Motoneurons In Segments Research Articles

GAP-43, aFGF, CCK and alpha- and beta-CGRP in rat spinal motoneurons subjected to axotomy and/or dorsal root severance.

The mRNA levels for growth-associated protein 43 (GAP-43), acidic fibroblast growth factor (aFGF), alpha- and beta-calcitonin gene-related peptide (CGRP), cholecystokinin (CCK) and choline acetyltransferase (ChAT) in rat lumbar spinal motoneurons were studied by in situ hybridization 1, 5 and 21 days and 20 weeks following unilateral peripheral nerve sectioning, ventral rhizotomy or dorsal rhizotomy. Furthermore, CGRP- and aFGF-like immunoreactivities in the ventral horn were studied using immunohistochemistry. One to 21 days after axotomy, GAP-43 and alpha-CGRP mRNAs increased in lesioned motoneurons, while the aFGF mRNA levels were marginally higher in motoneurons on the lesion side as compared to the control side. beta-CGRP, CCK and ChAT mRNA levels, on the other hand, decreased during the short-term response (1-21 days) to axotomy. After ventral rhizotomy, but not peripheral axotomy, there was complete disappearance of aFGF-like immunoreactivity in the ventral root proximal to the lesion. In animals subjected to long-term survival (20 weeks) after peripheral axotomy, the expression of all studied substances had returned to normal levels. Unilateral dorsal rhizotomy did not induce any substantial short- or long-term shifts in the cellular expression of the GAP-43, aFGF, CGRP and CCK peptides or their mRNAs in motoneurons of lesioned segments. These results indicate that peptides/proteins in motoneurons are expressed differentially after axotomy. Whereas alpha-CGRP and GAP-43 are up-regulated, CCK and beta-CGRP become down-regulated and aFGF is largely unaffected.

Facilitation of motor evoked potentials by somatosensory afferent stimulation

The effect of an electrically induced peripheral afferent volley upon electrical and magnetic motor evoked potentials (MEPs) from muscles of the upper and lower extremities was studied in 16 healthy volunteers. A standard conditioning-test (C-T) paradigm was employed whereby the test stimulus (transcranial electric or magnetic) was applied at random time intervals, from 10 msec prior to 90 msec after the conditioning stimulus (peripheral nerve stimulus). MEP amplitude facilitation was observed for the majority of the upper extremity muscles tested at two distinct periods, one occurring at short, and the other at long C-T intervals. This bimodal trend of MEP facilitation was found to be equally as prominent in the lower extremity muscles tested. The period of short C-T interval facilitation is consistent with modifications in the spinal excitability of the segmental motoneuron pool. On the other hand, the period of long C-T interval facilitation is suggested to be due to alterations in excitability of the motor cortex as a result of the arrival of the orthodromic sensory volley. Although most pronounced in muscles innervated by the nerve to which the conditioning stimulus was applied, this bimodal facilitatory effect was also observed in adjacent muscles not innervated by the stimulated nerve. Qualitatively, the conditioned MEPs from the upper and lower extremities responded similarly to both electrical and magnetic trans-cranial stimulation. In addition, our study demonstrates that the C-T paradigm has potential for use in the assessment of spinal and cortical sensorimotor integration by providing quantitative information which cannot be obtained through isolated assessment of sensory and/or motor pathways. This approach might be used for assessment of unconsciousness and/or comatose head injured patients.

Reinforcement of subliminal flexion reflexes by transcranial magnetic stimulation of motor cortex in subjects with spinal cord injury

A 20 msec train (500 Hz; 0.1–0.2 msec duration) of percutaneous electrical stimulation (ES) applied to the planar surface was used to condition muscle responses evoked in tibialis anterior (TA) by transcranial magnetic stimulation of the motor cortex in 8 subjects with traumatic spinal cord injury (SCI). The intensity of conditioning ES was adjusted to just subthreshold for evoking flexion reflexes in TA and was delivered at conditioning-test (C-T) intervals of 15–60 msec prior to cortical stimulation. Four subjects with clinically complete SCI revealed no muscle response to cortical stimulation or following combined subliminal percutaneous ES and cortical stimulation. Four subjects (3 clinically incomplete and 1 complete injury) demonstrated muscle response with a latency of 70–80 msec time-locked to the percutaneous ES when the conditioning subliminal stimulation was delivered at C-T: 15msec. These responses, resembling suprathreshold flexion reflexes, reflect the convergence of excitatory afferent and cortical inputs and provide evidence of preserved corticospinal innervation to the L4-5 segmental motoneuron or interneuron pools. In 3 of the subjects this preserved corticospinal influence was evident despite absence of motor evoked potentials (MEPs) following cortical stimulation. The effect of the combined electrical and cortical stimulation in yielding suprathreshold flexion reflexes, instead of the facilitated MEPs seen in control subjects, appears to be related to slowed central conduction, prolonged temporal dispersion of the motoneuron facilitation following cortical stimulation and segmental reflex changes associated with disrupted modulation of interneuronal pathways. The results show this conditioning paradigm to be useful in revealing preserved corticospinal innervation in some SCI subjects with absent MEPs.

Location and connectivity of abdominal motoneurons in the embryo and larva of Drosophila melanogaster

The soma location and peripheral connectivity of motoneurons in abdominal segments of the embryo and larva of the fruitfly, Drosophila melanogaster are described as an initial step in determining the mechanisms by which motoneurons make connections with their target muscles in a genetically accessible organism. Embryonic motoneuron somata were retrogradely labelled by application of the fluorescent dye, DiI, to the whole peripheral nerve or to its separate anterior or posterior fascicles in segments A5-A7 of late stage 15/early stage 16 embryos. This technique reveals a stereotyped, segmentally repeated population of 34 motoneurons per hemisegment, several of which can be individually identified from their soma position. The same set of motoneurons was revealed in third instar larvae of D. melanogaster by cobalt backfilling of abdominal peripheral nerves, although the positions of some of these neurons change during larval development. The peripheral connectivity and axon morphology of several of the abdominal motoneurons was determined by intracellular injection with Lucifer Yellow in stage 16 embryos. For the motoneurons with axons in the anterior fascicle there is no clear relationship between somata groupings and the muscle targets innervated: contrary to earlier claims, these motoneurons arborize over both ventral and dorsal muscles. Individual motoneurons possess a stereotyped pattern of terminal arborization.

NMDA receptors mediate poly- and monosynaptic potentials in motoneurons of rat embryos

We determined the contribution of glutamate receptor subtypes to developing excitatory synaptic transmission in isolated spinal cord of rat embryos. Using electrophysiological and morphological techniques, we studied the pattern of development of synapses between dorsal root afferents and motoneurons in lumbar spinal cords of 15- to 21-d-old rat embryos. Motoneuron dendritic fields and afferent projections onto motoneurons were identified by labeling with HRP. Afferents first entered the gray matter at Day 15 of gestation, and by Day 16 they terminated close to motoneuron dendritic trees. Afferent axons projected onto motoneuron dendritic fields at Day 17, when boutons were detected on motoneuron dendrites that were crossed by afferent axons. To determine the time course of formation of functional sensorimotor synapses and their pharmacological properties, a dorsal root was stimulated while recording intracellularly from segmental motoneurons. At Day 16, excitatory postsynaptic potentials (EPSPs) with long latencies, slow rates of rise, and long durations were recorded. The amplitudes of these EPSPs increased with membrane depolarization and in the absence of extracellular Mg2+. These EPSPs were blocked by D-2-amino-5-phosphonovalerate (APV) and ketamine, which are selective antagonists of N-methyl-D-aspartate (NMDA) receptors. These findings suggest that initial synaptic transmission in embryonic motoneurons is mediated solely by NMDA receptors. Short-latency EPSPs with fast rates of rise were first recorded in most motoneurons by Day 17. These EPSPs were composed of fast- and slow-rising potentials. The slow component was blocked by APV, while the fast component was eliminated by 6-cyano-7-nitroquinoxaline-2,3-dione and kynurenate. This indicates that the short-latency EPSPs are mediated by both NMDA and non-NMDA receptors. Dose-response curves of motoneurons to L-glutamate, NMDA, and kainate demonstrated that motoneurons are sensitive to these agonists prior to the formation of synapses between afferents and motoneurons. Motoneuron responses to NMDA and kainate increased immediately after the onset of short-latency EPSPs. This increased sensitivity could be due to extracellular factors influenced by growing sensory axons or intrinsic properties of differentiating motoneurons.

Regional variations in the extent and timing of motoneuron cell death in the lumbosacral spinal cord of the chick embryo

We have examined the distribution of motoneurons in different segments of the chick lumbosacral spinal cord before and after the period of motoneuron cell death. The extent of cell death was found to be greatest at the boundaries of the lumbosacral cord where over 60% of the motoneurons died and least in the central region where only 30% died. After cell death at stage 40 the number of motoneurons in each segment was linearly correlated with segment length, suggesting that growth of the segment and motoneuron numbers may be regulated by a common factor. The time of completion of motoneuron cell death exhibited a rostrocaudal gradient along the lumbar cord. Cell death was complete in the anterior segments by stage 35 but not until stage 38 in the caudal 4 segments. The regional variations in the extent and timing of motoneuron cell death suggest that the relative importance of the factors mediating cell death vary in different regions of the lumbar cord.

Recovery of postural control following chronic bilateral hemisections at different spinal cord levels in adult cats

Chronic cats with double hemisections of the spinal cord, first at a lower thoracic level followed by a contralateral midthoracic cord at intervals of 0 to 126 days (T-T preparations) or first at an upper cervical followed by a midthoracic at intervals of 15 to 74 days (C-T preparations), eventually recovered quadrupedal standing 7 to 53 days after the second hemisection. For about 7 days following the first hemisection at a lower thoracic level, floor reaction force (FRF) of the hind limb of the hemisected side decreased to 25 to 30% of the normal value, then recovered to the control value. A group of cats whose second hemisection was done within 7 days after the first hemisection needed 24 to 53 (mean 43) days to recover quadrupedal standing, whereas cats whose second hemisection occurred after 10 to 126 days needed 7 to 22 (mean 15) days. During the recovery period many unusual reflexes were elicited which eventually disappeared as the cats resumed standing and walking. Lateral stability of the double-hemisected cats deteriorated significantly, whereas segmental reflexes were augmented. These results indicate the importance of descending impulses over the segmental motoneuron pools to control standing posture and locomotion. It was assumed that the descending impulses were conveyed by polysynaptic pathways which had minimal functions before the hemisections.

Spinal control of pelvic floor muscles

A prevalent notion in the literature is that the pelvic floor muscles behave as a unitary mass. We examined this proposition experimentally. In spinal cats, we recorded EMG activity from the following pelvic floor muscles: the sphincter ani externus (SAE), the abductor caudae internus (coccygeus), and the levator ani (pubiocaudalis) muscles. The epaxial sacrocaudalis dorsalis lateralis muscle was also exposed and prepared for recording. Electrical stimulation of S2 ventral roots elicited twitch responses of the sphincter ani externus and of the sacrocaudalis dorsalis lateralis muscles. Stimulation of S3 and Cx1 ventral roots elicited responses in the other two muscles studied, the levator ani and abductor caudae internus. Thus a clear segregation of the segmental motor neuron pools innervating the different pelvic floor muscles was demonstrated. The various muscles of the pelvic floor region could be reflexly activated either individually or as a mass unit depending on the intensity of stimulation. Tactile or electrical stimulation of pudendal regions on either side of the body elicited responses of the sphincter ani externus. In contrast, activation of the levator ani and abductor caudae internus muscles could be lateralized: tactile or electrical stimulation of the dorsolateral surfaces at the base of the tail region elicited ipsilateral responses from these muscles. Section of one pudendal nerve did not alter the level of tonic activity (2 to 4s) of the sphincter ani externus. However, bilateral section of the pudendal nerve entirely abolished both tonic activity and phasic responsiveness of the SAE without affecting the activity of the levator ani and abductor caudae internus muscles. Pudendal nerve stimulation elicited only polysynaptic reflex responses from S2 ventral roots. The results presented show that the neural apparatus of the striated musculature of the pelvic floor is capable of activating individually the different muscles that make up the system, and that the sphincter ani externus from one side, and muscles that conform the diaphragm pelvis from the other, are subserved by different neuronal circuits.

EVOLUTION OF THE TELSON NEUROMUSCULATURE IN DECAPOD CRUSTACEA

The neuromusculature in the telsons of three macrurans (Pandalus platycerus, Procambarus clarkii, Upogebia pugettensis) and three anomurans (Munida quadrispina, Blepharipoda occidentalis, Emerita analoga) are compared to provide a framework for neurophysiological comparisons of their roles in the swimming behaviors of these decapods. The stereotypical arrangement in macruran telsons comprises a group of massive axial muscles and a trio of small appendage muscles (Fig. 1). The various arrangements of telson neuromusculature in the anomurans (Fig. 3) are interpreted in terms of specific modifications of particular macruran features. Homologies among muscles and nerve roots in the telsons of the six decapods are identified (Figs. 1, 3, 4, Table II) and homologies among particular axial and appendage motoneurons in the sixth abdominal ganglia are suggested (Fig. 5, Table III). The appendage neuromusculature in decapod telsons is inferred to be ontophyletically part of the seventh abdominal segment that was present in the ancestors of decapods. These muscles and their motoneurons, like most of the axial neuromusculature in the telson (Dumont and Wine, 1983, in prep.), may be serial homologs of muscles and motoneurons in abdominal segments.

Stepping by chronic spinal cats.

We investigated the consequences of spinal cord transection in adult cat. All became capable of stepping upon a treadmill. Two of six animals stepped in the first postoperative testing session, 1 week after surgery. It took up to 6 weeks for the others to start stepping. None of the cats became capable of hindquarter support within 2 months after surgery. There was permanent loss of fore-hindlimb coordination and increased variability in the duration of each step cycle and its components. Coordination between the hindlimbs was also impaired, as signaled by increased variability in the time between foot contacts with the belt. These changes may be due to the loss of timing signals coming down from supraspinal levels to the segmental step generators or - alternatively - to decreased excitability of segmental motoneurons leading to less precise execution of stepping commands from the hypothetical pattern generators.

Interactions among lumbar motoneurons on opposite sides of the frog spinal cord: morphological and electrophysiological studies.

Light and electron microscopy have been used to study the projections of dendrites from motoneurons in lumbar segments of the spinal cord of the frog following administration of horseradish peroxidase to cut ventral roots. Processes originating from motoneurons crossed to the opposite side of the spinal cord via the anterior commissure and made contact with dendrites and motoneuronal somata. Typically, in segments 6 to 8 the crossing dendrites showed irregular enlargements in diameter. Electrophysiological recordings were obtained both extracellularly from ventral roots and intracellularly from motoneuronal somata. In Ringer's solution containing 1 mM calcium, stimulation of a lumbar ventral root, elicited population responses with early and late components in the ventral root of the opposite side of the same segment. Only the early, short latency component remained in calcium-deficient Ringer's solution. In calcium-containing Ringer's solution, intracellular recording from an antidromically activated motoneuron showed an action potential with a short latency; this response was followed by excitatory postsynaptic potentials (epsps) from which action potentials could be generated. Contralateral ventral root stimulation also elicited in the same motoneuron a short latency action potential that was rarely followed by epsps. The short latency responses, that were elicited by stimulation of ventral roots of either side persisted in calcium-deficient Ringer's solution, but the epsps were abolished. Contralaterally elicited short latency responses were eliminated by section of the anterior commissure. We believe that electrically mediated crossed interactions among lumbar motoneurons may serve as a means of coordinating muscle groups of opposite sides that are used in movements that require bilateral synchronization, such as jumping and swimming.

The bulbo-spinal indoleaminergic pathway in the frog

The role of the bulbo-spinal indoleaminergic pathway of the frog was investigated. Using the isolated spinal cord preparation, responses of motoneurons of segments 9 and 10 to lateral column stimulation were recorded from ventral roots (LC-VRP) and by intracellularly placed microelectrodes. Control responses were compared to those obtained after addition of substances known to alter indoleaminergic synthesis or receptor activation. Responses from spinal cords of animals that were pretreated with indoleamine-depleting agents were compared to mean control Procedures that inhibited indoleamine synthesis or blocked indoleamine receptors reduced motoneuronal activity. This was manifested as an increase in mean latency and a decrease in amplitude of the monosynaptic LC-VRP, an increased duration of suppression following an LC conditioning stimulus, and a decrease in spontaneous activity. In contrast to intracellularly recorded responses from control cords, those recorded under these conditions typically showed single spikes with longer latencies. Additions of indoleamine precursors to normal cords or to cords depleted of monoamines by reserpine shortened mean latencies, increased amplitudes of LC-elicited responses and caused an increase in spontaneous activity. These observations were recorded both extracellularly and intracellularly. Our results suggest that the bulbo-spinal indoleaminergic pathway modulates the output of motoneurons of the frog spinal cord.

Development of the segmental innervation of the chick forelimb

A number of recent studies have shown that during embryonic development the initial innervation of a target structure may be made up, in part, by axons which do not form part of the mature innervation of that structure. In the present study we have examined the motor innervation of the major muscles of the chick forelimb at different stages of development using HRP-uptake-labelling of motoneurons, electrophysiological recording and measurement of muscle contraction. In the mature White Leghorn chick the major contribution to the motor innervation of the forelimb is from spinal segments 14, 15 and 16. Using the HRP-labelling technique we have shown that at stages 26-29 of development motoneurons in segments 12-17 have axon terminals in the presumptive biceps muscle. Between stages 30 and 35, however, the axon terminals arising from segments 12, 13, 16 and 17 are lost, leaving the mature innervation from segments 14 and 15. We have also observed the loss of innervation of the biceps muscle by segment 16 using electrophysiological recording of compound action potentials in the biceps nerve and by measurement of the local contraction of the biceps muscle in response to stimulation of the segmental nerves. Similar changes in the innervation of the triceps, extensor metacarpi radialis, flexor carpi ulnaris and flexor digitorum profundus muscles have also been observed. These results are discussed in relation to the hypothesis that (i) the motoneuron pools and muscles in the developing spinal cord and forelimb are matched, (ii) that some axons which arrive in a particular muscle during early development are unable to form a stable connexion and (iii) that the inability of an axon terminal to form a stable connexion in a muscle results in the death of the motoneuron. Intracellular recording from muscle cells at stage 35 shows that the synaptic site on each cell is innervated by about three separate axons. Over the next few stages, however, all but one of the innervating axons is lost. From our contraction studies it is clear that the removal of the excess axon terminals after stage 35 is not associated with the establishment of the mature segmental innervation pattern of the muscle.

Localization of motor neuron pools supplying identified muscles in normal and supernumerary legs of chick embryo.

Neuromuscular specificity has been investigated in chick embryos with a grafted supernumerary leg. The nerves of the lumbar plexus are divided between the two legs so that rostral nerves innervate the grafted leg and the caudal nerves supply the host's original leg. The basic topographic organization of the histologically definable motor neuron clusters of the lateral motor columns remains unchanged by the addition of a supernumerary leg. Intramuscular injections of identified leg muscles have been used to map the intraspinal location of specific motor pools in stage 38 (12-day) embryos. In the normal embryo, the gastrocnemius muscle is innervated by neurons in a central dorsal cluster of motor neurons in segments 26-29. In six experimental cases, the motor neurons supplying the gastrocnemius muscle of a rostrally placed grafted leg were consistently located in a specific medial cluster of neurons in segments 23-25. Motor neurons in this location never normally innervate a gastrocnemius muscle, even in the very young embryos during the period of naturally occurring cell death. This observation of a systematic mismatch between a particular motor cluster and an abnormally innervated muscle indicates the operation of a selective developmental process. A hierarchy of selective chemoaffinities may best explain our experimental results.

Monoaminergic inputs to frog motoneurons: An anatomical study using fluorescence histochemical and silver degeneration techniques

The monoaminergic input to spinal segment 10 of the frog, Rana pipiens, was investigated anatomically. Fluorescence histochemistry demonstrated a diffuse and overlapping distribution of catecholamine (CA) and indoleamine (IA) containing terminals within the spinal cord. Terminal fields containing high numbers of CA and IA varicosities included the marginal plexus and the anterior commissural region. IA terminals were also abundant in the dorsal portion of the dorsal horn, while CA terminals were found in high number in the ventral part of the dorsal horn. IA varicosities were regularly observed within the lateral motor nucleus, but CA varicosities were not. Throughout the medulla IA neurons were located exclusively within the raphe nuclei. From the level of the obex to the entry of cranial nerve X, CA neurons were scattered in the intermediolateral and intermediomedial reticular formation. In Fink-Heimer preparations of segment 10, raphe lesions produced a pattern of terminal degeneration that corresponded to the fluorescent histochemical picture of IA varicosities. All levels of the medullary raphe nuclei projected to the lateral motor column. Other spinal regions received a differential raphe input. It is concluded that motoneurons of segment 10 have a somal input from the medullary raphe nuclei that is indoleaminergic in nature. It is also probable that distal motoneuronal dendrites and interneurons are contacted by IA and CA terminals.

Synaptic effects evoked by supraspinal and intraspinal stimulation in lamprey motoneurons

In experiments onLampetra fluviatilis in response to electrical stimulation of bulbar reticulospinal neurons and descending fibers the postsynaptic potentials of segmental motoneurons and action potentials of single intraspinal axons were recorded intracellularly and the cord dorsum potentials were recorded by a surface electrode. Fast-conducting reticulospinal axons (Muller's axons) were shown to excite spinal motoneurons monosynaptically. Monosynaptic reticulo-motoneuronal EPSPs arise as the result of excitation of a limited number of descending fibers, they reproduce high frequencies of stimulation readily and, in some cases, they are divided into components of which the first may be attributed to an electrical, and the second to a chemical mechanism of transmission. Besides early monosynaptic EPSPs, late, probably polysynaptic, responses also are found.

Reticulospinal influences on segmental neurons inLampetra fluviatilis

By recording the activity of single Muller's axons extra- and intracellularly, the electrical properties of fibers of the descending tract were determined during stimulation of the vestibular system, optic nerve, and spinal cord. The amplitude of the action potentials (APs) is 60–80 mv, their duration 3 msec, the highest frequency of responses to rythymic stimulation 170–250/sec, and the velocity of conduction of excitation is 5.4±0.4 m/sec along fibers of the ventromedial tract and 2.4±0.4 m/see in the lateral tract. These velocities remain unchanged over segments of the spinal cord up to 20 cm long. In response to stimulation of the descending tract by single pulses, only 1 AP is generated in Muller's axons whatever the strength of stimulation. In response to synaptic stimulation of Muller's neurons, discharges of regular frequency and duration are generated in axons of the descending tract just as in the neurons themselves. Transmission of excitation from Muller's axons to segmental motoneurons can take place with segmental delay of less than 1–3 msec, indicating the existence of mono-synaptic connections between these structures. Motoneuron activity in response to stimulation of the descending tract consists of generation of 1–6 APs at intervals of 7–20 msec. The suggestion that there are connections between one motoneuron and several Muller's axons is discussed.

Monosynaptic excitory postsynaptic potentials in thoracic motoneurons under reticulospinal influences

We studied synaptic processes in motoneurons of thoracic segments (TIX-TXI) evoked by stimulation of the medial area of the giant-cell reticular nucleus in decerebrated cats. Monosynaptic EPSP were recorded in the majority of investigated motoneurons upon activation of the most rapidly conducting reticulospinal fibers. In some cells, such monosynaptic EPSP were accompanied by late EPSP or IPSP. Amplitude of monosynaptic EPSP attained 5 mV, but this value usually was insufficient for development of an action potential. Upon summation of single monosynaptic EPSP, the membrane potential reached the critical level and an action potential arose in the motoneuron. The efficiency of summary processes evoked by stimulation of the reticular formation exceeded the intensity of synaptic processes that arise in thoracic motoneurons on stimulating the nucleus of Deiters. Functional characteristics of reticular and vestibular monosynaptic EPSP are discussed in the work.

Monosynaptic control over spinal motoneurons at different levels

The results show that direct monosynaptic connections with segmental motoneurons are formed in mammals not only by fibers of the pyramidal tract, but also by axons of descending systems arising from different levels of the brain. A characteristic feature of these monosynaptic projections is their high conduction velocity and the relatively small number of conducting fibers of the descending system compared with the remaining thinner and more slowly donducting elements. This feature, clearly distinguishable in higher vertebrates up to primates, possibly reflects the form of organization of suprasegmental control which appeared originally in the most primitive vertebrates with a fully formed brain (cyclostomes). With progressive encephalization of the nervous system in the course of evolution, the number of direct connections between the brain and segmental motoneurons progressively increases, the main source of these connections shifting into the motor cortex. An essential feature of monosynaptic suprasegmental control in higher vertebrates is evidently the increasing specification of the connections with particular functional groups of α-motoneurons.

Inhibitory interaction between motoneurons of adjacent segments in the frog spinal cord

Inhibition of motoneurons through impulses in recurrent axon collaterals of neighboring motoneurons in the spinal cord of Xenopus laevis was investigated. The method of monosynaptic testing was employed. A slow dorsal root potential (A-DRP) was elicited by the same antidromic volley of impulses that caused recurrent inhibition of monosynaptic reflex responses. The presence of depolarization of dorsal root fibers indicates the operation of a presynaptic inhibitory mechanism. The time course of the inhibitory effect produced by a recurrent volley on the excitability of neighboring motoneurons closely resembled the time course of the A-DRP elicited by the same recurrent volley. The inhibitory effect of an antidromic stimulus depended on the pathway through which the motoneurons under investigation were excited. If recurrent inhibition was postsynaptic, the inhibitory effect would have been independent of the testing pathway. The results presented indicate that recurrent inhibition in the frog is a presynaptic process. The mechanism by means of which recurrent inhibition is effected in the frog, therefore differs from the postsynaptic Renshaw mechanism involved in recurrent inhibition of spinal motoneurons in mammals.