• The Journal of Neuroscience, December 1987, 7(12): 4195-4200

Sponsored Links

Download the ebook

The Journal of Neuroscience, December 1987, 7(12): 4195-4200
Neurotoxins from Wecffewys Spider Venom are Potent Presynaptic
Blockers in Drosophila
W. D. Branton, L. Kolton, Y. N. Jan, and L. Y. Jan
Howard Hughes Medical Institute and Department of Physiology, University of California, San Francisco, California 94143
Studies of presynaptic events in synaptic transmission may from Hololena have been characterized separately in a parallel
be facilitated through the use of specific ligands for func- study (Bowers et al., 1987). We found in the venom of the spider
tional components of the transmitter release mechanism and Plectreurys tristis inhibitory and excitatory neurotoxins with
through the use of genetics. For this purpose, neurotoxins essentially irreversible actions on the Drosophila larval neuro-
that affect neuromuscular transmission in Drosophila have muscular junction. We have purified inhibitory toxins that pro-
been identified and purified from Plecfreurys spider venom duce rapid and complete block of synaptic transmission at nano-
(PLTX). One class of toxins causes irreversible presynaptic molar concentrations. They may be potent blockers of calcium
block, probably by blocking calcium entry or by acting on entry or other processes involved in transmitter release.
other closely associated processes. These toxins have been
highly purified and are peptides of about 7 kDa in molecular Materials and Methods
weight. They specifically block transmitter release at nano- Venom. Crude Plectreurys venom (PLTX) was purchased from Spider
molar concentrations and may be useful in further biochem- Pharm. Black Canvon Citv. AZ. and from BioActives. Salt Lake Citv.
UT. The venom was pro&ced’by electrical milking procedures fro&
ical studies. collected and laboratory-reared spiders, and was specially selected to
be free of regurgitated digestive juices.
Potent biological toxins have been important in the isolation Electrophysiology. Standard electrophysiological techniques were used
and characterization of molecules involved in neurotransmis- to record from the Drosophila larval neuromuscular junction, as pre-
sion, such as sodium channels (Agnew, 1984; Catterall, 1986; viously described (Jan and Jan, 1976). The physiological saline con-
tained 120 mM NaCl, 2 mM KCl, 4 mM MgCl, 0.5 mM CaCl, 36 mM
Noda et al., 1986) and acetylcholine receptors (Noda et al., 1983; sucrose, and 5 mM HEPES buffer, pH 7.3. The nerve innervating a
Anholt et al., 1984). Many other components of neurotrans- single hemisegment was drawn into a stimulating suction electrode, and
mission, however, have not been characterized biochemically. one of the longitudinal muscles of the hemisegment was impaled with
Toxins that block certain potassium (Hughes et al., 1982; Miller a 20 Ma, 3 M KCl-filled glass microelectrode. The nerve was stimulated
once every 5 set (except where otherwise noted) with a 0.2 msec pulse
et al., 1985; Halliwell et al., 1986) and calcium (Kerr and Yo- at twice threshold for activation. Methods for stimulating and recording
shikami, 1984; Olivera et al., 1985) conductances in vertebrates the excitatory junction potential (ejp) in the experiments with mutant
have been found, and they may be useful tools in biochemical larvae (Figs. 4, 5) were the same, except that the saline contained 0.15
as well as physiological analyses. Similar studies in Drosophila mM CaCl. Extracellular recordings (Figs. 4, 5) were made from a small
loop of motor nerve pulled into a tight-fitting glass suction electrode at
melanogaster could be further facilitated by the use of mutants, a distance of about 200 pm from the nerve terminals (Wu et al., 1978).
and by the use of transformants carrying mutagenized genes. Toxins were applied with vigorous mixing to the bathing solution via
To identify neurotoxins that are potent in Drosophila, we screened small aliquots from concentrated stock solutions. Concentrations of
commercially available venom of various spider species. We toxins used in physiological studies were estimated by comparing the
were particularly interested in venoms containing toxins that absorbance peak at 2 14 nm on high-performance liquid chromatography
(HPLC) gel filtration to that of insulin. The validity of these estimates
might be good ligands for previously uncharacterized synaptic for alpha-PLTX II, the most abundant toxin found, has been corrob-
membrane channels or receptors. The venom of the spiders orated by preliminary amino acid analysis and microsequence data.
Agelenopsis aperta, Hololena curta, and Plectreurys tristes all Purification of toxins. All chromatography was done with a Waters
produced essentially irreversible presynaptic block at Drosoph- HPLC system. Water and acetonitrile-(ACN) were HPLC grade. Tri-
ila larval neuromuscular junction. Only Hololena and Plectreu- fluoroacetic acid (TFA) and heotafluorobutvric acid (HFBA) were Pierce
Sequanal reagent;. EfflLents w&e monitored for absorban& at 214 nm.
rys venoms were available to us in quantities sufficient for de- In the region of absorbance peaks, 0.2 ml fractions were collected,
tailed biochemical and physiological studies. Some of the toxins assayed for activity as in Figure 1, and active regions (Fig. 2, underline)
were pooled and lyophylized for further study. Initial HPLC gel filtration
was done with 7.5 x 300 mm TSK Bio Sil 125 and 250 columns
Received Mar. 20, 1987; revised June 3, 1987; accepted June 3, 1987 (BioRad) in series with each other, and a 7.5 x 75 mm precolumn. The
This work is supported by the Howard Hughes Medical Institute and NIH mobile phase was 32% ACN, 0.1% TFA, flow rate 1.O ml/min. Injection
Grant 15963. We would like to thank Dr. Leslie Timpe, who did the voltage- volume was 20 ~1. For reverse phase, appropriate fractions from gel-
clamp of pupal flight muscle, and Dr. Heidi Phillips, who did the anatomical filtration runs were pooled, lyophylized, and resuspended in 0.1% TFA.
binding, for allowing us to use their preliminary findings in the Discussion. We This material was loaded onto a 4.6 x 25 cm Cl 8 reverse-phase column
also thank Chuck Kristensen at Spider Pharm and Dr. Hunter Jackson at Bio (Vydac TP2 18), equilibrated with 0.1% TFA, and the column eluted at
Actives for their help and advice concerning venoms, Pamela Lee for help in
assaying materials for activity, Dr. Bany Rothman for advice on HPLC gel fil- a flow rate of 1.O ml/min with a linear gradient of O-60% ACN (0.1%
tration, and Phyllis Cameron and Lisa Schulte for expert help in preparation of TFA) in 60 min. Inhibitory peak II (alpha-PLTX II) was repurified on
the manuscript. the same reverse-phase column under isocratic conditions (approxi-
Correspondence should be addressed to Dale Branton, Howard Hughes Medical mately 45% ACN) using 0.05% HFBA in place of TFA, and was finally
Institute, 3rd & Pamassus, P.O. Box 0724, San Francisco, CA 94143. repurified on 2 TSK 250 gel-filtration columns in series without a pre-
Copyright 0 1987 Society for Neuroscience 0270-6474/87/124195-06$02,00/O column.
4196 Branton et al. - Presynaptic Toxins from Spider Venom
D - H
Figure 1. Inhibitory and excitatory activities found in the Plectreurys spider venom. The effect of the inhibitory toxin on excitatory junction
potentials (ejps) at the Drosophila larval neuromuscular junction is shown in oscilloscope tracings A-F. A, Control trace in the absence of toxin.
B-F, Sequential traces taken at 1 min intervals after the application of lo-9 M alpha-PLTX II. The block of transmission is progressive and complete
within 5 min. G-I, Effect of the excitatory activity on the larval neuromuscular preparation. G, Control trace showing a single ejp evoked in
response to a single nerve stimulus. H and I, Traces taken 10 min after the application of the excitatory toxin at a dilution of lo4 relative to crude
venom. In H there is a burst of ejps following the same stimulus delivered in G. I, Spontaneous activity in the absence of electrical stimulation.
Vertical calibration bar, 5 mV. Horizontal calibration bar: 5 msec (A-F); 0.2 set (G-I).
We have included analytical traces of PLTX under isocratic condi- presumed to be caused by an action on axonal sodium or po-
tions in Figure 2 in order to indicate the apparent purity that we have tassium channels, and have not been studied further.
achieved, but the retention time under these conditions can vary greatly The inhibitory activity was of particular interest because its
with the condition of the column and precolumn used, the precise con-
centration of ACN and buffer in the mobile phase, and the amount of effects were fully consistent with a specific and irreversible block
material loaded on the column. Generally we have found it necessary of presynaptic calcium channels. The activity had a M, of about
to verify the toxin peak at each step by bioassay on larval neuromuscular 6-7 kDa (Fig. 2.4, E) and was confined to a restricted region of
junction. reverse-phase HPLC gradients (Fig. 20. Three particularly po-
tent activities were identified in this region and the toxins des-
ignated alpha-Plectreurys toxin (PLTX) I, II, and III. All 3 were
Results purified. They showed an identical action on the neuromuscular
The Drosophila larval neuromuscular junction preparation (Jan junction. Alpha-PLTX II has been the most consistently abun-
and Jan, 1976) was used in analyzing the action of neurotoxins dant toxin in various samples of venom and consequently has
from spider venoms (Figs. 1, 3, 4, 5). A single suprathreshold been studied more thoroughly.
stimulus to the segmental nerve normally gives a single com- Purification of alpha-PLTX II was achieved in 4 steps: After
pound action potential (Wu et al., 1978). The action potential HPLC gel filtration (Fig. 2A) and reverse-phase gradient (Fig.
in the nerve, in turn, gives rise to a single excitatory junctional 20, a second solvent system was used to rechromatograph the
potential (ejp) in the muscle. Spontaneous miniature ejps (mejps) toxin on reverse-phase columns, and final purification was car-
are easily observed with intracellular recording of the postsyn- ried out on a high-resolution gel-filtration column. Analytical
aptic muscle cell. When axonal conduction is blocked by ad- chromatograms of the final fraction on reverse phase (Fig. 20)
dition of TTX, an ejp can still be generated if the length of the and gel filtration (Fig. 2E) showed a single, sharp peak of ab-
nerve between the stimulating electrode and the muscle is re- sorbance, corresponding to the biological activity of the toxin.
duced to a few hundred microns, thereby allowing direct stim- This toxin appeared to be composed of a single polypeptide,
ulation of the nerve terminal. This normal pattern of synaptic because reduced and alkylated toxin ran as a single peak on
transmission was altered by neurotoxins present in the spider HPLC gel filtration (not shown), with an apparent A4, of 7 kDa.
venom. Crude Plectreurys venom contains many different com- Preliminary amino acid analysis and microsequence data have
ponents that interfere with Drosophila neuromuscular trans- further corroborated that the purified toxin is a polypeptide of
mission. Although we have not exhaustively ruled out possible about 7 kDa. Alpha-PLTX II represented about 0.1% of total
secondary effects of the toxins described in this report, they were protein in crude venom, and accounted for about 20% of the
chosen for study because their effects appear to be specific, po- neuromuscular blocking activity of the venom.
tent, and not easily reversible. The purified toxin has very potent and specific physiological
The excitatory effect of the venom was manifested by multiple effects. When alpha-PLTX II was applied at a nanomolar con-
ejps in response to a single stimulus (Fig. 1H) and by the oc- centration in the saline bathing the neuromuscular preparation,
currence of frequent spontaneous ejps in the absence of nerve the ejp gradually disappeared over a period of about 5 min (Fig.
stimulation (Fig. II). These effects were not easily reversible; 1, A-F). The action of the toxin was irreversible; the block could
they persisted after the recording chamber was washed for an not be reversed by perfusing the chamber with saline for up to
hour with toxin-free saline. This activity had an apparent M, 2 hr. At higher doses of toxin ( 1O-8 M), the ejp could be blocked
of 8-10 kDa and was largely confined to a peak of material in less than 1 mitt, with very little effect on the postsynaptic
identified on reverse-phase HPLC gradients (Fig. 2A, B). Its muscle membrane potential or on the frequency or amplitude
effects were similar to those of scorpion venoms on the same of the spontaneous mejps (Fig. 3A). At low doses of toxin (c.
preparation, and were completely abolished by TTX under con- lOm*OM), release was nearly, but not completely, blocked. As
ditions where ejps could still be elicited by direct stimulation shown in Figure 3, B, C, the amplitude of the ejp varied dis-
of the nerve terminal. Therefore, the excitatory effects were continuously between 0 and a value predicted accurately by the
The Journal of Neuroscience, December 1987, 7(12) 4197
A. Gel Filtration A.
-50 . Resting potential
-40 3 AAAAAAAA).
MEJP amplitude
Retention time (min)
B. Reverse Phase .
.,..."..I., 60
EJP amplitude
Time (minutes)
Retention time (min)
C. Reverse Phase
~~“.‘.‘j-I 60-
0 10 20 30 40 50 60
Retention time (min)
Aprc Win 6.5 KD
D. E. sulin 5.7 KD O- d
0.05 - mV mV
0.04- Figure 3. Alpha-PLTX II causes a specific reduction in the quanta1
content ofthe ejp. A, Effect ofalpha-PLTX II on 4 parameters of synaptic
p 0.03- transmission measured continuously during rapid and complete trans-
r; mission block induced by application of lo-* M toxin. Toxin was applied
==c 0.02- at time 0. The nerve was stimulated once every 10 sec. Effects on
postsynaptic membrane potential (A), miniature excitatory junction po-
O.Ol- tential (mejp) amplitude (A), and mejp frequency (+) are small, while
the ejp (0) is quickly and completely blocked. B and C, The quanta1
,L &--JI L I I I I
nature of transmission is preserved during partial block of transmission
induced by application of approximately 1Om’o M alpha-PLTX II for
0-o 0 5 10 15 20
about 20 min. Mean quanta1 content was reduced from normal levels
Retention time (min) Retention time (min) (greater than 40) to approximately 0.5 by application of lo-lo M alpha-
PLTX II. A histogram of 200 nerve-evoked ejps, including 122 failures,
Figure 2. A, HPLC gel filtration separates crude venom into inhibitory is shown in B, and a histogram of 345 mejp amplitudes is shown in C.
(v) and excitatory (A) fractions, based on apparent M,. B, Gradient Mean quanta1 content, calculated by the method of failures, was 0.49,
elution on reverse-phase HPLC of excitatory fractions from the gel and was 0.44 when calculated by dividing the mean ejp amplitude by
filtration reveals a major peak of excitatory activity (A). C, Gradient the mean mejp amplitude. The distribution of the evoked potentials
elution on reverse-phase HPLC of inhibitory fractions from gel filtration closely parallels the mejp amplitudes as expected at very low quanta1
reveals a region of inhibitory activity that includes 3 major peaks (Z, content.
II, III). Peak II was repurified in a second solvent system on reverse
phase, and finally again on HPLC gel filtration under high-resolution
conditions. Analytical traces of the final material in these last 2 systems
are shown in D (reverse phase) and E (gel filtration). Arrows show the amplitudes of the spontaneous mejps; the quanta1 content was
elution positions of insulin (5.7 kDa) and aprotinin (6.5 kDa). reduced from greater than 40, before the application of toxin,
to about 0.5, demonstrating that the toxin acted by blocking
transmitter release.
Alpha-PLTX II does not act by blocking nerve conduction.
4198 Branton et al. - Presynaptic Toxins from Spider Venom
Figure 4. Recurrent terminal spikes
in eug.9~~~~~~~are not blocked by TTX. b TTX
Top truces are extracellular recordings
from a larval segmental nerve made near .
the nerve terminal. Bottom truces are
intracellular recordings from the post-
synaptic muscle. a, Characteristic train ,:*- 1;; //
of terminal spikes (arrows) closely as-
sociated with the prolonged ejp. b, Re-
sidual spikes still associated with the
ejp evoked by direct stimulation of the
terminal in the presence of 1O-5 M TTX.
(Calcium was 0.15 mM.) 20s
Figure 5. Alpha-PLTX II blocks electrical activity at or near the nerve terminal. A-F, Sets of simultaneous intracellular recordings (top truces)
from the larval muscle, and extracellular recordings (bottom truces) from the larval nerve near the terminal region in an abnormally excitable
mutant eug ShK0’20 in 0.15 mM calcium. A, Control traces showing a typical prolonged ejp associated with a train of extracellularly recorded spikes
(small arrows) originating at or near the terminal. The number of these recurrent spikes correlates well with the duration of the ejp. B, Recordings
made 2 min after application of 10m9 M alpha-PLTX II. Note the reduced ejp and concomitantly reduced number of terminal spikes. C, Recording
made after 5 min in the toxin. Both the ejp and terminal spikes are abolished. D-F, Recordings made at a faster time scale showing the presence
or absence of the propagated compound action potential in the nerve (large arrow). D, Control traces before application of toxin. E, Recordings
made 5 min after application of alpha-PLTX II. The compound action potential was still present after the ejp and associated terminal spikes were
completely blocked. (The shape of the compound action potential showed small variations during an experiment, but there was no consistent
alteration following the application of toxin.) F, Recordings made after application of 5 x 1O-6 M TTX. TTX blocked the remaining compound
action potential in the alpha-PLTX II-treated preparation, leaving only the stimulus artifact. Vertical calibration bar: 10 mV (top truces); 0.2 mV
(bottom truces). Horizontal calibration bar: 20 msec (A-C’); 5 msec (D-F).
The Journal of Neuroscience, December 1987, 7(12) 4199
The toxin did not block the compound action potential recorded matography, do not appear to share antigenic determinants, and
from the nerve (Fig. 5E). Furthermore, the toxin was still ef- preliminary cross-competition experiments in anatomical bind-
fective in blocking transmitter release when nerve conduction ing studies are consistent with separate binding sites. If, in fact,
was blocked by TTX and the nerve terminal was depolarized the binding sites are separate, it is possible that these toxins act
directly by local stimulation (not shown). These results dem- through completely separate mechanisms. They might also sim-
onstrate that alpha-PLTX II blocks transmitter release without ply act at separate sites on the same target molecule.
blocking nerve conduction. The toxin may block the entry of The venoms we have studied probably contain toxins that
calcium into the nerve terminal. Alternatively, it may act upon act on other targets and are specific for synapses in animals
subsequent biochemical processes involved in the release of other than insects. A presynaptic toxin that appears to act se-
transmitter. lectively on certain vertebrate central synapses has recently been
Indirect evidence, obtained through the use of Drosophila purified in another laboratory from the venom of Agelenopsis
mutants with defective potassium currents, suggests that alpha- aperta (Jackson et al., 1986).
PLTX II probably blocks calcium entry into the nerve terminal. We suspect that the alpha-Plectreurys toxins are either potent
The ether a go go (eag) mutation produces flies with reduced blockers of insect presynaptic calcium channels or that they
delayed-rectifier currents, while the Shaker (Sh) mutation elim- specifically affect processes closely linked to calcium entry at
inates the transient A current. Flies with both mutations thus the nerve terminal. As such, they may be useful anatomical and
have severely defective potassium conductances and therefore biochemical probes of neurotransmission. The availability of a
extremely excitable neuronal membranes (Salkoff and Wyman, battery of presynaptic toxins with similar actions in different
198 1; Ganetzky and Wu, 1982; Wu et al., 1983). In abnormally species and with different sites of action within species could
excitable mutants such as eag ShK0120, a single nerve stimulus contribute greatly to our understanding of the structure and
can cause a prolonged action potential in the nerve terminal function ofthe molecules involved in calcium-dependent release
and prolonged calcium-dependent release. The prolonged action of neurotransmitter.
potential originating at or near the motor nerve terminal is
associated with recurrent spikes that can be recorded extracel- References
lularly via a suction electrode placed on the nerve near the Agnew, W. S. (1984) Voltage-regulated sodium channel molecules.
terminal (Jan and Jan, 1979; Ganetzky and Wu, 1982) (Fig. 4a). Annu. Rev. Physiol. 46: 5 17-530.
When nerve conduction is blocked by TTX and the nerve ter- Anholt, R., J. Lindstrom, and M. Montal (1984) The molecular basis
of neurotransmission: Structure and function of the nicotinic acetyl-
minal is stimulated directly, recurrent spikes are reduced in
choline receptor. In Enzymes of Biological Membranes, vol. 3, A.
number, but still evident (Fig. 4b), suggesting that they arise at Martonosi, ed., pp. 335-401, Plenum, New York.
least in partial from a TTX-resistant mechanism (Jan and Jan, Bowers, C. W., H. S. Phillips, P. Lee, Y. N. Jan, and L. Y. Jan (1987)
1979). Since the recurrent spikes are blocked by Co or Cd, they The identification and purification of an irreversible presynaptic neu-
are most likely associated with calcium currents. When alpha- rotoxin from the venom of the spider, Hololena curta. Proc. Natl.
Acad. Sci. USA (in press).
PLTX II was applied to the mutant neuromuscular preparation, Catterall, W. A. (1986) Molecular properties of voltage-sensitive so-
the recurrent spikes were blocked simultaneously with the ejp dium channels. Annu. Rev. Biochem. 5.5: 953-954.
(Fig. 5). Therefore, this toxin affects electrical properties of the Ganetzky, B., and C. F. Wu (1982) Indirect suppression involving
nerve terminal. A likely target of the toxin would be the calcium behavioral mutants with altered nerve excitability in Drosophila mel-
anogaster. Genetics 100: 597-6 14.
channels in the presynaptic nerve terminal, although we cannot Halliwell, J. V., I. B. Othman, A. Pelchen-Matthews, and J. 0. Dolly
rule out the possibility that some other ionic mechanism is (1986) Central action of dendrotoxin: Selective reduction of a tran-
involved. sient K conductance in hippocampus and binding to localized accep-
tors. Proc. Natl. Acad. Sci. USA 83: 493-497.
Discussion Hughes, M., G. Romey, D. Duval, J. P. Vincent, and M. Lazadunski
(1982) Apamin as a selective blocker of the calcium-dependent po-
Presynaptic toxins often show a unique specificity of action. tassium channel in neuroblastoma cells: Voltage-clamp and biochem-
Omega-Conotoxin, for example, is a potent blocker of neuronal ical characterization of the toxin receptor. Proc. Natl. Acad. Sci. USA
calcium channels at the frog neuromuscularjunction, has a com- 79: 1308-1312.
plicated effect on mammalian CNS, and is inactive on both Jackson, H., M. Urnes, W. R. Gray, andT. N. Parks (1986) Presynaptic
blockage of transmission by a potent long lasting toxin from Agele-
mouse and Drosophila neuromuscular junctions (Kerr and nopsis aperta spiders. Sot. Neurosci. Abstr. 12: 730.
Yoshikami, 1984; Olivera et al., 1985). Apparently, alpha-PLTX Jan, L. Y., and Y. N. Jan (1976) Properties ofthe larval neuromuscular
II specifically affects the insect nerve terminal. Alpha-PLTX II junction in Drosophila melanogaster. J. Physiol. (Land.) 262: 189-
has not been found to be active in vertebrates. For example, it 214.
does not block the frog neuromuscular junction at 1O-8 M con- Jan, Y. N., and L. Y. Jan (1979) Genetic dissection of synaptic trans-
mission in Drosophila melanogaster in insect neurobiology and pes-
centration. In Drosophila, the toxin has no effect on the inward ticide actions. In Society of Chemical Industry, F. E. Rickett, ed., pp.
calcium or outward potassium currents of Drosophila pupal 161-168, London.
flight muscle, and it does not affect the divalent cation-depen- Kerr, L. M., and D. Yoshikami (1984) A venom peptide with a novel
dent action potentials generated in larval muscle in the presence presynaptic blocking action. Nature 308: 282-284.
Miller, C., E. Moczydlowski, R. Latorre, and M. Phillips (1985) Cha-
of 1 mM Sr. Therefore, this toxin does not block calcium chan- rabdotoxin, a protein inhibitor of single Ca*+ -activated K+ channels
nels in Drosophila muscle. from mammalian skeletal muscle. Nature 313: 3 16-3 18.
Hofolena toxin has an action similar to that of alpha-PLTX Noda, M., H. Takahashi, T. Tanabe, M. Toyosato, S. Kikyotani, Y.
II on fly neuromuscular junction, but is apparently unrelated to Furutani, T. Hirose, H. Takashima, S. Inayama, T. Miyata, and S.
Plectreurys toxins biochemically (Bowers et al., 1987). Hololena Numa (1983) Structural homology of Torpedo californica acetyl-
choline receptor subunits. Nature 302: 528-532.
toxin is a significantly larger molecule, apparently consisting of Noda, M., T. lkeda, H. Suzuki, H. Takashima, T. Takahashi, M. Kuno,
2 disulfide-bonded peptides, rat

Use: 0.3985