes normal 100
Lysed 104

In Experiment I, [1251]-ShN I binding was measured as
described under "Experimental Methods" in solutions in which the
choline chloride (130 mM) present in the standard incubation
medium was replaced either with NaCI (130 mM), KC1 (130 mM), or
with NaCI (130 mM) containing 10 pg/ml gramicidin D. The
results are presented as the percentage of the binding versus
the binding measured in the standard choline chloride (130 mM)
medium. In Experiment II, the binding was also determined as
described under "Experimental Methods", using either
synaptosomes or axolemma vesicles which had been lysed by prior
incubation in dH20 or with normal preparations. The
nonspecifically bound [1251]-ShN I was determined in the
presence of either 10 pM or 250 pM ShN I for axolemma and
synaptosomes, respectively, and subtracted from each value prior
to any calculation of percent bound.

Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EBZCMYQMJ_658P1C INGEST_TIME 2012-02-20T21:53:03Z PACKAGE AA00009088_00001