CHARACTERISTICS OF THE NEUROMUSCULAR JUNCTION IN A FLATFISH (Citharichthys sordidus) Roger Wobig Bio 175H Hopkins Marine Station June 15,1989 ABSTRACT The neuromuscular junction of Pacific Sanddabs was studied. Neither fast nor slow muscle fibers in the fin muscles of sanddabs produce action potentials. Muscle activation by indirect stimulation could be inhibited with TTI D-tubocurarine (curare), tetrodotoxin (11X), and GVIA omega-conotoxin. This indicated some similarities in the neuromuscular system of sanddabs compared to other vertebrates. Motor nerve and synaptic transmission are sensitive to sodium and calcium channel blocks, and acetylcholine is likely to be the neurotransmitter, as suggested by the curare sensitivity of neuromuscular transmission. However, direct stimulation of muscle fibers could be continued after the neuromuscular transmission had been blocked. Although they do not produce action potentials, the muscles do produce end plate potentials. A single end plate potential has proved adequate to produce a muscle twitch. INTRODUCTION Pacific Sandabs are interesting animals because they have fast twitch muscle fibers that do not generate action potentials, presumably, because they lack voltage-sensitive sodium channels (Gilly and Aladjem, 1987). This is a rare occurence in vertebrate skeletal muscle systems. Sanddabs also possess tonic fibers which hold contractions for extended periods of time (i.e. minutes). These fibers do not produce action potentials either. Other vertebrates use tonic fibers to control a wide range of specialized behaviors Instead of using action potentials, the tonic system is activated through several axons contacting the muscle fiber at multiple sites along its length. This produces localized depolarizations and enables graded force development. The sanddab also has intermediate fibers which contract at rates that are neither fast nor tonic (less than 1 second). The sanddab must rely on multiple innervation and summation of end plate potentials in order to control contractions of its fin muscle; this is accomplished through depolarizations along the entire course of its fibers. This makes the sanddab an intriguing experimental system. It would be interesting to understand more about this system and its relations to systems in other animals. Neuromuscular transmission has been extensively studied in amphibia and higher vertebrates, but work on fish is limited. At the end plate a motor axon makes contact with a muscle fiber, and a synapse exists between the two cellular elements. When an action potential in the axon invades the synaptic ter minals, neurotransmitter (acetylcholine) is secreted and quickly binds to receptors in the muscle cell. These receptors are cation-selective channels which open upon binding to ACh. Fatt and Katz (1951) first recorded end plate potentials (e.p.p.) in frog muscle by recording intracellularly from muscle fibers during indirect stimulation of the motor nerve. They found that one could reduce the size of the e.p.p. by adding curare to the preparation. This drug binds to the postsynaptic ACh receptor molecules. It was later established by Del Castillo and Katz (1954) that the e.p.p. was produced by a nuber of small quantal components. Further work was done by Martin in 1955 in which he described end plate potentials as the summation of many miniature potentials. It is now understood that these miniature e.p.p. are all-or-none. and therefore, correspond to packets of ACh being released either spontaneously or in response to nerve stimulation. In a frog muscle fiber, a single e.p.p. due to one action potential in the motor axon causes summation of many miniature e.p.p. and generally leads to the production of an all-or-none action potential in the muscle cell. However, sanddabs are able to produce muscle excitation without producing action potentials. Since sanddabs do not appear to normally have sodium channels in their muscle fibers, it would be interesting to see if they have the capabilities to express them under appropriate conditions. Denervation of skeletal muscles is well known to produce this effect (Keynes and Aidley, 1981). A severed nerve will cause the muscle it supplies to dramatically change its physiological properties. After a few days, the muscle will develop more sodium channels and ACh receptors. Moreover, the whole muscle cell surface becomes sensitive to ACh, which nor mally is effective just at the end plates. It would be interesting to see if sanddabs possess the capablities to produce these mor phological change: The present study focuses on three aspects of neuromuscular transmission. The first is to look at the the summation of e.p.p. in order to produce muscle fiber contraction. The second is to gain a better understanding of the phar macological aspects of the neuromuscular junction, and finally, to look at the question of functional changes due to denervation. MATERIALS AND METHODS Pacific Sand Dabs, Ctharicthys sordidus were caught with a hook and line at Monterey Bay. California. They were maintained in a flow-through sea water tank. They were killed with decapitation so that drugs or higher motor control centers would not influence results. Dissections were carried out to produce viable muscle and nerve preparations (see below and Figure 1). The preparations were maintained in Ringer solution and placed on a cooling plate (see below). They were secured with pins to the Sylgard (Dow Corning. Midland, Michigan) floor of an experimental chamber. The fin spines were tied with thread and connected to a force transducer which was supported by a micromanipulator. The main nerves and trunks running along the proximal ribs branch out less dramatically in more anterior regions of the fish. Results in this report are all obtained from the posterior end of the fish. Smaller fish were préferred because they possess significantly less connective tissue in the experimental region. FORCE MEASUREMENTS Contractile responses to indirect stimulation (via nerve) were studied. Motor nerves were stimulated with a polished glass suction electrode (approximately 250 um in diameter) which was supported by a micromanipulator. The electrode was gently placed against the nerve and a slight suction was applied. The same technique was also used to directly stimulate muscle fibers. Spurious stimulation of tissue away from the pipette tip by the indifferent wire outside of the glass electrode was avoided by placing the wire as far away from the muscle as possible, but still in the experimental chamber. Force was then measured with a transducer made from Pixie semiconductor elememts (Endevco, Pasedena, California; Gilly and Hui. 1980). Recordings were made on both a chart recorder and an oscilloscope. INTRACELLULAR RECORDINC In order to record voltage differences from muscle fibers, electrodes were filled with 3M KCI. Electrodes were 10-20 x 106 Ohms in resistance. They were carefully manipulated until they were near a muscle fiber, and then. upon a swift tap to the micromanipulator, they punctured the cell. Recordings were made from fin-muscle fibers only on the fin spine side of the end plate band (see figure 1). Contractile responses appeared to be strong and fast in this region. SOLUTIONS Composition of the ringer used was (mmol 1-D): 200 Nacl, 3 KCI, 4MgCI2, and 4 CaCl. This was buffered to pH 7.2 with SmM Tris. Ringer was kept refrigerated and the experimental chamber was maintained at approximately 20° C. Chemicals were obtained from Sigma Chemical (St. Louis, Missouri). Omega-conotoxin and the paralytic peptide were obtained from Dr. Baldomero Olivera, University of Utah. RESULTS ANATOMY AND DEVELOPMENT OF A NEUROMUSCULAR JUNCTION PREPARATION The pattern of the fin spine musculature of the sanddab is shown for the left side in Figure 1. The pattern is repeated for every dorsal and anal fin spine along the length of the animal on both left and right sides. The figure demonstrates that each fin spine is connected to two muscles on the left side; two identical muscles are present on the right as well. These muscles are connected to bone shafts with a triangular head. These bone shafts are connected to the proximal rib and to the fin spines with extensive connective tissue. Also connected to each fin spine is a skin muscle (not illustrated) which is connected in the middle of the fin spine and at a junction approximately halfway up the bone shafts. The skin muscles run between the two bone shafts and on top of the other two muscles. They were cut away so that the motor nerves could be exposed. Skin muscles also exist on the both sides of the fish, and the total is thus six muscles connected to each fin spine. In carrying out physiological studies of fin muscle, the skin muscles were dissected off of both sides, and the preparations were carefully cleaned from fin spine to proximal rib. Therefore, the preparations appeared as in Figure 1. except that the antagonistic muscles were also usually connected on the other side. Although there appeared to be no differences between the two sides of the fish, in order to maximize consistency of results, all preparations were taken ventrally from the right side of the fish. Previous work had employed left dorsal muscles. Motor nerves are indicated in Fig. I as running along the course of a proximal rib and between two triangular-headed bone shafts. Approximately midway down the length of the bone shafts, the nerve branches and turns to run laterally across the muscle surface. Stimulation applied to the nerve in this area was effective, and it appears that endplates lie in this middle portion of the muscles (see below). A large nerve présumably sensory in function, branches off the main nerve as shown in Fig. I, and runs between the skin muscles and courses down the length of the lin spine through its hollow interior. The nerves run in parallel down every spine on both left and right sides of the fish, but are physically seperated from one another, even the spine itself, One difficulty in producing good preparations is the large amount of connective tissue in the area used for stimulation (end plate band). There is also extensive connective tissue which forms part of the proximal termination of the skin muscles in this region. It was difficult to obtain a clean preparation without damaging the muscle fibers or nerves in this area; however, this procedure was necessary since a cleaner preparation was more readily excitable and was also suitable for intracellualr recording. It was also important to visually observe the muscle contractions and have an unobstructed contact between the nerve and the electrode. Several other difficulties were experienced in trying to produce viable preparations. Results are probably dependent on the fiber type being stimulated. Contractions were not reliably produced in the same area of the muscle fiber every time. Differences might be representitive of stimulating different fibers. Since the sanddab has three different muscle fiber types, it is important to stimulate the same fiber groups every time in order to control for variation, especially when measuring force from the whole muscle. Different regions of contractions could be representitive of different muscle fibers. The nerves also appeared to be extremely fragile, so that sucking a cut end of a nerve up into the electrode appeared to damage its conductive properties. This was allieviated by gently placing the electrode on the nerve (en passant) and applying a gentle suction. This technique is effective in producing indirect stimulation, but care must be exerted to avoid direct muscle stimulation in the immediate area of the electrode tip. Because of the difficulties associated with the fin muscle, an alternative preparation was sought. Extraocular muscle was found to be an excellent preparation, and it could be isolated with intact innervation. The preparation consisted of the superior oblique extraocular muscle and its intact innervation. The nerve could be sucked up into the electrode without apparent damage, and produce reliable contractions with repetitive stimulation. CHARACTERISTICS OF INDIRECT STIMULATION Indirect stimulation of fin muscle was carried out using single shocks, and the resulting contractile force was measured. We routinely found that the resting length of the muscle greatly influences contractile strength (see Figure 2). Over a small change in length (slack length- 1 cm), the muscle displays a large change in contractile strength. Force increases smoothly up to a 10x stretch beyond slack length, where the maximal contractile force is reached. A slight additional stretch (0.2 mm) causes a sharp drop in contractile force. All experiments were carried out near the maximum of the length-force curve which was determined for each preparation. When a nerve is excited, its axons fire in an all or none manner. As shown in figure 3a, extraocular muscle is indirectly excited to a maximal contractile strength over a very small range of applied voltages. The region of gradation in force probably reflects increasing excitation of more motor axons in the nerve over this range of stimulating voltages. This characteristic of a gradable response of the whole muscle is expected for indirect stimulation. Apparently, all of the motor axons in the oculomotoi nerve, which has been sucked up onto the stimulating electrode, can be uniformly excited at a shock of less than 1 volt. The fin spine muscles showed a much more gradable force response with shock strength (Fig. 3b), at least using single shock stimuli. A major difference between the two preparations is that the fin spine motor nerve has an extremely large amount of connective tissue around the stimulating region. Because the en passant configuration of nerve stimulation was used with fin spine preparations, it may have been impossible to equally excite the whole nerve, except with very strong shocks. Since the nerve was not cut and sucked up into the electrode, there is also the possibility of directly exciting muscle fibers with large shocks. The gradation of force with indirect stimulation of fin spine muscle is largely different from that seen with direct muscle stimulation. Direct stimulation shows that contractile strength is even more graded with voltage applied than the results shown in Fig. 3b. (Gilly and Aladjem. 1987.) Force elicited by direct muscle activation is also highly responsive to changes in the stimulus duration over a range that does not affect indirect stimulation (e.g. changing from 0.5 to 5 ms shocks. Data not illustrated). Direct stimulation is also not very responsive (if at all) to reverse polarity of the stimulus (i.e. inside of pipette positive), presumably because sanddab muscle does not have sodium channels. Nerves can be stimulated by this anode¬ break method, because the hyperpolarizing stimulus will induce Na channels to activate upon termination of the hyperpolarization and return of the membrane potential to its normal value. A final difference is that direct stimulation always produces visually detectable contractions im mediately in the region stimulated, whereas indirect stimulation often produces contractions far away from the stimulating site, e.g. from several groups of muscles. PRODUCTION OF AN END PLATE POTENTIAL AT THE NEUROMUSCULAR JUNCTION Intracellular recording revealed that an end plate potential (e.p.p.) is produced in response to indirect stimulation. This is shown in Figure 4a. The rise after the e.p.p. is due to the muscle twitch dislodging the microelectrode. As the electrode is moved closer to the end plate band described above, (in another muscle fiber), the e.p.p. becomes larger and faster and occurs at a shorter delay (Fig. 4b). Hence, the electrode has been moved closer to the endplate. Although the e.p.p. appears to be all or none, this arises from the all or none firing of the motor axon and not from the excitability of the muscle cell. Fig. 4c shows that the same e.p.p. is produced with shocks of 2.1. 2.2, and 2.3 V. However, a shock of 2.0 V. produced no e.p.p. at all. Nothing resembling an action potential is ever observed. The electrical signs of muscle activation in sanddab muscle differ markedly from more 'typical' vertebrate skeletal muscle. As a comparison, the action potential recorded intracellularly from tadpole myotome muscle is shown in Fig. 4d. Figure 5 shows some additional properties of the e.p.p. in fish. Even with repetitive stimulation, no action potential is formed. The e.p.p. depicted in Figure 4 were all produced in response to single shocks, which also produced brisk muscle twitches. This result differs from those in previously published work (Gilly and Aladjem, 1987). According to this paper, trains of stimuli were necessary to indirectly excite muscle. Also, figure 5 does not show the same pattern of summation of junction potentials to a plateau depolarization that was documented in the aforementioned paper (see Discussion). PHARMACOLOGY OE NEUROMUSCULAR TRANSMISSION Once it was clear that the contraction was accomplished through indirect stimulation, it was possible to attempt some phar macological experiments Figure 6 documents the block of the e.p.p. with 100 uM curare. This was accomplished in the course of three minutes. Since curare blocks ACh receptors, it is reasonable to assume that sanddabs use ACh as their neurotransmitter for motor neurons. This is supported by the fact that a strong contracture is produced when carbochol, an ACh analog, is added to the preparation. TTX is a toxin that blocks Na channels. Figure 7a shows the result of adding TTX to a muscle being stimulated indirectly with single shocks. Force is completely eliminated in 8 minutes (Fig. 7b). Presumably, conduction of the nerve’s Na channels are blocked so that electrical excitation of the nerve is stopped and no ACh can be released at the neuromuscular junction. TTX was effective at concentrations of 200 nM, although its time course ranged over à number of minutes. Because TTX blocks Na channels, it blocks direct muscle excitation in muscles with action potentials, e.g. tadpoles (not illustrated). However, sanddabs apparently do not have muscle Na channels and therefore, direct stimulation is still possible after indirect stimulation had been completely blocked by TTX. These phar macological results reinforce what is already known about sanddab muscle. The extraocular muscle preparation was used for further phar macological studies. These results are summarized in Figure 8. Stimulation was not done with à single shock, but rather with tetani of 200 Hz. for 0.5 sec. The shocks were at 0.5 V. and lasted for 0.4 msec. It appears as if 0.6 mM of the paralytic peptide (an uncharacterized conotoxin: obtained through personal communication with B. Olivera) had no effect. Either higher dosages or more time would be needed to document any effect this drug had. The next drug applied was 0.3 mM of GVIA omega-conotoxin. This drug blocks presynaptic calcium channels in frog neuromuscular junctions (Olivera et al. 1985), and therefore, blocks the release of ACh from the presynaptic terminal. This appeared to have an inhibitory effect in the present experiment. Finally, as controls, curare and TTX were used to induce comlpete loss of response, DENERYATION As à final experiment, it was attempted to denervate sanddab fin muscle A cut was made proximaf to the skin muscles. The incision was made laterally along one half of the fish on the right side. The muscle was pulled back to expose the main motor nerves diagrammed in Fig. 1. Several motor nerves on the right ventral side were cut proximal to the end plate band area nor mally used for stimulation. The fish were then allowed to recover for at least four days. After four days, one fish was dissected. Due to the operation, the damaged skin muscles appeared extremely white, tore easily and had lost their elasticity. Some of the experimental muscles had begun to take on this appearance also, but others which had clearly been denervated still appeared healthy and responded to direct muscle stimulation with repetitve shocks (not illustrated). Muscles on the left side of the fish maintained their normal appearance. It was anticipated that the denervated muscles would develop Na channels. This, along with increased sensitivity to ACh, has been documented in other animals. Unfortunately, any increase in muscle excitability was not detectable in this preliminary experiment. Perhaps more time is needed for these changes to develop or finer recording techniques may be necessary (e.g. intracellular recording). DISCUSSION Despite its relatively simple and highly repetitive anatomical layout, the sanddab is a rather difficult animal to study. The muscles and nerves appear to be rather fragile, and one must take great care during dissections so as not to damage the preparation. The muscles might also be easily damaged by stretching since there clearly is a maximal stretch which can be applied to the fin muscles. The sanddab's force vs. length curve (Fig. 2) is extremely more dramatic than other animals. Most animals show a slow loss of force after a maximal stretch due to the loss of overlap between the actin and myosin filaments (Keynes and Aidley. 1981). The sharp loss of force in the sanddab is a mystery. Perhaps multiply innervated systems are much more responsive to stretch and endplates are more easily disrupted. It appears that work might be easier to accomplish on the extraocular muscles. These muscles can be easily dissected out with a large intact nerve attached. In this system, stimulation could be accomplished with a cut nerve sucked up into the electrode. However. even in this preparation, some of the muscles did not appear responsive It appears as if there are some standard occurences at the neuromuscular junction of the sanddab. Indirect stimulation can be blocked to some extent by curare, 11A, and omega conatoxin. Presumably this is because sanddabs use ACh as their neurotransmitter, and their nerves have Na and Ca channels which can be blocked by appropriate toxins. However, some of the drugs appeared to take a long time to show an effect. Perhaps there is a problem with chemical diffusion in this fish, so that the endplates are extremely well insulated and perfect conditions are necessary to chemically alter behavior. Repeated studies need to be accomplished in order to understand the time course of the drugs used. With repeated experiments, perhaps a pattern would develop which would lead to greater insight. It would also be advantageous if the drugs were more responsive to washing out, so that work could be reliably reproduced on the same muscle. It is not clear why the preparations produced in these experiments were responsive to a single shock, while previous studies have needed to use repetitive stimuli to shock indirectly (Gilly and Aladjem. 1987). Perhaps this is caused by differences in the innervation patterns of fast twitch versus intermediate twitch muscle fibers. It would be interesting to see if there were significant differences between the neuromuscular junctions of the different fiber types. The previous study accomplished work on the left side of the fish, and there is the possibility that differences exist between the left and right sides of the sanddab. Differences were also observed in the summation of end plate potentials for probably the same reasons. Previous work showed a large summation of e.p.p. and then a plateau depolarization. Figure 5 shows a lack of dramatic facilitation, so that there is not a considerable amount of summation to a plateau. The first e.p.p. is also much larger in comparison to the second than those seen in the earlier study. This could explain why a single shock was a sufficient stimulus to produce muscular contraction in our present experiments. This paper is the first de monstration of the physiological effect of GVIA omega-conotoxin in fish. This is rather ironic since the snails are fish killing animals. Nevertheless, the toxin appears to block the neuromuscular junction of fish as expected. The paralytic peptide appeared to have no effect on peripheral nerve transmission or end plate transmisssion. However, it has been shown to have a paralytic effect when injected into an intact fish (Kuo, E., 1989). This implies that its effects are restricted to higher motor centers. The sanddab appears to be an intriguing experimental system. Two large sensory nerves can be seen supplying each fin spine. Presumably, these nerves are sensory in function. They might be linked to proprioception or substrate identification. The innervation of the skin muscle may also help to shed some light on the interesting muscles of the sanddab. The sanddab appears to have developed a complex system in order to finely orchestrate its movement. A more detailed characterization of this neuromuscular system should be interesting. ACKNOW LEDGEMENTS My sincere thanks and gratitude go to W. F. Gilly for his guidance and support throughout this project. REFERENCES Del Castillo,J. and Katz, B. (1954) Quantal components of the end plate potential. J Physiol. 124, 560-573. Fatt, P. and Katz. B. (1951) An analysis of the end plate potential recorded with an intracellular electrode. 1 Physiol 115, 320-369. Gilly, W. F. and Aladjem, E. (1987) Physiological properties of three muscle fibre types controlling dorsal fin movements in a flatfish. 1 Musc. Res Cell Mot. 8, 407-417. Gilly, W.F. and Hui, C.S. (1980) Mechanical activation in slow and twitch skeletal muscle fibers. Dependence on voltage and external calcium. J. Physiol 301, 137-156. Keynes and Aidley. Nerve and Muscle. Cambridge Univ. Press, Cambridge pg. 151 (1981) Kuo, E. (1989) Mauthner cell recording from Senorita, Oryjulis californica Biology 175H, Hopkins Marine Station of Stanford University, Pacific Grove, California (unpublished report). Martin A. R. (1955) A further study of the statistical components of the end plate potential. J Physiol 130, 114-122. Olivera, B.M. et al (1985) Peptide neurotoxins from fish-hunting cone snails. Science 230, 1338-1343. FIGURE LEGENDS FIG. 1. Schematic of fin spine preparation. This depicts a representation of the skeletal components, muscle, and nerves which control the fin spine. See text for details. FIG. 2. Relationship between force generated and muscle stretch. Single indirect shocks of 10 V. were applied for a duration of 0.4 msec each. FIG. 3. The relationship between force generated and voltage applied. (a) Indirect shocks lasting 0.4 ms were delivered at frequencies of 200 Hz. for 0.5 sec. to the extraocular preparation. (b) Single shocks of 0.4 ms were delivered to the fin spine muscles. FIG.Aa. Properties of end plate potentials. An end plate potential is produced by a 2.2 V. shock of 0.4 msec. (b). A picture of the e.p.p. produced as the electrode is moved closer to the endplate band. (c) This shows that the same end plate is produced at shocks of 2.1, 2.2, and 2.4 V. However, there is no production of an e.p.p. at 2.0 V (d) An action potential from a frog muscle produced by electrical stimulation. FIG. 5. Temporal summation of end plate potentials. End plate potentials are produced at the indicated frequencies. Indirect shocks consisted of 2.0 V. and were 0.4 msec. in duration. FIG. 6. Block of the end plate potential with D-tubocurarine. This represents the decay of the e.p.p. over 3 minutes. Block accomplished with 100 uM of curare. Shocks were 2.2 V. and lasted for 0.4 msec. FIG. 7. (a) Block of force generated with 100 nM TTX. Indirectly applied shocks consisted of 40 V. and lasted 0.4 msec. (b) Time course of 7a. FIG. 8 Phar macology of the extraocular preparation. There is a block of force due to 0.3 mM omega conatoxin, 200 uM curare, and 200 nM TTX. However, there is no apparent effect due to the application of 0.6 mM of paralytic peptide. Shocks were delivered indirectly at 0.6 V. for durations of 0.4 msec. They were delivered at tetani of 200 Hz. for 0.5 sec. FIGURE 1 MOTOR NERVE — MUSCLE 4 BONE S V PROXIMAL RIB END PLATE BAND — SENSORY NERVE T GEN SPNE 3 FORCE (mV) FIGURE 3 BELATIONSHIP BETWEEN FORCE GENERATED AND VOLTAGE APPLIED EXTRAOCULAR 40 20 FIN SPINE 40 20 — - — 40 20 VOLTAGE (V) 60 FIGURE 4 O m -60 mV O mV -60 mV END PLATE POTENTIALS 4 ms END PLATE POTENTIAL VS ACTION POTENTIAL 4 ms FIGURE 5 200Hz O I 50 Hz 40 mV REPETITIVE STIMULATION 4 ms FIGURE 6 O mV -60 mV LOSS OF EPP BY CURARE 4 ms FORCE (mV) .. . . FORCE (mV)