Regional differentiation of neuromuscular synapses in a lobster receptor muscle

The accessory flexor muscle in the walking legs of decapod crustaceans was des­cribed by Barth (1934) as part of a sense organ called the myochordotonal organ. The proprioceptive function of this receptor muscle, with its associated elastic strands and afferent and efferent innervation, was demonstrated by Cohen (1963 a, b) in the crab, Cancer magister. The muscle in Cancer is also interesting because its fibres are differentiated in a number of properties, such as sarcomere length, dia­meter, intracellularly recorded excitatory postsynaptic potential, membrane electrical properties and rate of tension development (Atwood & Dorai Raj, 1964; Dorai Raj, 1964). Previous studies, however, have not examined the quantal output of trans­mitter at the diverse synapses formed by the single excitor axon, nor have they correlated this synaptic diversity with the postsynaptic properties of the muscle fibre. We have therefore correlated the synaptic properties, including quantal release of transmitter, with the electrical and structural properties of the muscle fibres of the distal head of the accessory flexor muscle of the lobster Homarus americamis. We find that the single excitor axon gives rise to low- and high-output synapses that are regionally distributed. As the muscle fibres are largely similar in their electrical properties the synaptic diversity is due to specializations of the presynaptic elements only.

Adult lobsters, Homarus americanus (weight 450-680 g), were bought locally and held at 10-12 °C in aerated Instant Ocean (Aquarium Systems Inc., Eastlake, Ohio) sea water. The first and second walking legs were used in all the experiments. The distal head of the accessory flexor muscle (together with the leg nerve) was exposed in the meropodite by removing the main flexor and extensor muscles. The preparation was kept in physiological saline (NaCl, 472 mM; KC1, 10 mM; MgClj.6H₂0, 7 mM; CaClg, 16 mM; glucose, 11 mM; Tris-maleate, 10 mM; pH 7-4) and maintained at 12-15 °C by means of a cooling coil in the experimental chamber. Fresh chilled aerated saline was continuously perfused into the experimental chamber at a rate of about 200 ml/h.

The bundle of the leg nerve containing the axon innervating the accessory flexor muscle was stimulated with rectangular voltage pulses of 0.02-0.08 ms duration, delivered through thin silver-wire electrodes, insulated except at the tip. Excitatory post-synaptic potentials (EPSPs) were recorded intracellularly with glass micro­electrodes with tip resistances of 10-20 ΜΩ and filled with a saturated solution of methyl blue in 1 M potassium acetate solution. This permitted us to mark individual muscle fibres by iontophoresing the methyl blue into them (Thomas & Wilson, 1966) for subsequent examination of their sarcomere lengths. EPSPs were recorded at 1 Hz and 10 Hz stimulation. An index of facilitation F_e (Atwood & Bittner, 1971) was obtained by taking the ratio of EPSP amplitude at 10 Hz to that at 1 Hz and was used to compare synaptic properties of the different muscle fibres.

The quantal content of synaptic transmission was estimated from extracellularly recorded synaptic potentials (ERSPs) which were made with glass microelectrodes with tip resistances of 2-5 ΜΩ and filled with 2-3 m NaCl. Techniques described elsewhere (Dudel & Kuffler, 1961; Atwood & Johnston, 1968) were used in making these records, which served to relate the properties of the individual synaptic regions to the EPSPs.

Electrical membrane properties of individual muscle fibres were determined using the methods of square-pulse analysis (Hodgkin & Rushton, 1946; Fatt & Katz, 1953). Current pulses of 100-500 ms duration were passed through one micro­electrode inserted in the middle of the fibre while a second microelectrode was used to record the membrane voltage response at two or more locations along the fibre. All fibres behaved as ‘infinite cables’ with length constants (λ) equal to about one- tenth of the fibre length (Stefani & Steinbach, 1969, fig. 3). Equations for the ‘infinite cable’ model (Katz, 1948; Fatt & Katz, 1951; Weidmann, 1952) were used to deter­mine the membrane cable properties.

Sarcomere lengths (SL) were measured after fixing the entire muscle at rest for 48 h or more in Bouin’s fluid. Prior to fixation, the muscle was placed in an isotonic 7'5% solution of MgCl₂.6H₂O, diluted with an equal volume of perfusion fluid to relax the muscle and prevent local contractures. The fibre was teased into myofibrils on a glass slide in 70% alcohol and three consecutive sarcomeres were measured in three separate myofibrils. These nine measurements gave the mean sarcomere length of the fibre.

Muscle innervation was determined in two ways. First, the dissected preparation was left overnight at 5 °C in 1% methylene blue dissolved in perfusion fluid. Exami­nation the following morning usually revealed well stained axon branches which could be traced to their terminations in the muscle. Secondly, axonal iontophoresis of cobalt (Peters, 1976) was also used to reveal the innervation pattern. The cut nerve stump was drawn into a 0 5 M CoCl₂ solution which was allowed to diffuse down the nerve fibre for 6 h at 10 °C. After that, the cobalt was precipitated with ammonium sulphide (0-5% in saline) for 15.n1. Fixation was then in Bouin’s fluid or in 1% glutaraldehyde followed by dehydration in ethanol. Muscles were then dissected out and embedded in glycerine as whole mounts for examination. Peri­pheral nerves could be traced to their terminations in only a few favourable preparations. Both methods provided reliable results which were consistent with each other.

The gross morphology of the distal head of the accessory flexor muscle in Homarus is described elsewhere (Govind et al. 1978) and features pertinent to this study are summarized in Fig. 1. The single excitor axon (Wiersma & Ripley, 1952) gives rise to primary axonal branches (Fig. 1) whose number and distribution is constant from fcnimal to animal as revealed by methylene blue and cobalt chloride staining. A single primary branch innervated fibres in more than one of the five bundles con­stituting this muscle. The synapses formed by the single excitor axon are regionally differentiated across the surface of the muscle and the differentiation is particularly evident between synapses on fibres located at the proximal and the distal edges of the muscle (Fig. 1). This regional differentiation of synapses is described below and correlated with membrane electrical properties and sarcomere lengths.

Considerable variation in EPSP size and F_e values was found amongst fibres from different regions of the muscle. Maximal differences were found between fibres situated at the proximal and distal edge of the muscle. We have examined these in detail as they exemplify the extremes in synaptic types in the accessory flexor muscle. Table 1 compares the properties of proximal and distal fibres.

In the proximal fibres, EPSPs at 1 Hz were small, usually 0-2-4 *ⁿ amplitude but at 10 Hz these facilitated 3-8 times (Fig. 2, insert). By contrast, EPSPs from distal fibres were larger in amplitude ranging between 4-32 mV at x Hz, but did not facilitate appreciably at 10 Hz. Stimulation at higher frequencies often produced summation which resulted in a maintained depolarization of the membrane. Proximal fibres showed slightly lower levels of maintained depolarization than distal ones. Stimulation at frequencies greater than 40 Hz rarely produced levels of depolarization exceeding 35 mV in any fibre. Also at maximum levels of depolarization proximal and distal fibres never gave any form of active membrane response.

The relationship between EPSP amplitude and F_e was further examined by correlating these parameters by the least-squares regression analysis. An inverse correlation was found between EPSP amplitude and F_e for both proximal and distal fibres (Fig. 2). A hyberbolic function of the form Y = A + (B/X), where Y = EPSP amplitude and X = F_e provided the highest ‘index of determination’ (Snedecor & Cochran, 1967) among the various standard equations which were tested. A similar correlation has been made in other crustacean muscles including the limb muscles in crabs and crayfish (Atwood & Bittner, 1971; Sherman & Atwood, 1972; Sherman, 1977) and the stomach muscles of a crab (Govind, Atwood & Maynard, 1975).

Time constants of decay of EPSPs were similar for proximal and distal fibres (Table 1). Comparison of the mean EPSP rise times between proximal and distal fibres, however, showed significant differences. The reasons for this difference are unclear but may be related to differences in density of innervation. Thus the rise time of an EPSP will be influenced by the proximity of the recording electrode to an active synaptic site. This may explain the fact that very fast (4 ms) and very slow (7-5-8 ms) rise times were occasionally recorded in both proximal and distal fibres.

To compare the quantal output of transmitter between proximal and distal fibres, we made focal extracellular recordings at individual synaptic regions. This method of recording provides an estimate of transmitter release at single regions (Dudel & Kuffler, 1961; Atwood & Johnston, 1968). At low frequencies of stimulation (1 Hz or less) the externally recorded synaptic potentials (ERSPs) were usually small in the proximal fibres and large in the distal ones. An example of typical ERSP and EPSP records from proximal and distal fibres is given in the insert of Fig. 3. However, there is no consistent relationship between EPSP and ERSP amplitudes. For example, the largest EPSP in a series of evoked potentials often occurred simultaneous with a failure of the ERSP. Thus, the ERSP amplitude is not an accurate indicator of differences in transmitter output. This conclusion is consistent with earlier results on vertebrate (del Castillo & Katz, 1954) and invertebrate (Dudel & Kuffler, 1961; Bittner & Harrison, 1970) neuromuscular junctions.

Differences in quantal transmitter output between synapses on proximal and distal fibres were indicated by the number of transmission failures at 1 Hz, which were common in proximal but rare in distal fibres. The quantum content as estimated from the number of failures at 1 Hz (del Castillo & Katz, 1954) ranged from 1-20 to 5-61 (mean 2-9) in 18 synapses on 10 distal fibres and from 0-03 to 0-29 (mean 0-16) from 16 synapses on 10 proximal fibres. These values compare well with crayfish (Dudel & Kuffler, 1961; Bittner, 1968a; Bittner & Harrison, 1970; Bittner & Kennedy, 1970) and crab (Sherman & Atwood, 1972; Atwood, Govind & Jahromi, 1977) neuromuscular synapses.

Records of ERSPs from two or more sites along a single muscle fibre in 13 cases confirmed that distal fibres were innervated solely by high output synapses and proximal ones by low output synapses. These results suggest that individual synapti^ foci on any single fibre are similar to each other; a similar situation prevails in the proximal accessory flexor muscle of lobsters (Frank, 1973).

Quantal content of transmission was also estimated for a few synaptic foci by recording a number of ERSPs at each focus, measuring their amplitude, and dividing the mean ERSP amplitude by the mean amplitude of the spontaneous ERSPs ob­served during the stimulus train. Estimates of quantal content obtained in this manner were similar to those obtained for the same synaptic foci via the number of transmission failures. However, spontaneous ERSPs were too few in number to allow a detailed comparison of quanta! output using this method.

The relationship between quantal size and EPSP amplitude is graphically illus­trated in Fig. 3. The general result obtained from this analysis is that large amplitude EPSPs are always recorded from fibres which are preferentially innervated by high quantal output synapses. Thus, the intracellularly recorded synaptic potential (i.e. the EPSP) is an accurate reflexion of the type of synapse innervating that fibre.

Thus in the lobster accessory flexor muscle, synapses on the proximal fibres are low-output, highly facilitating types while those in the distal fibres are high-output, poorly facilitating types. The above observations on quantal content suggest that the differences in EPSP amplitudes are based on quantitative differences in quantal transmitter release of individual synapses.

 Fig. 4 illustrates a pair of curves obtained from current passing experiments on 22 proximal and 20 distal muscle fibres. The graphs show that the proximal fibres are characterized by a slightly non-linear current-voltage plot while the distal fibres have a more linear relationship. The slope of the current-voltage relationship is significantly greater at the 1% level for the distal fibres as judged by the t-test indi­cating a higher input resistance, R_in, for these fibres.

Rectification was minimal at low levels of hyperpolarization and input resistances were calculated using the voltages obtained by passing inward current of 100 nA across the membrane. Current-voltage data from an additional 15 distal and 13 proximal fibres were pooled with the data from Fig. 4 and are summarized along with other membrane electrical properties in Table 2. The differences between the input resistances of proximal and distal muscle fibres are significant at the 1% level for the grouped data as judged by the f-test. These differences were also significant when the two fibre populations were compared in individual muscles.

Comparison of the input resistance with EPSP amplitudes and F_e ratios are given in Figs. 5 and 6. Input resistance is significantly correlated with both EPSP amplitude and F_e, although there was more than one best fit curve for the data. However, when compared with Fig. 2, it is clear that a closer relationship exists between F_e and EPSP amplitude than between R_in and EPSP amplitude or F_t. However, since the latter correlations are statistically significant, this indicates that R_in may account for some of the relationship between EPSP amplitude and F_e. Therefore, a multiple regression analysis was performed to determine if the relationship between R_in and EPSP amplitude was weaker than that between EPSP and F_t as reported in Sherman & Atwood (1972), and Atwood & Bittner (1971) for other crustacean limb muscles. As shown in Table 3, the ‘correlation coefficient’, obtained for the curve that best fits the observed data for EPSP amplitude and F_e was not improved by the addition of R_in values to the analysis.

Therefore, even though R_in is significantly correlated with EPSP amplitude, the observed diversity in EPSP amplitudes cannot be attributed to differences in R_ln of the innervated fibres. In the present study, the observed diversity in synaptic types is related to differences in transmitter output as reflected by the quantal analysis and by the F_e. This conclusion was further confirmed by using partial regression methods.

Table 2 shows that proximal and distal muscle fibres do not differ significantly from each other with respect to several other membrane cable properties, excepting the specific internal resistances, R_v However, this difference became insignificant when the mean values within individual muscles were compared. The values for R_in and R„ compare well with those found in the stretcher muscle in lobsters by Werman & Grundfest (1961), who reported values of 2-0-40 x ίο³ Ω for Rm-

Since the membrane electrical properties are similar for proximal and distal fibres (Table 2) in the lobster distal accessory flexor muscle it is unlikely that these properties would be responsible for the synaptic differences between these two fibre types (Table 1). To test this we correlated the synaptic properties (EPSP and F_e) with the membrane electrical properties of R_m, T_m and C_m by a least-squares regression analysis. There is no significant correlation between these parameters. A similar conclusion was reached in the crayfish opener muscle (Bittner, 1968a). Therefore, synaptic differences between proximal and distal fibres in the lobster accessory flexor muscle reflect differences in transmitter output from nerve terminals on these fibre types.

To determine if muscle fibre structure is correlated with synaptic and membrane electrical properties we measured the sarcomere length (SL) of fibres in which these properties were already characterized. For these identified fibres the relation­ship between SL and F_e and SL and EPSP is graphically illustrated in Fig. 7. In these graphs the proximal and distal fibres form two separate populations because of a significant difference in SL between them (see also Govind et al. 1978). Regression analyses show a significant correlation between SL and F_e and SL and EPSP, for both populations with several functions showing best-fit correlations at a level of significance of 0-05 or better. The present sample does not allow us to choose between the alternatives. The graphs do, however, clearly illustrate that fibres with short sarcomeres are innervated by low-output, highly facilitating synapses whereas fibres with long sarcomeres are innervated bv high output, poorly facilitating synapses.

A plot of SL v. Rin (Fig. 8) failed to demonstrate significant correlations for either fibre population. This may be due to the small sample size (15 distal, 13 proximal fibres). The data does however show that fibres with long sarcomeres may have high and low Rm values, but short sarcomere fibres have only low Rm values. A similar relationship between SL and Rm has been shown for the crayfish opener (Bittner, 1968a) and Hyas stretcher (Sherman & Atwood, 1972) muscles. Sarcomere length was not correlated with other membrane electrical properties including T_m and C_m. The above data demonstrates that there is a correlation of the muscle fibre structure to its synaptic and membrane electrical properties.

The excitatory innervation of the accessory flexor muscle, a receptor muscle forming part of the limb proprioceptive myochodotonal organ, has been studied in crab (Dorai Raj, 1964), lobster (Frank, 1973) and crayfish (Angaut-Petite, 1977). The results of these studies show that synaptic potentials evoked by activity in the single excitor axon show marked regional diversity in amplitude and facilitation properties. Muscle fibre properties of the distal head of the crab accessory flexor muscle were further characterized by Cohen (1963a), Atwood & Dorai Raj (1964) and Dorai Raj (1964), who showed that the muscle was comprised of short sarcomere (2.4 /tm), large diameter (500 /im), fast-acting (phasic) fibres along the distal edge and long sarcomere (10.12 μm), small-diameter (100 μm), slow-acting (tonic) fibres along the proximal edge. Re-examination of this muscle, however, showed that although long sarcomere (10.12 μm) fibres are found on the proximal edge no short sarcomere fibres were seen (Govind et al. 1978). Instead, distal fibres had sarcomeres ranging from 6.8 μτη in length. Accordingly, all muscle fibres are of the intermediate to slow variety, and certainly not of the fast variety.

The results of the present study on the lobster distal accessory flexor muscle are consistent with the earlier results described above but go further in looking at the transmitter release properties and the membrane electrical and structural properties of the muscle fibres. In contrast to the crab distal accessory flexor muscle, we find that all muscle fibres are of uniform diameter. Furthermore distal fibres showed longer sarcomeres (mean 9.20 μm) than the proximal fibres (mean 7.14 μm).

Differences in input resistance, do exist between the distal and proximal fibres, 150-3 ± 42.2 × 10^s Ω s.d. and 112.4 ± 45.0 × 10^s Ω s.d. respectively. However, this difference cannot account for the 10- to 6o-fold differences in the EPSP ampli­tudes evoked by single stimuli. Large EPSP’s were often evoked in fibres with low Rm while small EPSPs were found in fibres with high R_in. Thus, it appears that EPSP amplitude is independent of Rm.

The apparent lack of correlation between EPSP amplitude and R_in could result from problems inherent in applying classical cable theory to crustacean muscle fibres (see Bittner, 1968b, for a full discussion). Thus, crustacean muscle fibres may not behave as true ‘cables’ due to any of several factors including: (1) inaccurate estimates of surface area due to the extensive sarcolemmal invagination; (2) possible large variations in Rm and R_t from fibre to fibre; (3) equivalent circuits for crustacean muscle fibres may be quite different from that assumed for vertebrate muscle on the basis of cable theory. In the absence of a rigorous test of the applicability of cable theory to crustacean muscle fibres, we can only assume for the present that crustacean muscle fibres behave like their vertebrate counterparts.

Differences in EPSP amplitudes and F_e ratios are instead related to differences in transmitter release properties. This is supported by several observations. First, the amount of transmitter released and degree of facilitation measured by focal extra­cellular recordings at individual synapses reveal marked differences between proximal and distal fibres. Thus, the lower quantal content of synapses on the proximal fibres is indicated by the higher number of transmission failures. Secondly, the synaptic diversity reflected in the correlation of EPSP amplitudes and the F_e ratios was not improved by the addition of R_in to the analysis in spite of the differences in R_in between proximal and distal fibres. Thirdly, since the muscle fibres are similar with respect to other membrane electrical properties, these cannot account for the large difference in EPSP amplitude between proximal and distal fibres. Instead these difference reflect variation in transmitter release between synapses on proximal and distal fibres.

Furthermore, this regional synaptic heterogeneity is correlated with differentiation of muscle fibre structure as reflected in sarcomere length. Thus, distal fibres which have long sarcomeres have high output, poorly facilitating, synapses and proximal ones which have shorter sarcomeres possess low output, highly facilitating, synapses. Thus the muscle fibre structure is selectively correlated to its synaptic properties. This selective correlation of nerve terminals to muscle fibres is identical to that established for other crustacean limb muscles which receive a single excitor axon (Atwood, 1965; Bittner, 1968a,b; Atwood & Bittner, 1971; Sherman & Atwood, 1972; Sherman, 1977). In all cases, it has been found that the more phasic (short sarcomere) fibres receive terminals which generate the smaller, more highly facilitating EPSPs. This correlation between nerve input and muscle fibre properties focuses attention on the interrelationship between nerve and muscle and the degree to which each influences the type of synapse formed.

Presently there are two hypotheses to account for the correlation of nerve terminals and muscle fibres. A neural hypothesis has been favoured for the differentiation of neuro-muscular systems in crustaceans, based largely on the correlative effect seen in singly innervated adult muscles (Atwood, 1973) and on a deterministic mechanism for synaptic differentiation in a regenerating crab-limb muscle (Govind, Atwood & Lang, 1973). In the latter case synaptic differentiation from a single excitor axon was correlated with the timing of innervation: neuromuscular contacts formed early in regeneration matured into high-release terminals while later contacts matured into low-release terminals. The time of innervation of muscle fibres is likely determined by the branching pattern of the axon. Thus it would appear that the muscle fibre has little influence on synaptic differentiation. However, it is likely that the innervating axon may influence muscle fibre development and in this way ensure the close match­ing between pre- and post-synaptic elements.

A myotypic hypothesis for synaptic differentiation was proposed for the lobster proximal accessory flexor muscle by Frank (1973). He showed that even though synapses varied between muscle fibres, all synapses along a single muscle fibre are uniform in their facilitation characteristics. This uniformity existed even in muscle fibres which received two or more separate primary axonal branches of the single excitor axon. He concluded that the muscle fibre governs the type of synapse that forma on it.

However, preliminary findings on the distal accessory flexor muscle show con­siderable variability in synaptic properties along the length of a single fibre (Meiss & Govind, 1979). Thus larger EPSPs and synapses with higher quantal content are found near the tendon end of the fibre while smaller EPSPs and synapses with lower quantal content occur at the opposite, or exoskeletal, end. This suggests a deter­ministic mechanism for the formation of neuromuscular synapses such as that due to the branching pattern of the axon. Indeed the excitor axon gives rise, at the tendon end, to large primary axonal branches which subsequently branch into smaller secondary and tertiary branches at the exoskeletal end. These findings suggest a correlation between the axonal branching pattern and properties of synapses.

Whatever mechanism exists for the differentiation of crustacean neuromuscular synapses the present study shows that the diversity of synapses from a single motor axon on fibres of the lobster distal accessory flexor muscle resides largely in differfentiation of the presynaptic (axonal) elements. Thus the different levels of depolariz­ation amongst the fibres is due to the different amounts of transmitter released by their respective synapses. Moreover these diverse synapses are regionally distributed and are therefore easily identifiable from preparation to preparation. These obser­vations provide a unique opportunity to examine the ultrastructural basis of trans­mitter release particularly as one can examine both high- and low-output synapses arising from a single excitatory axon. Preliminary studies with thin serial section electron microscopy has revealed that high output synapses in the lobster proximal accessory flexor muscle have a relatively larger area and number of presynaptic dense bodies than their low output counterparts. Characterizing the structural correlates of transmitter release such as presynaptic dense bodies in adults will permit tracing synaptic differentiation during development; which is probably when synaptic types are established.

In addition, these results will contribute to an understanding of the centrifugal control exerted by the accessory flexor muscle on the myochordotonal organ sensory discharge. Differentiation of synaptic response and muscle fibre properties may allow for precise control of the tension gradation of the accessory flexor muscle via a differential recruitment of fibres as the level of neural activity changes as suggested by Angaut-Petite (1977). It appears likely that consideration of structural and contractile properties of the fibres in the accessory flexor muscle are important in understanding the proprioceptive role of the myochordotonal organ.

We thank Drs H. L. Atwood and Fred Lang for reading and offering critical suggestions of this manuscript and Eva Yap Chung and Jenny King for able technical assistance.

D.E.M. was supported by a postdoctoral fellowship from the Muscular dystrophy Association of Canada. Research grants from the National Research Council of Canada and the Muscular Dystrophy Association of Canada supported this study.