Search

1. The effects of four different synaptic input systems on the recruitment order within a mammalian motoneuron pool were investigated using computer simulations. The synaptic inputs and motor unit properties in the model were based as closely as possible on the available experimental data for the cat medial gastrocnemius pool and muscle. Monte Carlo techniques were employed to add random variance to the motor unit thresholds and forces and to sample the resulting recruitment orders. 2. The effects of the synaptic inputs on recruitment order depended on how they modified the range of recruitment thresholds established by differences in the intrinsic current thresholds of the motoneurons. Application of a uniform synaptic input to the pool (i.e., distributed equally to all motoneurons) resulted in a recruitment sequence that was quite stable even with the addition of large amounts of random variance. With 50% added random variance, the recruitment reversals did not exceed 8%. 3. The simulated monosynaptic input from homonymous Ia afferent fibers generated a twofold expansion of the range of recruitment thresholds beyond that attributed to the differences in the intrinsic current thresholds. The Ia input generated a small reduction in the number of recruitment reversals due to random variance (6% reversals at 50% random variance). The simulated monosynaptic vestibulospinal input generated a twofold compression of the range of recruitment thresholds that exerted a modest increase in the number of recruitment reversals (12% reversals at 50% random variance). 4. In comparison with the modest effects of the two monosynaptic inputs, the simulated oligosynpatic rubrospinal excitatory input exerted a nine-fold compression in the recruitment threshold range that resulted in a recruitment sequence that was highly sensitive to random variance. With 50% added random variance, the sequence became nearly random (40% reversals). 5. Reciprocal Ia inhibition was simulated by a uniform distribution within the pool, but its effects on recruitment order were highly dependent on the distribution of the excitatory input. Reciprocal inhibition exerted only minor effects on recruitment order when combined with the Ia or vestibulospinal inputs. However, when the excitatory drive was supplied by the rubrospinal input, even small amounts of reciprocal inhibition were sufficient to completely reverse the normal recruitment sequence. 6. The simulated monosynaptic Ia input was highly effective in compensating for the disruptive effects of rubrospinal excitation on recruitment order. Even a small Ia bias combined with the rubrospinal excitation was sufficient to halve the effects of random variance and to restore the normal recruitment sequence in the presence of rather large amounts of reciprocal inhibition.(ABSTRACT TRUNCATED AT 400 WORDS)

1. We evoked steady-state synaptic potentials in triceps surae motoneurons of the cat by stimulating the hindlimb projection area of the contralateral magnocellular red nucleus at 200 Hz. We measured the effective synaptic currents (IN) underlying the synaptic potentials using a modified voltage-clamp technique. We also determined the effect of the rubrospinal input on the discharge rate of some of the motoneurons by inducing repetitive discharge with long injected current pulses during which the red nucleus stimulation was repeated. 2. At motoneuron resting potential, the distribution of IN from the red nucleus within the triceps surae pools was qualitatively similar to the distribution of synaptic potentials: 86% of the putative type F motoneurons received a net depolarizing IN from the red nucleus stimulation, whereas only 38% of the putative type S units did so. The mean values of IN were significantly different in the two groups [+4.1 +/- 5.0 nA (SD) for putative type F and -1.6 +/- 3.1 nA for putative type S]. 3. However, when the values of IN at threshold for repetitive firing were estimated, the distribution of IN from the red nucleus was quite different. At threshold, all of the putative type S units received hyperpolarizing IN but so did nearly half of the putative type F units. 4. As would be expected from the wide range of IN at threshold (-20 to +12 nA), the red nucleus input produced dramatically different effects on the discharge of different motoneurons.(ABSTRACT TRUNCATED AT 250 WORDS)

1. The tension produced by the combined stimulation of two to four single motor units of the cat tibialis posterior muscle was compared with the algebraic sum of the tensions produced by each individual motor unit. Comparisons were made under isometric conditions and during imposed changes in muscle length. 2. Under isometric conditions, the tension resulting from combined stimulation of units displayed marked nonlinear summation, as previously reported in other cat hindlimb muscles. On average, the measured tension was approximately 20% greater than the algebraic sum of the individual unit tensions. However, small trapezoidal movements imposed on the muscle during stimulation significantly reduced the degree of nonlinear summation both during and after the movement. This effect was seen with imposed movements as small as 50 microns. 3. The degree of nonlinear summation was not dependent on motor unit size or on stimulus frequency. The effect was also unrelated to tendon compliance because the degree of nonlinear summation of motor unit forces was unaffected by the inclusion of different amounts of the external tendon between the muscle and the force transducer. 4. Our results support previous suggestions that the force measured when individual motor units are stimulated under isometric conditions is reduced by friction between the active muscle fibers and adjacent passive fibers. These frictional effects are likely to originate in the connective tissue matrix connecting adjacent muscle fibers. However, because these effects are virtually eliminated by small movements, linear summation of motor unit tensions should occur at low force levels under nonisometric conditions.(ABSTRACT TRUNCATED AT 250 WORDS)

1. Steady-state inhibitory postsynaptic potentials (IPSPs) were evoked in tibialis anterior and extensor digitorum longus motoneurons of the cat by using tendon vibration to activate Ia-afferent fibers from the antagonist medial gastrocnemius muscle. 2. The effective synaptic currents (IN) underlying the steady-state IPSPs were measured by the use of a modified voltage-clamp technique. The amplitudes of the effective synaptic currents (1.62 +/- 0.66 nA, mean +/- SD; n = 20) extended over a fivefold range (0.5-2.7 nA) but were not correlated with the intrinsic properties of the motoneurons or with putative motor unit type. 3. We calculated the synaptic conductance (GS) underlying the steady-state Ia IPSPs from measurements of motoneuron input conductance during the activation of the Ia synaptic input. As was expected from Ohm's law, the Ia-inhibitory GS and IN were correlated (r = 0.49; P less than 0.05). Like IN, GS (175 +/- 202 nS, mean +/- SD; n = 20) was not correlated with the intrinsic properties of the motoneurons. 4. As has been reported previously for transient Ia IPSPs, the amplitudes of the steady-state IPSPs were correlated with motoneuron input resistance (r = 0.74; P less than 0.001) and homonymous Ia excitatory postsynaptic synaptic potential (EPSP) amplitude (r = 0.72; P less than 0.001). 5. The amplitudes of the steady-state Ia IPSPs and the homonymous Ia EPSPs were plotted on logarithmic axes. The slope (0.59) was significantly less than 1, which indicates that the gradient of Ia inhibition across the motoneuron pool is less steep than that of Ia excitation.(ABSTRACT TRUNCATED AT 250 WORDS)

1. Tension-frequency relations were measured in single fast-twitch motor units of the cat flexor digitorum longus (FDL) muscle before and after stimulating each unit with a series of 10-s trains at 20 Hz. The 20-Hz conditioning stimulation produced a combination of potentiating and fatiguing effects, similar to those previously reported to follow higher frequency stimulation of single motor units of the cat and maximal voluntary contractions in man. 2. The conditioning stimulation left three types of after effects: 1) short-term potentiation, 2) a somewhat longer lasting depression of maximal tension, and 3) a delayed depression of low-frequency responses (low-frequency fatigue). 3. The immediate potentiating effect of the conditioning stimulation was most prominent in fatigue-resistant (FR) motor units, whereas depression of maximal tension and low-frequency fatigue were most prominent in fatigue-intermediate (FI) and highly fatigable (FF) motor units. 4. On the basis of our results and those of other investigators, we propose that potentiation, depression of maximal tension, and low-frequency fatigue are independent phenomena, acting at distinct points in the excitation-contraction coupling process. 5. Our results suggest that both potentiation and low-frequency fatigue can result from rather modest amounts of preceding activity. Thus large changes in muscle force production are not unique to maximal contractions but are likely to follow sustained, submaximal contractions as well.

1. A pool of 100 simulated motor units was constructed in which the steady-state neural and mechanical properties of the units were very closely matched to the available experimental data for the cat medial gastrocnemius motoneuron pool and muscle. The resulting neural network generated quantitative predictions of whole system input-output functions based on the single unit data. The results of the simulations were compared with experimental data on normal motor system behavior in humans and animals. 2. We considered only steady-state, isometric conditions. All motoneurons received equal proportions of the synaptic input, and no feedback loops were operative. Thus the intrinsic properties of the motor unit population alone determined the form of the system input-output function. Expressing the synaptic input in terms of effective synaptic current allowed the simulated motoneuron input-output functions to be specified by well-known firing rate-injected current relations. The motor unit forces were determined from standard motor unit force-frequency relations, and the system output at any input level was assumed to be the linear sum of the forces of the active motor units. 3. The steady-state input-output function of the simulated motoneuron pool had a roughly sigmoidal shape that was quite different from those derived from previous recruitment models, which did not incorporate frequency modulation. Frequency modulation in combination with the skewed distribution of thresholds (low values much more frequent than high) restricted upward curvature to low input levels, whereas frequency modulation alone was responsible for the final gradual approach to the maximum force output. 4. Sensitivity analyses were performed to assess the importance of several assumptions that were required to deal with gaps and uncertainties in the available experimental data. The shape of the input-output function was not critically dependent on any of these assumptions, including those specifying linear summation of inputs and outputs. 5. A key assumption of the model was that systematic variance in motor unit properties was much more important than random variance for determining the input-output function. Addition of random variance via Monte Carlo techniques showed that this assumption was correct. These results suggest that the output of a motoneuron pool should be quite tolerant of random variance in the distribution of synaptic inputs and yet substantially altered by any systematic differences, such as unequal distribution of inputs among different motor unit types.(ABSTRACT TRUNCATED AT 400 WORDS)

1. Steady-state recurrent (Renshaw) inhibitory postsynaptic potentials (RIPSPs) were evoked in cat triceps surae motoneurons by stimulating the heteronymous muscle nerve at 100 Hz after dorsal root section. The effective synaptic currents (i.e., the net synaptic current measured at the soma, IN) underlying these inhibitory potentials were measured with a modified voltage-clamp technique. 2. The average value of the effective synaptic currents measured in medial gastrocnemius (MG) motoneurons was 0.4 nA. There was no significant correlation between the IN measured in individual cells and motoneuron input resistance (RN), rheobase (IR), duration of the spike afterhyperpolarization (AHPt1/2), or putative motor-unit type, although the steady-state inhibitory post-synaptic potential (IPSP) amplitudes were correlated with all of these parameters. 3. Steady-state recurrent inhibition was accompanied by a small (3.5%, on average) decrease in the resting input resistance of the motoneurons. The small magnitude of this measured change supports the hypothesis of Burke et al. that the site of synaptic contact between Renshaw cells and motoneurons is somewhat distal to the cell soma. 4. The absence of a differential distribution of the effective synaptic currents generated by Renshaw cells within the MG pool does not support the idea that recurrent inhibition mediates a selective reduction of the firing of small, low-threshold motoneurons by large, high-threshold motoneurons. The small amplitude of the effective synaptic currents we measured suggests that the contribution of recurrent inhibition to the direct modulation of motoneuron firing rate is subtle and that it is perhaps principally involved in the fine control and smooth production of muscle force.

The responses of Golgi tendon organs to single motor unit contractions were studied to determine whether receptors located in the same muscle region respond to a common set of motor units. In each of five experiments we isolated a large fraction (25-65%) of the motor units of the cat tibialis posterior muscle and determined to which of the units each of several tendon organs was responsive. Each tendon organ was excited by from two to fifteen of the isolated motor units, including units which produced very small forces. However, there was a much greater probability for large force units to excite a given receptor than for small force units to do so. The number of motor units which produced either an 'unloading' or an 'off response' exceeded, on average, the number of motor units which excited the same tendon organ. The extent to which single motor units excited both of a pair of tendon organs was examined statistically in relation to the mutual proximity of the receptors within the muscle. It was found, on average, that the closer were two receptors, the greater was the number of motor units that excited both of them. These results suggest that despite the extensive territories of individual motor units, the spike trains of tendon organs may still encode information about localized muscle activity.

The results of this study and its precedents suggest that the tetrapartite classification scheme might have universal applicability to at least the muscle units of cat hindlimb muscles and perhaps any mammalian muscle in which fiber typing reveals the presence of FG, FI, FOG, and type SO fibers. A possible exception to this generalization involves a small (n = 18) but thoroughly examined sample of muscle units from the first deep lumbrical muscle of the cat's foot, which led Kernell et al. (30) to conclude that the FF, FR, and S classification scheme was not directly applicable to that muscle. However, histochemical fiber typing is not yet available for that muscle. Furthermore, more extensive sampling, use of a different stimulation regime in the fatigue test, and a more detailed analysis of the sag property might well reveal that the tetrapartite classification scheme is indeed appropriate for units of cat foot muscles.

Experiments were performed to determine whether the topographic relationships between muscle spindles and their surrounding extrafusal fibers are preserved in the pattern of homonymous, monosynaptic connections from Ia-afferents to motoneurons. The medial gastrocnemius (MG) muscle of adult cats was chosen as a model system because previous studies have shown that its muscle nerve divides into several branches, each of which innervates a distinct muscle compartment (20, 27, 28, 38), and that the Ia-afferent fibers innervating muscle spindles within a compartment are found in the same nerve branch (27, 28). Thus, we could make intracellular recordings from MG motoneurons, determine which intramuscular compartment they innervated, and then compare the synaptic input they received from Ia-afferents innervating the same compartment with that which they received from Ia-afferents innervating different compartments. Our results indicate that homonymous Ia-afferent input is "topographically weighted" within the MG motor nucleus such that afferents innervating a given intramuscular compartment exert relatively greater synaptic effects in motoneurons that project to the same compartment than in other homonymous motoneurons. The degree of topographic weighting was quite variable in the different experiments but appeared more prominently in experiments in which a high proportion of the motoneurons studied were characterized by high-rheobase values and low input resistances. This suggests that topographic factors may exert more influence on the distribution of Ia-afferent input to large motoneurons than to small motoneurons. In the DISCUSSION, the extent of topographic weighting within the homonymous motoneuron pool is compared with weighting across synergist motoneuron pools, and alternative models of topographic weighting are proposed and evaluated.

Single muscle spindle afferent and motor unit EMG spike trains have been recorded simultaneously during periods of spontaneous motor activity in triceps surae muscles of decerebrate cats. The approximate time course and magnitude of the motor unit contractions were extracted from the whole muscle force record by spike-triggered averaging, and the functional interactions between motor unit contractions and spindle discharge were assessed by cross-correlating their respective spike trains. We have found that both spindle group Ia and II afferents are responsive to the contractions of single motor units in the presence of spontaneous motor activity, being strongly coupled to the activity of some motor units and indifferent to the contractions of others. Moreover, the cross-correlation analysis revealed modulation of a single motor unit's discharge pattern by the input of a single Ia afferent.

The neuro-mechanical properties and speed-force relations of 81 cat tibialis posterior motor units have been studied. Statistically significant correlations were found between alpha-axonal conduction velocity (CV) and average twitch tension (log10 TwT; r = 0.459, p less than 0.001), between CV and twitch contraction time (CT; r = -0.395, p less than 0.001) and between CT and log10 TwT (r = -0.277, p less than 0.02). The present correlations for the intermediate-sized tibialis posterior are stronger than those previously reported for large muscles such as soleus, medial gastrocnemius, plantaris and flexor hallucis longus. However, they are considerably weaker than those reported for the much smaller lumbrical muscles of the cat's foot. These findings support the contention that the spinal mechanisms governing an orderly recruitment of motor units according to the size of their muscle units must be more complex for large than for small muscles, at least in the cat hindlimb.

The longitudinal distributions of both phrenic and medial gastrocnemius motoneurons were quantitatively studied in the cat spinal cord. Both populations of motoneurons were retrogradely labeled by applying horseradish peroxidase (HRP) to the cut central ends of the appropriate peripheral nerves. The longitudinal positions of all labeled motoneurons in each motor column were determined; these data then were used to generate longitudinal distribution histograms and spatial interval distributions (SIDs), the latter being analyzed further by means of power spectra. In three of four cats, longitudinal clustering of phrenic motoneuronal cell bodies was revealed by the presence of a narrow central peak in the SID and the presence of subsidiary peaks. In the fourth cat, only a smaller central peak was observed. Power spectral analysis of the three SIDs having subsidiary peaks revealed that the mean longitudinal distance between clusters was 0.95 mm (range 0.52 to 1.22 mm). The analyses also revealed that on average a phrenic motoneuronal cluster contained 17 motoneurons, and the mean longitudinal length of a cluster was 450 microns. Using single, small-volume injections of HRP into the diaphragm, we concluded that not all the phrenic motoneurons within a single cluster innervate muscle fibers in a discrete region of the diaphragm. Similar quantitative analysis of the distribution of medial gastrocnemius motoneurons did not reveal clustering in this motor column. We suggest that the fundamental differences in the spatial distributions of motoneurons within these two motor columns may be related to differences in the functional organizations of motoneurons innervating axial versus appendicular musculature, i.e., diaphragm versus medial gastrocnemius.

Direct measurements of the sizes of and forces produced by single fibers of the cat soleus and medial gastrocnemius muscles were made to determine whether or not different fiber types have characteristically distinct specific tensions. Single fibers (5-mm lengths), whose sarcolemmas had been chemically removed using a 5-mM EGTA "skinning" solution, were attached to a photodiode force transducer. Each single fiber was first placed in "relaxing" solution (22 +/- 1 degrees C, pH 7.0, pCa 8), its sarcomere length set at 2.7 micron using its laser diffraction pattern, and its diameter measured using the calibrated graticule of a microscope eyepiece (+/- 2 micron). Subsequently, each fiber was transferred to an activating bathing solution (pCa 3.6) in which the fiber produced its maximum tension. The specific tension values for single soleus muscle fibers displayed a threefold range (1.19 to 3.53 kg/cm2) with a mean value of 2.30 +/- 0.61 (SD) kg/cm2 (N = 42). The medial gastrocnemius fibers studied had a fourfold range in specific tensions (1.05 to 4.47 kg/cm2) and a mean value of 2.42 +/- 0.61 (SD) kg/cm2 (N = 104). Many medial gastrocnemius fibers (N = 64) were type-identified using a standard actomyosin ATPase histochemical assay. Type I medial gastrocnemius fibers had mean specific tension values of 2.45 +/- 0.47 kg/cm2 (N = 18), whereas, type II single fibers had mean specific tension values of 2.43 +/- 0.67 kg/cm2 (N = 46). Our results suggest that there is no significant difference between the specific tensions of the different muscle fiber types within the cat medial gastrocnemius muscle.

To characterize the oligosynaptic group I afferent input to the cat medial gastrocneumius (MG) motoneuron pool, the medial branch of the tibial nerve (MTIB: flexor digitorum and hallucis longus, popliteus, tibialis posterior and interosseous nerves), the nerves to flexor digitorum and hallucis longus (FDHL), or the nerves to the quadriceps muscles (QUAD) were stimulated at submaximal group I strength while recording intracellularly from MG motoneurons. Since previous work indicates that stimulation of these nerves at group I strength produces no significant monosynaptic Ia excitation or Renshaw inhibition of MG motoneurons, group I effects were assumed to be predominantly, though not exclusively, due to the action of Ib-fibers. Evidence supporting this assumption is presented in the following paper. MTIB, FDHL, and QUAD postsynaptic potentials (PSPs) were most commonly inhibitory. Since the MTIB, FDHL, and QUAD nerves are composed predominantly of fibers innervating muscles with extensor action, their inhibitory effect on MG motoneurons is consistent with previous findings that stimulation of Ib-afferents in nerves to extensor muscles produces di- and trisynaptic inhibition of extensor motoneurons. However, excitatory effects were observed in about one third of the motoneurons, indicating that oligosynaptic group I input is not homogeneously distributed within the MG motoneuron pool. Variations in QUAD, FDHL, and MTIB PSP pattern and amplitude were correlated with variations in the PSP pattern evoked by stimulation of the sural nerve: excitatory oligosynaptic group I PSPs generally appeared in motoneurons receiving excitatory cutaneous (sural nerve) input, whereas inhibitory PSPs generally appeared in motoneurons receiving some inhibitory cutaneous input and were largest in motoneurons receiving predominantly inhibition from the sural nerve. These variations in QUAD, FDHL, and MTIB PSP pattern and amplitude were not due to variations in resting potential and were only partly due to variations in intrinsic motoneuron properties or motoneuron "type." Our results indicate that activation of these cutaneous and group I muscle afferents can exert similar effects on the MG motoneuron pool. Moreover, the presence of a strong correlation between the distributions of cutaneous and oligosynaptic group I PSPs within a single motoneuron pool is consistent with the results of previous studies that have shown that some of the input to motoneurons from these peripheral afferents is mediated through common interneurons.

In the experiments described in the preceding paper electrical stimulation of the quadriceps (QUAD), medial tibial (MTIB), and flexor digitorum and hallucis longus (FDHL) muscle nerves was used to evoke oligosynaptic group I postsynaptic potentials (PSPs) in medial gastrocnemius (MG) motoneurons. In the present study, we attempted to specify the types of afferent fibers which mediate that oligosynaptic activity (FDHL to MG only). In one series of experiments, isolated single flexor digitorum longus (FDL) and flexor hallucis longus (FHL) afferents were identified as Ia, Ib, or group II fibers according to their conduction velocities, responses to muscle contraction, and mechanical thresholds to small amplitude triangular stretches applied to the parent muscles. We also determined the electrical thresholds of the identified afferent fibers by applying graded electrical stimulation to their muscle nerve. These results were used as criteria to define the types of afferents that mediated the electrically and stretch-evoked FDHL oligosynaptic PSPs recorded in MG motoneurons during a second series of experiments. The amplitudes of the oligosynaptic PSPs evoked in MG motoneurons increased as the strength of the electrical stimuli applied to the FDHL muscle nerves was raised to activate greater numbers of Ia- and Ib-fibers, but showed little or no additional increase when the stimulus intensity was raised further to include the majority of group II fibers. On this basis, a significant contribution by group II fibers to these oligosynaptic PSPs was considered unlikely. Simultaneous electrical activation of both Ia- and Ib-fibers produced distinct oligosynaptic PSPs in MG motoneurons, but these were likely due primarily to Ib-afferent activity, since selective activation of Ia-afferents (by stretch) rarely produced oligosynaptic PSPs in the same motoneurons. There was, however, evidence for some Ia contribution to these oligosynaptic PSPs. This is consistent with the demonstration that Ia- and Ib-afferent fibers converge onto common interneurons and that selective activation of Ia-fibers can produce PSPs similar to those evoked by concurrent stimulation of Ia- and Ib-fibers. On the basis of the present results and those of several related studies it is argued that the oligosynaptic PSPs evoked in MG motoneurons by submaximal group I stimulation of the FDHL, MTIB, or QUAD muscle nerves can be ascribed predominantly to the activation of Ib-afferent fibers, with only minimal Ia and probably no group II contribution.

Results from a wide variety of recent studies on the architecture and innervation of skeletal muscles, the neuromechanical characteristics of motor units, and the properties and spinal reflex actions of muscle proprioceptors present a number of challenges to conventional views of the functional organization of the segmental motor system. To illustrate the nature of these challenges, studies directed toward several specific issues are reviewed. These include the functional subdivision of single muscles into two or more neuromuscular compartments; the patterns of synaptic input from peripheral afferent fibers to motoneurons innervating muscle units of different "type;" and the convergence in the segmental reflex pathways from muscle spindles and tendon organs to motoneurons.

Motor units of soleus and medial gastrocnemius (MG) muscles were studied in pairs during stretch reflexes in the decerebrate cat to determine the relation between their recruitment orders and axonal conduction velocities. In 97% of soleus pairs, the motor unit with the lower axonal conduction velocity was recruited first. Since the soleus is a homogeneous muscle in the cat, differences in motor-unit type are, therefore, not a sine qua non for orderly recruitment nor is recruitment random within homogeneous populations of motor units, as recently proposed (28). In the medial gastrocnemius, a heterogeneous muscle, the same high correlation (97%) between recruitment sequence and conduction velocity was observed. Thus, the factors that determine recruitment order in heterogeneous muscles are as closely correlated with axonal diameter as they are in homogeneous muscles. Comparison of axonal conduction velocities in our sample of MG units with those in three samples of type-identified MG units studied by other investigators also suggests that motor-unit type is not the critical factor controlling the sequence of activation in heterogeneous muscles. It is concluded that the combined effects of all presynaptic and postsynaptic factors that determine susceptibility to discharge in motoneurons during stretch reflexes are strictly correlated with their axonal conduction velocities, as predicted by the size principle.

The spike-triggered averaging technique (26) was used to determine whether the synaptic input from medial gastrocnemius (MG) Ia-afferent fibers to homonymous motoneurons is "topographically weighted" (22) by means of differences in projection frequency, excitatory postsynaptic potential (EPSP) amplitude, or a combination of both factors. Motoneurons were classified as either "same branch" or "other branch," depending on whether a Ia-afferent fiber and motor axon were contained in the same or different intramuscular nerve branches. No difference was found in the projection frequency of Ia-afferents to the same branch and other branch motoneurons (95 versus 94%, respectively). The mean EPSP amplitude was larger in the same branch group of motoneurons (92 +/- 8 (SE) microV; n = V; n = 97) than in the other branch group (77 +/- 7 microV; n = 79). This difference was most striking in high-rheobase (greater than or equal to 10 nA) motoneurons, for which the mean EPSP amplitude in the same branch group was 82 +/- 12 microV (n = 48), whereas that in the other branch group was 52 +/- 5 microV (n = 37). In 60 cases it was possible to compare the EPSPs produced by a same branch afferent and an other branch afferent in the same motoneuron. The same branch afferent produced the larger EPSP in 73% (44/60) of the cases. Moreover, the mean ratio of the same branch to the other branch EPSP amplitudes was 1.7, which was both statistically significant and consistent with analogous results from our preceding study of aggregate EPSPs (22). Mean rise times and half-widths of EPSPs in the same branch group were not significantly different from those in the other branch group. Furthermore, no significant differences in rise times or half-widths between the two groups were evident when motoneurons were segregated according to their rheobase values. This suggests that the segregation of Ia-afferent and motor axons across the intramuscular nerve branches is not reflected in the locations of Ia terminals on the motoneuron somadendritic surface and that other factors must account for observed EPSP amplitude differences. Our data suggest that the topographic weighting of homonymous Ia-afferent input to cat MG motoneurons is mediated by a gradient of EPSP amplitude rather than by a gradient of Ia connectivity and also suggest that the effect is most prominent in high-rheobase motoneurons.

The spike trains of afferent fibers innervating muscle spindles and Golgi tendon organs in the medial gastrocnemius muscle were recorded during spontaneous contractions in either decerebrate cats or decapitate cats treated with L-dopa. For each afferent fiber, the approximate location of its receptor within the muscle was determined. Cross-correlation histograms were compiled from the simultaneously recorded spike trains of pairs of afferent fibers (Ia, Ib, spindle II) to determine if the degree of temporal correlation in their discharge was related to the mutual proximity of the receptors they innervated within the muscle. The frequency of occurrence and degree of correlated activity between pairs of muscle afferents, regardless of receptor type, was much greater in the decerebrate preparations than in the decapitate-L-dopa preparations. However, in all cases, the extent of correlated activity appeared to be unrelated to the relative locations of the receptors. The results suggest that the degree to which the discharge patterns of muscle receptors display temporal correlations and thereby potentially reinforce the "sensory partitioning" (5) of their parent muscle is strongly dependent on the type of preparation used, and thus by inference, dependent on the central state of the animal.

A broad survey of muscle unit properties in 14 muscles of the cat hind limb is presented which emphasizes some general features of unit properties in mammalian muscles. A more detailed analysis of muscle unit properties in three muscles of the posterior compartment of the lower leg is then presented using Burke's tetrapartite (FF, FI or F (Int.), FR, and S) unit classification scheme. Our data on the properties of motor units in cat tibialis posterior (TP) have been compared to those generated by Burke and colleagues on units in flexor digitorum longus (FDL) and medial gastrocnemius (MG). In all three muscles, twitch contraction time was distinctly slower for type S units and specific tension outputs were substantially greater for type FF units than for type S units. The innervation ratios of type FR units were slightly lower than for type S units but the specific tension of the FR units was closer to FF units than to type S units. The FF units controlled 70-74% of the cumulative force output of each muscles, indicating a substantial capacity for powerful rapid contractions of all three of these muscles despite their differences in "size," action, and force generation. Distinctive features of the three muscles included differences in the unit types' force producing capabilities and in the relative representation of "nonfatigable" type FR and S units in each muscle. In particular, TP is endowed with some unusually powerful type FF units and a high percentage (42%) of type S units. In contrast, FDL has units that develop relatively little force and an unusually high representation (56%) of type FR units. The possible relationships between these muscle features and their presumed role in posture and locomotion is discussed.