The effects of five phospholipases A2 from the venom of king brown snake, Pseudechis australis, on nerve and muscle

0041-0101(94)W115-4

Tackon. Vol. 32. No . 12, pp . 1559-1572 . 1994 Science Ltd Printed in Great Britain . All rights reserved aa1-0101/94 $7.00+ .00

THE EFFECTS OF FIVE PHOSPHOLIPASES AZ FROM THE VENOM OF KING BROWN SNAKE, PSEUDECHIS AUSTRALIS, ON NERVE AND MUSCLE M. FATEm, 1 E. G. ROWAN,' A. L. HARvEY'* and J. B. HARRIS2

'Department of Physiology and Pharmacology, University of Strathclyde, Glasgow GI 1XW, U.K.; and 'Muscular Dystrophy Group Research Laboratories, Newcastle General Hospital, University of Newcastle upon Tyne, Newcastle upon Tyne NE4 6BE, U.K. (Received 10 March 1994 ; accepted 5 August 1994)

M. FATEM, E. G. ROWAN, A. L. HARVEY and J. B. HARRLs. The effects of five phospholipases A2 from the venom of king brown snake, Pseudechis australis, on nerve and muscle . Toxicon 32, 1559-1572, 1994 .-The effects on vertebrate neuromuscular function of five homologous phospholipases A2 (PLA 2) (Pa-3, Pa-8, Pa-9C, Pa-1 OF and Pa-12B) from the venom of the Australian king brown snake, Pseudechis australis, were determined . These isoenzymes (0 .2-1 .6 .UM) reduced, with different potencies, responses of chick biventer cervicis preparations to nerve stimulation and to exogenously applied acetylcholine, carbachol and KCl in a time- and concentration-dependent way but with different potencies . They also blocked twitches of mouse hemidiaphragm preparations evoked by nerve and by direct muscle stimulation. Pa-8 was the most active and Pa-9C was the least potent . There was a strong correlation between the enzymatic activity and the effect of toxins on the responses of mouse hemidiaphragm to direct muscle stimulation, but weak correlation between the effects on indirect responses and enzymatic activity. Intracellular recording from endplate regions of mouse triangularis sterni nerve-muscle preparations showed that Pa-lOF and Pa-12B at 0.2 uM significantly reduced quantal content after 10 min. Pa-8 (0.2,u M) reduced the amplitude of endplate potentials by about 25% and abblished miniature endplate potentials within 15 min. Pa-3 (0 .2 p M) and Pa-9C (0 .8,u M) also significantly reduced quantal content by about 30% of control after 30 min. Among these toxins, Pa-3 and Pa-8 at 0.2 p M depolarised mouse muscle fibres after 30 min. Extracellular recording of action potentials at motor nerve terminals of mouse triangularis sterni preparations indicated that these isoenzymes reduced the waveforms associated with both Na+ and K+ conductances . Since no facilitatory effect on the release process has been observed, the apparent blockade of K+ conductance by some of these toxins may not be a selective action on K+ channels, but may be secondary to membrane depolarisation. An in vivo study with Pa-8 and Pa-lOF demonstrated myotoxic effects. Light microscopic examination showed a degeneration of mouse and rat skeletal muscle fibres caused by Pa-8 and Pa-lOF . For the in vivo study, rats received 80 hg/kg of the toxins s.c . and mice were injected i.m . with the toxins (40 hg/kg) . Myotoxicity appears to be the predominant effect of these five toxins. *Author to whom correspondence should be addressed . 1559

1560

M. FATEHI et al. INTRODUCTION

SEVERAL PLA2 isoenzymes have been purified from the venom of the Australian king brown snake, Pseudechis australis (NISHWA et al., 1985). These PLA2 consist of a single chain of 118 amino acids and contain 14 half-Cys residues. The amino acid sequences of eight of these enzymes has been reported (TAKACAKr et al., 1990b) . The enzymatic activity and lethality of these PLA2 were determined by TAKASAKI et al. (1990x). Their enzymatic activities against 1,2-dipahnitoylglycerophosphocholine varied from 557 U/mg (Pa-9C) to 8750 U/mg (Pa-8) and their LD50 values measured by i.v . injections in mice ranged from 0.11 ug/g for Pa-8 to 4.7 Rg/g for Pa-9C (TAKAsAKI et al., 1990x) . The neuromuscular activity of seven PLA2 from P. australis was reported by RowAN et al. (1989) and GEH et al. (1992). The aims of present study were to characterise the neuromuscular activity of five different PLA2 from P. australis (Pa-3, Pa-8, Pa-9C, Pa-10F and Pa-12B) and to determine any correlation between enzymatic activity and the pharmacological activity of these toxins . Preliminary accounts of some of the results have been presented at the Xth European Meeting on Toxinology (FA7'EHI et al., 1993) and the 3rd Asia-Pacific Congress on Animal, Plant and Microbial Toxins (FATEtn et al ., 1994). MATERIALS AND METHODS Toxins Pure phospholipase A2 toxins (Pa-3, Pa-8, Pa-9C, Pa-10F and Pa-12B) were gifts from Drs N. TAWYA and C. TAKAsAm (Department of Chemistry, Tohoku University, Sendai, Japan) . The isolation and purification of these toxins have been reported by TAKAum et at. (1990x). Chick biventer cervicls preparations Biventer cervicis muscles and associated nerves (Gu4sBaRo and WAxanvea, 1960) were dissected from 7-10-day-old chicks killed by exposure to COz. Two muscles were mounted in 2ml tissue baths containing modified Krebs-Henseleit solution maintained at WC, pH 7.3-7.4, and bubbled with 95% 02 + 5% C02- The composition (mM) of modified Krebs-Henseleit solution was as follows: NaCl, 118.4; KH 2PO4, 1.2 ; glucose, 11 .1 ; NaHC03, 25 ; CaC12, 2.5; MgSO4, 1.4 and KCI, 4.7 . All conditions for twitch tension recording were the same as those described previously (RowAN et al., 1989). Twitches were evoked by stimulating the motor nerve at 0.1 Hz with pulses of 0.2 msec duration and a voltage greater than that which produced a maximal response . To detect any changes in postsynaptic sensitivity, responses to exogeuously applied acetylcholine (1-2 mM), carbachol (30-40 p M) and KCl (20-40 mM) were recorded prior to the addition of toxin and at the end of the experiment. In control preparations (after 5 hr) there was no significant change in twitch height or in responses to exogenously applied acetylcholine, carbachol or KCI. Mouse phrenie nerve-hemtrdiaphragm preparations Hemidiaphragms and attached phrenic nerves were dissected from male mice (20-25 g, Bantin and Kingman A strain) and mounted in 5 ml tissue baths containing physiological solution at 37°C and pH 7.3-7 .4. The other conditions were the same as those used for chick biventer cervicis preparations . The preparations were mounted on an electrode that enabled both nerve and direct muscle stimulation. The mouse hemidiaphragms were stimulated alternately indirectly via the phrenic nerve and directly at frequency of 0.05 Hz. For indirect stimulation, the phrenic nerve was stimulated by pulses of 0.2 msec duration and a voltage greater than that required to produce a maximal response . For direct stimulation, twitches were evoked by pulses of 2 msec duration and a voltage greater than that which produced a maximal response. Mouse Manguiaris sterni nerve-muscle preparations The left tniangular is sterni nerve-muscle preparations (McAaDLE et at., 1981) isolated from 20-25 g male mice (Bantin and Kingman A strain) were used for electrophysiological experiments . Mice were killed by exposure to a lethal concentration C0,. The rib cage was removed and transferred to a constant-flow Petri dish superfused with physiological (modified Krebs-Henseleit) solution at a rate of 20-25 ml min- '. During dissection, the solution was aerated with a mixture of 95% 03/5% C02. After dissection, the preparations were pinned thoracic side downwards to the base of a 5 ml chamber perfused at a rate of 10-15 ml min-' with the physiological

Pseudechis atrstralts PLA7 on Nerve and Muscle

1561

solution . This solution (30ml) was aerated with a mixture of 95% 07/5°/. C07 and recycled throughout the experiment, by a peristaltic pump. The intercostal nerves were stimulated via a suction electrode with pulses of 0.5 Hz frequency, 20 psec duration and sufficient strength to produce a maximal twitch . MgCl7 (5-7 mM) was added to the solution to prevent muscle twitching. Experiments were performed at room temperature (23-25°C). Intracellular recording. Resting membrane potentials, endplate potentials (epps) and miniature endplate potentials (mepps) were measured by standard microelectrode techniques (FATT and KArz, 1951). The glass microelectrodes (resistance: 10-'20 mQ) were filled with 3 M KCI. Epps and mepps with rise times of 1 msec or less were recorded continuously from one muscle cell before and after application of toxin. Recording sites were rejected if the membrane potential was less than -60 mV on the initial impalement or the membrane potential varied by more than 10% during the first 20 min of control recording. While adopting this protocol, stable recordings of over 180 min could be achieved . Extracellular recordheg. Perineural waveforms were recorded by inserting a glass microelectrode (resistance: 10-15 MR filling solution: 2 M NaCl) into the perineural sheath of one of the branches of an intercostal nerve close to endplate areas (MALLART, 1985). Since the shape of the perineural waveform was dependent on the electrode position (PsNxBt and DnsveR, 1986), signals were monitored continuously from one site throughout the experiment. If the amplitude of signal decreased by more than 10% within the first 30 min before addition of toxins, the recordingsite was considered to be unsuitable and was rejected. While adopting this protocol, stable recordings of over 180min could be achieved. Electrophysiological data analysis The amplified signals were recorded on video tape using a video tape recorder (JVC HR-D530 EK) and' Sony PCM-701ES adapted for wide bandwidth (DC-20 kHz) analogue signal recording. The Synaptic Current Analysis (SCAN) program written by J. Dempster at Strathclyde University was used to visualise and analyse the digitised signals (DtJwr18t, 1988). At each time period 100-200epps, 200-400 mepps and 50-60 perineural waveforms were stored . Abnormal signals due to electrical interference, summation of events or spiking were rejected, and accepted signals were averaged. To calculate quantal content of epps, the amplitudes of epps and mepps were corrected to a normalised potential of -80 mV, so that comparisons of potentials could be made at different membrane potentials. Quantal content was then calculated as the average epp divided by the average mepp . Because of the small size of the endplate in Mg7+ paralysed preparations, corrections for non-linear summation were not required (McLAcxt-AN and MARTIN, 1981). Statistics The results are expressed as the mean f S.E . The statistical significance was evaluated by the Mann-Whitney U-test and Student's meat . Values of P < 0.05 were taken as significant. Histological methods Rats (250-300 g) and mice (23-25 g) were anaesthetised with halothane/N70/07 (5 : 65 : 30). Rats were injected s.c. with normal saline (0.2 ml) and with toxin (80,ug/kg in 0.25 ml) at the demarcation line between the soleus and gutrocnemius muscles into the right and left hind limbs, respectively. The technique was described in full by HARats and JomvsoN (1978) . Mice received toxins (40 pg/kg in 0.1 ml, i.m .) and normal saline (0.1 ml) into the tibialis anterior muscles of right and left hind limbs, respectively. The animals were sacrificed 24 hr later, and muscle tissues were carefully dissected and then either frozen in liquid nitrogen or fixed by immersion in 10% formalin and embedded in wax. Longitudinal and transverse sections (5 pm) were prepared from frozen or wax-embedded muscles and stained with hematoxylin and eosin. RESULTS Chick biventer cervicis preparations

Twitch responses of chick biventer cervicis preparations (CBC) evoked by nerve stimulation were reduced in the presence of the five toxins in a time- and concentrationdependent manner (see Figs 1 and 2). The order of potency was: Pa-8 = Pa- IOF > Pa12B > Pa-3 > Pa-9C. There was no significant correlation between the enzymatic activity and the ability of these toxins to reduce twitches evoked by nerve stimulation. Figure 3

M.

1562

ô c 0 U t

. ca ô z u 3 H

FATEHI et ai.

110 100 90 90 70 90 50 40 30 20 10

0

Time (Min)

Ma. 1 .

EFFEcis OF PHOSPHOLIPASBS A, (0 .8 p M) PUAIF® FROM THE VENOM of PseudeCIdS aWtralia ON RESPONSES OF CHICK BIVEN7ER CERVICIS PREPARATIONS TO NERVE STaWLATION. (/) Pa-3, (AL) Pa-8, (V) Pa-9C, ( " ) Pa-10F, and (*) Pa-12B. Points represent the means of values obtained from four experiments and S .E. are indicated by the bars unless smaller than symbols.

shows that the correlation coefficient is 0.60 for enzymatic activity and the effect on indirect twitches of CBC calculated from linear regression . In the presence of the toxins, responses to exogenous acetylcholine, carbachol and KCl were also reduced. Figure 4 shows agonist responses after exposure to concentrations of toxins that caused equal amounts of twitch blockade (50% block in 140-150 min). Although Pa-3 (1.6 pM), Pa-8 (0.2 juM), Pa-10F (0.2 juM) and Pa-12B (0.8 pM) caused

t 270 n

F 0

e Â

C

_ m O

>K 0

240 210ISO150 120

90 6030-

E 10 -7 10

i-

-'

10-'

Concentration (M)

FIN. 2 . EFFECI30FDn'PERENT OONCENTaA7toNsoF (N) Pa-3, (A) Pa-8, (V) Pa-9C, (#) Pa-10F, AND (*) Pa-1213 ON Tm TO 50% BLOCKADE OF Twn'CHBS OF CHICK BIVBNTER CERVICIs PREPARATIONS

EVOKED BY NERVE STIMULATION . Points show the means of four values and S.E . are indicated by the bals unless smaller than symbols.

Pseudechts australis PLA2 0 c

100-

U

so-

0

M 0 t

u

soso-

0

30-

11,

20-

O

c

70-

40-

m

1563

go-

3 FF wr

on Nerve and Muscle

a

10 0-

r10-1

10°

Enzymatic Activity (KU/mg)

FYo . 3 . RELATION OF ENzvmATIc Acrtvi m of (a) Pa-9C, (b) Pa-12B, (c) Pa-10F, (d) Pa-3, AND (e) Pa-8 AND THEIR mmsiTORY HFFECIS AT (0.814M) ON RESPONSES of cmcK BImNm cERvicis PREPARATIoNs To NERvE sTUruLATIoN AFTER 150 min (n - 4). Phospholipase A2 activity (kU/mg) was determined using the titrimetric method, against l,2dipahnitoylglycerophosphocholine at pH 8.0 and 37°C . The velocities of reactions were calculated from the slope of the linear part of the curve. kU, kilounit . One unit was defined as the amount of the enzyme that released 1 umole of fatty acid/min under above conditions (data from TAxAs ju et al., 1990x). The straight line was obtained by linear regression analysis of the data shown .

50% block of twitches in almost the same time, they produced different amounts of block of responses to agonists . Pa-8 (0 .2 juM) induced a general decrease of 60-70% in responses .to agonists, but Pa-9C, even at a higher concentration (4.8 uM), did not decrease these responses significantly . The other toxins showed a pattern between these extremes as follows: Pa-3 > Pa-IOF = Pa-12B . _

110 100

`o U

so

--

70 s0

w

so

0 a ô Q

so

c

20

40 30

ACh

Carb

KCI

FYo. 4. EFPEcT3 OF THE EQuIEymcnvE coNcENTaAnoNs of (p) Pa-3 (1 .6PM), ( ") Pa-8 (0 .2,u M), (0) Pa-9C (4.8pM), (®) Pa-10F (0.2pM), AND (®) Pa-12B (0.8 /AM) ON CONTRACrIJRESSOFCIBCK BIvENTER cERvias FREPAmTtoNs IN RESPONSE To ACgnrLcHoLm (ACh) AT 1-2 mM, cARBACHOL (Carb) AT 30-40pM AND KCI AT 20-4OmM . Each column represents the mean of values obtained from four experiments and bars show S .E.

1564

M.

FATEHI et al.

. d r u

H 0w 0

u A W C O

F+

V

no. 5.

Concentration (M)

TtE MAxmule coNTRAcnjRE of cmcx arvENzEa cEavrcrs MUSCLES nvnucm By ( ") Pa-3, (A) Pa-8, (V) Pa-9C, ( " ) Pa-10F, AND (*) Pa-12B . Contractures have been compared to the control responses to nerve stimulation . Points are means of values obtained from four experiments and S .E. are indicated by the bars unless smaller than symbols.

All toxins caused a slow and sustained contracture of the preparations but with greatly varying potency (Fig. 5). The order of potency for causing contracture was: Pa-8 > Pal OF = Pa-3 > Pa-12 > Pa-9C. Times to maximum contracture of CBC caused by toxins were not the same . The values for time to maximum contracture produced by -Pa-3 (1 .6pM), Pa-8 (0 .2,uM), Pa-9C (4.8 juW, Pa-LOF (0 .2,uM) and Pa-12B (0 .8pM) were 100 f 5 min, 210 f 10 min, 300 t 15 min, 200 f 15 min and 190 t 10 min, respectively. Mouse hemidiaphragm preparations

All five PLAZ toxins reduced responses to direct and indirect muscle stimulation in a concentration-dependent manner (Fig. 6a, b). Pa-3 and Pa-9C reduced indirect and direct A

B ..

110 100 90 s0 70 60

110 100

so

90 s0 70 60, 30

20 10

20 10

40 30

40 30

0

10-5

a

0 1ô-7

1

0-6

Concentration (M) Concentration (M) IF1o . 6. E1mEcn op ( ") Pa-3, (A) Pa-8, (V) Pa-9C, ( " ) Pa-10F, AND (*) Pa-12B on RESPONSES or MousE 1uouPHRAom P1e>PARAnons To (A) NERVE s71MU1 .AT1on, AND (B) nnRECr MUSCLE srnrul.Arlon .

lâ7

Pmudechis australes PLA2 on Nerve and Muscle

1565

â

â ô

â 0

70

.a

60 504030 ~-

bI

20

10-1

ei

100 Enzymatic activity (KU/mg)

dI e~

ô 8

90 8o 70 60 50 40 30 20

1o1

3

10-1

100

101

Enzymatic activity (KU/mg)

FIG. 7 . RELATIONS of PIIosPxoLIPAsE A2 Acnvnus AND EPPEc,-Is of (a) Pa-9C, (b) Pa-12B, (c) Pa-10F, (d) Pa-3, AND (e) Pa-8 AT 0 .8 ,UM ON RESPONSES of MOUSE iEMIDIAPHRAGM PREPARA71ONS To (A) NERvE sTnnULAT1oN AND (B) DIRECT MU3ctE srosuLATION .

Bars represent S.E . (n = 4) . Analysis as for Fig . 3 .

twitches almost in parallel, but reductions to indirect responses induced by Pa-8, Pa- IOF and Pa-12B were much steeper than reductions to direct responses. The order of potency at 0 .81~ M for blocking of indirect muscle responses of mouse hemidiaphragm preparations was: Pa-12B > Pa-8 = Pa- IOF > Pa-3 > Pa-9C, but in the case of direct responses the order was: Pa-8 > Pa-3 > Pa- IOF > Pa-12B > Pa-9C . There was a significant correlation (r = 0.98) between enzymatic activity and the action ofthe toxins on directly stimulated twitches, but the overall correlation (r = 0.64) between enzymatic activity and block of responses to indirect stimulation was not significant (Fig. 7a, b). Electrophysiological results Effects on the resting membrane potential. Muscle fibres of mouse triangularis sterni were

depolarised by exposure to Pa-3 (0.2pM), membrane potential falling to 91 f 2%, 89 f 2% and 82 t 1 % of control after 10, 30 and 60 min, respectively. Pa-8 (0.2 juM) reduced membrane potential to 87 f 1 % and 73 f 8% of control after 10 and 30 min, respectively . Neither Pa-9C (0.8 pM), Pa-IOF (0.2 pM) nor Pa-12B (0.2 pM) caused significant depolarisation after exposure for 30 min (Fig. 8a), but the toxins decreased the membrane potential to 89 f 1%, 87 t 1 % and 85 f 2% of control after 60 min, respectively (Fig. 8b). Effects on endplate potentials. Pa-3 (0 .2 pM) decreased the amplitude of endplate potentials (epps) to 87 f 2%, 67 f 7% and 47 f 10% of control after 10, 30 and 60 min, respectively . There was, however, no significant effect on the time constant of decay of epps . Pa-3 also reduced the amplitude of miniature endplate potentials (mepps) after 10 min. Pa-3 showed a biphasic action on the frequency of mepps: a transient increase of about 15% of control after about 10 min was followed by a reduction of about 40% in spontaneous release. Pa-3 reduced the quantal content of epps to 77 f 4% of control after 30 min, and to 65 f 17% of control after 60 min (Fig. 8a, b).

M . FATEHI er ai.

1566 A

B

g

control Fra "

Pa-8 "

" :1

0

Pa-9C Pa-1012 Pa-128 11

I

"

. " A010

M

a:

" . .I : .I :'

I

control

Pa-3

110101110

ii

Pa-9C

"

:1 " '

Pa-10F

Pa-10F

""

"'

' "

C4

AND (B) 60 min. except ept Pa-9C (0 .8 y M) . Each column represents the mean S .E . of values' obtained from five experiments. Significantly different from control: *P < 0.05. I

Pa-8 (0.2,uM) diminished the amplitude of epps to 79 f 5% and 26 f 7% of control after 10 and 30 min, respectively, and reduced the time constant of decay of epps by about 20% . Pa-8 also reduced the size and frequency of mepps by about 70% and 60% within 10 min. No mepps could be detected after about 15 min exposure of preparations to Pa-8 (0.2 p M) and, hence, effects on quantal content could not be determined . Pa-9C (0.8 juM) reduced the amplitude of epps to 93 f 2%, 74 f 3% and 54 t 6% of control after 10, 30 and 60 min, respectively, without changing the time constant of decay of epps significantly . Pa-9C had no effect on either amplitude or the frequency of mepps within 30 min. After 60 min there was a decrease in the amplitude of mepps to 83 f 7% of control but mepp frequency remained unchanged. The quantal content of epps. was reduced by Pa-9C to 69 f 4% and 54 f 4% of control after 30 and 60 min, respectively (Fig. 8a, b).

1

402D I 10

I 30

I 60

0

10

Time (min)

30

Time (ta ia)

9 . EFFECra OF (CI) Pa-3 (0.2pM), (A) Pa-8 (0 .2pM), (0) Pa-9C (6.4pM), (O) Pa-10F (0 .2 JIM), AND (yr) Pa-12B (0 .2 yM) ON THE FIRST (A) AND THE SECOND (B) NEGATIVE COMPONENTS FIG .

OF THE PERINEURAL WAVEFOR1dS RECORDED FROM MOUSE TRIANGULARIS STERN1 PREPARATIONS .

Each point represents the mean f S .E . (n = 3) .

60'

Pseudechis australls PLA Z on Nerve and Muscle

B

1567

C

1ms

Emscrs OF (A) KCl AT (*) 4.7 mM, 6 mM AND 7 mM ; (O+) 10 mM ; (*) 12 MM, AND ( " ) 3 min AR1ER ADDmoN of KCI (12 mM), (B) Pa-3 (0.2 pM), (C) Pa-8 (0.2,uM), (D) Pa-9C (6.4,u M) . (E) Pa-10F (0.2 uM), AND (F) Pa-12B (0.2 PM) ON THE Pt3twEURAL wAvtpoRHs REeoRDED FRom MO. 10 .

HOUSE TRIANOULARLS SPERM PREPARATIONS .

a: Control, 0: 10 min after toxin, b; 30 min after toxin, and c: 60 min after toxin. The upper calibration bar is related to (A) and (B) and the lower calibration bar is for (C}-(F).

Pa-1OF (0.2 pM) reduced the amplitude of epps to about 60%, 50% and 35% of control after 10, 30 and 60 min, respectively . Pa- IOF did not significantly affect the size of mepps within 30 min, but it increased the frequency of mepps by about 30% of control after 30 min. However, the frequency of mepps returned to control levels after 60 min. This toxin decreased quantal -content of epps to 74 f 5%, 52 f 5% and 50 f 6% of control after 10, 30 and 60 min, respectively. Pa-12B (0.2 pM) reduced the amplitude of epps to about 60%, 50% and 35% of control after 10, 30 and 45 min, respectively. It did not affect the time constant of decay of epps. Pa-12B did not affect the size and frequency of mepps within 30 min, but there was about 15% reduction in both the amplitude and frequency of mepps after almost 45 min. Pa-12B reduced the quantal content of epps to about 65%, 50% and 40% of control after 10, 30 and 45 min, respectively . Effects on the nerve terminal action potential. When an electrode is inserted through the perineural sheath of a motor nerve a waveform composed of two negative spikes can be recorded upon nerve stimulation. The first negative spike is associated with inward Na+ current at the last nodes of Ranvier in the axonal trunk. The second negative spike represents the local circuit current generated by the outward current of K+ and an inward Cat+ current at motor nerve terminals. The second spike can be reduced by K+ channel blockers such as 3,4-diaminopyridine (3,4-DAP) and tetraethylammoniurn (TEA) (BRiGANT and MALLART, 1982; MALLART, 1985). By monitoring perineural waveforms, it is

1568

M . FATEHI et al.

possible to investigate the effects of substances on the ionic conductances generating nerve terminal action potentials . All five PLAZ reduced the amplitude of both negative spikes of perineural waveforms (Fig. 9a, b). Pa-3 (0.2 pM) reduced the amplitude of the first negative spike by less than 8% within 30 min, and by 16 t 6% of control after 60 min. Pa-3 caused about 30% reduction in the second spike after 60 min. Pa-8 (0.2,u M) did not affect the perineural waveform within 10 min, but it decreased the first spike by 12 f 4% and 37 t 4% of control after 30 and 60 min, respectively . The second spike was reduced by Pa-8 by about 47 f 3% ofcontrol after 60 min. After 60 min, Pa-9C (6 .4 juM) decreased the first and the second spikes by 63 f 10% and 49 t 5% of control after 60 min, respectively. The

FYo . 11 . SBcnoNs of rxozEN musmx~ 5-6IIM TmcK, sretNED wrm Beeereroxvt .na AM Borna . Normal soleus muscle of the rat (A) in transverse section consists of muscle fibres grouped into fascicles . The inoculation of Pa-8 (B) results, within 24 hr, in widespread destruction of the muscle . Many fibre profiles exhibit hypercontraction (arrowheads) and others the loss of internal architecture (arrows). In longitudinal section (C) the muscle fibres exhibit hypercontraction (arrowheads) areas where sarcomeres are torn apart by hypercontraction (arrows) . Calibration bar : 100 'Um .

Pseudechis austratis PLA2 on Nerve and Muscle

1569

depressive effect of Pa- IOF on the perineural waveform was observed within 10 min and developed very quickly. After 60 min, Pa- IOF (0.2 pM) decreased the amplitude of the first and the second spikes by 53 f 2% and 77 t 4%, respectively . Pa-12B (0 .2,uM) did not change the first spike within 10 min, but decreased the amplitude of the second spike by about 15% . However, Pa-12B reduced the first and the second spikes by about 60% -after 60 min. Figure 10 shows some examples of the effects of toxins on perineural waveforms recorded extracellularly from mouse triangularis preparations . In general, the second spike ofperineural waveform was reduced more than was the first negative spike by these toxins . The first negative spike of perineural waveform, however, was eventually reduced by all of the toxins. It would seem, therefore, that the toxins did not appear to show specific effects on either K+ or Na+ channels in motor nerves . The effects on perineural waveforms may be due to a generalised depolarisation of the nerve, i .e. the toxins may cause a localised release of K+ either directly or indirectly from damaged muscle . In an attempt to explore this possibility the effect of graded depolarisations produced by increasing the extracellular concentration of KCl was studied on perineural waveforms. Hence, any changes to the perineural waveform resulting from KCl-induced depolarisation can be compared with the effects of the PLA2 toxins . Increasing the extracellular KCl to around 4.7-6 .0 mM affected neither the resting membrane of mouse muscle fibres nor the perineural waveforms. KCl (7 mM) reduced muscle fibre resting membrane potential to 94% of control, but there was no effect on the perineural waveform . KCl (10 mM) depolarised muscle cell membrane to 87% of control and reduced the first and the second negative spikes by about 13% and 30% of control, respectively. KCl at a higher concentration (12 mM) initially abolished the second spike and decreased the first spike by about 60%, and then after about 3 min abolished the waveform (Fig. 10). Histological studies Many phospholipases A2 of snake venoms are potent myotoxins, and it was of interest to determine whether the phospholipases of P. australis were myotoxic . Only fractions Pa-8 and Pa-lOF were available in sufficient quantity to extend the studies into this area . Both toxins were myotoxic . In the rat soleus, histological examination of muscles 24 hr after the injection of 80 hg/kg of either Pa-8 or Pa-lOF revealed that both toxins caused vasodilation, the coagulation of erythrocytes, and hypercontraction of the muscle fibres . In many cases, the hypercontraction resulted in the tearing of myofibres. There was no evidence of significant haemorrhage. The main features of myotoxic damage are illustrated in Fig. 11 . Similar studies were made on the tibialis anterior muscles ofmice. Each toxin (40 pg/kg) caused extensive necrosis, with characteristic hypercontraction, and disruption of the myofibres. Blood vessels (capillaries, venules and arteriols) were dilated. There was little evidence of haemorrhage, although occasional extravascular erythrocytes could be identified. Many damaged fibres were infiltrated with polymorphs and other inflammatory cells . Intramuscular nerve bundles and muscle spindles were largely undamaged. DISCUSSION

The venom of the king brown snake, Pseudechis australis, contains several phospholipases A2 (PLA2) with a wide range of enzymatic activity and lethality (TAKASAKI et al.,

157 0

M. FATEHI et al.

1990x). These PLAZ show almost 80% homology in their amino acid sequences (TAKMAK1 et al., 19906). As such homologous enzymes could provide an appropriate set of molecules for studies ofstructure-activity relationships of toxic PLAZ, we wanted to characterise their effects on nerve and muscle . In this study, we have shown that the phospholipases Pa-3, Pa-8, Pa-9C, Pa-lOF and Pa-12B from P. australis have diverse pharmacological activities . All of the toxins tested decreased twitch responses of chick biventer cervicis nerve-muscle preparations and induced a muscle contracture . Some of the toxins significantly reduced postsynaptic sensitivity to acetylcholine and carbachol, whereas others caused only minor changes to postsynaptic sensitivity. Thus, on CBC some of the toxins seem to have predominantly presynaptic activity and others postsynaptic activity (this is especially clear for Pa-9C) . The changes to postsynaptic sensitivity could be due to changes in receptor sensitivity or to changes in muscle contractility, following damage to muscle fibres . The latter case may account for some of the observed effects, e.g. responses to KCl were reduced by some of the toxins, and responses of mouse hemidiaphragm preparations to direct stimulation were also reduced. In an attempt to clarify the presynaptic and postsynaptic effects of the toxins on mammalian neuromuscular transmission, transmitter release studies were carried out on mouse triangularis sterni preparations. These studies unequivocally demonstrated that all of the toxins tested have postsynaptic activities, i.e. they cause a modest reduction in muscle fibre resting membrane potential (the largest decrease is about 30%) and, with the exception of Pa- IOF, they reduce the amplitude of mepps after 60 min . There was an apparent relationship between enzymatic activity and myotoxicity of these phospholipases. For example, there was a significant correlation between enzymatic activity and the effect of the toxins on responses of CBC preparations to KCl (r = 0.91); between enzymatic activity and contracture of CBC preparations caused by the tôxins (r = 0.87); between enzymatic activity and the effect on direct responses of MHD preparations (r = 0.98); and between enzymatic activity and effect of the toxins on resting membrane potential of mouse muscle fibres (r = 0.87). Although the results obtained from experiments in vitro and in vfvo experiments confirmed the myotoxicity of Pa-8 and Pa-lOF from the venom of P. australis, there is evidence for additional neuromuscular activity of these enzymes. For example, Pa-9C, Pa-lOF and Pa-12B blocked nerve-evoked responses of MHD preparations faster than responses to direct muscle stimulation. The blockade of nerve-evoked twitches induced by Pa-lOF and Pa-12B was greater than the reduction in responses of CBC preparations to exogenous acetylcholine and carbachol. In addition, intracellular recording showed that all of the toxins reduced the amplitude and the quantal content of epps, indicating that they act presynaptically to decrease the release of acetylcholine . This reduction in quantal content may be due to block or reduction of the nerve action potential and/or disruption of exocytosis . In an attempt to investigate the mechanism of action, the effects of the toxins on the ion channels controlling nerve terminal action potentials were examined . Pa-8, Pa-lOF and Pa-12B reduced the amplitude of both parts of the extracellular waveform generated by the nerve action potential. The reduction of the second part of the waveform (which is associated with the efflux of potassium ions at the nerve terminal) was faster and more marked than the reduction of the first part of the waveform which is associated with inward Na+ currents . Since these phospholipases did not increase quantal content, they are probably not blocking potassium channels. It is more likely that the changes in perineural waveform are as a result of a decrease in the

Pseudechis australis PLA2 on Nerve and Muscle

157 1

ability of the action potential to depolarise the nerve terminal . In fact, BRAGA et al. (1992) have shown that blockade of a fraction of the Na' channels (by either reducing the driving force for Na+ or by applying low concentrations of tetrodotoxin) reduces the spread of depolarisation to the nerve terminals and prevents voltage-dependent K+ channels being activated at nerve terminals. Hence, the reduction in the amplitude of the second part of the perineural waveform is an indirect effect of the toxins . Possible mechanisms for the effects of these toxins on the perineural waveform could be: (1) direct nerve terminal depolarisation following membrane damage, (2) blocking the conduction of the action potential to the terminals, or (3) release of K+ from damaged muscle fibres leading to indirect nerve terminal depolarisation . Very similar changes to the perineural waveform were seen after exposure of mouse triangularis sterni preparations to KCl (10-12 mM) (see Fig. 10), which produced a 10 mV depolarisation of the membrane potential of muscle fibres . Hence, terminal depolarisation (either directly or indirectly) may account for some of the effects on the perineural waveform . The mechanism responsible for the reduction in quantal content remains as yet unknown. The lack of a significant correlation between enzymatic activity and changes in nerve-evoked responses of CBC preparations and those of MHD preparations, and the decrease in quantal content of epps, suggest that neurotoxicity is not necessarily associated with the phospholipase activity of these enzymes. In addition to the previously described effects, these toxic PLA2 exerted some striking selectivity in terms of susceptible species. For instance, they were more potent and fast acting when applied to mouse preparations, but much slower in their action against chick preparations . Acknowledgements-We are grateful to Prof. N. TAAuYA and DR C. TAKASAKI who provided us with the toxins . We also wish to thank the technical staff in the Muscular Dystrophy Research Group of Newcastle General Hospital who prepared slides for histological study .

REFERENCES BRAGA, M. F. M., ANDERSON, A. J., HARvEY, A. L. and ROWAN, E. G. (1992) Apparent block of K+ currents in mouse motor nerve terminals by tetrodotoxin, p-conotoxin and reduced external sodium . Br . J. Pharmac. 106, 91-94. BAIGANT, J. L. and MALLART, A. (1982) Presynaptic currents in mouse motor endings. J. Physiol. 333, 619-636. DE1U'sm, J. (1988) Computer analysis of electrophysiological signals. In: Microcomputers in Physiology : A Practical Approach, pp . 51-93 (FRAZrR, P. J., Ed .) . Oxford : IRL Press. FATErn, M., ROWAN, E. G. and HARvEY, A. L. (1993) Effects of two toxic phospholipases Az from venom of the king brown snake, Pseudechis australis, on neuromuscular transmissi on . Toxicon 31, 519. FATEHI, M., ROWAN, E. G., HARvEY, A. L. and HARRIs, J. B. (1994) Characterization of toxic effects on nerve and muscle induced by three phospholipases A2 from the venom of Pseudechis australis. Toxicon 32, 546. FATT, P. and KATZ, B. (1951) An analysis of the endplate potential recorded with intracellular electrodes . J. Physiol. 115, 320-330. GEH, S. L., ROWAN, E. G. and HARvEY, A. L. (1992) Neuromuscular effects of four phospholipases A2 from the venom of Pseudechis australis, the Australian king brown snake. Taxiton 30, 1051-1057. GwswRG, B. L. and WARRrmR, J. N. (1960) The isolated chick biventer cervicis nerve-muscle preparation. Br. J. Pharmac. 15, 410-411 . HARRIs, J. B. and JOHNSON, M. A. (1978) Further observations on the pathological responses of rat skeletal muscle to toxins isolated from the venom of the Australian tiger snake, Notechis scutatus scutatus. Clin . exp. Pharm. Phys. 5, 587-600. MALLART, A. (1985) Electric current flow inside perineurial sheaths of mouse motor nerves . J. Physiol. 368, 565-575. WARDLE, J. J., ANGAUT-PETrr, D., MALLART, A., BOURNAÜD, R., FAILLE, L. and BRIGANT, J. L. (1981) Advantages of the triangularis sterni muscle for investigation of synaptic phenomena . J. Neurosci . Meth . 4, 109-116.

1572

M . FATEHI et al.

McL.Actu.AN, E . M . and MARTIN, A. R . (1981) Non-linear summation of end-plate potentials in the frog and mouse. J. Physiol. 311, 307-324 . NImwA, S .,1ERASimaA, M ., SinmAzu, T ., TAKASAxi, M. and TAwA, N . (1985) Isolation and properties of two phospholipases A2 from the venom of an Australian elapid snake (Pseudechis australis) . Toxicon 23, 73. PENNER, R . and DREYER, F . (1986) Two different presynaptic calcium currents in mouse motor nerve terminals. PJklgers Arch . 4W 190-197. ROWAN, E. G., HARvEY, A. L., TAKA&Am , C. and TAUayA, N . (1989) Neuromuscular effects of three phospholipases A2 from the venom of the Australian king brown snake Pseudechis australis . Toxicon 27, 551-560. TACAsAri, C., Suzuxi, J. and TAIIQYA, N . (1990x) Purification and properties of several phospholipases As from the venom of Australian king brown snake, Pseudechis australts . Toxicon 28, 319-327 . TACAsA m , C ., YuTANi, F. and Kmiymrmn, T . (1990b) Amino acid sequences of eight phospholipases Az from the venom of Australian king brown snake, Pseudechis australfa . Toxicon 23, 329-339 .