Respiratory muscle control during vomiting

The changes in thoracic and abdominal pressure that generate vomiting are produced by coordinated action of the major respiratory muscles. During vomiting, the diaphragm and external intercostal (inspiratory) muscles co-contract with abdominal (expiratory) muscles in a series of bursts of activity that culminates in expulsion. Internal intercostal (expiratory) muscles contract out of phase with these muscles during retching and are inactive during expulsion. The periesophageal portion of the diaphragm relaxes during expulsion, presumably facilitating rostral movement of gastric contents. Recent studies have begun to examine to what extent medullary respiratory neurons are involved in the control of these muscles during vomiting. Bulbospinal expiratory neurons in the ventral respiratory group caudal to the obex discharge at the appropriate time during (fictive) vomiting to activate either abdominal or internal intercostal motoneurons. The pathways that drive phrenic and external intercostal motoneurons during vomiting have yet to be identified. Most bulbospinal inspiratory neurons in the dorsal and ventral respiratory groups do not have the appropriate response pattern to initiate activation of these motoneurons during (fictive) vomiting. Relaxation of the periesophageal diaphragm during vomiting could be brought about, at least in part, by reduced firing of bulbospinal inspiratory neurons.     

Influences from laryngeal afferents on expiratory bulbospinal neurons and motoneurons

The purpose of this study is to analyze the reflex effects of laryngeal afferent activation on respiratory patterns in anesthetized, vagotomized, paralyzed, ventilated cats. We recorded simultaneously from the phrenic nerve, T10 internal intercostal nerve, and single bulbospinal expiratory neurons of the caudal ventral respiratory group (VRG). Laryngeal afferents were activated by electrical stimulation of the superior laryngeal nerve (SLN) or by cold-water infusion into the larynx. Both types of stimuli caused inhibition of phrenic activity and facilitation of internal intercostal nerve activity, indicating expiratory effort. The activity of 46 bulbospinal expiratory cells was depressed during SLN electrical stimulation, and 13 of them were completely inhibited. In 44 of 56 neurons tested, mean firing frequency (FFmean) was decreased in response to cold-water infusion and 8 others responded with increased FFmean; in the remaining 4 neurons, FFmean was unchanged. Possible reasons for different neuronal responses to SLN electrical stimulation and water infusion are discussed. We conclude that bulbospinal expiratory neurons of VRG were not the source of the reflex motoneuronal expiratory-like activity produced by SLN stimulation. Other, not yet identified inputs to spinal expiratory motoneurons are activated during this experimental condition.     

Ventrolateral medullary respiratory network participation in the expiration reflex in the cat.

The expiration reflex is a distinct airway defensive response characterized by a brief, intense expiratory effort and coordinated adduction and abduction of the laryngeal folds. This study addressed the hypothesis that the ventrolateral medullary respiratory network participates in the reflex. Extracellular neuron activity was recorded with microelectrode arrays in decerebrated, neuromuscular-blocked, ventilated cats. In 32 recordings (17 cats), 232 neurons were monitored in the rostral (including B?tzinger and pre-B?tzinger complexes) and caudal ventral respiratory group. Neurons were classified by firing pattern, evaluated for spinal projections, functional associations with recurrent laryngeal and lumbar nerves, and firing rate changes during brief, large increases in lumbar motor nerve discharge (fictive expiration reflex, FER) elicited during mechanical stimulation of the vocal folds. Two hundred eight neurons were respiratory modulated, and 24 were nonrespiratory; 104 of the respiratory and 6 of the nonrespiratory-modulated neurons had altered peak firing rates during the FER. Increased firing rates of bulbospinal neurons and expiratory laryngeal premotor and motoneurons during the expiratory burst of FER were accompanied by changes in the firing patterns of putative propriobulbar neurons proposed to participate in the eupneic respiratory network. The results support the hypothesis that elements of the rostral and caudal ventral respiratory groups participate in generating and shaping the motor output of the FER. A model is proposed for the participation of the respiratory network in the expiration reflex.     

Influences of superior laryngeal afferent stimulation on expiratory activity in cats

The effects of superior laryngeal nerve (SLN) stimulation on the activity of the expiratory muscles and medullary expiration-related (ER) neurons were investigated in 24 pentobarbital-anesthetized cats. In some experiments the animals were also paralyzed and artificially ventilated. Sustained tetanic stimulation of SLN consistently caused an apneic response associated with the appearance of tonic CO2-dependent activity in the expiratory muscles and in ER neurons located in the caudal ventral respiratory group (VRG) and the B?tzinger complex. Single shocks or brief tetani at the same stimulation intensities failed to evoke excitatory responses in the expiratory muscles and in the vast majority of ER neurons tested. At higher stimulation strengths, single shocks or short tetani elicited excitatory responses in the expiratory muscles (20- to 35-ms latency) and in the majority of ER neurons of the caudal VRG (7.5- to 15.5-ms latency). These responses were obtained only during the expiratory phase and proved to be CO2 independent. On the contrary, only inhibitory responses were evoked in the activity of B?tzinger complex neurons. The observed tonic expiratory activity most likely represents a disinhibition phenomenon due to the suppression of inspiratory activity; activation of expiratory muscles at higher stimulation intensities appears to be a polysynaptic reflex mediated by ER neurons of the caudal VRG but not by B?tzinger complex neurons.     

Ventrolateral medullary respiratory network and a model of cough motor pattern generation

The primaryhypothesis of this study was that the cough motor pattern is produced,at least in part, by the medullary respiratory neuronal network inresponse to inputs from "cough" and pulmonary stretch receptorrelay neurons in the nucleus tractus solitarii. Computer simulations ofa distributed network model with proposed connections from the nucleustractus solitarii to ventrolateral medullary respiratory neuronsproduced coughlike inspiratory and expiratory motor patterns. Predictedresponses of various "types" of neurons (I-DRIVER, I-AUG, I-DEC,E-AUG, and E-DEC) derived from the simulations were tested in vivo.Parallel and sequential responses of functionally characterizedrespiratory-modulated neurons were monitored during fictive cough indecerebrate, paralyzed, ventilated cats. Coughlike patterns in phrenicand lumbar nerves were elicited by mechanical stimulation of theintrathoracic trachea. Altered discharge patterns were measured in mosttypes of respiratory neurons during fictive cough. The resultssupported many of the specific predictions of our cough generationmodel and suggested several revisions. The two main conclusions were asfollows: 1) TheBötzinger/rostral ventral respiratory group neurons implicated inthe generation of the eupneic pattern of breathing also participate inthe configuration of the cough motor pattern.2) This altered activity ofBötzinger/rostral ventral respiratory group neurons istransmitted to phrenic, intercostal, and abdominal motoneurons via thesame bulbospinal neurons that provide descending drive during eupnea.

     

Electrophysiological evaluation of the intensity of protective respiratory reflexes

The impulse activity of bulbar respiratory neurons and the electrical activity of main respiratory muscles were studied stereotaxically and electromyographically on 21 male and female cats anesthetized with pentobarbital (40 mg/kg, i.p.) during defensive respiratory (expiratory and coughing) reflexes. During stimulation of laryngopharyngeal and tracheobronchial receptors, a pronounced focus of excitation appears in the bulbar respiratory centre, its peripheral manifestation being powerful electrical activity of expiratory muscles (expiratory reflex) or of both expiratory and inspiratory muscles (coughing). Respiratory defensive reflexes are very powerful and stable and are retained in hypercapnia and hypoxia.     

Activity of the laryngeal abductor and adductor muscles during cough, expiration and aspiration reflexes in cats

We studied the temporal relationships and the patterns of electromyographic activities of the posterior cricoarytenoid and thyreoarytenoid muscles (laryngeal abductor and adductor), the diaphragm and abdominal muscles in anesthetized cats during mechanically induced tracheobronchial and laryngopharyngeal coughs, expiration and aspiration reflexes. The posterior cricoarytenoid muscle activity reached the maxima just before the peak of diaphragmatic activity in both types of cough and aspiration reflexes and slightly before the top of abdominal muscle activity in coughs and the expiration reflex. Thus, this muscle contributes to the inspiratory phase of coughs and aspiration reflex and also to the expulsive phase of coughs and the expiration reflex. The thyreoarytenoid muscle presented strong discharges in the compressive phase of coughs and expiration reflex (during the rising part of the abdominal muscle activity) and in the subsequent laryngoconstriction (following the diaphragmal and/or abdominal muscle activity) in all four reflexes. This muscle was also slightly activated at the beginning of the aspiration reflex. The existence of four phases of the cough reflex is also discussed.     

Microinjection of DLH into the region of the caudal ventral respiratory column in the cat: evidence for an endogenous cough-suppressant mechanism.

The caudal ventral respiratory column (cVRC) contains premotor expiratory neurons that play an important role in cough-related expiratory activity of chest wall and abdominal muscles. Microinjection of d,l-homocysteic acid (DLH) was used to test the hypothesis that local activation of cVRC neurons can suppress the cough reflex. DLH (20-50 mM, 10-30 nl) was injected into the region of cVRC in nine anesthetized spontaneously breathing cats. Repetitive coughing was elicited by mechanical stimulation of the intrathoracic airways. Electromyograms (EMG) were recorded bilaterally from inspiratory parasternal and expiratory transversus abdominis (ABD) and unilaterally from laryngeal posterior cricoarytenoid and thyroarytenoid muscles. Unilateral microinjection of DLH (1-1.5 nmol) elicited bilateral increases in tonic and phasic respiratory ABD EMG activity, and it altered the respiratory pattern and laryngeal motor activities. However, DLH also decreased cough frequency by 51 +/- 7% compared with control (P < 0.001) and the amplitude of the contralateral (-35 +/- 3%; P < 0.001) and ipsilateral (-34 +/- 5%; P < 0.001) ABD EMGs during postinjection coughs compared with control. The cough alterations were much less pronounced after microinjection of a lower dose of DLH (0.34-0.8 nmol). No cough depression was observed after microinjections of vehicle. These results suggest that an endogenous cough suppressant neuronal network in the region of the cVRC may exist, and this network may be involved in the control of cough reflex excitability.     

Medullary raphe neuron activity is altered during fictive cough in the decerebrate cat.

Chemical lesions in the medullary raphe nuclei region influence cough. This study examined whether firing patterns of caudal medullary midline neurons were altered during cough. Extracellular neuron activity was recorded with microelectrode arrays in decerebrated, neuromuscular-blocked, ventilated cats. Cough-like motor patterns (fictive cough) in phrenic and lumbar nerves were elicited by mechanical stimulation of the intrathoracic trachea. Discharge patterns of respiratory and nonrespiratory-modulated neurons were altered during cough cycles (58/133); 45 increased and 13 decreased activity. Fourteen cells changed firing rate during the inspiratory and/or expiratory phases of cough. Altered patterns in 43 cells were associated with the duration of, or extended beyond, the cough episodes. The different response categories suggest that multiple factors influence the discharge patterns during coughing: e.g., respiratory-modulated and tonic inputs and intrinsic connections. These results suggest involvement of midline neurons (i.e., raphe nuclei) in the cough reflex.     

Medullary expiratory activity: influence of intercostal tendon organs and muscle spindle endings

Studies were conducted to determine the effects of intercostal muscle spindle endings (MSEs) and tendon organs (TOs) on medullary expiratory activity in decerebrate cats. Impeded intercostal muscle contractions, elicited by electrical stimulation of the peripheral cut end of the T6 ventral root, were used to stimulate intercostal TOs without MSEs. Impeded contractions of the intercostal muscles augmented expiratory laryngeal motoneuron activity, and either had no effect on or reduced the activity of bulbospinal expiratory neurons. Vibration was used to stimulate intercostal MSEs. Intercostal MSEs had no effect on medullary expiratory neuron activity. It is concluded that both external and internal intercostal TOs have an excitatory effect on expiratory laryngeal motoneuron activity and an inhibitory effect on a subpopulation of expiratory neurons driving intercostal and/or abdominal muscles, and intercostal MSEs have no direct influence on medullary expiratory activity.     

The effects of cold-induced muscle spindle secondary activity on monosynaptic and stretch reflexes in the decerebrate cat.

The effects of muscle spindle secondary ending activity on the stretch reflex were studied in unanesthetized decerebrate cats. Activation of secondary endings was accomplished by reducing the muscle temperature. This has been shown to cause a sustained asynchronous discharge from secondary endings. Cooling of the medial gastrocnemius or lateral gastrocnemius-soleus muscles caused an increase in the phasic and tonic components of their stretch reflexes. Cooling of the relaxed medial gastrocnemius muscle caused similar increases in the components of the stretch reflex of the synergistic lateral gastrocnemius-soleus muscle and an increase in its monosynaptic reflex. It was concluded that the facilitatory autogenetic and synergistic effects of muscle cooling on the stretch and monosynaptic reflexes were brought about by activity in group II afferents from muscle spindle secondary endings and could not be ascribed to any other type of muscle receptor. These results support the concept of an excitatory role for the secondary endings of the muscle spindle in the stretch reflex of the decerebrate cat.     

Calcium as a counteractive agent to streptomycin induced respiratory depression: an in vivo electrophysiological observation.

Effect of streptomycin on respiratory function in cats was studied. It was observed that streptomycin at a dose of 40 mg/kg body weight intravenously (i.v.) caused respiratory failure or streptomycin induced respiratory depression (SIRD). This respiratory failure is not linked with Herring-Breuer stretch receptors because the effect remained unaltered in artificially ventilated cats. The involvement of central structures in SIRD can be discarded since intracarotid and intraventricular administration of streptomycin failed to produce any change in respiration. Studies on monosynaptic reflex, dorsal and ventral root activities of spinal phrenic and intercostal nerves, and on fusimotor and alpha-motor neuron activities of spinal intercostal and phrenic nerves in decerebrated cats indicated clearly that respiratory depression is not only due to blockade at neuromuscular junction but due to functional depression at the level of muscle receptors and spinal cord motor neurons. The respiratory depression induced by streptomycin was more or less completely reversed when calcium was administered intravenously from external source. It is speculated that streptomycin induced respiratory depression may be mediated through calcium inhibition which can be treated with external calcium in conjunction with artificial respiration.     

Reflex effect of esophageal distension on respiratory muscle activity and pressure

The electrical activity of the respiratory skeletal muscles is altered in response to reflexes originating in the gastrointestinal tract. The present study evaluated the reflex effects of esophageal distension (ED) on the distribution of motor activity to both inspiratory and expiratory muscles of the rib cage and abdomen and the resultant changes in thoracic and abdominal pressure during breathing. Studies were performed in 21 anesthetized spontaneously breathing dogs. ED was produced by inflating a balloon in the distal esophagus. ED decreased the activity of the costal and crural diaphragm and external intercostals and abolished all preexisting electrical activity in the expiratory muscles of the abdominal wall. On the other hand, ED increased the activity of the parasternal intercostals and expiratory muscles located in the rib cage (i.e., triangularis sterni and internal intercostal). All effects of ED were graded, with increasing distension exerting greater effects, and were eliminated by vagotomy. The effect of increases in chemical drive and lung inflation reflex activity on the response to ED was examined by performing ED while animals breathed either 6.5% CO2 or against graded levels of positive end-expiratory pressure (PEEP), respectively. Changes in respiratory muscle electrical activity induced by ED were similar (during 6.5% CO2 and PEEP) to those observed under control conditions. We conclude that activation of mechanoreceptors in the esophagus reflexly alters the distribution of motor activity to the respiratory muscles, inhibiting the muscles surrounding the abdominal cavity and augmenting the parasternals and expiratory muscles of the chest wall.     

Internal intercostal nerve discharges in the cat: influence of chemical stimuli

We studied the influence of central and peripheral chemoreceptor stimulation on the activities of the phrenic and internal intercostal (iic) nerves in decerebrate, vagotomized, and paralyzed cats with bilateral pneumothoraces. Whole iic nerves of the rostral thorax (T2-T5) usually discharged during neural inspiration, whereas those of the caudal thorax (T7-T11) were primarily active during neural expiration. Filaments of rostral iic nerves that terminated in iic muscles generally discharged during expiration, suggesting that inspiratory activity recorded in whole iic nerves may have innervated other structures, possibly parasternal muscles. All nerves were phasically active at hyperoxic normocapnia and increased their activities systematically with hypercapnia. Isocapnic hypoxia or intra-arterial NaCN injection consistently increased phrenic and inspiratory iic nerve activities. In contrast, expiratory iic nerve discharges were either decreased (10 cats) or increased (7 cats) by hypoxia. Furthermore, expiratory responses to NaCN were highly variable and could not be predicted from the corresponding response to hypoxia. The results show that central and peripheral chemoreceptor stimulation can affect inspiratory and expiratory motoneuron activities differentially. The variable effects of hypoxia on expiratory iic nerve activity may reflect a relatively weak influence of carotid body afferents on expiratory bulbospinal neurons. However, the possibility that the magnitude of expiratory motoneuron activity is influenced by the intensity of the preceding centrally generated inspiratory discharge is also discussed.     

Bradykinin stimulates respiratory drive by activating pulmonary sympathetic afferents in the rabbit.

We recently identified a vagally mediated excitatory lung reflex by injecting hypertonic saline into the lung parenchyma (Yu J, Zhang JF, and Fletcher EC. J Appl Physiol 85: 1485-1492, 1998). This reflex increased amplitude and burst rate of phrenic (inspiratory) nerve activity and suppressed external oblique abdominal (expiratory) muscle activity. In the present study, we tested the hypothesis that bradykinin may activate extravagal pathways to stimulate breathing by assessing its reflex effects on respiratory drive. Bradykinin (1 microg/kg in 0.1 ml) was injected into the lung parenchyma of anesthetized, open-chest and artificially ventilated rabbits. In most cases, bradykinin increased phrenic amplitude, phrenic burst rate, and expiratory muscle activity. However, a variety of breathing patterns resulted, ranging from hyperpnea and tachypnea to rapid shallow breathing and apnea. Bradykinin acts like hypertonic saline in producing hyperpnea and tachypnea, yet the two agents clearly differ. Bradykinin produced a higher ratio of phrenic amplitude to inspiratory time and had longer latency than hypertonic saline. Although attenuated, bradykinin-induced respiratory responses persisted after vagotomy. We conclude that bradykinin activates multiple afferent pathways in the lung; portions of its respiratory reflexes are extravagal and arise from sympathetic afferents.     

Reductions of neural activities to upper airway muscles after elevations in static lung volume.

We evaluated the hypothesis that the tonic discharge of pulmonary stretch receptors significantly influences the respiratory-modulated activities of cranial nerves. Decerebrate and paralyzed cats were ventilated with a servo-respirator, which produced changes in lung volume in parallel with integrated phrenic activity. Activities of the facial, hypoglossal, and recurrent laryngeal nerves and nerves to the thyroarytenoid muscle and triangularis sterni were recorded. After a stereotyped pattern of lung inflation, tracheal pressure was held at 1, 2, 4, or 6 cmH2O during the subsequent ventilatory cycle. Increases in tracheal pressure caused progressive reductions in both inspiratory and expiratory cranial nerve activities and progressive elevations in triangularis sterni discharge; peak levels of phrenic activity declined modestly. Similar changes were observed in normocapnia and hypercapnia. We conclude that the tonic discharge of pulmonary stretch receptors is an important determinant of the presence and magnitude of respiratory-modulated cranial nerve activity. This reflex mechanism may maintain upper airway patency and also regulate expiratory airflow.     

Pattern of expiratory muscle activation during lower thoracic spinal cord stimulation.

Large positive airway pressures (Paws) can be generated by lower thoracic spinal cord stimulation (SCS), which may be a useful method of restoring cough in spinal cord-injured patients. Optimal electrode placement, however, requires an assessment of the pattern of current spread during SCS. Studies were performed in anesthetized dogs to assess the pattern of expiratory muscle recruitment during SCS applied at different spinal cord levels. A multicontact stimulating electrode was positioned over the surface of the lower thoracic and upper lumbar spinal cord. Recording electromyographic electrodes were placed at several locations in the abdominal and internal intercostal muscles. SCS was applied at each lead, in separate trials, with single shocks of 0.2-ms duration. The intensity of stimulation was adjusted to determine the threshold for development of the compound action potential at each electrode lead. The values of current threshold for activation of each muscle formed parabolas with minimum values at specific spinal root levels. The slopes of the parabolas were relatively steep, indicating that the threshold for muscle activation increases rapidly at more cephalad and caudal sites. These results were compared with the effectiveness of SCS (50 Hz; train duration, 1-2 s) at different spinal cord levels to produce changes in Paw. Stimulation at the T9 and T10 spinal cord level resulted in the largest positive Paws with a single lead. At these sites, threshold values for activation of the internal intercostal (7-11th interspaces) upper portions of external oblique, rectus abdominis, and transversus abdominis were near their minimum. Threshold values for activation of the caudal portions of the abdominal muscles were high (>50 mA). Our results indicate that 1) activation of the more cephalad portions of the abdominal muscles is more important than activation of caudal regions in the generation of positive Paws and 2) it is not possible to achieve complete activation of the expiratory muscles with a single electrode lead by using modest current levels. In support of this latter conclusion, a two-electrode lead system results in more uniform expiratory muscle activation and significantly greater changes in Paw.     

Responses of the anterolateral abdominal muscles during cough and expiratory threshold loading in the cat.

The present study was conducted to determine the pattern of activation of the anterolateral abdominal muscles during the cough reflex. Electromyograms (EMGs) of the rectus abdominis, external oblique, internal oblique, transversus abdominis, and parasternal muscles were recorded along with gastric pressure in anesthetized cats. Cough was produced by mechanical stimulation of the lumen of the intrathoracic trachea or larynx. The pattern of EMG activation of these muscles during cough was compared with that during graded expiratory threshold loading (ETL; 1-30 cmH(2)O). ETL elicited differential recruitment of abdominal muscle EMG activity (transversus abdominis > internal oblique > rectus abdominis congruent with external oblique). In contrast, both laryngeal and tracheobronchial cough resulted in simultaneous activation of all four anterolateral abdominal muscles with peak EMG amplitudes 3- to 10-fold greater than those observed during the largest ETL. Gastric pressures during laryngeal and tracheobronchial cough were at least eightfold greater than those produced by the largest ETL. These results suggest that, unlike their behavior during expiratory loading, the anterolateral abdominal muscles act as a unit during cough.     

Neural Control of Swimming in the Leech

Leeches swim by undulating; they alternately form crests thentroughs at their anterior end and move Them backward, therebyproducing forward thrust. These movements are accomplished byalternating contractions of dorsal and ventral longitudinalmuscles in each of the 21 body segments. These contractionsare caused by bursts of impulses in groups of excitatory andinhibitory motor neurons. Connections among motor neurons helpto coordinate these bursts: synergistic muscle excitors areelectrotonically coupled, which aids in keeping their burstsnearly synchronous; muscle inhibitors also inhibit the excitorsto the same muscles, and it is this inhibition which keeps theexcitors from being tonically active during swimming. Neuronssensitive to either dorsal or ventral body wall stretch producereciprocal stretch reflexes to the muscle excitors, probablyvia the inhibitors. That these stretch reflexes may be involvedin generating the periodic bursts is supported by the resultsof both behavioral and electrophysiological experiments.     

Central action of tachykinins on activity of expiratory pumping muscles.

The central effects of tachykinins (substance P, neurokinin A, and neurokinin B) on the distribution of the motor activity to rib cage and abdominal expiratory muscles were studied in anesthetized tracheotomized spontaneously breathing dogs and cats. Intracisternal application of substance P (11 dogs) in doses of 10(-5) to 10(-4) M caused diaphragm electrical activity to change insignificantly from 19.3 +/- 1.9 to 24.8 +/- 3.2 units (P greater than 0.05), produced a moderate increase of triangularis sterni activity from 12.6 +/- 2.2 to 19.2 +/- 2.2 units (P less than 0.05), and stimulated a large increase of transversus abdominis activity from 9.4 +/- 2.7 to 28.5 +/- 2.6 units (P less than 0.01). Comparable effects were seen with similar doses of neurokinin A (8 dogs) and neurokinin B (3 dogs) administered intracisternally. Local application of substance P to the ventral medullary surface (5 dogs and 4 cats) also caused expiratory muscle activity to increase more than diaphragm activity, and in addition transversus abdominis activity increased to a larger extent than triangularis sterni activity. Furthermore, administration of the substance P antagonist [D-Pro2,D-Trp7,9]-SP to the ventral medullary surface decreased respiratory motor output, with expiratory muscles activity being attenuated to a greater extent than diaphragm activity. Application of neurotensin and N-methyl-D-asparate to the ventral surface of the medulla produced responses similar to those observed as a result of central administration of tachykinin peptides. The results suggest that 1) mammalian tachykinins are involved in the regulation of thoracic and abdominal expiratory muscle activity, 2) these muscles manifest substantial differences in their electrical responses to excitatory neuropeptides acting centrally, and 3) inputs from modulatory neurons located in this vicinity of the ventral medullary surface seem to be distributed unevenly to different expiratory premotor and/or motoneurons.