Phrenic nerve stimulation in patients with spinal cord injury

Phrenic nerve pacing (PNP) is a clinically useful technique to restore inspiratory muscle function in patients with respiratory failure secondary to cervical spinal cord injury. In this review, patient evaluation, equipment, methods of implementation, clinical outcomes, and the complications and side effects of PNP are discussed. Despite considerable technical development, and clinical success, however, current PNP systems have significant limitations. Even in patients with intact phrenic nerve function, PNP is successful in achieving full-time support in ～50% of patients. Inadequate inspired volume generation may arise secondary to incomplete diaphragm activation, reversed recruitment order of motor units, fiber type conversion resulting in reduced force generating capacity and lack of coincident intercostal muscle activation. A novel method of pacing is under development which involves stimulating spinal cord tracts which synapse with the inspiratory motoneuron pools. This technique results in combined activation of the intercostal muscles and diaphragm in concert and holds promise to provide a more physiologic and effective method of PNP.     

Ventilatory support by pacing of the conditioned diaphragm in quadriplegia

We provided full-time ventilatory support in five patients with respiratory paralysis accompanying quadriplegia by continuous electrical pacing of both hemidiaphragms simultaneously for 11 to 33 months through the application to the phrenic nerves of a low-frequency stimulus. The strength and endurance of the diaphragm muscle increased with pacing. Biopsy specimens taken from two patients who had uninterrupted stimulation for 6 and 16 weeks showed changes suggestive of the development of fatigue-resistant muscle fibers. When we compared these results with those of our earlier experience with intermittent unilateral stimulation of the diaphragm in 17 patients with respiratory paralysis, we found that continuous bilateral pacing using low-frequency stimulation appeared to be superior because of more efficient ventilation of both lungs, fewer total coulombs required to effect the same ventilation, and absence of myopathic changes in the diaphragm muscle. For patients with respiratory paralysis and intact phrenic nerves, continuous simultaneous pacing of both hemidiaphragms with low-frequency stimulation and a slow respiratory rate is a satisfactory method of providing full-time ventilatory support.     

Extra-segmental reflexes derived from intercostal afferents: phrenic and laryngeal responses

1. Phrenic and recurrent laryngeal efferent responses were evoked by brief tetani or single shocks to the cut external intercostal nerves of anaesthetized cats. The reflexes derived from middle thoracic segments (T5 and 6) were compared with those emanating from caudal thoracic segments (T9 and 10).2. During inspiration, middle intercostal nerve stimulation transiently inhibited the spontaneous discharge in both efferent neurograms, whereas stimulation of caudal intercostal nerves facilitated phrenic discharge and usually inhibited recurrent laryngeal activity.3. During expiration, stimulation at either thoracic level enhanced recurrent laryngeal discharge while provoking little or no phrenic response.4. Superficial lesions of the lateral cervical cord, ipsilateral to the stimulus sites, above or below the phrenic outflow, eliminated all reflex responses except the phrenic response to caudal thoracic stimuli. Similarly, in the spinal animal, middle intercostal afferents could not be shown to decrease phrenic excitability. Caudal intercostal afferents cause phrenic excitation by a spinal reflex.5. Group I afferents of the mid-thoracic segments and group II afferents of the caudal thoracic segments initiate these extra-segmental reflexes.6. The recurrent laryngeal responses manifest, for the most part, changes in the discharge of fibres innervating the posterior cricoarytenoid muscle. The responses fit the overall pattern of response to middle intercostal nerve stimulation, namely, inhibition of inspiratory muscles and excitation of expiratory muscles. Intercostal afferent stimulation also activated the laryngeal adductor muscles.7. The results support the view that intercostal mechanoreceptors initiate an array of extra-segmental respiratory reflexes, including spinal and supraspinal arcs. The simplest way to account for the various responses to stimulation of middle intercostal afferents is to postulate a reflex involving supraspinal respiratory neurones.8. The observed reflexogenic differences correlate with anatomical differences between the middle and caudal ribs. Possible functional implications of this relationship are discussed.     

Hypothesis that vagal reinervation of diaphragm could sensitise it to electrical stimulation

The hypothesis proposed is that restoration of functional capacity of denervated diaphragm may be achieved by reinervating it with vagus nerve. Following trauma, carcinomatose infiltration, and/or large thoracic surgery and neck surgery, phrenic nerve is frequently injured. Reinervation even in the most favourable conditions would not follow and diaphragm would rest permanently denervated and paralysed. This results in unilateral or bilateral paralysis of diaphragm. In principle, intermittent electrical stimulation of the phrenic nerve or diaphragm could elicit regular diaphragm contractions and maintain satisfactory respiration. While this technique could be used in upper motor neurone injury, in lower motor neurone injury and denervated diaphragm, that imposes too high electrical resistance, direct diaphragm pacing is practically impossible. In these cases, long term artificial ventilation is often necessary. Nevertheless, those patients are at high risk to suffer from atelectasis and respiratory infections. We project a hypothesis that reinervation of denervated diaphragm by vagus nerve could re-establishes its sensitivity to intramuscular electrical stimulation and may allow stimulation of the diaphragm by implanted pace-maker electrodes. An appropriate electrical stimulation might then be possible and diaphragm pacing could replace prolonged artificial ventilation in those patients. Restoration of functional capacity of denervated diaphragm could open a perspective for long term diaphragm pacing in patients with irreversible phrenic nerve injury and diaphragm paralysis.     

Electrical stimulation of arterial and central chemosensory afferents at different times in the respiratory cycle of the cat: II. Responses of respiratory muscles and their motor nerves

The response patterns of the electrical activity of the respiratory motor nerves and muscles to brief electrical stimulation of the arterial and the intracranial chemosensory afferents were studied in anesthetized cats. Stimulation during inspiration increased the activity of phrenic nerve and the inspiratory muscles (intercostal, diaphragm) with a latency of 15–25 ms, whereas expiratory muscle activity in the following expiration remained almost unaltered. Stimulation during expiration increased the activity of expiratory nerves and muscles (intercostal, abdominal) after a delay of 80–120 ms. The later the stimulation occurred in the insor expiratory period the larger the increase in amplitude and in steepness of rise of the respective integrated activity in respiratory nerves and muscles. Stimulation in early inspiration shortened the discharge period of inspiratory muscles, whereas excitation in early expiration caused an earlier onset and prolonged the activity in the expiratory muscles. Stimulation in the late phase of ins- or expiration prolonged the discharge of the respective nerves and muscles. Both the arterial (carotid sinus nerve, CSN, and aortic nerve, AN) and intracranial chemosensory (VM) afferents stimuli were able to affect both the inspiratory and the expiratory mechanisms. The restriction of the effects to the phase of the stimulus suggests a mechanism by which these afferents, when activated during inspiration, effectively project only to inspiratory neurones, and vice versa for expiration.Supported by the Deutsche Forschungsgemeinschaft, SFB 114 Bionach     

The role of the fusimotor system with respect to the contribution of the diaphragm and the intercostal muscles to the respiratory tidal volume

Summary The efferent electrical activity in the phrenic nerve can be quantified in such a way that it gives a good correlation to tidal volume. After administration of the drug benzoctamine this relationship changes: more phrenic nerve activity is needed for the same tidal volume. No changes were found in the neuro-muscular transmission from the phrenic nerve to the diaphragm. There was no alteration in dynamic compliance of the lungs or in airway resistance. The afferent phrenic nerve activity from proprioceptors in the diaphragm did not change. It seems unlikely that respiratory neurons in the brainstem were affected since the sensitivity of the respiratory system to CO₂ did not change.It is known that the tonic fusimotoneuron activity is suppressed at a supraspinal level by benzoctamine. Since intercostal muscles have muscle spindles and the diaphragm hardly has any, the intercostal muscle activity will be affected more than diaphragmatic activity by benzoctamine. This could actually be shown by quantifying the electromyogram of inspiratory external intercostal muscles.The tidal volume regulation is controlled by the vagal feedback loop. In order to reach a certain tidal volume after administration of benzoctamine, the contribution of the diaphragm has to increase because the activity of the intercostal muscles is diminished.     

Augmentation of muscle nociceptive respiratory reflex facilitation by vagal afferents

Vagal influence on the facilitation of phrenic neural activity during respiratory phase-locked, gastrocnemius muscle nerve nociceptive electrical stimulation was examined in anesthetized, glomectomized, paralyzed, and artificially ventilated cats. (1) In the vagi-intact state, respiratory reflex facilitation was characterized by a sharp rise in peak amplitude, maximum rate of rise or slope, and mean rate of rise of integrated phrenic nerve activity. This was greater during inspiratory phase-locked (T1-locked) muscle nerve electrical stimulation than during expiratory phase-locked (TE-locked) muscle nerve electrical stimulation. "Evoked post-inspiratory phrenic activity" during the early expiratory phase was also observed during TE-locked muscle nerve electrical stimulation. (2) Bilateral vagotomy significantly attenuated the respiratory facilitation during both T1- and TE-locked muscle nerve electrical stimulation. In particular, the "evoked post-inspiratory phrenic activity" during TE-locked muscle nerve electrical stimulation was also attenuated or almost completely abolished. (3) Conditioning electrical stimulation of the vagus nerve revealed facilitatory reflexes which co-exist with inspiratory inhibitory reflexes. (4) The "evoked post-inspiratory phrenic activity" during TE-locked muscle nerve electrical stimulation, which was attenuated or abolished after vagotomy, was restored after vagal T1-locked conditioning stimuli combined with TE-locked muscle nerve electrical stimulation. The results suggest that vagal facilitatory reflexes augment the respiratory reflex facilitation during muscle nociceptive stimulation.     

Activation of respiratory afferents by resistive loaded breathing modifies somatosensory evoked potentials to median nerve stimulation in humans

The cortical projections of respiratory afferents (vagus and respiratory muscle nerves) are well documented in humans. It is also shown that their activation during loaded breathing modifies the perception of tactile sensation as well as the motor drive to skeletal muscles. The effects of expiratory or inspiratory loaded breathing on somatosensory evoked potentials (SEPs) elicited by median nerve stimulation were studied in eight healthy subjects. No significant changes occurred in latencies of N20, N30 and P40 throughout the expiratory loading period, except for a significant lengthening in P1 latency compared with unloaded breathing. However, inspiratory loading induced a significant increase in peak latency of N20, N30 and P40 components. We suggest that projections of inspiratory afferents from the diaphragm and the intercostal muscles, activated by inspiratory loading, could be responsible for the lengthened latency of median nerve SEP components. Thus, respiratory afferents very likely interact with pathways of the somatosensory system.     

Expiratory activity of the inspiratory muscles during cough.

To investigate the neural mechanism of the expiratory activity of the inspiratory muscles during a cough, EMG of the respiratory muscles were recorded in anesthetized and tracheostomized dogs. A laparoscope was used to minimize injury to the abdominal muscles for implantation of the electrodes into the costal diaphragm. During the expulsive phase of a cough, the diaphragm was active in 7 of 12 dogs and the external intercostal muscle was active in 3 of 6 dogs. During a cough, the expiratory activity of the diaphragm, after the termination of its inspiratory activity, started at 52.9 +/- 24.6 ms, and that of external intercostal muscle started at 51.1 +/- 20.5 ms. The expiratory activity of the internal intercostal muscle and of the transversus abdominis started at 34.3 +/- 13.0 and 27.8 +/- 15.2 ms, respectively. The onset of expiratory activity of the inspiratory muscles is significantly later than that of expiratory muscles. Continuous activity in the expiratory muscles evoked by airway occlusion, i.e., Hering-Breuer reflex, was suppressed during the inspiratory phase of a cough, but not suppressed during the expulsive phase even when the expiratory activity of the diaphragm was observed. We concluded that the expiratory activity of inspiratory muscles is controlled independently of both expiratory activity of the expiratory muscles and inspiratory activity of the inspiratory muscles.     

Role of upper cervical inspiratory neurons studied by cross-correlation in the cat

Summary Axonal projections and synaptic connectivity of upper cervical inspiratory neurons (UCINs) were investigated in anaesthetised cats to clarify their role as propriospinal respiratory interneurons. Antidromic mapping showed axonal collaterals near phrenic and intercostal motonuclei. Of the UCINs tested, 37% had collaterals at T3-4; 55% had ipsilateral projections and 45% had contralateral projections. Ipsilateral or contralateral cross-correlations of the activity of pairs of UCINs (one on each side of the spinal cord) with the discharge of internal intercostal, external intercostal (T3-4) or phrenic nerves revealed similar features. Those with the internal intercostal and phrenic nerves were interpreted as evidence for shared or oligosynaptic excitation, those with the external intercostal nerve as shared excitation and inhibition. No evidence for monosynaptic connections was found. Monosynaptic connections could also not be demonstrated between inspiratory intercostal neurons located near (< 0.5 mm) the UCINs collateral arborizations in T3-4, examined by cross-correlation. Afferent feedback from internal intercostal nerves (T3-4) was investigated by cross-correlating nerve stimulation with UCINs activity. Ipsilateral and contralateral cross-correlograms had similar features, providing evidence for excitation in some cases and inhibition in others. Finally, cross-correlations between ipsilateral UCINs and cervical sympathetic nerves were featureless. The results suggest that the role of UCINs as part of a respiratory propriospinal control system analagous to forelimb motor control is untenable, although they may be part of an intercostal afferent feedback loop.     

Chemical activation of caudal medullary expiratory neurones alters the pattern of breathing in the cat.

1. The purpose of this work was to ascertain whether the activation of caudal expiratory neurones located in the caudal part of the ventral respiratory group (VRG) may affect the pattern of breathing via medullary axon collaterals. 2. We used microinjections of DL-homocysteic acid (DLH) to activate this population of neurones in pentobarbitone-anaesthetized, vagotomized, paralysed and artificially ventilated cats. Both phrenic and abdominal nerve activities were monitored; extracellular recordings from medullary and upper cervical cord respiratory neurones were performed. 3. DLH (160 mM) microinjected (10-30 nl for a total of 1.6-4.8 nmol) into the caudal VRG, into sites where expiratory activity was encountered, provoked an intense and sustained activation of the expiratory motor output associated with a corresponding period of silence in phrenic nerve activity. During the progressive decline of the activation of abdominal motoneurones, rhythmic inspiratory activity resumed, displaying a decrease in frequency and a marked reduction or the complete suppression of postinspiratory activity as its most consistent features. 4. Medullary and upper cervical cord inspiratory neurones exhibited inhibitory responses consistent with those observed in phrenic nerve activity, while expiratory neurones in the caudal VRG on the side contralateral to the injection showed excitation patterns similar to those of abdominal motoneurones. On the other hand, in correspondence to expiratory motor output activation, expiratory neurones of the Bötzinger complex displayed tonic discharges whose intensity was markedly lower than the peak level of control breaths. 5. Bilateral lignocaine blockades of neural transmission at C2-C3 affecting the expiratory and, to a varying extent, the inspiratory bulbospinal pathways as well as spinal cord transections at C2-C3 or C1-C2, did not suppress the inhibitory effect on inspiratory neurones of either the ipsi- or contralateral VRG in response to DLH microinjections into the caudal VRG. 6. The results show that neurones within the column of caudal VRG expiratory neurones promote inhibitory effects on phrenic nerve activity and resetting of the respiratory rhythm. We suggest that these effects are mediated by medullary bulbospinal expiratory neurones, which may, therefore, have a function in the control of breathing through medullary axon collaterals.     

Analysis of activity of motor units in the biceps brachii muscle after intercostal-musculocutaneous nerve transfer

We examined respiratory activity of motor units (MUs) in the internal intercostal nerves (IICNs)-transferred biceps brachii muscle (IC-biceps) in cats. MUs of IC-biceps showed respiratory discharges in inspiratory and expiratory phases, and these were enhanced by CO2 inhalation. Narrowing the airway also enhanced inspiratory and expiratory MUs activity. A mechanical load to the thorax immediately enhanced inspiratory MUs activity and weakened expiratory MUs activity. We analyzed the cross-correlation of MUs activity in interchondral muscle and IC-biceps to characterize the respiratory spinal descending inputs to motoneurons. We confirmed the short-term synchronization from interchondral muscles indicating divergence of a single respiratory presynaptic axon to thoracic motoneurons, but could not find synchronization from IC-biceps. The motor axonal conduction velocity (axonal CV) of IC-biceps MUs was lower than that of interchondral muscles. There was no correlation between the respiratory recruitment order of IC-biceps MUs and their axonal CV. These results indicate that IC-biceps shows the respiratory activities and afferent inputs from intercostal muscle spindles in the neighboring segments remain influential on activity of IC-biceps. In addition, the short-term synchronization from IC-biceps could not be found, suggesting that the intercostal nerve transfer alters the respiratory spinal descending inputs to thoracic motoneurons.     

Diaphragmatic pacing stimulation in spinal cord injury: anesthetic and perioperative management

OBJECTIVE:

The standard therapy for patients with high-level spinal cord injury is long-term mechanical ventilation through a tracheostomy. However, in some cases, this approach results in death or disability. The aim of this study is to highlight the anesthetics and perioperative aspects of patients undergoing insertion of a diaphragmatic pacemaker.

METHODS:

Five patients with quadriplegia following high cervical traumatic spinal cord injury and ventilator-dependent chronic respiratory failure were implanted with a laparoscopic diaphragmatic pacemaker after preoperative assessments of their phrenic nerve function and diaphragm contractility through transcutaneous nerve stimulation. ClinicalTrials.gov: {"type":"clinical-trial","attrs":{"text":"NCT01385384","term_id":"NCT01385384"}}NCT01385384.

RESULTS:

The diaphragmatic pacemaker placement was successful in all of the patients. Two patients presented with capnothorax during the perioperative period, which resolved without consequences. After six months, three patients achieved continuous use of the diaphragm pacing system, and one patient could be removed from mechanical ventilation for more than 4 hours per day.

CONCLUSIONS:

The implantation of a diaphragmatic phrenic system is a new and safe technique with potential to improve the quality of life of patients who are dependent on mechanical ventilation because of spinal cord injuries. Appropriate indication and adequate perioperative care are fundamental to achieving better results.     

Intercostal and abdominal muscle afferent influence on caudal medullary expiratory neurons that drive abdominal muscles

Summary Our objective was to determine if caudal ventral respiratory group (VRG) expiratory (E) neurons that drive abdominal expiratory motoneurons in the lumbar cord respond to intercostal and lumbar nerve afferent stimulation. Results showed that 92% of medullary E-neurons that were antidromically activated from the upper lumbar cord reduced their activity in response to stimulation of external and internal intercostal and lumbar nerve afferents. We conclude that afferent information from intercostal and abdominal muscle tendon organs has an inhibitory effect on caudal VRG E-neurons that drive abdominal expiratory motoneurons.This study was supported by National Heart, Lung, and Blood Institute grant RO1-HL-17715     

Neurotization of the phrenic nerve with accessory nerve: a new strategy for high cervical spinal cord injury with respiratory distress

The prevalence of high cervical spinal cord injury has been rising and the life quality of these survivors remains poor. Even though mechanical ventilation prolongs their lifespans, the complications of mechanical obstruction and infection always perplex the doctors and patients. While phrenic nerve pacing was developed to improve the survival quality of them and have an analogous negative pressure mechanism. Herein we postulate that a potential physiological respiration may be resulted from neurotization of the phrenic nerve with accessory nerve. Once the potential strategy can be succeeded in the clinical application, patients will acquire remarkable survival profit.     

Experimental evaluation of the optimal tidal volume for simultaneous pacing of the diaphragm and respiratory muscles

Because of problems with pacing devices, surgical procedures, and diaphragm fatigue in pacing therapy of the phrenic nerve, we performed simultaneous pacing of the diaphragm alone and of multiple respiratory muscles in dogs and evaluated the optimal tidal volume. After intravenously anesthetizing 20 dogs with an average weight of 11kg, their tidal volume was measured with a spirometer to obtain control values. In the first 4 dogs, electrodes were sutured to the diaphragm and the optimal voltage, pulse width, and output to maximize tidal volume were determined. In the remaining 16 dogs, we stimulated individual canine respiratory muscles, i.e., the diaphragm, the rectus thoracis, and intercostal muscles 3-5 and simultaneously stimulated the diaphragm and the rectus thoracis; the diaphragm and intercostal muscles; the rectus thoracis and the intercostal muscles; or the diaphragm, rectus thoracis, and intercostal muscles. We compared a group in which a counterelectrode was positioned in each muscle group (group A) with a group in which no counterelectrode was used (group B). The best tidal volume was obtained at 10V, 50Hz, and a pulse width of 1.0ms. All the respiratory muscle pacings yielded better tidal volumes in group B than in group A. The greatest tidal volume was obtained with the rectus thoracis and intercostal muscle combination, suggesting the possibility of being able to reduce diaphragm fatigue by alternate pacing of these muscles and the diaphragm.     

The output from human inspiratory motoneurone pools

Survival requires adequate pulmonary ventilation which, in turn, depends on adequate contraction of muscles acting on the chest wall in the presence of a patent upper airway. Bulbospinal outputs projecting directly and indirectly to 'obligatory' respiratory motoneurone pools generate the required muscle contractions. Recent studies of the phasic inspiratory output of populations of single motor units to five muscles acting on the chest wall (including the diaphragm) reveal that the time of onset, the progressive recruitment, and the amount of motoneuronal drive (expressed as firing frequency) differ among the muscles. Tonic firing with an inspiratory modulation of firing rate is common in low intercostal spaces of the parasternal and external intercostal muscles but rare in the diaphragm. A new time and frequency plot has been developed to depict the behaviour of the motoneurone populations. The magnitude of inspiratory firing of motor unit populations is linearly correlated to the mechanical advantage of the intercostal muscle region at which the motor unit activity is recorded. This represents a 'neuromechanical' principle by which the CNS controls motoneuronal output according to mechanical advantage, presumably in addition to the Henneman's size principle of motoneurone recruitment. Studies of the genioglossus, an obligatory upper airway muscle that helps maintain airway patency, reveal that it receives simultaneous inspiratory, expiratory and tonic drives even during quiet breathing. There is much to be learned about the neural drive to pools of human inspiratory and expiratory muscles, not only during respiratory tasks but also in automatic and volitional tasks, and in diseases that alter the required drive.     

Spinal integration of segmental, cortical and breathing inputs to thoracic respiratory motoneurones

1. The spinal integration of cortical, segmental and breathing inputs to thoracic motoneurones was studied in anaesthetized, paralysed cats: the breathing input was intensified by underventilation or abolished by hyperventilation.2. In apnoeic animals, low intensity stimulation of an internal intercostal nerve evoked a brief latency polysynaptic reflex discharge of expiratory motoneurones (direct response) in several adjacent segments with no or little response of the inspiratory motoneurones.3. A similar direct response of expiratory motoneurones occurred with brief tetanic stimulation of the trunk area in the contralateral sensorimotor cortex.4. Conditioning of an intercostal-intercostal test reflex by a prior stimulus to an intercostal nerve or to the cortex gave conditioning curves showing facilitation of transmission to expiratory motoneurones at short intervals (5-25 msec) and inhibition at long intervals (25-200 msec).5. The direct response of expiratory motoneurones to the cortical or segmental inputs was depressed during the inspiratory phase when the animal was underventilated; conversely the spontaneous activity of the inspiratory motoneurones was inhibited for a period that corresponded with the direct response or to the phase of facilitated transmission to expiratory motoneurones. During the expiratory phase, the cortically or segmentally induced direct response was facilitated but the inhibition of inspiratory motoneurone activity was concealed by the absence of spontaneous activity.6. It was possible with discrete lesions of the spinal cord to differentiate between the pathways subserving the responses to cortical stimulation and the spontaneous activity due to the breathing input.7. To account for the results a working hypothesis is proposed utilizing a segmental interneuronal network which transmits mutual reciprocal inhibition between inspiratory and expiratory motoneurones.     

Rhythmic phrenic,intercostal and sympathetic activity in relation to limb and trunk motor activity in spinal cats

During L-DOPA-induced fictive spinal locomotion rhythmic activities in nerves to internal intercostal and external oblique abdominal muscles and in phrenic and sympathetic nerves were observed which were always coordinated with locomotor activity in forelimb and hindlimb muscle nerves. A periodicity with longer lasting tonic phases could be induced by cutaneous nerve stimulation or asphyxia. This activity was observed in limb motor nerves as well as in respiratory motor and sympathetic nerves. A slow independent activity of the phrenic and intercostal nerves or the sympathetic nerves, which could be related to a normal respiratory rhythm or independent sympathetic rhythms was not observed. The findings indicate that during fictive spinal locomotion the activity of spinal rhythm generators for locomotion also projects onto respiratory and sympathetic spinal neurones.     

Intercostal and abdominal respiratory motoneurons in the neonatal rat spinal cord: spatiotemporal organization and responses to limb afferent stimulation

Respiration requires the coordinated rhythmic contractions of diverse muscles to produce ventilatory movements adapted to organismal requirements. During fast locomotion, locomotory and respiratory movements are coordinated to reduce mechanical conflict between these functions. Using semi-isolated and isolated in vitro brain stem-spinal cord preparations from neonatal rats, we have characterized for the first time the respiratory patterns of all spinal intercostal and abdominal motoneurons and explored their functional relationship with limb sensory inputs. Neuroanatomical and electrophysiological procedures were initially used to locate intercostal and abdominal motoneurons in the cord. Intercostal motoneuron somata are distributed rostrocaudally from C₇–T₁₃ segments. Abdominal motoneuron somata lie between T₈ and L₂. In accordance with their soma distributions, inspiratory intercostal motoneurons are recruited in a rostrocaudal sequence during each respiratory cycle. Abdominal motoneurons express expiratory-related discharge that alternates with inspiration. Lesioning experiments confirmed the pontine origin of this expiratory activity, which was abolished by a brain stem transection at the rostral boundary of the VII nucleus, a critical area for respiratory rhythmogenesis. Entrainment of fictive respiratory rhythmicity in intercostal and abdominal motoneurons was elicited by periodic low-threshold dorsal root stimulation at lumbar (L₂) or cervical (C₇) levels. These effects are mediated by direct ascending fibers to the respiratory centers and a combination of long-projection and polysynaptic descending pathways. Therefore the isolated brain stem-spinal cord in vitro generates a complex pattern of respiratory activity in which alternating inspiratory and expiratory discharge occurs in functionally identified spinal motoneuron pools that are in turn targeted by both forelimb and hindlimb somatic afferents to promote locomotor-respiratory coupling.