The Effects of Curare in the Cockroach: II. Blockage of Nerve Impulses by dTC

Although for many years insects were believed to be insensitive to curare (Harlow, 1958; Hopf, 1952; Krupp, Lendle & Stapenhorst, 1952; Roeder, 1948, 1953; Roeder & Weiant, 1950; Usherwood, 1963), the recent report of Larsen, Miller & Yamamoto (1966) has demonstrated that insects are, indeed, sensitive to curare. Subsequent investigations (McCann, 1966; Flattum, Friedman & Larsen, 1967; Friedman, 1967) have been unable to elucidate the site or mechanism of insect curarization. However, studies of leg contraction in the cockroach (Friedman & Carlson, 1969) led us to the conclusion that curare most probably causes failure of contraction by preventing conduction of action potentials in motor nerve fibres. We furthermore proposed that blockage by curare of action potential conduction is the mechanism of insect curarization. This report supports our proposal by presenting electrophysiological evidence which indicates that curare (d-tubocurarine, dTC) blocks the conduction of action potentials in nerve fibres of the cockroach, Periplaneta americana.

1. The giant-fibre preparation (Fig. 1) was used to assess the effects of various test solutions on the ability of giant fibres within the nerve cord to conduct action potentials. The cockroach was first decapitated and its legs and wings were removed. The animal was then pinned on to Plasticine, ventral surface uppermost. The sterna overlying the nerve cord in the abdominal segments, A1−A2 and A5−A6 were dissected away. The nerve cord connectives between A5 and A6 were freed and a pair of platinum stimulating electrodes was placed under them. The nerve cord was then cut posterior to A1, and the A1–A2 connectives were dissected free of peripheral connexions. The freed connectives were then picked up in a suction electrode. Test solutions were applied to the A5−A6 region of the preparation with a Pasteur pipette.

2. The cercus-cercal nerve preparation (Fig. 3) was used to assess the effect of curare on a sensory system. An animal was first decapitated and its legs and wings were removed. It was then pinned on to Plasticine, dorsal surface upwards. After exposing the nerve cord both cereal nerves were dissected of their attachments and severed from the sixth abdominal ganglion. The freed ends of the cereal nerves were picked up in suction electrodes. To introduce test solutions into this preparation the last six segments of the cerci were first removed. Small drops of test solutions on the needle tip of a microlitre syringe were then applied by touch to the cerci. The solutions were drawn into the cerci by capillary action. In each experiment one cercus received a known volume of curare solution while the contralateral cercus received an equal volume of saline. Curare solution was always applied to the cercus whose cereal nerve yielded the higher initial endogenous activity.

3. The cercus-nerve cord preparation (Fig. 4) was used to assess the post-synaptic effect of curare application to the cercus. The preparation was made by first decapitating the animal and removing its legs and wings. The animal was then pinned on to Plasticine, ventral surface uppermost. The sterna overlying abdominal segments A1−A2 were removed. The exposed nerve cord was freed and cut below A1. The nervecord connectives were then picked up in a suction electrode. Test solutions were introduced into the preparation in the same manner as used for the cercus-cercal nerve preparation described above. Activity in each cercus was induced by passing an air current over the cerci.

4. The cereal nerve-nerve cord preparation (Fig. 5) was used to assess the effect of curare when the drug was applied to the cereal nerve. The cockroach was decapitated and its legs and wings removed. The animal was then pinned on to Plasticine, ventral surface uppermost. The sterna overlying the A4-cerci area were removed. One cereal nerve and the A4–A5 nerve cord connectives were freed of their surroundings. A pair of platinum stimulating electrodes was placed under the cereal nerve. The nerve cord was cut below A4, and the A4–A5 connectives were picked up in a suction electrode. Test solutions were applied to the A6-cerci region of the preparation with the aid of a Pasteur pipette.

A Grass S-4 stimulator was used to generate stimuli. The stimuli were fed through a Grass stimulus-isolation unit and delivered to the preparation with platinum hook electrodes. Stimuli were always of the same duration (0·1 msec.) and frequency (0·5 pulses/sec.). The voltage varied between 1·5 and 6·0 volts, as indicated in the Results. Electrode polarity was reversed to obtain consecutive stimuli of alternate polarity. Nerve impulses were recorded with a suction electrode. The suction electrode was composed of fine polyethylene cannula tubing and silver wire 0·1 mm. in diameter. The solution used to relay impulses to the silver wire within the tubing was the saline used for the dissection of the preparation. Nerve impulses were amplified by a Grass P-9 pre-amplifier. The approximate amplification gain used in these experiments was 2400. The output was displayed on an oscilloscope (Tektronix 502A) and photographed with a Grass Kymograph camera.

The saline solution used was Pringle’s (1938) insect saline. The test solutions included saline containing 0·032 M d-tubocurarine, 0·033 M sucrose and 0·032 M aspartic acid respectively. All solutions were buffered with Tris buffer to pH 7·4.

Stimulation of the giant-fibre preparation with external electrodes evoked a compound action potential followed by smaller, trailing potentials (Figs, 1, 2). Threshold stimulating voltages for such preparations were found never to exceed 2·0 V. Preparations maintained in saline were observed for as long as 330 min. without any changes attributable to deterioration (not shown). However, the irrigation of giantfibre preparations with a 0·032 M curare solution caused a gradual increase in the voltage needed to evoke a compound action potential. As shown in Fig. 1D, 120 min. after the irrigation of a giant-fibre preparation with 0·032 M curare solution, a fourfold increase in the stimulating voltage failed to evoke an action potential. Failure to obtain a compound action potential with a stimulating voltage of 6·0 V. was taken to indicate blockage of conduction since it was initially observed that preparations which did not respond to stimulation at 6·0 V. would not respond to stimulation at higher voltages.

The conduction blockage seen after the application of 0·032 M curare solution to the giant-fibre preparation is due to the effect of the curare molecules per se. A number of experiments were undertaken to rule out other possible causes. Fig. 2 A, for example, demonstrates that the giant-fibre preparation maintains the ability to conduct impulses at threshold voltage 120 min. after initiating saline irrigation. The conduction blockage seen 120 min. after the application of curare solution cannot be due to the composition of the saline solution. Conduction blockage is also not attributable to an osmotic imbalance caused by the addition of 0·032 moles of solute per 1. of saline. As shown in Fig. 2B, the addition of 0·033 moles of sucrose per 1. of saline does not produce conduction block. Finally, dTC is an acidic molecule, and an additional amount of Tris had to be added to hold the pH to 7·4. That this additional amount of Tris in the solution is not responsible for the conduction block was demonstrated by showing that a 0·032 M aspartic acid solution buffered with additional Tris does not cause conduction blockage (Fig. 2C). In contrast, 0·032 M curare solution produced conduction block even when the preparation was stimulated with a voltage of 4 × threshold (Fig. 2D).

The cereal nerves of the cockroach conduct sensory impulses from the cerci to the sixth abdominal ganglion (A 6). The extreme sensitivity of the cerci to air currents produces low-amplitude, high-frequency action potentials which can be recorded from the cereal nerves of an undisturbed cercus-cercal nerve preparation (Fig. 3 A). The application and subsequent uptake by capillary action of 1·5 µl. of 0·032 M curare solution by a cercus blocked these action potentials. In contrast, the uptake of 1·5 µl. saline did not block such activity (Fig. 3B).

The cerci of the cockroach are connected via the cereal nerves and their synapses in A 6 to the giant fibres of the nerve cord. Thus, passage of an air current over the region of the cerci increases the frequency of firing of the giant fibres. It is therefore possible to monitor the post-synaptic giant-fibre response to the application of curare to one cercus, and the response to the application of saline to the contralateral cercus. A diagram of the preparation used for such studies appears in Fig. 4. If the application of curare to one cercus blocks impulses in its cereal nerve, then the giant-fibre response seen when the cerci are stimulated (by the passage of air) should be due to the cereal nerve input from the non-curarized cercus. Fig. 4 demonstrates that this was the case. The removal of the curare-treated cercus did not noticeably alter the giant-fibre response to the passage of an air current over the region of the cerci (compare Fig. 4B with Fig. 4A). The removal of the non-curarized cercus, however, prevented a giantfibre response to the passage of an air current (Fig. 4C).

The application of 0·032 M curare solution to the cereal nerve–A6 region of the cockroach prevented the giant fibres from responding to electrical stimulation of the cereal nerve. As shown in Fig. 5 A, the giant fibres responded to electrical stimulation of a cereal nerve with a burst of impulses. The giant-fibre response did not vary so long as the cereal nerve-A6 region of the preparation was irrigated with saline. The substitution of 0·032 M curare solution for saline, however, prevented the giant-fibre response. Blockage of the giant-fibre impulses in response to electrical stimulation of the cereal nerve did not always occur within as short a time interval as occurred in the experiment shown in Fig. 5. Blockage did always occur within 120 min. of the application of 0·032 M curare solution.

In a previous study (Friedman & Carlson, 1969) evidence was given to suggest that curare blocks motor nerves in the cockroach. The results of this investigation demonstrate that it can block both sensory (cereal) and central (giant) axons. Giant fibres of the cockroach nerve cord fail to respond to electrical stimulation 120 min. after exposure to 0·032 M curare solution. Since the fibres are capable of conducting action potentials when exposed to 0·032 M aspartic acid solution, neither the presence of extra Tris or the presence of a 0·032 M solution can be held responsible for the conduction block seen after the application of 0·032 M curare solution. Curare may act to block conduction in the cereal nerve because when a solution of this drug was applied to a cercus activity in the cereal nerve terminated. This possibility is further strengthened by the fact that application of dTC to the cereal nerve-A6 region of a cockroach blocks the post-synaptic giant-fibre response to cereal nerve stimulation.

Curare-induced blockage of action potentials has been previously reported for both vertebrate and invertebrate nerve fibres. Walsh & Deal (1959) described curare-induced blockage in the frog sciatic nerve, while Dettbarn (1960) described blockage of single nerve fibres of the same animal. Dettbarn (1963) achieved blockage of the walking leg nerve in the lobster, while Rosenberg and his colleagues have achieved conduction blockage in the giant axons of squid. All of these reports, however, involved some sort of pre-treatment of the fibre: exposure to venom and/or detergent, teasing out of individual fibres, or exposure to abnormal ionic conditions. In the current investigation fibres were not pre-treated in any manner. The studies which involve pre-treatment of nerve fibres to obtain a curare-induced blockage have been rightfully viewed with caution. However, the establishment of blockage without pretreatment in this work leaves little doubt that curare can induce blockage.

If one accepts the blockage of conduction as a mechanism of curare action, some of the conflicts in previous reports concerned with the mechanism of insect curarization are resolvable. McCann (1966) was unable to inhibit muscle action potentials by directly applying dTC to the neuromuscular junction, yet inhibited these potentials by injecting curare. It seems reasonable that direct application of dTC to the neuromuscular junction was ineffective because dTC does not act at the insect neuromuscular junction. It also seems reasonable that dTC when injected intra-abdominally was effective because the drug was able to reach the motor axons and produce blockage.

Suga & Katsuki (1961) demonstrated that perfusion of the prothoracic ganglion of the grasshopper with 0·03 % dTC blocked auditory impulses in the central nervous system. They argue that dTC penetrated the ganglion and acted on synapses inside the ganglion. This mechanism of action seems unlikely in view of the known inaccessibility of insect central nervous system synapses to drugs (Hoyle, 1965; Roeder, 1953) and the low concentration of dTC used. More probably, curare caused blockage in the tympanic nerve as it enters the prothoracic ganglion.

Friedman (1967) and Friedman & Carlson (1969) have shown that failure of leg contraction results when curare is injected into the abdomen, is injected into a leg or is applied to the metathorax of a cockroach. Mechanical studies revealed that the contraction failures observed in response to all three of these modes of curare administration are similar in nature. Thus, the contraction failures all seem to be due to one common mechanism of curare action. The one common mechanism is most probably the blockage by curare of action potential conduction in the motor nerve fibres.

The demonstration and proposal that curare acts in insects by blocking actionpotential conduction in nerve axons seems correct for two other reasons: (1) it does not challenge the theory or the evidence which indicate that glutamate is the transmitter at insect excitatory neuromuscular junctions, and (2) it preserves the theory that curare acts at cholinergic neuromuscular junctions and at no other type of neuromuscular junction.

This work constitutes part of a Ph.D. dissertation submitted to the State University of New York at Stony Brook by K. Friedman. The authors with to thank Professor Edward Eisenstein for his invaluable help during the course of the work.