- Email: [email protected]
Selective neuromuscular blocking properties of α-conotoxins In vivo
ooal-0lol~9o s3.oo+ .oo (f) 1990 Pergamoo Press plc
rorrean vol. zs, No . z ~. zsl-z3a,199o. Printed in Great Britain .
SHORT COMMUNICATIONS SELECTIVE NEUROMUSCULAR BLOCKING PROPERTIES OF aCONOTOXINS IN VIVD I. G.
M~RSt-tntt,
and A. L.
HARVmr*
Department of Physiology and Pharmacology and The Strathclyde Institute for Drug Research, University of Strathclyde, Glasgow G1 1XW, U .K . (Acceptedfor publication 18 July 1989)
I. G. MARSFIALI, and A. L. HARVEY . Selective neuromuscular blocking properties of a-conotoxins in vivo. Toxicon 28, 231-234, 1990.-Synthetic a-conotoxins GI and MI (based on the natural polypeptides isolated from Conus geographus and C. magus, respectively) were tested for effects on newomuscular transmission and on the cardiovascular system in anaesthetised cats. Both produced neuromuscular blockade at doses of 20-80 kg/kg. Conotoxin GI was about 2.5 times more potent than MI . Both were rapidly reversed by the anticholinesterase neostigmine. At doses that abolished neuromuscular transmission, the a~onotoxins had no effects on arterial blood pressure, heart rate, or responses to vagal and preganglionic stimulation, indicating their specificity for nicotinic receptors at the neuromuscular junction . of the marine snails of the genus Conus contain a variety of polypeptide toxins (for review, see OI,IVERA et al., 1985). The a-conotoxins are 13-15 residue polypeptides that bind to nicotinic acetylcholine receptors of skeletal muscle . They cause death by paralysis of respiratory muscles (for example, LD sp of conotoxin GI is 12 hg/kg i .p. in mice; CRUZ et al., 1978) and they block neuromuscular transmission in vitro (McMnxus and Muslcx, 1985 ; Hnsl-nMOTO et al., 1985). Although HnsHIMOTO et al. (1985) reported that some a-conotoxins were without effects on ganglionic nicotinic cholinoceptors or on muscarinic cholinoceptors in vitro, little is known of the selectivity of a-conotoxins for the neuromuscular junction, especially in vivo. In the present study, we have compared the effects of the a-conotoxins GI and MI with those of the non-depolarizing muscle relaxant vecuronium in the anaesthetised cat. In addition to assessing their neuromuscular potency, we simultaneously determined their effects on sympathetic ganglionic neurotransmission and on cardiac muscarinic cholinoceptors . Synthetic conotoxins GI and MI were obtained from the Peptide Institute (Osaka, Japan) . These are identical to the natural toxins isolated from Conus geographus and C. magus, respectively . Vecuronium was a gift from Organon Scientific Development Group (Newhouse, U.K .) and neostigmine methyl sulphate was from Sigma Chemicals VENOMS
~To whom wrt~espondence should be addressed . 231
232
Short Communications
0 100
S'
~~~~i~l~l~ti~ Î~
5~ I~^Y
~1 VEC
FIG .
l.
EFFECTS OF VECURONIUM (VEC
G1
3Spg/kg)
AND CONOTOXIN TISED CAT .
GI (3Opg/kg)
IN AN ANAFSTI-II:-
Records show arterial blood pressure (BP), heart rate (HR), contraction (upward deflections) of the nictitating membrane (NM) to preganglionic nerve stimulation, and contractions (upward deflections) of the tibialis anterior (TIB) and soleus (SOL) muscles in response to sciatic nerve stimulation . The right vagus nerve was stimulated every (00 sec to produce a decrease in heart rate (downward deflections of the HR trace) accompanied by a transient reduction in arterial pressure . Note that both vecuronium and conotoxin GI produce neuromuscular block, but minimal effects on BP, HR, NM or responses to vagal stimulation . Time bar indicates S min .
Ltd. Two female cats (each 2.0 kg) were anaesthetised with a-chloralose (80 mg/kg) and pentobarbitone (2.5 mg/kg) given i .p. The twitch responses of the tibialis anterior and soleus muscles of one hind leg in response to stimulation of the sciatic nerve (0.1 Hz, 0.2 msec, voltage greater than needed to elicit maximal contractions) were recorded throughout the experiment . Contractions of the left nictitating membrane were recorded in response to preganglionic stimulation of the left cervical sympathetic nerve (once every 100 sec with 10 sec trains at 5 Hz, 0.5 msec duration). Blood pressure was recorded from a carotid artery and the arterial pressure pulse triggered a heart rate monitor. Both vagi were ligated, and every 100 sec the right vagus nerve was stimulated (5 Hz for 10 sec) to induce bradycardia. Drugs and toxins were injected into a brachial vein . Further details of the preparation are given by DURANT et al. (1977) . Vecuronium (35-60 ~g/kg, 54-92 nmoles/kg, in bolus doses) produced an 83-98% block of the twitches of the tibialis anterior and soleus muscles (Fig .1) . When tested by cumulative additions in the same animals, conotoxin GI did not alter the twitch responses when given at 10 ~g/kg (6.9 nmoles/kg), but at 20 leg/kg it caused about a 10% decrease in the twitches of the soleus muscle and a slight increase in twitches of the tibialis anterior muscle . Similar augmentation of contractions of the tibialis had been observed with subblocking doses of vecuronium . At a total of 40 ~g/kg, conotoxin GI reduced responses of the tibialis anterior muscle by 80% and abolished twitches of the soleus muscle within 4 min. Twitch responses began to recover spontaneously after about 6 min, with full recovery after 15 min (tibialis) and about 40 min (soleus) . Conotoxin MI also reduced twitches : 20 Fig/kg (i.e. 13 .4 nmoles/kg) was below threshold; a total of 40 ~g/kg caused a slight augmentation of twitches of the tibialis anterior and a slight reduction of those of
Short Communications
23 3
BP mm Hp
5 HR
~
NM
~illl~lll r
I20p
~
~.
~
! g
~
(
.
I
I
~.
~
~
~
I
I
I
FiG . 2 . EFFECTS OF CONOTOXIN MI (75 Nglkg) IN AN ANAFSTHEi'ICFn CAT . Recordings are as in Fig . 1 . Note that, like vecuronium and conotoxin GI in Fig . 1, conotoxin MI produced only minimal effects on BP, HR, NM or responses to vagal stimulation at the neuromuscular blocking dose shown . Neostigmine (NEO 40 Rg/kg) produces an immediate reversal of the neuromuscular blocking effects of conotoxin MI on both TIB and SOL muscles .
the soleus muscle ; a total of 80 hg/kg abolished twitch responses of both muscles within 3~ min. Twitches began to recover after 15 min and full recovery took about 45 min. A single bolus dose of 30 ~g conotoxin GI per kg caused 99% block of the tibialis anterior and 100% block of the soleus muscles (Fig. 1). In the same experiment, a bolus dose of conotoxin MI (75 ~g/kg) also caused 100% block of both muscles in about 2 min (Fig. 2). After twitches of the tibialis anterior had recovered to 50%, an injection of neostigmine (401cg/kg) rapidly restored twitch height to control levels . In a different experiment, 20 hg conotoxin GI per kg blocked twitches by about 80% ;this block was also rapidly reversed following injection of neostigmine (40 hg/kg) . Throughout the neuromuscular blockades induced by conotoxins GI and MI (10-80 hg/kg), there were no changes in blood pressure or resting heart rate, or in the bradycardia induced by vagal stimulation, and only very small changes in the contractions of the nictitating membrane (Figs 1 and 2). Because of shortage of material, it was not possible to increase the doses of the conotoxins above those that abolished neuromuscular transmission in order to quantify the autonomic/neuromuscular selectivity of the compounds. Our results confirm that conotoxins GI and MI block neuromuscular transmission in vivo . Both toxins are as rapid in onset as vecuronium . The neuromuscular blocks wear off spontaneously, with durations similar to those of the longer-acting muscle relaxants tubocurarine and pancuronium, and are also reversed by neostigmine . The toxins are also potent neuromuscular blockers : on a molar basis, they are two times (for MI) and five times (for GI) more potent than vecuronium, which in turn is about 10 times more potent than tubocurarine . Conotoxin GI was about 2.5 times as potent in the cat as conotoxin MI, in agreement with in vitro findings (HASHIMOTO et al., 1985). However, this contrasts
234
Short Communications
with the finding that conotoxin MI was about 2.5 times more potent than conotoxin GT following i.p. injection in mice (GRAY et al., 1983) . We have also demonstrated that the two a-conotoxins are specific for the neuromuscular junction : at doses abolishing contractions of skeletal muscle, there were no obvious effects on the cardiovascular system or on responses mediated by pre-ganglionic sympathetic nerve stimulation or by vagal stimulation . Hence, like the highly selective muscle relaxant vecuronium, the a-conotoxins appear to lack the ganglion blocking properties or affinity for cardiac muscarinic cholinoceptors that result in the common side effects of clinically used muscle relaxants such as tubocurarine and pancuronium (MARSHALL, 1980 ; I3owMAN, 1982). Acknowledgement-The authors thank C[flusntve PA~reesoN for her assistance with the experiments.
REFERENCES Bow[~ux, W. C. (1982) Non-relaxant properties of neuromuscular blocking drugs. Br . J. Anaesth. 54, 147-160. CRUZ, L. Z., GRAY, W. R. and OL[veR~, B. M. (1978) Purification and properties of a myotoxin from Conus geographus. Archs Biochem. Biophys. 190, 539-548. Duw~xr, N. N., Bowau[v, W. C. Snd MARCHAI L, I. G. (1977) A comparison of the neuromuscular and autonomic blocking activities of (+)-tubocurarine and its N-methyl and O,O, N-trimethyl analogues. Eur. !. Pharmac. 46, 297-302. Gw+v, W. R., RIVER, J. E., GALYEAN, R., CRUZ, L. J. and OL[veRe, B. M. (1983) Conotoxin MI . Disulfide bonding and conformational states. J. biol. Chem . 258, 12247-12251 . HxsF[[MO~ro, K., UCHIDA, S ., YosIIro~, H., N[sxrucx[, Y., SAKAx1e~RA, S. and YUKAR[, K. (1985) Structureactivity relations of conotoxins at the neuromuscular junction . Eur. J. Pharmac. 118, 351-354. McM~r[us, O. B. and Mus[c[c, J. R. (1985) Postsynaptic block of frog neuromuscular transmission by conotoxin GI . J. Neurosci. S, 110-116. MARSHALL, I. G. (1980) Actions of non-depolarizing blocking agents at cholinoceptors other than at the motor endplate. In: Curares and Curarization, pp. 257-274 (Conseillier, C., Desmonts, J. M., Glaser, P., Montagne, J., Nahas, G. G., Sahunagne, J. C., Viars, P. and Vourc'h, G., Eds) . Paris: Libraire Arnette. OL[veR~, H. M., GRAY, W. R., Zmcus, R., Mclxrosx, J. M., V~RCe, J., R[v>ER, l., De Sxrrras, V. and CRUZ, L. J. (1985) Peptide neurotoxins from fish-hunting cone snails. Science, N. Y. 230, 1338-1343.
rorrean vol. zs, No . z ~. zsl-z3a,199o. Printed in Great Britain .
SHORT COMMUNICATIONS SELECTIVE NEUROMUSCULAR BLOCKING PROPERTIES OF aCONOTOXINS IN VIVD I. G.
M~RSt-tntt,
and A. L.
HARVmr*
Department of Physiology and Pharmacology and The Strathclyde Institute for Drug Research, University of Strathclyde, Glasgow G1 1XW, U .K . (Acceptedfor publication 18 July 1989)
I. G. MARSFIALI, and A. L. HARVEY . Selective neuromuscular blocking properties of a-conotoxins in vivo. Toxicon 28, 231-234, 1990.-Synthetic a-conotoxins GI and MI (based on the natural polypeptides isolated from Conus geographus and C. magus, respectively) were tested for effects on newomuscular transmission and on the cardiovascular system in anaesthetised cats. Both produced neuromuscular blockade at doses of 20-80 kg/kg. Conotoxin GI was about 2.5 times more potent than MI . Both were rapidly reversed by the anticholinesterase neostigmine. At doses that abolished neuromuscular transmission, the a~onotoxins had no effects on arterial blood pressure, heart rate, or responses to vagal and preganglionic stimulation, indicating their specificity for nicotinic receptors at the neuromuscular junction . of the marine snails of the genus Conus contain a variety of polypeptide toxins (for review, see OI,IVERA et al., 1985). The a-conotoxins are 13-15 residue polypeptides that bind to nicotinic acetylcholine receptors of skeletal muscle . They cause death by paralysis of respiratory muscles (for example, LD sp of conotoxin GI is 12 hg/kg i .p. in mice; CRUZ et al., 1978) and they block neuromuscular transmission in vitro (McMnxus and Muslcx, 1985 ; Hnsl-nMOTO et al., 1985). Although HnsHIMOTO et al. (1985) reported that some a-conotoxins were without effects on ganglionic nicotinic cholinoceptors or on muscarinic cholinoceptors in vitro, little is known of the selectivity of a-conotoxins for the neuromuscular junction, especially in vivo. In the present study, we have compared the effects of the a-conotoxins GI and MI with those of the non-depolarizing muscle relaxant vecuronium in the anaesthetised cat. In addition to assessing their neuromuscular potency, we simultaneously determined their effects on sympathetic ganglionic neurotransmission and on cardiac muscarinic cholinoceptors . Synthetic conotoxins GI and MI were obtained from the Peptide Institute (Osaka, Japan) . These are identical to the natural toxins isolated from Conus geographus and C. magus, respectively . Vecuronium was a gift from Organon Scientific Development Group (Newhouse, U.K .) and neostigmine methyl sulphate was from Sigma Chemicals VENOMS
~To whom wrt~espondence should be addressed . 231
232
Short Communications
0 100
S'
~~~~i~l~l~ti~ Î~
5~ I~^Y
~1 VEC
FIG .
l.
EFFECTS OF VECURONIUM (VEC
G1
3Spg/kg)
AND CONOTOXIN TISED CAT .
GI (3Opg/kg)
IN AN ANAFSTI-II:-
Records show arterial blood pressure (BP), heart rate (HR), contraction (upward deflections) of the nictitating membrane (NM) to preganglionic nerve stimulation, and contractions (upward deflections) of the tibialis anterior (TIB) and soleus (SOL) muscles in response to sciatic nerve stimulation . The right vagus nerve was stimulated every (00 sec to produce a decrease in heart rate (downward deflections of the HR trace) accompanied by a transient reduction in arterial pressure . Note that both vecuronium and conotoxin GI produce neuromuscular block, but minimal effects on BP, HR, NM or responses to vagal stimulation . Time bar indicates S min .
Ltd. Two female cats (each 2.0 kg) were anaesthetised with a-chloralose (80 mg/kg) and pentobarbitone (2.5 mg/kg) given i .p. The twitch responses of the tibialis anterior and soleus muscles of one hind leg in response to stimulation of the sciatic nerve (0.1 Hz, 0.2 msec, voltage greater than needed to elicit maximal contractions) were recorded throughout the experiment . Contractions of the left nictitating membrane were recorded in response to preganglionic stimulation of the left cervical sympathetic nerve (once every 100 sec with 10 sec trains at 5 Hz, 0.5 msec duration). Blood pressure was recorded from a carotid artery and the arterial pressure pulse triggered a heart rate monitor. Both vagi were ligated, and every 100 sec the right vagus nerve was stimulated (5 Hz for 10 sec) to induce bradycardia. Drugs and toxins were injected into a brachial vein . Further details of the preparation are given by DURANT et al. (1977) . Vecuronium (35-60 ~g/kg, 54-92 nmoles/kg, in bolus doses) produced an 83-98% block of the twitches of the tibialis anterior and soleus muscles (Fig .1) . When tested by cumulative additions in the same animals, conotoxin GI did not alter the twitch responses when given at 10 ~g/kg (6.9 nmoles/kg), but at 20 leg/kg it caused about a 10% decrease in the twitches of the soleus muscle and a slight increase in twitches of the tibialis anterior muscle . Similar augmentation of contractions of the tibialis had been observed with subblocking doses of vecuronium . At a total of 40 ~g/kg, conotoxin GI reduced responses of the tibialis anterior muscle by 80% and abolished twitches of the soleus muscle within 4 min. Twitch responses began to recover spontaneously after about 6 min, with full recovery after 15 min (tibialis) and about 40 min (soleus) . Conotoxin MI also reduced twitches : 20 Fig/kg (i.e. 13 .4 nmoles/kg) was below threshold; a total of 40 ~g/kg caused a slight augmentation of twitches of the tibialis anterior and a slight reduction of those of
Short Communications
23 3
BP mm Hp
5 HR
~
NM
~illl~lll r
I20p
~
~.
~
! g
~
(
.
I
I
~.
~
~
~
I
I
I
FiG . 2 . EFFECTS OF CONOTOXIN MI (75 Nglkg) IN AN ANAFSTHEi'ICFn CAT . Recordings are as in Fig . 1 . Note that, like vecuronium and conotoxin GI in Fig . 1, conotoxin MI produced only minimal effects on BP, HR, NM or responses to vagal stimulation at the neuromuscular blocking dose shown . Neostigmine (NEO 40 Rg/kg) produces an immediate reversal of the neuromuscular blocking effects of conotoxin MI on both TIB and SOL muscles .
the soleus muscle ; a total of 80 hg/kg abolished twitch responses of both muscles within 3~ min. Twitches began to recover after 15 min and full recovery took about 45 min. A single bolus dose of 30 ~g conotoxin GI per kg caused 99% block of the tibialis anterior and 100% block of the soleus muscles (Fig. 1). In the same experiment, a bolus dose of conotoxin MI (75 ~g/kg) also caused 100% block of both muscles in about 2 min (Fig. 2). After twitches of the tibialis anterior had recovered to 50%, an injection of neostigmine (401cg/kg) rapidly restored twitch height to control levels . In a different experiment, 20 hg conotoxin GI per kg blocked twitches by about 80% ;this block was also rapidly reversed following injection of neostigmine (40 hg/kg) . Throughout the neuromuscular blockades induced by conotoxins GI and MI (10-80 hg/kg), there were no changes in blood pressure or resting heart rate, or in the bradycardia induced by vagal stimulation, and only very small changes in the contractions of the nictitating membrane (Figs 1 and 2). Because of shortage of material, it was not possible to increase the doses of the conotoxins above those that abolished neuromuscular transmission in order to quantify the autonomic/neuromuscular selectivity of the compounds. Our results confirm that conotoxins GI and MI block neuromuscular transmission in vivo . Both toxins are as rapid in onset as vecuronium . The neuromuscular blocks wear off spontaneously, with durations similar to those of the longer-acting muscle relaxants tubocurarine and pancuronium, and are also reversed by neostigmine . The toxins are also potent neuromuscular blockers : on a molar basis, they are two times (for MI) and five times (for GI) more potent than vecuronium, which in turn is about 10 times more potent than tubocurarine . Conotoxin GI was about 2.5 times as potent in the cat as conotoxin MI, in agreement with in vitro findings (HASHIMOTO et al., 1985). However, this contrasts
234
Short Communications
with the finding that conotoxin MI was about 2.5 times more potent than conotoxin GT following i.p. injection in mice (GRAY et al., 1983) . We have also demonstrated that the two a-conotoxins are specific for the neuromuscular junction : at doses abolishing contractions of skeletal muscle, there were no obvious effects on the cardiovascular system or on responses mediated by pre-ganglionic sympathetic nerve stimulation or by vagal stimulation . Hence, like the highly selective muscle relaxant vecuronium, the a-conotoxins appear to lack the ganglion blocking properties or affinity for cardiac muscarinic cholinoceptors that result in the common side effects of clinically used muscle relaxants such as tubocurarine and pancuronium (MARSHALL, 1980 ; I3owMAN, 1982). Acknowledgement-The authors thank C[flusntve PA~reesoN for her assistance with the experiments.
REFERENCES Bow[~ux, W. C. (1982) Non-relaxant properties of neuromuscular blocking drugs. Br . J. Anaesth. 54, 147-160. CRUZ, L. Z., GRAY, W. R. and OL[veR~, B. M. (1978) Purification and properties of a myotoxin from Conus geographus. Archs Biochem. Biophys. 190, 539-548. Duw~xr, N. N., Bowau[v, W. C. Snd MARCHAI L, I. G. (1977) A comparison of the neuromuscular and autonomic blocking activities of (+)-tubocurarine and its N-methyl and O,O, N-trimethyl analogues. Eur. !. Pharmac. 46, 297-302. Gw+v, W. R., RIVER, J. E., GALYEAN, R., CRUZ, L. J. and OL[veRe, B. M. (1983) Conotoxin MI . Disulfide bonding and conformational states. J. biol. Chem . 258, 12247-12251 . HxsF[[MO~ro, K., UCHIDA, S ., YosIIro~, H., N[sxrucx[, Y., SAKAx1e~RA, S. and YUKAR[, K. (1985) Structureactivity relations of conotoxins at the neuromuscular junction . Eur. J. Pharmac. 118, 351-354. McM~r[us, O. B. and Mus[c[c, J. R. (1985) Postsynaptic block of frog neuromuscular transmission by conotoxin GI . J. Neurosci. S, 110-116. MARSHALL, I. G. (1980) Actions of non-depolarizing blocking agents at cholinoceptors other than at the motor endplate. In: Curares and Curarization, pp. 257-274 (Conseillier, C., Desmonts, J. M., Glaser, P., Montagne, J., Nahas, G. G., Sahunagne, J. C., Viars, P. and Vourc'h, G., Eds) . Paris: Libraire Arnette. OL[veR~, H. M., GRAY, W. R., Zmcus, R., Mclxrosx, J. M., V~RCe, J., R[v>ER, l., De Sxrrras, V. and CRUZ, L. J. (1985) Peptide neurotoxins from fish-hunting cone snails. Science, N. Y. 230, 1338-1343.