ABSTRACT MORPHOLOGICAL AND PHYSIOLOGICAL STUDIES OF THE WALKING LEG MUSCLES OF THE HORSESHOE CRAB, LIMULUS POLYPHEMUS (LINNAEUS) BY Charles Russell Fourtner The gross morphology and innervation of each of the muscles of the walking legs of Limulus polyphemus (Linnaeus) are described. Each muscle is composed of at least two heads and each muscle, with the exception of the extensor of the dactylopodite, appears to be innervated solely by either the internal pedal nerve or the external pedal nerve. The sarcomere lengths of single muscle fibers from all of the muscles of the walking legs were measured. The mean sarcomere lengths for muscle fibers in each of the muscles ranged from 7.6 to 8.3u; and the total range for all muscle fibers was 5.9 to 10.5u. No correlation was found between the muscle fiber diameters and their sarcomere lengths. The ultrastructure of the muscle fibers of several of the walking leg muscles was also examined and compared favor- ably with the ultrastructure of other arthrOpod skeletal muscle fibers: thin to thick ratio of 5 to 6:1; diads located at the A—I junction; multinucleate; mitochondria and Charles Russell Fourtner nuclei located primarily near the periphery of the fiber. The neuromuscular synaptic area in limulus was located in the arm-like extensions of the sarc0plasm. These extensions may consist of sarcoplasm from one or a few muscle fibers. Glia surrounds the nerve processes except at the neuromuscular synapse. In the presynaptic terminal there were two types of synaptic vesicles, smaller clear vesicles and larger dense— core vesicles. In addition a very extensive sarcoplasmic complex consisting of many tubules and vesicles was found in the sarcoplasm of the synaptic area. Physiological investigations were performed on each of the three distal groups of the mero—carpopodite flexor. The mechanical activity, the external electrical activity and the intracellular activity from single muscle fibers of each group were recorded and analyzed. All three muscle groups displayed similar mechanical responses: rates of rise and decay of the twitch (single stimulus) response; summation to multiple stimulation with a mean tetanus to twitch ratio of 90:1; marked facilitation to paired pulse stimulation. At pulse pair intervals of 10 and 25 msec the mechanical response evoked by the pulse pair may be as great as six times that evoked by a single stimulus. In all three groups at least eleven different stimulus- strength-dependent thresholds could be evoked by indirect stimulation. No inhibitory input was found to these muscle groups. Charles Russell Fourtner All three muscle groups display similar external electri- cal responses to single stimuli and multiple stimulation. At least eight different stimulus-strength-dependent thres- holds could be recorded from the muscle surface with a suction electrode (tip diameter 200u). The intracellular activity evoked by a single supra- threshold stimulus and recorded from an individual muscle fiber was either an excitatory postsynaptic potential (55 percent of the fibers) or a graded-spike (45 percent of the fibers). However, paired pulse stimulation evoked graded— spikes in those muscle fibers which displayed EPSPs to a single stimulus. From two to six stimulus-strength—dependent increases in intracellular activity could be recorded, thus implying that the muscle fibers were innervated by at least from two to six different axons. No IPSPs were ever recorded. From this study it was concluded that the muscle fibers composing the walking leg muscles of limulus are similar as to their structure and their physiological responses. It was also shown that the walking leg muscles receive a greater number of excitatory axons than are found innervating the skeletal muscles of other arthropods. The phylogenetic relationships between limulus and other arthropods are discussed concerning the evolution of a physiological structure from a less specialized to a more specialized system. l. ' ' v - _-~‘.“__ -_. __. ‘. _ .. ‘1 | . ' " ‘ ' .:.nn-r_- MORPHOLOGICAL AND PHYSIOLOGICAL STUDIES OF THE WALKING LEG MUSCLES OF THE HORSESHOE CRAB, LIMULUS POLYPHEMUS (LINNAEUS) BY Charles Russell Fourtner A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Zoology 1971 ‘éflj/J/CQI To Mo'e ACKNOWLEDGMENTS I wish to express my gratitude and appreciation to Dr. Ralph A. Pax for his guidance and valuable criticism throughout the course of this project. I also wish to thank Drs. R. N. Band, G. L. Gebber and G. I. Hatton for their advice and their critical read- ing of this manuscript. Special thanks are reserved for my colleagues in this laboratory, Charles Drewes and Vincent Palese for allowing me the privilege of bouncing my problems off their brains. The electron micrographs presented in this paper were taken at the University of Toronto, Department of Zoology. Dr. Robert G. Sherman must be credited for producing these fine micrographs. Finally, I wish to thank Annie, to whom this thesis is dedicated, for her patience and her typing skills, both of which abbreviated this study by at least twenty years. iii TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . . . . . LIST OF FIGURES. . . . . . . . . . . . . . INTRODUCTION . . . . . . . . . . . . . . OBJECTIVES . . . . . . . . . . . . . . . . MATERIALS AND METHODS. . . . . . . . . . . Source and Maintenance of Animals . . Saline. . . . . . . . . . . . . . . . Histological Techniques for Sarcomere Ultrastructure Studies. . . . . . . . Nerve—Muscle Preparation. . . . . . . Mechanical Recordings . . . . . . . . Studies External Recording of Muscle Electrical Activity . . . . . . . . . . . . Intracellular Recordings. . . . . . . RESULTS. . . . . . . . . . . . . . . . . . Morphology. . . . . . . . . . . . . . Anatomy. . . . . . . . . . . . . Histological Investigation . . . Ultrastructure . . . . . . . . . Physiology. . . . . . . . . . . . . . Mechanical Recordings. . . . . Thresholds of Mechanical Response . The Twitch Response . . . . Response to Repetitive Stimulation. Facilitation of the Twitch Response External Electrical Recordings . Response to a Single Stimulus . . . Response to Repetitive Stimulation. Response to Paired Pulses . Intracellular Recordings . . . . iv Page vi vii \l \Dmflfl 10 13 14 16 16 16 23 28 46 46 46 49 51 58 64 65 75 80 TABLE OF CONTENTS--C0ntinued Page Resting Potentials. . . . . . . . . . 81 Spontaneous Activity. . . . . . . . . 81 Excitatory Postsynaptic Potentials. . 84 Graded—Spikes . . . . . . . . . . . . 91 Paired Pulse Stimulation. . . . . . . 96 DISCUSSION . . . . . . . . . . . . . . . . . . . . . 104 Muscle Morphology . . . . . . . . . . . 104 Synaptic Area Morphology. . . . . . . . . . . 107 Innervation Patterns. . . . . . . . . . . . . 109 Physiological Responses . . . . . . . . . . lll Comparisons to Other Chelicerates . . . . . . 115 Phylogenetic Considerations . . . . . . . . 116 SUMMARY. . . . . . . . . . . . . . . . . . . . . . . 118 LITERATURE CITED . . . . . . . . . . . . . . . . . . 119 TABLE II. III. IV. VI. LIST OF TABLES Page Sarcomere lengths. . . . . . . . . . . . . . . 27 Mechanical responses of muscle groups of mero- carpopodite flexor . . . . . . . . . . . . . . 50 Characteristics of the external electrical activity . . . . . . . . . . . . . . . . . . . . 69 Innervation of single muscle fibers. . . . . . 85 Characteristics of the excitatory postsynaptic potential. . . . . . . . . . . . . . . . . . . 90 Characteristics of the graded-spike. . . . . . 95 vi LIST OF FIGURES FIGURE 1. 11. 12. 13. 14. 15. Drawing of the fourth walking leg of limulus and its corresponding muscles. . . . . . . . . Photomicrograph of a transverse section through limulus skeletal muscle. . . . . . . . Longitudinal section through a muscle fiber. Transverse section showing a pair of diads . . Transverse section showing a number of muscle fibers and a synaptic area . . . . . . . . . . Transverse section through part of the synap- tic area . . . . . . . . . . . . . . . . . . . Transverse section showing two types of synaptic vesicles. . . . . . . . . . . . . . . Transverse section showing two neuromuscular synapses O O O O I O O O O O O O O O O O O O 0 Transverse section of the postsynaptic complex Recording of mechanical activity showing threshold dependent increases in tension . . . Recording of mechanical activity resulting from different stimulus frequencies. . . . . . Percent of tetanic tension is plotted against the stimulus frequency . . . . . . . . . . . . Response to paired pulse stimulation . . . . . The ratio between the tension developed by a paired pulse and that developed by a single pulse is plotted against the interval between the pulse pairs. . . . . . . . . . . . . . . . External electrical activity evoked by a single stimulus. . . . . . . . . . . . . . . . vii Page 18 26 3O 32 32 36 39 41 48 54 57 61 63 67 LIST OF FIGURES--continued FIGURE 16. Electrical record showing six thresholds. . . 17. Effect of stimulus frequency on the external electrical activity . . . . . . . . . . . . . 18. The ratio of the fifteenth pulse (F pulse) of a stimulus train to the first response of the train is plotted against the frequency. . . . 19. External electrical response to paired pulses 20. Spontaneous intracellular activity. . . . . . 21. Miniature excitatory postsynaptic potentials. 22. Examples of EPSPs . . . . . . . . . . . . . . 23. Examples of graded-spikes . . . . . . . . . . 24. Graph in which the rate of rise of the EPSP is plotted against the amplitude of the EPSP. 25. Intracellular recordings of electrical activ- ity evoked by paired pulses . . . . . . . . . 26. Graph in which the ratio between the ampli- tude of the response to the second stimulus of a pair and the amplitude of the first response of a pair is plotted against the interval between the pulse pair . . . . . . . viii Page 67 73 77 79 83 83 88 88 93 98 102 INTRODUCTION Among the Arthropoda, the neuromuscular junction has been extensively studied in the appendages of only a few Mandibulates, more specifically in the decapod crustaceans (Atwood, 1967) and the orthOpteran insects (Hoyle, 1965a; Usherwood, 1967, 1969). Thorough studies on the Chelicerates are almost totally lacking. In the Mandibulates several types of muscle fibers may occur in one muscle. Atwood (1963; 1965; 1967; Atwood and Dorai Raj, 1964; Atwood, HOyle and Smythe, 1965; Atwood and Morin, 1970) has given several criteria by which these muscle fibers may be characterized: 1) sarcomere length, 2) amount of sarcoplasmic reticulum, 3) actin-myosin ratio, 4) fiber diameter, 5) membrane cable properties, i.e., space constant and time constant, 6) membrane capacitance, 7) mechanical and electrical facilitative capabilities, 8) time course of the junction potentials, and 9) synaptic terminals. The interpretation of these criteria allows for a large variation in the number of muscle fiber types of which a muscle may be composed. In the Mandibulates also at least three types of axons. fast, slow and inhibitory, innervate the appendage muscula- ture. Wiersma (1961) describes these axons as follows: stimulation of the fast axon produces rapid contractions of muscle and little or no facilitation of either tension development or junction potentials; stimulation of the slow axon produces a slow contraction and a large facilitation in both the tension developed and the junction potentials; stimulation of the inhibitory axon alone produces no visible change in tension development but does produce inhibitory postsynaptic potentials in the muscle fibers it innervates. In addition fast axons are thicker and have faster conduc- tion velocities than slow axons. Among the Chelicerates the only thorough study of a neuromuscular system has been conducted by Rathmayer and his colleagues on the tarantula, Dugesiella hentzi (Rathmayer, 1965; 1967; Zebe and Rathmayer, 1968). These workers have found fast and slow axons innervating the appendage muscula- ture but could find no evidence for inhibitory axons. They have also described the muscle fibers as being of a single type--both morphologically and neurophysiologically. Recently Shenman (personal communication) has found morpho— logically different muscle fibers in the skeletal muscles of the tarantula, Q, hentzi. In a single report on the closer muscle of the pedipalp claw-in the scorpion, Leiurus guinguestriatus, Gilai and Parnas (1970) found two types of motor axons, a fast and a slow, and no inhibitory axon. They reported that the muscle fibers appear to be of a single type. Both Rathmayer (1965) and Gilai and Parnas (1970) studied muscles which had no antagonistic pairs and found no inhibitory innervation. This apparently substantiates Hoyle's and Smyth's (1963) hypothesis that the inhibitory axons function as a counter-balance force between antagonis— tic pairs of muscles and would probably not be found inner- vating a muscle lacking an antagonistic mate. In Limulus polyphemus, the horseshoe crab, only two physiological studies (Hoyle, 1958; Parnas gt al., 1969) have been reported on neuromuscular phenomena in appendage muscles. Both of these studies were on the flexor (closer) of the dactylopodite (claw). Hoyle reported finding that at least two motor axons, a fast and a slow, innervate the closer. He also reported finding no inhibitory innervation. The findings of Parnas £5 31. (1968) differ greatly from those of Hoyle. They found that the closer muscle is innervated by at least six motor axons, which appear to be very similar in that each axon evokes similar types of junction poten— tials. They also found at least one inhibitory axon innervat- ing the flexor of the dactylopOdite. The structure of individual muscle fibers in the append- age musculature of limulus has not been investigated. However, a few studies have been published which deal with the tail musculature and the opisthosomal extensor and flexor muscles. Jordan (1917) did a histological investigation of the tail musculature and described the individual fibers as being 1amme1ar around the periphery and mottled toward the center; however, he gave no information concerning fiber diameter nor the sarcomere conformation, except to state that the muscle fibers he observed were striated. De Villafranca (1960) and de Villafranca and Philpott (1961), examined the opisthosomal extensor muscle using elec- 'tronmicrosc0pic techniques, however these workers were more «zoncerned with the actin-myosin interactions than with Ipossible structural differences among the individual muscle iiibers. However, they did state that the length of individual sarcomeres in resting muscles have an average length of 7.5u. .Nkare recently Ikemoto and Kawaguti (1968) investigated the tilstrastructure of the Opisthosomal flexor (retractor) muscle iri an oriental horseshoe crab, Tachypleus tridentatug. The muscle fibers in this animal have diameters of 20-50;; and sarcomere lengths of approximately 8.0u. OBJECTIVES The present knowledge of neuromuscular phenomena in limulus appendage muscle is extremely sketchy. The only two studies of physiological processes in limulus appendage muscle (Hoyle, 1958; Parnas §t_31., 1969) give contradictory results concerning the number and types of axons innervating a particular muscle fiber. In addition both studies give mainly a qualitative description of neuromuscular phenomena, such as facilitation, summation and the time course of synaptic events. Although some muscles have been viewed histologically, no complete study has been done to determine whether there are structural differences among various muscle fibers in the horseshoe crab. Only one thorough examination of the ultra— structure of the striated muscle of the horseshoe crabs has been published and that was on the Opisthosomal flexor of the oriental species, Tachypleus tridentatus (Ikemoto and Kawaguti, 1968). Furthermore, to the best of my knowledge, nothing has been published concerning the ultrastructure of neuromuscular synapses in horseshoe crab straited muscle. Needless to say, a further and more complete study is neces— sary to attain a better understanding of the morphological and physiological properties of limulus neuromuscular sys- tems. In this study I have redescribed the gross morphology and innervation of each of the muscles controlling movement in the distal segments of the limulus walking legs. I have described the histology and ultrastructure of the muscle fibers of various muscles. Finally, I have studied the physiological prOperties of one specific muscle, the distal flexor of the mero-carpopodite, of the walking leg of limulus. I have chosen this particular muscle for two reasons: first, it consists of three distinct muscle groups which can be studied separately to determine if there are physiological differences among the groups; second, since this muscle has :no antagonistic muscle (Ward, 1969), one has the Opportunity to test the Hoyle-Smyth hypothesis concerning inhibitory input to arthropod skeletal muscle. Through these investigations a better understanding of the mechanism of neuromuscular transmission and muscular responses in limulus is achieved. Also this information‘ allows one to speculate on the phylogenetic significance of neuromuscular systems in the arthrOpods. MATERIALS AND METHODS Source and Maintenance of Animals Horseshoe crabs, Limulus polyphemus (carapace width 18 to 23 cm), were obtained from the Gulf Specimen Co., Panacea, Florida. The animals were stored in an artificial sea water aquarium (Dayno Co., Model 703) and were main— tained at a temperature of 13 to 17°C. Saline The physiological saline used was a 10 percent limulus blood saline; nine parts Millecchia (Millecchia and Mauro, 1969) saline, 424 mM NaCl, 9 mM KCl, 20 mM CaClz, 20 m_M_ M9C12, 25 mpg M9504 to one part limulus blood. The limulus blood was obtained by inserting a hypodermic needle into the ventral sinus and applying negative pressure by means of a syringe. The blood was allowed to clot and the clot was removed. The blood was then filtered and stored at 0°C. Before use the saline was refiltered and the fil- trate was added to the Millecchia saline. The pH of this Saline was always between 7.0 and 7.2. Histological Techniques for Sarcomere Studies The leg segments containing the leg muscles to be investigated were pinned out on a piece of balsa wood. In all cases the legs were placed in such a way that the muscle to be studied would be held at its maximum _i_n_ git_u length. As much of the exoskeleton, as possible, of the segment in which the muscles were located was removed to facilitate penetration of the fixative. Two different procedures were used to fix and prepare the tissue for the sarcomere studies. The first technique was used exclusively on the distal part of the mero- carpopodite flexor. In this technique, the preparation was Placed in a high Mg++ saline (30 percent M9804 and 70 Percent NaCl) for three hours. This procedure greatly reduced contraction of the muscle fibers which occurred when the muscles were placed directly in the fixative. After treatment with the high Mg++ saline, the muscles were then fixed in Carnoy's solution, dehydrated in alcohol, Cleared in xylene and embedded in paraffin (Gray, 1964). The tissues were cut either longitudinally or transversely at 6p, mounted on slides and stained with Delafield' s hematoxylin and eosin. In the second procedure for preparing muscle fibers for sar<=<>mere studies, the muscles were fixed in a 4 percent glutaraidehyde solution for two hours and then washed and Stored in 70 percent alcohol. This procedure fixes the contractile apparatus of muscle fibers quite well. A small number of muscle fibers (1 to 5) were then teased from the main muscle bundle, placed on a glass slide in 70 percent alcohol and observed with a 430x microscope. Ultrastructure Studies The muscles to be examined were prepared E s_i__t_u as described above for determination of sarcomere length. The muscles were fixed in a 4 percent glutaraldehyde—buffer-sa1t solution: 8 m1 of 50 percent glutaraldehyde, 92 ml 0.1M sodium cacodylate buffer (pH 7.1 to 7.3) and, 29 NaCl and O -29 CaClz. Fixation lasted for 2 hours with the fixative being changed every 5 minutes for the first 15 minutes. The tissue was then washed for 30 minutes in the buffer-salts solution, which was changed frequently throughout the wash— ing period. Following washing, the tissue was post—fixed for one hour in a 1 percent osmium tetroxide—buffer solution. During this hour the muscle tissue turned black. After the Post-fixation period the tissue was again washed for 30 minutes in the buffer solution. Small pieces of muscle (3 mm in length and 1 mm in Width) were dissected free. Only tissue that was black was used. These pieces were placed in small glass test tubes and Prestained with uranyl acetate (saturated solution in 50 Percent alcohol) for two hours. Following the prestaining treatment the tissue was then dehydrated; twice in 50 percent 10 alcohol, 10 min each; 70 percent alcohol, 10 min; 95 percent alcohol, 10 min; twice in 100 percent alcohol, 10 min each. The tissue was then placed in a one to one mixture of 100 percent alcohol and Spurr Low-Viscosity Embedding Media (Polyscience Inc., Paul Valley Industrial Park, Warrington, Penna. 18976) for two hours at room temperature. The tis- sue was then placed in 100 percent Spurr and allowed to set at room temperature overnight. Finally, the tissue was placed in a plastic beaker or embedding boats containing 100 percent Spurr and placed in an oven at 50°C for 18 to 24 hours to harden. Tissues were then cut into thin sections on an LKB microtome and floated onto c0pper grids. The sections were stained with uranyl acetate and lead citrate (Reynolds, 1963) and observed with a Philips 200 electron microsc0pe (Department of Zoology, University of Toronto). Nerve-Muscle Preparation The third and fourth pairs of walking legs were used in these experiments. A leg) was removed from the body by severing the leg just distal to the coxopodite. A plug was then inserted in the coxopodite to stop excessive bleeding. The ventral articular membrane between the mero- carpopodite and the ischiopodite was carefully severed expos— ing both the internal pedal nerve (IPN) and the external pedal nerve (EPN) . Each nerve was ligated with a thread and 11 cut distal to the ligature. Two parts of the flexor of the propodite originate on the dorsal exoskeleton of the ischio- podite (Group III of propodite flexor) . These muscles were carefully cut as near their origin as possible. The pro— podite and the mero—carpopodite were removed. The prepara- tion now consisted of the basipodite and the ischiOpodite. Two cuts were made along the ventral lateral margins of the basipodite and the ventral exoskeleton was lifted and removed thus exposing the IPN and the EPN. The nerves were ligated near their proximal ends and then were carefully lifted and dissected free from the surrounding tissue. The basipodite was then renoved by cutting along the articular membrane joining the basipodite to the ischiopodite. The exoskeleton of the isChiopodite was cut along its entire dorsal midline. The preparation was then placed in a Paraffin tray and covered with saline. The exoskeleton was very carefully pinned back exposing the distal muscle of the mero-carpodite flexor" (see Figure l) . The remnants of the propodite flexors were very carefully dissected away from the Group I and Group II muscles. Also the muscles (flexors and extensors of the ischiopodite) in the proximal part of the ischiopodite were carefully removed. The prepa— ration now consisted of the exoskeleton of the ischioPOdite. the three muscle groups of the distal flexor of the mero- carpopodite, the IPN, the EPN and the apodeme. 12 The particular muscle group to be used was isolated. The IPN and EPN were carefully dissected free from the two muscle groups to be discarded. These muscles were then cut at their insertions along the apodeme, and removed from the preparation. The distal part of the apodeme was dissected free from the articular membrane, thereby leaving the ,apodeme attached to only the muscle group to be studied. 11 small hole was then drilled through the apodeme. The Exreparation now consisted of the muscle group to be investi— gated, the apodeme, the IPN, the EPN and a small part of the exoskeleton of the ischiOpodite from which the muscle group originated. The IPN and the EPN are surrounded by the leg blood vessel (Dumont gt _a_l., 1965). Therefore before any stimula- tzican was applied this blood vessel at times was removed. This was accomplished by carefully cutting the vessel just distal to the thread, then carefully cutting the vessel dis- tally toward the muscle and finally pulling the vessel away from the nerve. The vessel was left intact around the nerve approximately 1 cm proximal to the muscle. In most cases, however, the blood vessel was left intact. This provides good protection to keep the nerve from drying when being stimulated in air . Mechanical Recordings A small hook, with attached thread, was inserted into the hole which had been bored in the apodeme. The thread was 13 also attached to a mechanotransducer, either a Narco Bio- systems Inc. microdisplacement transducer (F-50) or a Grass FT .03 C transducer. The preparation was allowed to rest for thirty minutes before experiments were performed. The mechanical activity was either recorded on a ,Narco Biosystems Inc. Physiograph IV or displayed on a frektronic 502A oscilloscope and photographed with a Grass C-4 oscillograph camera. External Recording of Muscle Electrical Activity The electrical activity occurring at the external sur- face of the muscles was recorded using two different techniques. Either a suction electrode or a glass micro— electrode filled with NaCl was used. The suction electrode consisted of a thin Teflon tube connected at one end to a syringe. Approximately 3 cm from time: other end a small hole was cut in the tube, and a short length of silver wire was inserted through the small hole and pushed to within 2 mm of the end of the tube. The small hOle was then sealed with epoxy glue so that there was no 1eakage around the silver wire. The wire exposed in the tube was then chlorided. The inner diameter of the suction electrode varied from 150 to 2000.. Electrical activity rECOrded with the suction electrode was amplified using a Grass P—5 preamplifier and monitored on an oscilloscope. 14 A11 recordings were single-ended against a silver—silver chloride ground electrode placed in the bath. The second technique utilized smaller tipped glass electrodes (10 to 20p inner diameter). These electrodes were filled with a l.0_M_ solution of NaCl. The electrodes were pulled and filled in a manner similar to the technique used to fill the KCl microelectrodes (see section on intracellular recordings). The microelectrodes were then placed on a small silver—silver chloride wire (gauge 34) and connected to a Grass P-15 preamplifier; the signal was monitored on an oscilloscope. Stimulation of the nerves was electrical. The nerve to be stimulated was lifted in air and placed on a pair of platinum wire electrodes. The electrodes were connected to a Grass Stimulator Isolation Unit (SIU 478 A) of a. Grass S-4 stimulator. The stimulus was always in the form of a square wave. Intracellular Recordings ‘Intracellular activity was recorded via glass micro— electrodes. Kimax glass tubing (0.8 to 1.0 mm in diameter) was drawn out using a Narishi’ge horizontal microelectrode Puller. The electrode tip diameter was less than lu in Cliameter. The, electrodeswere held on a glass slide with a ru‘bberband, placed in a filter flask and submerged in a hot 3 L4 KCl solution. A controlled vacuum was applied which 15 induced a gentle boiling in the KCl solution for approximately 15 min. Then the tube from the vacuum source to the filter flask was closed with a clamp and the electrodes were left in the flask for an additional 15 min. The vacuum was then removed and the electrodes were removed from the filter flask and stored in a coplin jar filled with 3M KCl. The electrodes were usable for at least two weeks after they had been filled. Electrodes with resistance of 3M) to ZOMQ were used. The electrodes were placed on a silver-silver chloride wire (Gauge 34) which was clamped to a microelectrode probe by a pin plug and a microgator clip. The intracellular activity was recorded with either a WP-4 or a WP—4A electrom- eter (WP Instruments) and was monitored and filmed using the apparatus described previous ly . RESULTS Morphology Anatomy The gross anatomy, innervation and function of the rnusculature of the walking legs of limulus were first described by Lankester (1885) and later by Patten and Redenbaugh (1899) . These investigators gave a general over- view of the walking leg musculature; origins, insertions, functions and innervation. The gross morphology of the muscles was studied further by Vachon (1945) . However, these investigators considered most of the muscles as single morphological units, and gave no description of the different groups (or heads) of each of the muscles, nor did they use vital staining techniques to investigate the innervation cxf ‘these muscles. In the following section the morphology <>f"the individual muscles of the distal segments, from the ba810podite to the dactlepodite of the limulus walking legs Will be described (see Figure l) . The terminology used by Patten and Redenbaugh (1899) in their treatise on the gross morphology of limulus will be used throughout this descrip- tiOn, 16 17 Figure 1. Drawing of the fourth walking leg of limulus and its corresponding muscles. A is a dorsal View with the dorsal exoskeleton removed and the extensor muscles stripped away. B is a ventral view with the ventral exoskeleton removed and the flexor muscles stripped away. The surface towards the lettering in both A and B is the posterior surface. The dotted line in the mero- carpopoditeis the M-C groove. Dactylopodite, da; propodite, pr; mero-carpopodite, me; ischiopodite, is; flexor, fl; extensor, ex. The Roman numerals designate the different muscle groups for each functional muscle. III/W \ I is H \ me I IV BA \\.\ Tm ///////////// ‘- (((«((((((((((« a \\\\ \\\\x \\\\\\\\R \\ //////// —';, ‘ p / gy/ 19 The extensor of the dactylopodite is a small muscle composed of two distinct muscle groups. Group I originates along the dorsal surface of the propodite and Group II cxriginates on the anterior surface near the large condyles (If the propodite. Both groups insert on a single apodeme vflnich is attached to the dactylopodite. Both groups are iJrnervated by the external pedal nerve (EPN), the smaller (If the two leg nerves. A process of the internal pedal Imerve (IPN), the larger leg nerve also penetrates the more (distal muscle group (Group I). This has been described previously by Hayes and Barber (1967) . I have some physio— logical evidence (electrical stimulation) which favors the view that the IPN sends at least one excitatory axon to the extensor of the dactylopodite. The flexor of the dactylopodite (claw closer) is like- wise composed of two groups of muscle fibers. However, these groups are not as clearly defined as those of the extensor. Group I originates on the ventral surface of the propodite and Group II on the posterior surface of the pro— podite. Both groups insert on a single apodeme which is joined to the dactylopodite. This muscle is innervated solely by the IPN. The extensor muscle of the prOpodite is composed of two associated muscles. The two muscles are symmetrical, one being located in the anterior half of the mero-carpopodite and the other in the posterior half. Therefore, a description 20 (Df the anterior extensor will also hold for its symmetrical pair. The anterior extensor consists of two groups of nuiscles. Group I is a thin mass of tissue which originates along the dorsal anterior surface and dorsal surface of the merocarpopodite distal to the M—C groove, which is the area of fusion between two embryonic segments-~the mer0podite axud the carpopodite. Group II is a more dense muscle mass arud originates on the anterior surface of the mero-carpopo- dite proximal to the M-C groove. Both groups insert on a ILarge single apodeme Which is attached near the distal edge <3f the dorsal surface of the propodite. Methylene blue staining reveals that the extensor muscle receives processes from both the EPN and the IPN. Stimulation of the EPN causes a contraction of the extensor muscle; this is in contrast to the report of Pringle (1956). Stimulation of the IPN produces no measurable contraction of the extensors. However, this does not rule out the possibility of inhibitory fibers from the IPN innervating the extensor muscle. The flexor muscle of the pr0podite is likewise a paired symmetrical muscle with anterior and posterior muscle bundles. Each of these bundles consists of three distinct and large muscle groups. The most distal group (I) originates along the entire ventral lateral surface of the mero- carpopodite distal to the M—C groove. The second group (II) originates on the dorsal-lateral surface of the mero-carpopo- dite proximal to the M-C groove. The third group (III) 21 originates on the dorsal surface of the distal end of the ischipodite. The two more proximal groups (II and III) of either flexor pair insert via numerous long apodenes near the large condyles of the pr0podite. The most distal group (I) likewise inserts on the prOpodite via these apodemes; in addition however, this particular group also inserts all along the lateral surface of the propodite. This is of functional significance since the prOpodite of limulus is capable of promotor and remotor actions as well as flexion and extension. Patten and Redenbaugh (1899) claim that the promotor and remotor action of the propodite is due to a coordinated contraction of the extensors and flexors of the PrOpodite. Vachon (1945) stated that the propodite is cap- able of only flexion and extension. The innervation of the prOpodite flexor is via the IPN. TWO large nerves (one to each symmetrical pair) emerge from the IPN at the proximal margin of the mero-carpopodite. These nerves traverse the surface of the three muscle groups. NO Processes of the EPN are found innervating the flexor m“Scle. Patten and Redenbaugh (1899) apparently mistook the Grollp III flexor of the prOpodite for the extensor of the meIre—carpopodite. However Vachon (1945) and Ward (1969) have shown that there is no extensor muscle of the mero- car:Pepodite and have also demonstrated that extension of the meITO-carpopodite is a mechanical process. Ward has shown 22 that extension of the mero—carpopodite is the result of three parameters: 1) weight of distal segments, 2) contrac— tions of flexor muscles of ischiOpodite and basipodite, and 3) contact between the claw and some object. Flexion of the mero-carpOpodite is accomplished by a pair of associated muscles; the proximal one originates in the basipodite and the distal one originates in the ischio- podite. Both these muscles insert on a single long and very large apodeme which extends from the basiOpodite to the mero—carpopodite. As is seen in Figure l the distal muscle of the flexor of the mero-carpopodite is composed of three distinct groups (I, II, III) of muscle fibers. Group I originates on the anterior dorsal surface of the ischiopodite and Group II originates on the posterior dorsal surface of the ischiopodite. Both of these groups insert on the apodeme. Although these two groups appear Symmetrical, Group I is somewhat more massive than Group II. Group III originates on the posterior ventral surface of the ischiopodite and also inserts on the apodeme. The proximal flexor of the mero-carpopodite (Group IV) originates on the ventral surface of the basipodite and also inserts on the large apodeme. Innervation of these four muscle groups is accomplished solely by the IPN. The extensor of the ischiOpodite is an interesting muscle in that it originates in the segment on which it acts. This muscle is composed of three groups of muscles. 23 The second group originates on the ventral surface of the ischiopodite and the first and third groups are symmetrical pairs and originate on the anterior and posterior surfaces of the ischiopodite. All three muscle groups insert on the distal dorsal margin of the basipodite. These groups are innervated by the EPN. The flexor of the ischiOpodite is an extremely large muscle and consists of two groups. Group I originates on the anterior surface of the basipodite and inserts on the posterior ventral surface of the ischiopodite. Group II originates on both the anterior and posterior surfaces of the basipodite and inserts on the large condyle on the posterior surface of the ischiopodite. Innervation of both groups is accomplished by the IPN. Histological Investigation Since studies by Atwood and Dorai Raj (1964) and Dorai Raj (1964) have shown that crustacean muscle fibers can be categorized morphologically as well as functionally by the length of their individual sarcomeres, I have surveyed the muscle fibers in each of the muscles of the distal segments 0f the walking legs of limulus to see if such morphological categories are present in limulus. Muscle fibers from every muscle group of each of the muscles were examined. The muscles were fixed i_q situ at normal resting length. 24 Figure 2 shows a typical longitudinal section through the muscle fibers of a Group II muscle. These muscle fibers possess the usual A, I and Z bands one would expect to see in straited muscle. The sarcomere lengths (the distance between the beginning of one A band to the beginning of the adjacent A band) of individual muscle fibers were measured. The length of individual sarcomeres was determined by count- ing the number of sarcomeres per 100u and dividing lOOu by the number present. The results of these studies are given in Table I. The table is composed of information from 1723 muscle fibers. As is seen in the table, the total range of sarcomere lengths measured was 5.9 to 10.5u. However, the greatest range of sarcomere lengths measured for a single muscle group was 3.1u. Therefore, from the muscle fibers investigated, it is evident that the muscle fibers of the walking legs are very similar in regard to their sarcomere lengths. It has been shown in crustacean systems that a correla- tion exists between the muscle fiber diameters and the sarcomere lengths. Therefore, to determine if such a rela- tionship exists in limulus, a linear regression analysis was Performed on 35 muscle fibers. These fibers varied in diameter from 15 to 50p. and varied in sarcomere length from 6.7 to 10.0u. In this analysis it was found that the muscle fiber diameters and the sarcomere lengths are not linearly correlated; the correlation coefficient being only -0.08. 25 Figure 2. Photomicrograph of a transverse section through limulus skeletal muscle. 26 Figure 2 - lun~ any-uuu .N. 3.4.V...Av.v-ucmh n N lUd.P~E..H. 27 Ame oeH m.OH ou m.o 6.0 H m.m HHH 6cm H mmsouwv muHoomoHsumH roe mme m.OH 0» m.m m.o n H.m H>H psm HHH .HH .H mmsouwv muwpomomumolouoz Ame «Hm m.OH on m.o m.o a m.m AHHH can HH .H mmsoueo muapomoum rec «mm o.OH on m.o m.o h H.H HHH can H hunches evapomoamuoma QQHUQDE HONOHW Ame mmH m.m on m.o m.o H H.H .HHHH can HH .H mmsouwc ouHeomoHsomH rec mmm o.OH 0» H.o m.o h H.H AHH can H hunches muflpomoum Ame mHH m.m on m.o m.o H o.H IHH can H mmsouec opapomoaapomn mesonse Homsmuxm “modems: mo Honeszo 1 mmcmm Hmuoa Om H use: mumnHH Ho Hmpsuz Hayes 1 numcmq mumsouumm .nnumcwd oumeouumm .H manna 28 ultrastructure Longitudinal sections of a single muscle fiber of limulus skeletal muscle show that it consists of several bundles of myofilaments (range from 0.2 to 1.0u in width) separated by areas of sarcoplasm and sarcoplasmic reticulum (Figure 3). The bundles of myofilaments consist of a dense Z band (0.2 to 0.311 in length), a light I band (actin) and a dense Aband (myosin) which is 3.0 to 3.5u in length. In Figure 3 the muscle is partially contracted so that the I hand is much shorter than that seen in a resting length fiber; the sarcomeres in this particular section are between 5-0 and 6.0u. No H zones were found in these muscle fibers. H zones likewise were not found in the opisthosomal extensor (de \Iillafranca and Philpott, 1961) nor in the Opisthosomal flexor (Ikemoto and Kawaguti, 1968) . In longitudinal sections mitochondria are dispersed throughout the sarcoplasm but occur more frequently near the 2 bands. There are also several diads visible in Figure 3. The diads are located near the junction of the A and I bands of the myofilaments. Figure 4 is a high—power electron micrograph showing a transverse section of a pair of diads. A diad consists of a clear membrane bound area which is part of the transverse tubule system ("T system"), and a second membrane bound Strueture which contains an electron dense material. This Be(“311d structure is a part of the sarc0plasmic reticulum. 29 Figure 3. Longitudinal section through a muscle fiber. Diad, D; sarcoplasmic reticulum, SR; mitochondrion, M; I band, I; A band, A; Z line, Z. Calibration, 1.0u. X 11,000. 30 m... Figure 3 31 Figure 4. Transverse section showing a pair of diads. Transverse tubule system, T; sarcoplasmic reticulum, SR. The white circle surrounds one thick filament and its corresponding thin filaments. Calibration, 0.2u. X 67,600. Figure 5. Transverse section showing a number of muscle fibers and a synaptic area. Nucleus, N; blood cell, BC; Diads, D; mitochondria, M; synaptic area, SA. Calibra- tion, 2.0u. x 5,700. .3 2 en d a m.Sl its luscle , BC; bra- 11 Eillci 5 Figures 33 Tflne diads are approximately 0.3u in length and 7508 in width. Tflne space between the two components of the diads is between 90 and 1203. There are also several electron dense "bridges" vihich are always found in this space. These dense areas are <:alled septate junctions (Lowenstein, 1966). Septate junc- 1:ions have also been found in other arthropod diads (Hoyle and McNeill, 1968) and in vertebrate triads (Kelly, 1969) . Also present in Figure 4 are the thick (myosin) and ‘thezthin (actin) myofilaments which make up the contractile apparatus. The white circle in the figure is drawn around (a single thick filament and the surrounding thin filaments, ‘which are believed to interact with the thick filament. 9There are approximately 10 to 12 thin filaments surrounding each thick filament. Assuming that every thin filament is shared by two thick filaments, the thin to thick ratio is 5 or 6 to one (Toselli and Pepe, 1968). This ratio is that usually found in arthropod slow skeletal muscle systems (Sherman and Atwood, 1971). Figure 5 shows a cross sectional view of parts of six different muscle fibers, four large cyanocytes (see Fahrenbach, 1969 for description of limulus blood cells) and a small ovoid structure in the lower center of the figure, the synap— tic area. The muscle fibers are surrounded by a fibrous basement membrane and are multinucleate; the fiber in the upper right hand corner of Figure 5 has two distinct nuclei. The nuclei are generally located near the periphery of the fibers. 34 Mitochondria are dispersed throughout the muscle fibers but are found in the greatest numbers near the periphery of the muscle fibers. The sarcoplasmic membrane (sarcolemma) surrounding the muscle fibers sends long invaginations into the muscle fibers. These invaginations form the "T system" which runs throughout the muscle fibers (Peachy, 1970) . Several diads can also be seen in the areas where these invaginations occur. The synaptic area seen in Figure 5 is composed of a sarcoplasmic evagination from at least one muscle fiber, several nerve processes which may be derived from a single axon or from several axons, glial elements surrounding the nerve processes, and numerous nerve terminals. Figures 6, 7. 8, and 9 are enlargements of the synaptic areas. It is possible that sarcoplasmic evaginations from more than one muscle fiber may be present in one synaptic area since the synaptic areas found, thus far, appear not to be located inside the basement membrane of any one muscle fiber and appear to be centrally located among several muscle fibers. In Figure 6 there is one large glial element from which a much smaller process is emerging. Several smaller glial elements are also seen surrounding the nerve processes. There are nine areas in which there are heavy accumulations 0f sYnaptic vesicles suggesting the presence of a synapse in the near vicinity. This type of multiterminal synaptic 35 Figure 6. Transverse section through part of the synaptic area. 0 is a single nerve process in which there are three (1, 2 and 3) condensations of synaptic vesicles. Glial, G; mitochondria, M; synaptic vesicles, SV; sarco— plasm, SP; invaginations of the sarcolemma, SL. Calibra— tion, 0.5u- x 34,500. 36 ,.L “V5 mm 1! 5. TCO' bra- Figure 6 37 area has recently been found in crustacean skeletal systems (Atwood and Morin, 1970). In one small nerve process labeled Q, there are three distinct regions of synaptic vesicles which suggest that a single nerve process may have several synapses within a few micra of one another. Mitochondria are also found in the nerve processes near the nerve terminals (Figure 8). Figure 7 is a high—power electron micrograph of several nerve terminals showing the synaptic vesicles. At least two types of vesicles are present, a smaller, clear or slightly electron Opaque, vesicle and a larger very dense vesicle. The smaller vesicles are much more numerous than the larger more dense vesicles (ratio of approximately 20:1). The smaller vesicles range in diameter from 350 to 450 3, and the larger dense vesicles range from 700 to 1000 A in diam- eter. These dense core vesicles have recently been observed in several neuromuscular synapses (Atwood 23 gl., 1971). One also sees microtubules in the cytoplasm of these nerve processes. 1 In Figure 8 there are two nerve terminals which appear to form true neuromuscular synapses. The arrows located in the glial elements designate the boundaries of the synaptic area. The glial elements surrounding the nerve terminal appear to end abruptly; no longer separating the nerve fiber from the muscle cell. There is a close association between the nerve terminal membrane (presynaptic membrane) and the 38 Figure 7. Transverse section showing two types of synaptic vesicles. Clear synaptic vesicles, SVc; dense synaptic vesicles, SVd; microtubules, MT. Calibration, O-Zu- x 67,600. 39 Figure 7 40 Figure 8. Transverse section showing two neuromuscular synapses. Arrows indicate the synaptic boundaries. Glia, G; sarcoplasm, SP; mitochondrion, M; synaptic vesicles, SV. Calibration, 0.2u. X 67,000. 41 Figure 8 In.“ 42 sarcolemma (postsynaptic membrane). The synaptic cleft is approximately 150 R in width. There is also a condensation of synaptic vesicles near the presynaptic membrane. Another neuromuscular synapse can also be seen in Figure 9 in the nerve process designated number one. The synaptic cleft in this synapse appears to be filled with an gm, electron dense material. At least three other nerve pro- cesses are present in this section, and in all four processes neurotubules are present. Large numbers of microtubules are also present in the glial elements. In the center of Figure 9 and located in the sarcoplasm is a complicated dense area of large membrane-bound vesicles and short tubules. A few of these sarcoplasmic complexes are found in each synaptic area. In fact these sarcoplasmic complexes constitute a major part of the sarcoplasmic evagi— nations. As seen in cross sections, they make up approxi- mately 25 percent of the sarcoplasmic region of the synaptic area. Note the deep invaginations of the sarcolemma in the lower part of the figure. The postsynaptic region of many neuromuscular systems has been shown to include clear vesicles which compare favor- ably in size to the presynaptic vesicles (see Smith, 1960). However, larger vesicles have'been found in the muscle sarcoplasm near the postsynaptic membrane in only a few systems, e.g., insect muscle (Smith, 1960) and leech muscle (Tulsi and Coggelshall, 1971). A sarcoplasmic complex of 43 Figure 9. Transverse section of the postsynaptic complex. Four nerve processes are visible in this section (1 to 4). The small arrows indicate a neuromuscular synapse at nerve process 1. Basement membrane, BM; glia, G; sarcolemma, SL; sarcoplasmic complex, SC. Calibration, 0.5m. X 34,500. 44 Figure 9 r ‘ \‘péu~ ~\. , «.7... s 45 large vesicles and tubules found in the postsynaptic sarco- plasm of limulus skeletal muscle is much more extensive than those described previously and indeed is a major constituent of the sarc0plasmic evaginations in the synaptic areas. This complex may be composed of sarcolemma via pinocytotic invaginations and/or may also be composed of sarcoplasmic reticulum. The functions of this complex are unknown, but since the sarcoplasmic evaginations of the synaptic area are surrounded by hemolymph it is possible that extracellular fluid may be ingested via a pinocytotic process. The ultrastructure of limulus skeletal muscle compares well with other arthropod skeletal muscle; thin to thick ratios of 5 or 6:1, diads located at the A-I junction, multinucleate, mitochondria primarily located near the periphery of the muscle fibers, and a large number of nerve terminals located in an extensive synaptic area. However, the synaptic areas in limulus skeletal muscles appear to be a good distance from the contractile elements and have been found in sarcoplasmic evaginations of the muscle fibers, whereas in most other arthropods thus far investigated the synapses are located along the surface of the muscle fibers near the contractile apparatus (Atwood and Johnston, 1969). In addition, in the sarcoplasm of limulus synaptic regions, there is found a very extensive sarcoplasmic complex of vesicles and tubules. A similar system of large post- synaptic vesicles has been reported by Smith (1960) in the 46 flight muscle of Tenebrio molitor. The function of this sarcoplasmic complex is unknown. Physiglogy The three distal muscle groups of the mero-carpopodite flexor were used in the following experiments. Each group was investigated separately to determine both its mechanical response and its externally recorded electrical activity following indirect stimulation. The stimulus was applied as either a single stimulus, a paired stimulus or a train of stimuli. Individual muscle fibers from each of the three groups were also investigated to determine their responses to a single stimulus and to a paired stimulus. From the results recorded below these three muscle groups appear quite similar in their physiological responses. Mechanical,Recording§ Thregholds of Mechanical Response.-—A single stimulus (0.05 to 0.10 msec) applied to the internal pedal nerve (IPN) produces a twitch in each of the three muscle groups. The amplitude of the twitch is dependent on the stimulus strength. A series of experiments was performed to determine the number of changes in the twitch amplitude and thereby to determine the number of axons (or groups of axons) innervat- ing the muscle group. Figure 10 shows the results of two such experiments in which stimuli were applied at a frequency 47 Figure 10. Recording of mechanical activity showing threshold dependent increases in tension. Top trace is a record of a Group III muscle showing at least eleven threshold dependent increases in tension-~indicated by arrows. The bottom trace is a recording from a Group II muscle also displaying eleven different thresholds. Time mark, 30 sec. 48 Figure 10 <— <— 49 of one per three seconds. In these experiments the stimulus strength was slowly increased. The arrows indicate distinct increases in tension. The top trace is of a Group III muscle displaying at least eleven distinct increases in tension. The greatest number of increases in tension develop— ment recorded in any experiment was eleven. The mean number ::u of tension thresholds of the Group I muscles was 9 (Range 7 to 11); of the Group II muscles, 7 (Range 3 to 11); and of the Group III muscles, 8 (Range 3 to 11) (see Table II). At no time in any of these muscle preparations was there recorded a decrease in the tension developed that was a func- tion of increasing stimulus strength. A single stimulus applied to the external pedal nerve (EPN) never evoked a recordable response in any of the three muscle groups, nor did stimulation of the EPN simultaneously with stimulation of the IPN produce any change in the response of the muscle groups to the IPN stimulation. The Twitch Re§p0n§§,—-The muscle response to a single stimulus (twitch response) was analyzed as to the total tension, the rise time, the decay time and the duration. Single stimuli were applied at a supramaximal threshold with at least a 15 sec interval between the stimuli. Supramaximal thresholds were used due to the difficulty of maintaining the same threshold values when stimulating at intermediate thresholds. For each muscle preparation five twitch responses were recorded. 50 anew mm mN.o H mh.N mm mm.o H mH.~ mm No.0 «.m~.o “Hmm o» mmo am mm H HmH Hm.o on m.oc am m.~ u o.m HHHnmc m Huhoa wo.m -H mm 00.0 mm om.o u mm.~ mm eo.o « m~.o rmmm ou Hmc am as H mom :6 3 Too om e.H H m.H “HHIMV h Hump mm mo.o H m®.H mm mo.o « HH.H mm 8.0 u 3.0 lows on mOH. mm mm H mHm lo.m on ¢.Hc am m.~ H H.¢ radius 0 owpmm nouwza ou mssmuma 0mm Awmoon mo $50 ouo sowumusa 00m havoc mo Rho 00m mmwm mo Rho encommmm SouHBB “modems .Hm sowmcma vacuums “mosses .Hm sowmsms nouHBB “madame mpaonmmnna mo .02 HHH Hague HH msouw H msouw .Hoxoam ouwpomomusoloume mo unsoum odomse mo noncommmu HmUHcmsumz .HH oHsmH 51 The mean peak tension developed by the Group I muscles was 4.3 gr; by the Group II muscles 1.9 gr; and by the Group III muscles 3.0 gr (see Table II). The rise time for the twitch response is defined as the time necessary to reach the 67 percent level of the peak tension of the response. The mean rise time for a Group I twitch response is 240 msec; for Group II twitch response 290 msec; and for a Group III twitch response 230 msec (see Table II). The decay time of the twitch response is defined as the time necessary for the response to decay to 67 percent of its peak tension. The mean decay time for the Group I twitch response is 1.11 sec; for the Group II response 2.28 sec; and for the Group III response 2.18 sec (see Table II). The response to a single stimulus could be recorded for up to five hours without deterioration when the 9:1, saline to limulus blood, solution was used. When artificial seawater 100 percent Millecchia saline was used the twitch response would deteriorate and become very small after three hours. Rggponge to Repetitive Stimulation.-—In arthropod striated muscle the tension develOped by the muscle is dependent on the frequency of the applied stimulus. Experi- ments were performed to determine to what extent the response of each of the three muscle groups of mero-carpopodite flexor of limulus is dependent on the frequency of stimulation. 52 Figure 11 shows the results of one such experiment using a Group I muscle. vThe experiment consists of two parts. First, the stimuli were applied at frequencies of l, 2, 3, 4, 5, and 6 pulses/sec. Second, the sensitivity of the recording device was changed and the stimuli were applied at frequencies of 5, 10, 15, 20, 30, 50, and 100 pulses/sec. Each stimulation lasted for at least 15 sec, and a supramaximal threshold stimulus was used. An interval of at least five min was allowed between the low—frequency stimulus trains; and of at least 15 min between the high-frequency stimulus trains (lo/sec and higher). As seen in Figure 11 repetitive stimulation of the IPN produces two very obvious effects as the frequency of the stimulation is increased. First, the rate of contraction increases. In all experiments a frequency of 50 pulses/sec was sufficient to produce a maximum rate of contraction, that is, stimulation at frequencies greater than 50 pulses/sec, for example 100 or 200 pulses/sec, did not produce a rate greater than produced by a 50/sec stimulation. In most cases a frequency of 30 pulses/sec was sufficient to produce a maximum rate of contraction. Secondly, as the frequenCy of stimulation is increased, the tension developed is increased. Generally, after a 15 to 20 sec stimulation at the lower frequencies, the response approaches complete tetany. At the higher frequencies, 53 Figure 11. Recording of mechanical activity resulting from different stimulus frequencies. In the bottom record the applied frequencies are (bottom trace to top); 1, 2, 3, 4, 5 and 6 Hz. In the top record the applied frequencies are (bottom trace to tOp trace); 5, 10, 15, 20, 30, 50 and 100 Hz. The responses to 50 and 100 Hz fall along the same line. The bar at the bottom of the figure indicates the stimulus period. Vertical calibration; bottom record, 30 gr; tOp record, 50 gr. Horizontal calibration; 1 sec. 55 20 pulses/sec or higher, the response usually attains tetany 10 to 15 sec following the onset of stimulation. The frequency-response characteristics of each of the three muscle groups following repetitive stimulation are given in Figure 12. In this figure the frequency of the applied stimulus is plotted against the percent tetanic h tension. The tension developed at each frequency was measured 15 sec after the onset of stimulation, and the percent of the tetanic tension was then calculated. It is evident from this graph that the tension developed is dependent on the fre- quency of the applied stimulus and that all three groups of muscles display similar frequency-response characteristics. It is evident from Figures 11 and 12 that the striated muscle of limulus is capable of producing a high degree of summation. An index of the degree of summation is given by the tetanus to twitch ratio, that is, the comparison of the total tension a muscle is capable of developing to the tension developed by a single stimulus. The mean tetanic tension developed by the Group I muscles is 341 gr; by the Group II muscles, 205 gr; and by the Group III muscles, 191 gr. The mean tetanus to twitch ratio for the Group I muscles was 73:1 (9 preparations); for the Group II muscle 107:1 (7 preparations); and for the Grefip III muscle 64:1 (8 prepara— tions). Although the mean tetanus to twitch ratio was some— what higher for the Group II muscle the range of the tetanus to twitch ratio in all three groups is similar (see Table II). 56 .moHOGMHHu on» .HHH msosu “moaouao peace was .HH QDOHO “mx man an pmusommHQmH mH H msoum .msoaumummmum pswuommqp 0>Hm unwed um Eoum came map mucmmmummu ucflom 30mm mSHSEfium may pmcfismm pmuuoam we cosmcmu Oflsmumu mo unouumm .wocwsvoum .NH wusmHm 57 on NH musmHm Umw\mmm.5a on ‘X. ‘X. X. on oo— snnual JO 1N3383d 58 No frequency-dependent decrease in mechanical response was ever recorded when stimulating the IPN other than a decrease which occurred when the frequency was higher than 300 pulses/sec. This decrease is probably due to a conduc- tion block in some of the motor axons innervating the muscles. I: Stimulation of the EPN at the above frequencies did not produce tension development in any of the three muscle groups. High—frequency stimulation of the EPN simultaneously with stimulation of the IPN did not produce any changes in the muscle response to the IPN stimulation. Facilitation of the Twitch Respon§Q.--In addition to displaying summation, arthropod striated muscle generally is characterized by some degree of facilitation. However since the response to a single stimulus in the arthropods thus far investigated is extremely susceptible to fatigue and deteri— oration, it has not been possible to do a thorough study to determine to what extent the twitch response can facilitate. Facilitation as defined by Kennedy (1966) is the case in which the response to the second of a pair of stimuli is greater than the response to the first stimulus. The twitch response recorded from the limulus muscle does not fatigue or deteriorate but can be recorded for at least five hours after removal of the leg from the animal. Therefore, the following experiment was performed to determine 59 to what extent the twitch response can facilitate in the striated muscle of limulus. Paired pulses were applied using various intervals between the pulses. Each of the paired stimulations was preceded by a single stimulus. All stimuli were suprathreshold. A 30 sec interval was allowed between each stimulation. The mechanical response to the paired pulses was then compared to the mechanical response to the single pulse which immediately preceded it. Figure 13 is an example of one such experiment in which a Group II muscle was used. The intervals between the paired pulses are as follows: 1.0 sec, 500 msec, 250 msec, 100 msec, 50 msec, 25 msec, 10 msec, 5 msec, 2.5 msec and 1.0 msec. It is evident from this figure that at the intermediate intervals the response to the paired pulses is much greater than that Observed for the single stimulus. Figure 14 is a graphic demonstration of these experi- ments. In this figure the ratio of the response to the paired pulses to the response evoked by the preceding single stimulus is plotted against the interval between the pulse pairs. The data were plotted for each of the three muscle groups (Group I, n=7; Group II, n=12; Group III, n=14). The mean and the range at each interval is plotted. The dashed-line at the 2:1 level represents the highest ratio one would expect if the responses to each of the pulses of the paired pulses were merely additive. However, it is 60 Figure 13. Response to paired pulse stimulation. The record is of a Group III muscle. Each pulse pair is preceded by a single stimulus. The pulse pair intervals used are as follows; 1 sec (second response), 500 msec, 250 msec, 100 msec, 50 msec, 25 msec, 10 msec, 5 msec, 2.5 msec and 1.0 msec. The last response is the result of a single stimulus. Time mark, 30 sec. 61 Figure 13 62 Figure 14. The ratio between the tension developed by a paired pulse and that develOped by a single pulse is plotted against the interval between the pulse pairs. The points represent the means calculated from at least seven different preparations. The solid lines represent the ranges at each interval. The dashed line at the 2:1 level is the amount of tension expected if the mechanical responses merely summated. RATIO IAIIO RATIO 4:1 2:! 8:] 6:! 4=I 2:I Oil 4:! 2:1 63 GROUP l ‘l—u—l *- INTERVAI. BETWEEN PAIRED PULSES IMSEC) Figure 14 l ' g”, l as; ' u———J GROUP" ,1 I 0 9 I) o l J l I GROUPIII (’0 H l 1 n 1 l _jfll l l; I ,, I O 50 100 250 500 1000 64 obvious from the graphs that in all experiments at intervals from 250 msec to 5 msec the responses were always greater than 2:1. Facilitation was greatest at intervals of 5 msec, 10 msec and 25 msec at times reaching ratios of 7:1, thereby, demonstrating that facilitation always occurred at frequencies of 4, 10, 20, 40, 100 and 200 pulses/sec. Although the mean ratios of the three muscle groups do not coincide, the range of Groups I and III fall completely within the range of Group II. It is apparent that all three muscle groups are similar as to the degree of facilitation of the twitch response. The low ratios at 500 msec and 1.0 sec may be due to the fact that a twitch reaches its maximum amplitude within 500 msec and therefore the second twitch probably occurs on the falling phase of the first response. The low ratios at 2.5 msec and 1.0 msec are most probably due to either a conduction block in the motor axons at the stimulating electrodes or failure of the second nerve impulse to invade the nerve terminals. External Electrical Recordings To record the electrical activity occurring on the surface of the muscle, following stimulation of the IPN, a suction electrode was placed at various spots on the surface of the three muscle groups of the distal mero—carpopodite flexor. Uhless otherwise stated the recording was single- ended against a ground electrode in the bath. The suction 65 applied was sufficient so that contact was maintained even when the muscle contracted maximally. In some cases micrOpipettes filled with l M_NaC1 were also utilized for recording the activity along the surface of the muscle. Eggponse to a Single Stimulus.——The activity recorded from the muscle surface following stimulation of the IPN usually consisted of a biphasic wave. The first phase is a large positive going potential which is followed by a slower and much smaller negative potential. Figure 15D and E give examples of the biphasic wave form. Figure 15A to C show ; the positive phase at a faster speed. The maximum amplitude of the positive phase was 5 mV but in most cases was approxi- mately 2 mV. In a few cases triphasic wave forms were recorded and in these cases the normal biphasic wave was pre- ceded by a very small negative wave. By varying the strength of the applied stimulus it was possible to evoke several distinct thresholds of recorded electrical activity at any given spot on the surface of the muscle. At any one spot on the muscle surface the range of recordable thresholds was three to eight. Figure 16 is an example of a Group III muscle in which there are six dis- tinct threshold—dependent levels of recordable electrical activity. The electrode was located in the center of the ‘most proximal muscle mass of the Group III muscle. The rise time to 67 percent of the total amplitude was measured for each group of muscles. The electrode was 66 Figure 15. External electrical activity evoked by'a single stimulus. A, B and C are records in which the electrical activity was recorded at a high speed. D and E are simultaneous recordings of electrical (lower traces) and mechanical (upper traces) activity. Vertical calibra- tion: A, 400 uV; B, 500 hv; C, 1.0 mv; D and E (lower traces), 2.0 mV; D and E (upper traces), 3.0 gr. Horizontal calibration: A, B and C, 50 msec; D and E, 250 msec. Figure 16. Electrical record showing six thresholds. The six thresholds are obtained at different stimulus strengths. Vertical calibration, 500 mv. Horizontal calibration, 20 msec. 67 :\ .Bk ff”\ 68 always placed midway between the origin and insertion of a muscle but at various locals in the proximal to distal plane. The mean 67 percent rise time for the Group I muscles (5 preparations) was 7.3 msec; for the Group II muscles (6 preparations) was 8.7 msec; and for the Group III muscles (6 perparations) was 9.2 msec (data included in Table III). ; The duration of the positive phase was also measured. The mean duration of the positive phase for the Group I muscle was 83.6 msec (7 preparations); for the Group II muscles was 110.8 msec (7 preparations); and for Group III muscles was 127.4 msec (5 preparations) (data included in Table III). Simultaneous records were obtained of the external electrical activity and the twitch response resulting from a single stimulus. Figure 15D and E gives two examples of such recordings. From these records it is possible to measure the delay between the onset of electrical activity and the onset of the measurable mechanical contraction. The mean delay for Group I was 63.2 msec (n of 9); for Group II 69.3 msec (n of 6); and for Group III 77.2 msec (n of 5). The electrical activity along the surface of the muscle ‘was also monitored by placing the electrode on the surface of the muscle but without applying any suction. In this case the wave forms differed somewhat from those recorded While suction was applied. If the electrode was placed on the rmuscle surface near either the insertion or the origin, 9 6 Ame mm owns m.~H H «.HH Ame mm UOmE o.HH H ¢.hma loo mm owns s.o H ~.m .mowuw>wuom Hmowcmnooe was HMOHHH roe mm come o.m H m.mo Ame mm owns ¢.~ H m.mo nomHm Ho ummco smm3pmm amamn and mm ome m.mH H m.oaa “no mm come N.OH H o.mm soHumusQ Hmuoe roe mm owns H.H H H.m Ame mm owns H.H H m.H wsHs mmHm see manna o>HuHmom HHH msoHO HH msoHO H muons .m»«>wuom HMOHHuOOHO Hmsumuxm mo noeumwuouomumnu .HHH manna 70 a usual biphasic potential was recorded with the first phase being a positive going potential. The onset of activity occurred approximately 25 msec after the stimulus artifact. When the electrode was placed midway between the origin and insertion of a group of muscle fibers, the phases Of the usual biphasic wave were reversed, that is, a biphasic potential was recorded but the negative phase occurred first. The onset of the negative potential was recorded approxi- mately 18 msec following the stimulus artifact. Furthermore, if the electrode was placed at the halfway point between either the origin and the middle or the insertion and the middle of a group of fibers, the biphasic potential recorded may have either the negative or the positive potential occurring first. The delay between stimulus artifact and muscle activity was approximately 20 msec. If suction was then applied in the areas of the muscle where the negative potential was recorded prior to the positive potential, one would once again observe the normal biphasic potential, i.e., a positive wave followed by a negative wave. However, on careful examination of these recordings, each of the large positive potentials were pre- ceded by a very small (less than‘lO percent of the total amplitude of the positive potential) negative potential which lasted for less than 5 msec. Since suction is used to maintain a solid contact with the muscle surface during contraction of the muscle, this 71 explains why positive—going potentials are seen in the experiments reported in this section. However, the tech- nical explanation of this reversal potential is somewhat of a mystery. Josephson (1960) has reported the same phenomenon when recording with suction electrodes from the body surface of hydra; that is by placing a suction electrode on the body surface without applying suction, he Obtained potentials of polarity opposite from those recorded when suction was applied. When using a NaCl filled microelectrode and moving the electrode along the surface of the muscle, I obtained records similar to the records obtained when the larger tipped suction electrode was placed on the muscle surface without applying suction. gggponge to Repetitive Stimulation.-—To determine the effect of different frequencies of stimulation on the externally recorded electrical activity, the following experiments were performed. Stimuli were applied at fre- quencies of 1, 2, 3, 4, 5, 6, 10, 20, and 50 per sec for twenty seconds. At least a five minute rest was allowed between each stimulus train. The results of one such experi- ment are shown in Figure 17A. It is evident from this figure that as the frequency of the stimulus is increased from l/sec to 5/sec, there is a definite increase in the size of the positive going potential. A measure of this increase is the ratio of the amplitude of the 15th response 72 Figure 17. Effect of stimulus frequency on the external electrical activity. A. The electrical activity was recorded with an a.c. preamplifier. Applied frequencies (a to f) are: l, 3, 5, 10, 20 and 50 Hz. Vertical calibration, 1.0 mV. Horizontal calibration, 1.0 sec. B. The electrical activity was recorded with a d.c. preamplifier. Applied frequencies (a to d) are: l, 5, 10 and 20 Hz. Vertical calibration, 500 uV. Horizontal calibration, 1.0 sec. 73 rLLLLLLLdeU/LLLLLLL rummmmmmmm(mmmmmmmmu/ 74 (F pulse) to the amplitude of the first response of the stimulus train. In Figure 17A the ratios are: 1.02 at l/sec; 1.38 at 3/sec; 1.60 at 5/sec and 1.58 at lO/sec. However, at the higher frequencies the amplitude of the positive potential evoked by each impulse instead of increasing begins decreasing so that the ratio of the F pulse to the first pulse in the train is less than one. The ratio at 20/sec is 0.90, and the ratio at 50/sec approaches zero. This decrease is further evident in Figure 17A when one looks at the amplitude of the response following the full 20 sec stimulation. This decrease in amplitude could be explained as a result of fatigue. However, since the mechanical responses at these frequencies do not show any trace of fatigue, it is unlikely that this is a simple case of fatigue. The decrease in amplitude could be a result of a general depolarization of the muscle fiber membranes. The depolari- zation evoked by the individual pulses may have attained their maximum amplitude, thereby giving the appearance of a fatigued response. Since these experiments were recorded through an a.c. preamplifier, any sustained depolarization would not be recorded. Two experiments were done using a d.c. preamplifier (Argonaut 1043) and the recordings of one of these experiments are shown in Figure 178. These record- ings show that there is indeed a sustained potential differ- ence at the muscle surface when the IPN is stimulated at the higher frequencies. 75 In Figure 18 the ratio of the F pulse to the first pulse of the different frequencies was plotted against the frequency for each of the three muscle groups. The mean of each of the frequencies is plotted. Each point is the mean of at least five different preparations. This graphically demonstrates an increase in the F pulse: first pulse ratio t for frequencies of 1 to 6/sec implying a gradual facilita- tion of the junctional activity if one assumes that the activity recorded for these muscles is similar to other arthr0pod systems and that the electrical activity is the sum of electrical events at several neuromuscular junctions. Rggponge to Paired Pul§§§.--From the preceding experi- ment it is obvious that, at least following low-frequency stimulation, there is some degree of facilitation of the externally recorded electrical activity. To further investi- gate the phenomenon of facilitation, a series of experiments were performed in which paired pulses were delivered to the IPN. The intervals used between the pulse pairs were; 1.0 msec, 2.5 msec, 5.0 msec, 10 msec, 25 msec, 50 msec, 100 msec, 250 msec, 500 msec. In these experiments a single stimulus preceded each of the pulse pairs. A rest of at least 15 sec and usually 30 sec was allowed between stimula- tions. An example of one such experiment using a Group II muscle is shown in Figure 19. This figure shows simultaneous recordings of the externally recorded electrical activity and the mechanical activity. 76 .mmHmcmHHu may .HHH msouw “mOHOHHU ammo was .HH QSOHO HmOHUHHO UHHom 0:» an pmucmmmnmmu m« H QSOHO .msoHumummOHm me ummma um mo same mes muswmmnmmH usHom nomm .moswsvmum may umsHmmm OOuHOHm mH sHmuu may «0 OmHsm pmHHH men on chuu msHSEHum m Ho AOmHsm mo Omasm ApsOOHHHH 03¢ mo OHpmu use .mH wusmHm 77 ON ma musmHm a: .6233: O— ‘I|*O 4|(3 4.!) “010 ‘00. 1 O. E em OHM! 78 .OOmE OOH .GOHumHnHHmo HmusouHuom .>E O.H .COHHMHHHHmo HMOHuHO> .Omme m .3 “come OH .0 “come mm .H “come Om .0 rooms OOH .O HommE 0mm .0 “OOmE OOm .Q “msHSEHum mHmsHm .m "OHM mHm>HOusH HHmm mmHsm O39 .s3O£m mum ammomuu Hmmmso wuH>Huom HMOHsmeooe map paw Amoomuu HOBOHV >HH>Huom HMOHHHOOHO men Ho mmsHOHoowH msoocmuHseHm .mmesm OOHHMQ on Omsommmu HMOHHUOOHO Hmcumuxm .mH ousmHm (5 1 1 17: )3 EH 80 In 70 percent of all preparations (14 of 20) a ratio of greater than 2:1 was recorded with paired pulse intervals of 10 msec and 25 msec. Therefore, these responses are not merely the summation of the activity evoked by each of the two stimuli. In the remaining 30 percent of the prepara- tions the maximum ratios attained were between 1.8:1 and 2:1. The possible reasons for these lower ratios will be discussed later. At longer pair intervals (50 to 250 msec) at least a slight increase in amplitude of the second pulse of a pair was always recorded. These experiments further demonstrate that the externally recorded activity from limulus muscle is capable of being facilitated. Intracellular Recordings Single muscle fibers from each of the three distal groups of the mero-carPOpodite flexor were examined. In all three groups the muscle fibers could be penetrated more easily when approaching the ventral surface of the muscle rather than the dorsal surface. A thickened connective tis- sue layer, the presence of many blood vessels and small nerve bundles, as well as the presence of the IPN along the dorsal surface of the muscle groups are the major reasons making penetration of the muscle fibers at the dorsal surface more difficult. Therefore, in the studies reported here the muscle groups were approached from the ventral surface. 81 .Rgsting,Potentiglg.—-Determination of the resting mem- brane potentials of single muscle fibers is rather difficult since the potential drOp, as the electrode enters the fiber, is not the sharp drop one would expect to see. Instead the decrease in the potential occurs in a stepwise fashion (voltage drOps of 10 to 30 mV) as the electrode is advanced into the fiber. This phenomenon has been discussed by Fatt and Katz (1953) in their investigation of crustacean muscle fibers. Therefore, the range of the resting membrane potentials recorded from the muscle fibers in this study is quite large (30 to 70 mV). However, the mean resting potentials of the fibers in each of the three muscle groups falls between 50 and 51 mV (see Table V). SpontaneouggActiyity.-—Two types of spontaneous activity were recorded from the muscle. The first type, as shown in Figure 20, was a regular, repetitive, spike-like discharge which occurred in approximately 25 per cent of the fibers investigated and was recorded immediately after the electrode had penetrated the fiber. The frequency of discharge was greatest immediately following penetration and then gradually decreased until the discharge subsided. This activity could subside in a few seconds or could occur for as long as five minutes. The amplitude of these potentials also varied greatly, ranging from 2 to 50 mV. It is assumed that this activity is due to a local damage of the muscle membrane by the penetrating electrode. 82 Figure 20. Spontaneous intracellular activity. Activity is usually recorded immediately following penetration of the cell. A is from a Group I muscle; B, Group II muscle. Vertical calibration: A, 10 mV; B, 5.0 mV. Horizontal calibration: 400 msec. Figure 21. Miniature excitatory postsynaptic potentials. A is from a Group I muscle; B, Group II; C, Group III. Vertical calibration: 1.0 mV. Horizontal calibration: A, 1.0 sec; B, 4.0 sec; C, 2.0 sec. 83 .JL .1 20 21 84 The second type of spontaneous activity is the typical miniature postsynaptic potentials one normally sees when studying neuromuscular junctions (Katz, 1966). Figure 21 gives three examples of the spontaneous miniature post- synaptic potentials recorded frOm limulus muscle fibers. The maximum amplitude of the miniature potentials I have recorded is 5.0 mV, however the amplitude of most of the potentials is less than 1.0 mV. These miniature potentials are usually recorded in fresh preparations (less than one hour old), but will also appear in Older preparations follow- ing high-frequency stimulation. Exgitatory Pogtsynaptic Potentigl§.--In this report I will follow the nomenclature of Atwood (1963) and designate the small local neuromuscular events excitatory postsynaptic potentials (EPSPs). These potentials have also been termed excitatory junction potentials (EJPs) in crustacea (Hoyle and Wiersma, 1958) and end plate potentials (EPPs) in crustacea (Fatt and Katz, 1953) and vertebrates (Eccles and O'Connor, 1939). Several thresholds of postsynaptic activity were re- corded in each muscle fiber following stimulation of the IPN; these levels of activity depended on the strength of the applied stimulus. Table IV gives a complete description of the number of muscle fibers investigated and the number of thresholds found in each fiber of the three different muscle groups. The percentage of muscle fibers displaying 85 m NN ve mm b I mHOQHm Hmuou mo Hamoumm He m m mH m m mHmHHH mHomss mo Hmsssz HHH msouo 0H mH mm mH H , mHmQHH Hmuou Ho HchHmm OH H o Hm o m mHmnHm mHomss mo Honssz HH macho H mm mm HH 0 mHmnHH Hmuou mo ucmuuom om H mm mm OH O mHoHHH mHumss mo Hmnesz H QSOHO Hmuos m m e m m HOQHH OHomse Hem mOHonmmusv mo Honesz .mumnHH OHomsE mamsHm mo sOHum>HoacH .>H OHHMB 86 the different thresholds is also given. (The data in this table also include some muscle fibers which gave the graded-spike response.) The majority of the fibers investigated displayed at least 4 or 5 distinct thresholds, thereby implying that these fibers are innervated by at least 4 or 5 different axons. The greatest number of thresholds Observed in any muscle fiber was 6. However, it is possible that more than six axons can innervate a muscle fiber since several axons may be excited at the same stimulus strength. In the fibers innervated by only two axons, generally, the amplitude of the low—threshold activity was small (1 to 5 mV), whereas, the amplitude of the high-threshold activity would be 3 to 5 times greater than the low-threshold activity. This corresponds to the fast and slow responses recorded by Hoyle (1958) in the limulus closer muscle. .Hoyle uses this as the basis for assuming that limulus has both fast and slow axons. However, it is possible that the second level of activity is evoked by stimulation of at least two axons which could not be separated by changing the stimulus strength, thereby, giving the appearance of a typical crustacean fast response. Figure 22 gives several examples of EPSPs recorded from various muscle fibers of the three muscle groups. As seen in the figure the EPSPs are composed of a rapid depolarization followed by a much slower repolarization. 87 Figure 22. Examples of EPSPs. Vertical calibration, 5.0 mV. Horizontal calibration, 100 msec. Figure 23. Examples of graded—spikes. A, B and C are from different muscle fibers. D and E are consecutive recordings from the same muscle fiber. Vertical calibra- tion: 5.0 mV; B, 25 mV; C, D and E, 12.5 mV. Horizontal calibration: 100 msec. 89 To more clearly quantify the EPSPs, at least five potentials from several muscle fibers of the three groups were recorded, and the amplitude, the 67 percent rise time and the 67 per- cent decay time of each EPSP were measured. The rate of rise of each EPSP was calculated. The means of the above parameters for each muscle fiber was then determined. In all three muscle groups the mean amplitudes of the EPSPs was between 9.0 and 10 mV (Table V). The mean time required for the EPSP to attain 67 percent of its maximum amplitude was likewise quite similar in all three groups, ranging from 8.4 to 9.0 msec. These rise times correspond quite favorably to the 67 percent rise times recorded extra- cellularly from the muscle surface (see Tables III and V for comparison). The length of time necessary for an EPSP to decay 67 percent of its maximum amplitude was extremely variable. The total range for all muscle groups was 48.0 msec to 325 msec (see Table V). Part of this large vari- ability is probably due to movement artifact since the micro— electrode is extremely sensitive to bending. The rate of rise was determined by dividing the maximum amplitude of each EPSP by the time elapsed from the onset of the EPSP to the attainment of the maximum amplitude. The rate of rise for all muscle fibers studied varied from 0.24 to 0.92 V/sec, and the mean rate of rise of muscle fibers in the three groups were quite similar: for Group I 0.51 V/sec; Group II 0.52 V/sec; Group III 0.47 V/sec (see Table V). 90 Hon.o OH ¢N.ov Heme 60.0 H H0.0 +| rams 0.0 0.HOH Ammo 6.0 H 0.6 m.m +| IHNO 6.0 A05 Op emv HoHO 0.0 H 0.06 "00.0 cu 6~.0O Isms 00.0 H «6.0 AmHo ®.OH H m.mm o.m +| Haws 6.0 H0~O 6.0 H 0.0 Hos ou mmc IoHc 6.0 H 6.06 HHH.0 0» H~.0O HHmO No.0 H H6.0 Immc 6.mH H 0.6mH "H6O 6.0 H «.6 erO 0.0 H 0.0 Amp 0» emv r6HO 0.0 H 0.06 “Omsmmv mm H umm\> mem mo mumm mm H come OEHB haven $50 mm H UOmE OEHB OmHm fine 66 H >5 oesHHHms« mmmm Awmsmmo mm H >8 mHmHucmuom 0cHummm HHH mucus HH QUOHO H moose .HsHusouom OHuQMGMmumom huoumuHuxm we» no mOHumHuouomumgv .> OHQMB 91 There is also a direct relationship between the mean rate of rise and the maximum amplitude of the EPSPs (Figure 24) (r = 0.77, p >.0.01, Spearman Rank Test). Graded-Spikes.-—The second type of electrical activity recorded from muscle fibers is the graded-spike. This potential always arises from an EPSP and can occur on either the rising or the falling phase of the EPSP. However, it is not all "all-or—nothing" response. For example, in any particular muscle fiber displaying the graded spike, it may be absent, it may have an amplitude of only a few millivolts or, as is generally the case, it may be large enough to approach zero potential (see Figure 23). In three fibers the spikes had "overshoots" of 5 to 10 mV, however, this small "overshoot" may be due to a false evaluation of the resting potential due to bending of the microelectrode. In a few cases negative—going after—poten- tials were observed but in most fibers the recordings of after—potentials were marred by movement artifact. The graded—spikes occurred rather frequently in the muscle fibers of the three muscle groups. In Group I 44 percent (63 of 142) of the fibers investigated produced the graded-spikes; in Group II, 48 percent (55 of 115); in Group III, 39 percent (48 of 124). The wave form of the graded-spike was analyzed to determine the total amplitude of the response (i.e., the EPSP and the spike), the 67 percent decay time of the total 92 .mOHomsE HH msouw Hmum>mm EOHH pwchuQO wum3 ammum mHnu CH concomoum sump OLE .HOQHH mHomsE mHmsHm m EOHH mmmmm pmxo>m mHHmEmeEmumsm Ho mesmewusmmoe O>HH no game may mH usHom some .mmmm may mo mpsuHHmEm was umCHsmm pmuuon mH mmmm may Ho omHH Ho mumu may SOH£3 CH namuo .¢~ musmHm 93 15 IO EPSP, LO- ”VA ‘asm so 31sz InV Figure 24 94 response and the rate of rise of the spike. At least three responses, and in most fibers five responses were recorded and the means for these responses were determined. Graded- spikes with amplitudes of greater than 10 mV were used in this analysis since there is some difficulty in measuring the above parameters in fibers displaying graded—spikes of only a few millivolts. The mean amplitude of the total response in each of the muscle groups were between 29.0 and 35.0 mV (see Table VI). However, the mean amplitude for individual fibers ranged from 22.8 to 50.8 mV. The 67 percent decay times of the total graded-spike response was extremely short compared to the decay times of the EPSPs. In Group I the mean decay time was 24.6 msec; Group II, 16.6 msec; Group III, 23.5 msec (see Table VI). In some fibers the 67 percent decay times were less than 10 msec; suggesting that some active response may be responsible for the rapid decay of the graded-spike. The rate of rise of the graded-spike was determined by measuring the amplitude of the spike from the spike base (potential at which the spike arises from the EPSP) to the peak of the spike. This amplitude was then divided by the spike rise time (time elapsed from the onset of the spike to its peak). The rates of rise of the graded-spikes in the individual muscle fibers can vary by more than 2.0 V/sec and the mean rates of rise of all fibers investigated range 95 Amh.m 0» HH.HV ”no NN.O H Oh.H “we m.h H m.mm HH.H6 0» 6.0mc Ass 6.H H ~.0~ x00.~ 0» 60.00 r-.0 0» m0.HO rm0c6mc H0HO 6H.0 H 60.H HHHO 0H.0 H m0.H mm H omm\> =6memn6666H0= 0:» Ho mmHm «0 mumm AmHO 6.H H 6.6H H6HO 6.m H 6.H~ mm H owns uncommmm Hmuoa no known Rho r6.6¢ 0H 6.-O “0.06 0» 0.NNO .mmcmmv r0HO 0.H H 0.06 AHHO 0.m H 0.0m mm H >5 mmcommmm Hmuos Ho mcsuHHmsm HHH msouw HH msouw H msouw .Omeuupmanm we» no mOHumHuouomumro .H> OHnma 96 from 0.93 V/sec to 4.22 V/sec (see Table VI). However, the mean rate of rise for each of the muscle groups was quite similar; Group I, 1.83 V/sec; Group II, 1.95 V/sec; Group III, 1.70 V/sec (see Table VI). The mean rates of rise of the graded—spikes are 3.5 to 4.0 times greater than those of the EPSPs. But, the rates of rise of the graded—spikes in limulus muscle fibers are only 10 percent of the rates of rise of the crustacean and insect muscle spikes (15.0 to 20.0 V/sec), and much less than the rate of rise of the frog muscle action potential (500 to 700 V/sec) (Spector, 1956). Paired Pulse Stimulation.-—The effect of paired ‘pulse stimulation on the intracellular activity was also examined. Pulse pairs with intervals of 10, 25, 50, 100, 250, and 500 msec were used in these experiments. In most cases a paired pulse delivered to the IPN evoked a sufficient contraction to dislodge the microelectrode. However, in a few fibers the effect of pulse pairs could be recorded. The effect of the pulse pairs on non-spiking and on graded- spiking fibers will be described separately. Figure 25 shows the effect of paired pulse stimulation on the intracellularly recorded EPSPs. It is obvious from this figure that the response to the second stimulus of the pair is sufficient to produce spikes at the shorter pair intervals, i.e., 10, 25 and 50 msec. At longer pair 97 .OOmE OO¢ .0 rooms OON .m lemma OOH .e "COHHMHQHHMO HmusouHHom .>E Om .O “>5 6H .m can e HGOHumunHHmo HMOHuHm> .omme Omm .H “Dome OOH .m momma Om .p HOOmE mm .0 “come OH .9 “msHsEHum OHOCHm .m "m3OHHOH mm OHM mHm>kusH HHMQ OmHsm .AO pom m .o MHH>HuOm HMOHHHOOHO Ho mmsHpHOUOH HMHsHHOOMHusH .mm OHDOHH H _.| jl/IIJAI4I4I/HIH (HHH/(Heed... 99 intervals, the second response was generally greater than the first. The pulse pair interval necessary to evoke a spike response differed in individual fibers. Thirty-six fibers of Group III muscles were investigated to determine which pulse pair intervals would evoke a spike to the second stimulus. A11 36 fibers produced a spike when a pulse pair with a 10 msec interval was applied; 31 fibers produced spikes at 25 msec; 19 fibers at 50 msec; 7 fibers at 100 msec; and one fiber produced a spike at 250 msec. In one Group I fiber a spike was evoked by the second pulse when a pulse interval of 500 msec was used. Also in only one muscle fiber, of 93 fibers investigated (Group II, Prep 2, fiber 7), was there a failure of the second pulse to produce a spike using any of the pair intervals. However, the second re- sponse in this fiber did show a marked degree of facilitation. To better demonstrate the marked increase in the total depolarization (facilitation) of the intracellularly recorded potential, the ratio of the amplitude of the second response of a pair to the amplitude of the first response was calcu- lated and then plotted against the pulse pair interval. Since the first response of 10 msec pair could not be direct- ly measured, the ratio of the amplitude of a single stimulus response and the total amplitude of the 10 msec pair minus the amplitude of the single response was calculated. 100 Figure 26 is a graphic demonstration showing the dramatic increase in potential as recorded from single fibers of a Group I muscle. The facilitation of the mechanical response (Figure 13) and the EPSP (Figure 25) is quite marked, with pulse pair intervals of 10 and 25 msec producing the greatest facilita- tion. The recordings of the gross muscle electrical activity, however, do not show the degree of facilitation that is seen in either the mechanical recordings or the EPSP (the degree of facilitation depends on conductance changes which occur at the neuromuscular junction) recordings. In fact it would appear that only simple summation of the electrical responses is occurring. This perhapSWCan be explained in part if one assumes that the intracellularly recorded graded-spikes compose the major portion of the external potential recorded from the muscle surface. If this is the case and since approximately 45 percent of the muscle fibers produce spikes and approximately all of the muscle fibers produce a graded- spike to a paired pulse of 10 msec, i.e., an increase of approximately 50 percent, then the expected externally recorded activity would necessarily be doubled, thereby giving “ the impression of only simple summation of the externally recorded activity to paired pulse stimulation. In those fibers displaying a graded—spike to a single stimulus, the response to the second stimulus of a pulse pair is variable. In approximately 50 percent (12 of 27) of the spiking fibers investigated, the response to the second lOl .Hm>HmuCH some as mmCsH on» OHMCmHmop mush was pCm Hm>HOuCH some as mCmmE on» uCOmOHme muCHom OCH .mOHUmCE H QCOHO EOHH mHmQHH OHOmCE HmHm>Om EOHH COxMu mum sump snore .HHmm OmHsm may Cmm3umn Hm>HOuCH esp umCHmmm.pmuHOHm mH HHMQ m HO OmCommOH umHHH on» no OOCuHHmEm OCH pCm HHmm m Ho mCHsEHpm OCOOOm wnu on mmCommOH may no opsuHHmEm on» Cmozuon OHHMH use HOHAB CH somnw .ON OHCOHH 102 onu 6m 6H00HH use. .._<>¢u._.z. CO— on mm o— HUN :V OIlVII 103 pulse at intervals greater than 25 msec was also a graded- spike. At an interval of 25 msec a small spike was usually seen on the decaying phase of the first spike. At an intere val of 10 msec no change in potential was observed unless the graded-spike to a single stimulus did not approach zero potential. In these fibers there was an increase in the spike amplitude when a 10 msec pulse pair was delivered. In the remaining fibers (15 to 27) the response to the second stimulus was a smaller graded—spike or at times there was only an EPSP recorded. This decrease in or "instability" of the spiking activity has been reported previously by Hoyle (1958) in the limulus dactlepOdite flexor. An implication from these paired pulse experiments is that all the muscle fibers appear to be quite similar. Ostensibly one could classify limulus muscle fibers into two groups, those producing only an EPSP (slow) to a single stimulus and those producing a graded—spike (fast). However, since the fibers producing EPSPs can produce a graded-spike .to paired pulse stimulation, it would appear that all fibers are similar and differ only in the amount of stimulation (be it due to transmitter release or membrane depolarization) necessary to produce graded-spike activity. DISCUSSION The evidence presented in the above result section suggests quite strongly that the muscle fibers of the walk- ing leg musculature of Limulug polyphemus are morpho— logically similar to one another. In addition, physiological - In?“ luuttmw evidence suggests that the muscles and muscle fibers, at least in the three distal groups of the merO-carpopodite flexor, are very similar in their responses to indirect stimulation. In the following discussion I will compare the skeletal muscles in limulus to those in other arthropod groups, particularly the Mandibulates (the crustaceans and insects) as to their morphology, their innervation and their physiological responses to indirect stimulation. MuscleWMorphology One of the parameters Atwood (1963) has used to classify the muscle fibers of decapod crustaceans is sarcomere length. He groups muscle fibers in three categories: Type A (phasic), sarcomere lengths of 2.0 to 3.0u; Type B (tonic), sarcomere lengths of 8.0 to 12.0u; and Type C (intermediate), sarcomere lengths of 5 to 10p. In the insects the sarcomere lengths of muscle fibers, thus far studied, range from 2.5 to 7.0m (Hoyle, 1965a; Usherwood, 1967). 104 105 In addition to categorizing muscle fibers according to sarcomere lengths, several ultrastructural parameters have also aided in further defining the differences among muscle fibers. The ultrastructure of fast and slow crusta- cean skeletal muscles has been studied.by several investi- gators (Jahromi and Atwood, 1967, 1969; Cohen and Hess, 1967; Fahrenbach, 1967). Likewise the ultrastructure of insect muscles has been investigated (Smith, 1960, 1968; Hoyle, 1966b: Jahromi and Atwood, 1969). There is one basic similarity between the fast and slow systems. Diads and infrequently triads are formed between the sarcoplasmicsreticulum and the transverse tubule sys- tems; these diads are found adjacent to the A bands at the A—I junctions. However, there are several differences be— tween the fast and slow muscle fibers in both crustaceans and insects. The slow fibers lack H zones, have a rather high thin to thick myofilament ratio (5 or 6:1) and have broken or uneven Z lines. The fast fibers have an H zone, a smaller thin to thick myofilament ratio (3 or 4:1) and a Z line which appears fairly straight across a muscle fiber (Sherman and Atwood, 1971). The Z line is also somewhat ‘thinner in the fast fibers (Jahromi and Atwood, 1969). With the exception of the fast fibers of the deep abdominal ex- tensors Of the crayfish Procambarus clarkii and Orconectes virilis (Jahromi and Atwood, 1967), fast fibers have a greater amount of sarc0plasmic reticulum than the slow fibers. 106 This is similar to the differences found between fast and slow muscle fibers in the vertebrates (Peachy, 1970). In the muscles of the walking legs of limulus, the range of sarcomere lengths was 5.9 to 10.5u. However, in any one muscle or muscle group the range was never greater than 3.1u. This suggests that the muscle fibers of limulus skeletal muscle are all of one type as far as the sarcomere length is concerned. This is supported by previous works of de Villafranca (1961) and de Villafranca and Philpott (1961) on the limulus opisthosomal extensor in which the sarcomere lengths were approximately 7.5u and of Ikemoto and Kawaguti (1968) on the opisthosomal flexor (retractor) of Tachypleus Epidentatus which had sarcomere lengths of 8.0u. Therefore, the sarcomere lengths measured for limulus skeletal muscle fibers, although somewhat larger than those of insects, fall in the range of the intermediate fibers of crustacean systems. In this study and in those of de Villafranca and Philpott (1961) and Ikemoto and Kawaguti (1968), it has been shown that the ultrastructure of the skeletal muscle of the horseshoe crab is quite similar in most respects to the slow muscles of the crustaceans and insects. There are diads, usually two per sarcomere, adjacent to the A bands at the A-I junction. There iS'no H zone dividing the A band. The Z line of the myofilament bundles do not form a straight line in each of the muscle fibers, but they are not as uneven as 107 those seen in crustacean tonic muscle fibers. The Z lines in limulus are approximately equal to those of the tonic superficial abdominal extensors of the crayfish (Jahromi and Atwood, 1969). Finally the amount of sarc0plasmic retic- ulum present in limulus muscle appears to be considerably less than that found in most crustacean and insect fast muscle fibers but more well developed than the sarc0plasmic reticulum in the slow fibers. These data suggest that the skeletal muscle fibers of limulus are structurally of an intermediate type when compared to the various types of muscle fibers found in crustacean and insect systems. Synaptic Area Morphology In crustacean and insect neuromuscular systems the synapses between motor neurons and the muscle fibers are of three basic types. A synapse may be at the surface of the muscle where a small evagination from the muscle surface contacts a nerve process; it may be partially embedded in the sarcoplasm, below the muscle surface but distant from the contractile apparatus, or it may be located in the muscle fiber adjacent to the myofilaments (Atwood §£_§l,, 1969; Atwood and Morin, 1970; Peterson and Pepe, 1965). In one case, the superficial abdominal extensor muscle of the crayfish, arm-like extensions of the sarcoplasm surround the nerve processes and the synapses are located in these 108 sarc0plasmic extensions (Jahromi and Atwood, 1969). At least three structural types of synaptic vesicles have recently been described in crustaceans: small, round, clear vesicles; small, ovoid, clear vesicles; and large, dense-core vesicles (Atwood and Morin, 1970; Atwood 33 al., 1971). Only the small, round, clear vesicles have been found in insects (Usherwood, 1967), but the dense-core vesicles have been observed in other invertebrate neuro- muscular synapses (Barrantes, 1970). In limulus the organization of the synaptic area is quite similar to synaptic areas found in other arthropods. The synapse is located in extensive evaginations of the sarcoplasm and the nerve processes are located in these sarcoplasmic evaginations. The nerve processes are completely sheathed by glial elements except in small areas in which there are large accumulations of synaptic vesicles. These areas are presumed to be the regions of synaptic contact between motor nerves and muScles. Therefore, the synapses are greatly removed from the contractile apparatus. The synaptic vesicles in.limulus skeletal muscle consist of at least two types; smaller, more numerous, clear vesicles and larger, dense-core vesicles. A large number of groups of synaptic vesicles are found in a small region of the synaptic area. In one nerve process three distinct groups of vesicles were observed within a few micra of one another. This suggests that the innervation of limulus is an extranely 109 intricate system in which a single nerve process is capable of forming a large number of synaptic contacts with a muscle fiber; it is also very probable that a single axon gives rise to many of the smaller nerve processes. Innervation Patterns In the crustaceans and the insects, the skeletal muscles are innervated by only a few excitatory neurons, less than five. These axons can be classified as either fast or slow depending on their conduction velocity and the response they evoke in muscle (Wiersma, 1941; Hoyle, 1965a). One or two inhibitory neurons innervate crustacean skeletal muscle (Wiersma, 1941), however, inhibition has only recently been found in insects (Usherwood and Grundfest, 1965; and Pearson and Iles, 1969). Each axon branches many times and sends several nerve processes to many muscle fibers and also several processes to a single muscle fiber (Van Harreveld, 1938; Bittner and Kennedy, 1970). Therefore, arthropod muscle is said to be polyneuronally and multiterminablly innervated. In limulus Parnas §§_§l, (1968) have reported finding at least six excitatory axons and at least one inhibitory axon innervating the dactylopodite flexor but with never more than three excitatory axons innervating a particular muscle fiber. Hoyle (1958) had previously worked on the flexor muscle of the dactlepodite of limulus and reported 110 finding only two motor axons, a fast and a slow, and no inhibitory innervation. However, in a recent review Hoyle (1969) implies that he found at least three excitatory axons and one inhibitory axon innervating the claw flexor muscle, but to the best of my knowledge this data has not been published. Parnas gt El. (1968) separated the internal pedal nerve (IPN) and stimulated single motor axons innervating the dactylopodite flexor. They found very little difference in the electrical activity evoked in a single muscle fiber when stimulating the different“excitatory axons; thus, implying that all the excitatory axons innervating the dactlepodite flexor were very similar. In limulus recordings of the mechanical and the external- ly recorded electrical activity from the three groups of the distal flexor of the mero—carp0podite have shown that the individual groups may be innervated by as many as eleven excitatory axons from the IPN. No inhibitory input from either the IPN or the EPN was observed. Intracellular record— ings from single muscle fibers show that from two to six axons may innervate any particular muscle fiber. The above evidence suggeSts that limulus skeletal muscle is similar to other arthrOpod muscle in that it is polyneuronal- ly and multiterminally (from ultrastructure) innervated. However, in limulus the neuromuscular system differs somewhat from other arthrOpod systems studied in that the muscles are innervated by a larger number of axons. Furthermore, these 111 axons appear to be rather similar in function (Parnas gt 31., 1968) and cannot be classified as being fast and slow. Parnas gt a1. (1968) demonstrated inhibitory innerva- tion to the flexor of the dactylopodite and Hoyle (1969) also implies that he found an inhibitory input to this muscle. I found no inhibitory axons innervating the distal flexors of the mero—carpopodite. Since this muscle has no antagonist (Ward, 1969), this study supports the Hoyle-Smyth hypothesis (1963) that one would expect to find inhibitory axons inner- vating only those muscles which have an antagonist. Physiological Responses One of the purposes of choosing the three distal groups of the flexor of the mero-carpopodite for this study was to determine whether there were any physiological differences among the groups. From the data summarizing the mechanical (Table II) and the electrical (Table III) activity recorded from each of the muscle groups, all of the groups appear similar as to their gross physiological responses- Likewise, from the data on intracellular events of individual muscle fibers in each of the muscle groups, all three groups appear similar. All three groups are composed of fibers showing EPSPs to a single supramaximal stimulus (55 percent of total fibers) and fibers displaying graded-spikes (45 percent of total fibers). Therefore, in the following discussion of 112 single stimulus response (twitch response) and summation, these three muscle groups will be treated as being identical. A mechanical response to a single stimulus could always be recorded in these muscles even at the lowest threshold; this twitch response could be recorded throughout the course of an experiment. Parnas gt a1. (1968) found that each of the six excitatory axons innervating the flexor of the dacty— lopodite of limulus produces a measurable amount of tension following a single stimulus. This differs markedly from crustacean skeletal muscle systems in that the twitch follow- ing stimulation of the fast axon is either easily fatigued (Lucas, 1917) or usually lost after an hour of experimenta- tion (Atwood gt al., 1965). Apparently the twitch response in some insect muscles is much more stable than that in crustaceans (Wilson, 1954; Pearson and Iles, 1971). In other insect muscles the twitch can fatigue quite rapidly (Hoyle, 1955; Becht ££.él-: 1960; usherwood, 1962). The twitch response in both crustaceans (Wiersma, 1961) and insects (Usherwood, 1962) is much more rapid (rise times less than 15 msec) and of shorter duration (half-decay time less than 20 msec in insects and 50 msec in crustaceans) than that reported here for limulus. It should also be noted that a single stimulus applied to the slow axons innervating crust- acean and insect muscle produce virtually no measurable tension. 113 In limulus the intracellular response of a single muscle fiber to a single maximum threshold stimulus would evoke either an EPSP (55 percent of the fibers) or a graded- spike (45 percent of the fibers). The EPSPs have a slower rise time, rate of rise and decay time than EPSPs evoked by the fast axons innervating crustacean muscle (Hoyle and Wiersma, 1958), and the EPSP evoked by both the fast and slow axons innervating insect muscle (Hoyle, 1955; Pearson and Iles, 1971). The graded-spike response to a single stimulus is found much more commonly in limulus skeletal muscle fibers than in crustacean fibers since stimulation of the fast axon in crustacean system usually evoked large EPSPs instead of a graded—spike (Wiersma, 1961). The graded—spike found in limulus is, however, quite similar to that found in crusta- ceans in that it is most likely a local response (due to the extreme variation of spike heights, and also due to the fact that the spike amplitude at its greatest value only attains zero potential plus or minus 10 mV). The spike seen in insect muscle has been shown to be "all-or-none" (Wilson, 1954; del Castillo gt g1., 1953). The rate of rise of the graded-spike in limulus muscle is much slower than that found in either the fast system of crustaceans (20 V/sec) (Fatt and Katz, 1953) or of insects (34 to 40 V/sec) (Spector, 1956). Summation of mechanical and electrical responses is of common occurrence in all arthropod muscle with the exception 114 of a few fast systems in crustaceans (Wiersma, 1961) and insects (Usherwood, 1962). In the experiments reported here limulus skeletal muscle response shows a high degree of summation with a tetanus to twitch ratio of 90:1. Tetany can be attained with stimulus frequencies between 20 and 50 Hz in the flexor of the mero-carpopodite, where Hoyle (1958) reported that a frequency of 200 Hz is necessary to produce tetany in the dactylopodite flexor of limulus. However, his value was for stimulation of the slow axon, and he was not using a suprathreshold stimulation for all axons innervating the muscles. The tetanus to twitch ratio cannot be determined for the crustacean slow syStem since no response is evoked by stimulating the slow axon. The tetanus to twitch ratios given for the fast system are rather tenuous since the twitch responses evoked by the fast axons are rather labile. For insects the ratios for stimulation of the fast system is only 1 to 10:1 (Usherwood, 1962; Becht and Dresden, 1956) which is much lower than that found in limulus. Tetany can be produced in the fast system of crust— aceans by stimulating the fast axons at frequencies of at least 20 Hz. Frequencies Of at least 100 Hz and at times of 200 Hz are necessary to evoke tetanus in muscle innervated by the slow system. In insects little summation is usually seen in muscles supplied with fast axons, but frequencies of 150 Hz will produce tetany in muscles supplied with slow axons . 115 Compari§0n§_to Other Chelicerates The preceding discussion concentrated on comparing the neuromuscular system of limulus to that of the Mandibu- late groups, the crustaceans and the insects. In the Chelicerate group only two other organisms besides limulus have been investigated; the tarantula, Dugesiella'hentzi (Rathmayer, 1965, 1967; zebe and Rathmayer, 1968) and the scorpion, Leiurus quinquestriatus (Gilai and Parnas, 1970). In the spider, there are only a few motor axons (less than five) innervating the leg muScles and these are classified as either fast or slow. Recently Sherman (personal communi- cation) has found different morphological types of muscle fibers in the skeletal muscle of the spider. In the scorpion claw closer, there are only two motor axons, a fast and a flow, innervating the muscles, but the muscle fibers appear to be similar in their physiological responses. From the above comparison of physiological responses between the limulus and other arthropods, it would appear that the twitch response in limulus is intermediate between the fast and slow system of the other arthropods. However, it must be stated that theSe comparisons are between whole muscles as far as mechanical activity is concerned. Atwood (1963) and Atwood gt El- (1965) have shown that there is a large variation in the mechanical responses of individual muscle fibers in crustacean muscles. The muscle fibers in the muscles of the limulus leg appear similar in structure, 116 and the fibers of the flexor of the mero—carpopodite and dactylopodite flexor appear similar in their physiological responses and once again intermediate between the spiking and non-spiking fibers of crustacean and insect muscle. Finally, the evidence of Parnas gt 31. (1968) shows that the excitatory axons of limulus also appear similar, i.e., they cannot be designated as either fast or slow. Therefore, this information suggests that the neuromuscular systems in the walking leg of limulus is rather a non-specialized system when compared to its counterpart in other arthropods. Phylogenetic Considerations Limulus polyphemus is one of the most ancient arthropods living today. Horseshoe crabs first appeared in the Jurassic era, and remains of the modern species were observed first in the Triassic era (Moore 33 21., 1952). Simpson (1953) has stated that the horseshoe crab has existed without any notable structural change for some 200,000,000 years and that it must have evolved rapidly at first but its evolutionary rate must have dropped nearly to zero thereafter. If one assumes that the evolution of internal structures and physiological relationships have likewise remained static for the last 200,000,000 years, then there are some interest- ing phylogenetic relationships that may be drawn between limulus and the more modern arthrOpods. In the more modern members of Arthropoda there is a smaller number of motor 117 axons innervating the skeletal muscle than are found inner- vating limulus muscle. There is a specialization of motor axons in the modern arthropods while the motor axons in limulus appear to be of a single type. There is specializa- tion in the muscle fibers of modern arthropods in both structure and function while in limulus the muscle fibers appear similar in structure and function. Therefore, assum- ing that limulus is indeed an ancient representative of the arthropods, the above relationship suggests a type of struc- tural and physiological evolution from an unspecialized neuromuscular system to a more specialized system. SUMMARY 1. The morphology of the musculature of the walking legs of Limulus polyphemug is described, and the sarcomere lengths of single muscle fibers from each of the muscles are examined. 2. The ultrastructure of the muscle fibers and the synaptic area is described. 3. The mechanical and external electrical activity are recorded from the distal muscle groups of the mero— carpopodite flexor. 4. The intracellular activity (EPSPs and graded-spikes) recorded from single muscle fibers is described. 5. The muscles are innervated by a large number of axons, which appear similar in function. The muscles are composed of fibers which appear similar in structure and function. 6. The phylogenetic relationships between limulus and other arthropod groups are discussed. 118 LITERATURE CITED LITERATURE CITED Atwood, H. L. 1963. Differences in muscle fiber properties as a factor in "fast" and "slow" contractions in Carcinus. Comp. Biochem. Physiol. 19:17—32. Atwood, H. L. 1965. Excitation and inhibition in crab muscle fibers. Comp. Biochem. Physiol. 16:409-426. Atwood, H. L. 1967. Crustacean neuromuscular mechanisms. Amer. 2001. 1:527-521. Atwood, H. L. and B. S. Dorai Raj. 1964. Tension develop- ment and membrane responses in phasic and tonic muscle fibers of a crab. J. Cell. Comp. Physiol. 64:55-72. I Atwood, H. L. and H. S. Johnston. 1968. Neuromuscular synapses of a crab motor axon. J. Exp. Zool. 167:457—470. Atwood, H. L. and W. A. Morin. 1970. Neuromuscular and axo- axonal synapses of the crayfish opener muscle. J. Ultrastruc. Res. 32:351-369. Atwood, H. L., G. Hoyle and T. Smyth. 1965. Mechanical and electrical responses of single innervated crab muscle fibers. J. Physiol. 180:449-482. Atwood, H. L., T. Smyth and H. S. JOhnston. 1969. Neuro- muscular synapses in the cockroach extensor tibiae muscle. J. Insect Physiol. 15:529-535. Atwood, H. L., A. R. Luff, W. A. Morin and R G. Sherman. 1971. Distribution of synaptic vesicles in crustacean neuromuscular junctions. Exper. (in press). Barrantes, F. J. 1970. The neuromuscular junctions of a pulmonate mollusc. I. Ultrastructural study. Ziet. Zell. Mik. Anat. 104:205-213. Becht, G. and Dresden. 1956. Physiology of the locomotory muscles in the cockroach. Nat. (London). 177:836-837. 119 120 Becht, G., G. Hoyle and P. N. R. Usherwood. 1960. Neuromuscular transmission in the coxal muscles of the cockroach. J. Insect. Physiol. 4:191-201. Bittner, G. D. and D. Kennedy. 1970. Quantitative aspects of transmitter release. J. Cell Biol. 41:585-592. del Castillo, J., G. Hoyle and X. Machne. 1953. Neuro- muscular transmission in a locust. J. Physiol. 121:539—547. Cohen, M. J. and A. Hess. 1967. Fine structural differences in "fast” and "slow" muscle fibers of the crab. Amer. J. Anat. 121:285-304. Dorai Raj, B. S. 1964. Diversity of crab muscle fibers innervated by a single motor axon. J. Cell. Comp. Physiol. 64:41-54. Dumont, J. N., E. Anderson and E. Chomyn. 1965. The anatomy of the peripheral nerve and its ensheathing artery in the horseshoe crab Xiphosura (Limulus) polyphemus. J. Ultrastruc. Res. 13:38-64. Eccles, J. C. and W. J. O'Connor. 1939. Responses which nerve impulses evoke in mammalian striated muscle. J. Physiol. 21:44-102. Fahrenbach, W. H. 1967. The fine structure of fast and slow crustacean muscles. J. Cell Biol. 35:69-79. Fahrenbach, W. H. 1970. The cyanoblast: Hemocyanin forman tion in Limulus polyphemus. J. Cell Biol. 44:445—453. Fatt, P. and B. Katz. 1953. The electrical properties of crustacean muscle fibers. J. Physiol. 120:171-204. Gilai, A. and I. Parnas. 1970. Neuromuscular physiology of the closer muscle in the pedipalp of the scorpion, Leiurus guinquestriatus. J. Exp. Biol. 52:325-344. Gray, P. 1964. "Handbook of Basic Microtechnique." McGraw Hill, New York. 302 pp. van Harreveld, A. 1939. The nerve supply of doubly and triply innervated crayfish muscles related to their function. J. Comp. Neurol. 10:267-284. Hayes, W. F. and S. B. Barber. 1967. Proprioceptor distribu- tion and properties in Limulus walking legs. J. Exp. Zool. 165:195-210. 121 Hoyle, G. 1955. Neuromuscular mechanisms of a locust skeletal muscle. Proc. Roy. Soc. Lond. Series B. 143:343-367. Hoyle, G. 1958. Studies on neuromuscular transmission in Limulus. Biol. Bull. Woods Hole. 115:209-218. Hoyle, G. 1965a. Neural control of skeletal muscle. In: "Physiology of Insecta," Ed. M. Rockstein. Vol. II:408-449. Academic Press, New York. Hoyle, G. 1965b. Nature of the excitatory sarcoplasmic reticular junction. Science. 149:70-72. Hoyle, G. 1969. Comparative aspects of muscle. Ann. Rev. Physiol. 31:43-84. Hoyle, G. and P. A McNeill. 1968. Correlated physio- logical and ultrastructural studies on specialized muscles. Ib. Ultrastructure of the white and pink fibers of the levator of the eyestalk of Podophthalmus vigil (Weber). J. Exp. Zool. 161:487-521. Hoyle, G. and T. Smyth. 1963. Neuromuscular physiology of giant muscle fibers of a barnacle, Balanug nubilus Darwin. Comp. Biochem. Physiol. 10:291—314. Hoyle, G. and C. A. G.‘Wiersma. 1958. Excitation at neuro- muscular junctions in crustacea. J. Physiol. 143:403-425. ’ Ikemoto, N. and S. Kawaguti. 1968. Electron microscopy on the organization of myofilaments in the retractor muscle of the horseshoe crab. Biol. J. Okayama Univ. 14:21—33. Jahromi, S. S. and H. L. Atwood. 1967. Ultrastructural features of crayfish phasic and tonic muscle fibers. Can. J. 2001. 45:601-606. Jahromi, S. S. and H. L. Atwood. 1969. Structural features of muscle fibers in the cockroach leg. J. Insect Physiol. 15:2255—2262. Jordan, H. E. 1917. The microscopic structure of striped muscle of Limulus. Carnegie Inst. Wash. Publ. 251:275-289. Josephson, R. K. 1961. Repetitive potentials following brief electrical stimuli in a hydroid. J. Exp. Biol. 38:579-594. 122 Katz, B. 1966. "Nerve, Muscle and Synapse." McGraw-Hill, New York. 193 pp. Kelly, D. E. 1969. The fine structure of skeletal muscle triad junctions. J. Ultrastruc. Res. 22:37—49. Kennedy, D. 1966. The comparative physiology of inverte— brate central neurons. Adv. Comp. Biochem. Physiol. 2:117-184. Lankester, E. R. 1885. On the muscular and endoskeletal systems of Limulus and Scorpio; with some notes on the anatomy and generic characteristics of scorpions. Trans. 2001. Soc. Lond. .112311-384. Lowenstein, W. R. 1966. Permeability of membrane junctions. Ann. N. Y. Acad. Sci. 137:441—472. Lucas, K. 1917. On summation of prOpagated disturbances in the claw of Astacus, and on the double neuromuscular system of the adductor. J. Physiol. 51:1-35. Millecchia, R. and A. Mauro. 1969. The ventral photore- ceptor cells of Limulus. II. The basic photoresponse. J. Gen. Physiol. 54:310-330. Moore, R. C., C. G. Lalicker and A. G. Fisher. 1952. "Invertebrate Fossils." McGraw-Hill, New York. 766 pp. Parnas, I., B. C. Abbott, B. Shapiro and F. Lang. 1968. Neuromuscular system of Limulus leg closer muscle. Comp. Biochem. Physiol. 26:467-478. Patten, W. and W. A. Redenbaugh. 1899. Studies on Limulus. II. The nervous system of Limulus polyphemu§_with observations upon the general anatomy. J. Mbrph. 16:91-190. Peachy, L. D. 1970. Form of the sarc0plasmic reticulum and T system of striated muscle. In: "Physiology and Biochemistry of Muscle as a Food," Eds. E. J. Briskey, R. G. Cassens and B. B. Marsh. Vol. 2:273-310. Pearson, K. G. and J. F. Iles. 1971. Innervation of the coxal depressor muscles in the cockroach, Periplaneta americana. J. Exp. Biol. 54:215-232. Peterson, R. P. and F. A. Pepe. 1965. The relationship of the sarcoplasmic reticulum to the sarcolemma in crayfish stretch receptor muscle. Amer. J. Anat. 109:277—297. 123 Pringle, J. W. S. 1956. Proprioception in Limulus. J. Exp. Biol. 33:658-667. Rathmayer, W. 1965. Neuromuscular transmission in a spider and the effect of Calcium. Comp. Biochem. Physiol. 14:673—687. Rathmayer, W. 1967. Elektrophysiology untersuchung der membraneigenschaften eines spinnenmuskels. Verh. Dtsch. Zool. Ges. Heidelberg. Zool. Anz. Supplement, i1: 616-622 . Reynolds, E. S. 1963. The use of lead citrate at high pH as an electron—opaque stain in electron microsc0py. J. Cell Biol. 11:208—212. Sherman, R. G. and H. L. Atwood. 1971. Structure and neuro— physiology of a newly discovered muscle in the walking legs of the lobster, Homarus americanus. J. Exp. 2001. 116:461-474. Simpson, G. G. 1953. "The Meaning of Evolution." Yale Press, New Haven. 364 pp. Smith, D. S. 1960. Innervation of the fibrillar flight muscle of an insect: Tenebrio molitor (Coleoptera). J. Biophy. Biochem. Cytol. 85447-466. Smith, D. S. 1968. "Insect Cells. Their Structure and Function." Oliver and Boyd, Edinburgh. 372 pp. Spector, W. S. 1956. "Handbook of Biological Data." Wright Air Development Center, Wright-Patterson Air Force Base, Ohio. 584 pp. Toselli, P. A. and F. A. Pepe. 1968. The fine structure of the ventral intersegmental abdominal muscles of the insect Rhodnius prolixus during the molting cycle. I. Muscle structure at molting. J. Cell Biol. 31:445-461. Tulsi, R. S. and R. E. COggeshall. 1971. Neuromuscular junctions on the muscle cells in the central nervous system of the leech, Hirudo medicinalis. J. Comp. Neurol. 14151-16. ’ Usherwood, P. N. R. 1962. The nature of "slow" and "fast" contractions in the coxal muscles of the cockroach. J. Insect Physiol. 8:31-52. Usherwood, P. N. R. 1967. Insect neuromuscular mechanisms. Amer. 2001. 1:553—582. 124 Usherwood, P. N. R. 1969. Electrochemistry of insect muscle. Adv. Insect Physiol. 6:205-278. Usherwood, P. N. R. and H. Grundfest. 1965. Peripheral inhibition in skeletal muscle of insects. J. Neuro- physiol. ‘2§:497—518. Vachon, M. 1945. Remarques sur 1es appendices du prosoma des Limulus et leur arthrogenese. Arch. 2001. Exp. Gen. 84:271-300. de Villafranca, G. W. 1961. The A and I hand lengths in stretched or contracted horseshoe crab skeletal muscle. J. Ultrastruc. Res. 5:109-115. de Villafranca, G. W. and D. W. Philpott. 1961. The ultra-. structure of striated muscle from Limulus polyphemus. J. Ultrastruc. Res. [5:151-165. Ward, D. V. 1969. Leg extension in Limulus. Biol. Bull. Woods Hole. 136:288-300. Wiersma, C. A. G. 1941. The efferent innervation of muscle. Biol. Symp. 3:259-289. "The Physiology of Crustacea." Ed. T. H. Waterman Vol. ;;:191-240. ' Wiersma, C. A. G. 1961. The neuromuscular system. I : Wilson, V. J. 1954. Slow and fast responses in cockroach eleg muscle. J. Exp. Biol. 31:280-290. Zebe, E. and W. Rathmayer. 1968. Elektronenmikroskopische untersuchungen an spinnenmuskeln. Zeit. Zell. Mik. Anat. 22:377-387. TATE UNIV R ”'Tl’r'lflijlmflu 1 mi (WWII? 293 03056