[Clinical analysis and laboratory diagnosis of three cases with infantile botulism caused by Clostridium botulinum type B].

Objective: To summarize the clinical characteristics and laboratory diagnostic methods of infant botulism caused by Clostridium botulinum type B. Methods: Clinical data of 3 infants with type B botulism who were admitted to Children's Hospital Affiliated to Capital Institute of Pediatrics from May to November 2018 were retrospectively analyzed. Botulinum toxin was detected in fecal samples or fecal enrichment solution of the patients, and Clostridium botulinum was cultured and isolated from fecal samples. Results: The age of onset of the patients (two boys and one girl) was 3, 3 and 8 months old, respectively. Two cases had the onset in May and one case had the onset in November. There were two cases with mixed feeding and one case with breast feeding. One case's family members engaged in meat processing. All of them were previously healthy. All the children presented with acute flaccid paralysis, cranial nerve involvement and difficult defecation. Two cases had secondary urinary tract infection. Electromyograms of two cases showed that action potential amplitude of the motor nerve were lower than those of their peers. After treatments including intravenous human immunoglobulin, respiratory tract management, urethral catheterization, nasal feeding, etc., three cases recovered completely 2 to 4 months later. Type B botulinum toxin was detected in the fecal diluent of one patient, and the TPGYT enrichment solution and cooked meet medium of the feces of 3 patients, respectively. Clostridium botulinum B was identified from the feces of 3 infants after culture, isolation and purification. Conclusions: Combined with typical clinical manifestations including acute flaccid paralysis, cranial nerve involvement symptoms and difficult defecation examination, infant botulism can be clinically diagnosed. The detection of fecal botulinum toxin and the culture and isolation of Clostridium botulinum are helpful for the diagnosis.

Up-regulation of LncRNA MEG3 inhibits cell migration and invasion and enhances cisplatin chemosensitivity in bladder cancer cells.

It has been proven that maternally expressed 3 (MEG3), a long non-coding RNA (LncRNA), is down-regulated and inversely correlated with prognosis in various types of cancer, including bladder cancer (BC). Nevertheless, the role of MEG3 in BC has not been fully identified. Herein, we found that MEG3 expression was reduced in 21 BC tumor tissue samples compared to corresponding adjacent tissues. We then established T24 and 5637 cells with a stably integrated expression of MEG3 by G418 resistance screening, and data revealed that the BC cells over-expressing MEG3 displayed weaker migration and invasion ability than control cells. The expression and activity of matrix metalloproteinase (MMP)2 and MMP9 were down-regulated when MEG3 was over-expressed. Moreover, MEG3 over-expression sensitized BC cells to the chemotherapy drug cisplatin (DDP). DDP treatment significantly induced cell apoptosis, down-regulated bcl2 expression, and up-regulated cleaved-caspase-3 and bax expression in BC cells with MEG3 over-expression. MEG3 and p53 can also stimulate mutual expression in BC cells, thus indicating a potential positive feedback loop of MEG3 and p53. Our combined results suggest that over-expression of MEG3 inhibits migration and invasion and enhances DDP chemo-sensitivity in bladder cancer cells.

[Characteristics of cognitive impairment of different infarct locations among patients after acute ischemic stroke].

OBJECTIVE: To investigate the characteristics of cognitive function changes of different infarct sites among patients after acute ischemic stroke, so as to provide theoretical basis for preventing and treating vascular cognitive dysfunction. METHODS: One hundred and five cases of acute ischemic stroke within fourteen days meeting the standard set were enrolled, and they were tested by Mini-Mental State Examination (MMSE) and Montreal Cognitive Assessment (MoCA). And the characteristics of cognitive changes with different infarction sites were analyzed. RESULTS: Patients with acute stroke suffered cognitive impairment.The significantly impaired cognitive domains in MMSE were: graphics execution in patients with left hemisphere infarction (P=0.027); verbal repetition in frontal infarction (P=0.003); short memory (P=0.04) and verbal repetition (P=0.007) in parietal infarction.The significantly impaired cognitive domains in MoCA were: language (P=0.002), naming (P=0.011), attention (P=0.028) and time orientation (P=0.031) in frontal infarction; delayed memory (P<0.001), attention (P=0.041), language (P=0.049) and visual space and executive ability (P=0.049) in parietal infarction; attention in temporal infarction (P=0.045); language (P=0.009) and time orientation (P=0.026) in basal ganglia region infarction. CONCLUSION: Most ischemic stroke patients at acute phase suffered cognitive impairment and the characteristics of cognitive changes differed according to different infarction sites.Comprehensive assessment of cognitive impairment after acute stroke is of great importance.

Calbindin-immunoreactive neurons in the reticular formation of the rat brainstem: catecholamine content and spinal projections.

Calbindin-D28k (calbindin) is a calcium-binding protein that is distributed widely in the rat brain. The localisation of calbindin immunoreactivity in the medulla oblongata and its colocalisation with adrenaline-synthesising neurons [phenylethanolamine-N-methyltransferase-immunoreactive (PNMT-IR)] was examined (Granata and Chang [1994] Brain Res. 645:265-277). However, detailed information about the distribution of calbindin-IR neurons in the reticular formation of the medulla oblongata in particular is lacking. In this report, the authors address this issue with an emphasis on the quantitation of calbindin-IR neurons, catecholamine neurons [tyrosine hydroxylase (TH)-IR, or PNMT-IR], and spinally projecting neurons in the ventral brainstem. Rats received injections of the retrograde tracing agent cholera toxin B (CTB) into the thoracic spinal cord or into the superior cervical ganglion. Immunocytochemistry was used to reveal calbindin, TH, PNMT, and CTB immunoreactivity. Ten calbindin-IR cell groups were identified within the pontomedullary reticular formation. Seven previously undescribed but distinct clusters of calbindin-IR neurons were found. Within the ventral pons, a population of calbindin-IR neurons occurred dorsal but adjacent to the A5 cell group. These calbindin-IR neurons did not contain either TH or PNMT immunoreactivity, and few if any of these neurons projected to the spinal cord. A distinct group of calbindin-IR neurons was present in the ventral medulla. Seventy-five percent of these calbindin-IR neurons contained TH immunoreactivity, 45% contained PNMT immunoreactivity, and 21% were spinally projecting neurons. Spinally projecting, calbindin-IR neurons were a subpopulation of PNMT-IR cells. In the caudal ventral medulla, no TH-IR or PNMT-IR cells were calbindin-IR. In the intermediolateral cell column, close appositions of calbindin-IR terminals on identified sympathetic preganglionic neurons as well as calbindin-IR synapses indicated that these neurons may affect directly the sympathetic outflow. The results demonstrate for the first time the existence of a new subpopulation of spinally projecting, PNMT-IR neurons in the rostral ventrolateral medulla.

Spinal GABA(A) receptors do not mediate the sympathetic baroreceptor reflex in the rat.

Activation of baroreceptors causes efferent sympathetic nerve activity (SNA) to fall. Two mechanisms could account for this sympathoinhibition: disfacilitation of sympathetic preganglionic neurons (SPN) and/or direct inhibition of SPN. The roles that spinal GABA and glycine receptors play in the baroreceptor reflex were examined in anesthetized, paralyzed, and artificially ventilated rats. Spinal GABA(A) receptors were blocked by an intrathecal injection of bicuculline methiodide, whereas glycine receptors were blocked with strychnine. Baroreceptors were activated by stimulation of the aortic depressor nerve (ADN), and a somatosympathetic reflex was used as control. After an intrathecal injection of vehicle, there was no effect on any measured variable or evoked reflex. In contrast, bicuculline caused a dose-dependent increase in arterial pressure, SNA, phrenic nerve discharge, and it significantly facilitated the somatosympathetic reflex. However, bicuculline did not attenuate either the depressor response or sympathoinhibition evoked after ADN stimulation. Similarly, strychnine did not affect the baroreceptor-induced depressor response. Thus GABA(A) and glycine receptors in the spinal cord have no significant role in baroreceptor-mediated sympathoinhibition.

Identification of posterior cricoarytenoid motoneurons in the rat.

The posterior cricoarytenoid (PCA) muscle is the sole abductor of the larynx and is controlled by motoneurons located in the nucleus ambiguus. These motoneurons receive inputs from a variety of interneurons, including those that impart respiratory modulation, and are responsible for the phasic inspiratory activity of the PCA muscle. Identification of PCA motoneurons is therefore an essential initial step in understanding the mechanisms responsible for coordinated vocal cord abduction. We identified PCA motoneurons in the rat model by retrograde labeling, and following antidromic activation. A total of 194 neurons were identified by retrograde labeling with cholera toxin B subunit (CTB). Labeling was exclusively ipsilateral where the contralateral vagus and superior laryngeal nerves had been divided. The neurons were multipolar, with dimensions of 33.2 +/- 6.4 microm (mean +/- standard deviation) in length and 22.4 +/- 3.4 microm in width. The neurons were located within a range of 0.6 to 2.4 mm caudal to the caudal pole of the facial nerve, 1.2 to 1.7 mm lateral to the midline, and 1.5 to 2.3 mm deep to the dorsal surface of the medulla. The PCA motoneurons were antidromically activated by focal stimulation of the PCA muscle. The extracellular field was recorded in 5 rats, and the PCA motoneurons were found within a range of 0.8 to 1.7 mm caudal to the caudal pole of the facial nerve, 1.5 to 2.0 mm lateral to the midline, and 1.9 to 2.4 mm deep to the dorsal surface of the medulla. The mean conduction velocity ranged from 37.0 +/- 5.8 to 68.6 +/- 5.0 m/s. An extracellular antidromic field potential, which corresponds to the distribution of the PCA motoneuron pool demonstrated by retrograde labeling with CTB, can be reliably obtained in a rat model following focal PCA muscle stimulation.

Intracellular recording from posterior cricoarytenoid motoneurons in the rat.

The ability to maintain coordinated vocal cord abduction and upper airway patency is dependent on the integrity of the posterior cricoarytenoid (PCA) motoneurons and their multiple neural connections. Study of the PCA motoneurons represents the initial step in understanding the complex mechanisms responsible for coordinated vocal cord abduction and may provide an insight into the possible pathological processes underlying the various clinical presentations of vocal cord dysfunction. Intracellular recordings were made from 11 PCA motoneurons in Sprague-Dawley rats, which all showed an inspiratory augmenting discharge pattern that is also characteristic of phrenic nerve activity. The resting membrane potential was -56+/-11 mV. Two PCA motoneurons were injected with Neurobiotin to demonstrate neuronal morphology, which was found to be similar to that obtained by retrograde labeling with cholera toxin B subunit. The technique described for intracellular recording of PCA motoneurons should allow more detailed morphological, electrophysiological, and immunohistochemical information to be obtained, to thereby identify some of the factors responsible for maintaining normal function of the PCA muscle.

Substance P-immunoreactive boutons closely appose inspiratory protruder hypoglossal motoneurons in the cat.

In anesthetized cats, we recorded intracellularly from 26 hypoglossal motoneurons which were antidromically activated following electrical stimulation of either the medial or lateral branches of the hypoglossal nerve. Twenty-one of these neurons were protruder motoneurons 6 of which had inspiratory activity. Three of the protruder motoneurons with inspiratory activity were filled with Neurobiotin and found to be closely apposed to substance P-like immunoreactive nerve terminals.

The pre-Bötzinger complex and phase-spanning neurons in the adult rat.

To characterise respiratory neurons in the pre-Bötzinger complex of adult rats, extracellular recordings were made from 302 respiratory neurons in the ventral respiratory group of sodium pentobarbitone anaesthetised adult rats. Neurons were located 0 to 1.6 mm caudal to the facial nucleus, and ventral to the nucleus ambiguus. The pre-Bötzinger complex comprised expiratory neurons (22%, 22/100), inspiratory neurons (37%, 37/100) and phase-spanning neurons (41%, 41/100). In contrast, 80% (125/157) of Bötzinger neurons were expiratory, and 80% (36/45) of rostral ventral respiratory group neurons were inspiratory. Rostrocaudally, the pre-Bötzinger complex extended about 400 microns, starting at the caudal pole of the nucleus ambiguus compact formation. The pre-Bötzinger complex was also characterised by a predominance of propriobulbar neurons (81%, 13/16). Furthermore, 68% (33/48) of expiratory-inspiratory neurons found were located within the pre-Bötzinger complex. The variety of neuronal subtypes in the pre-Bötzinger complex, including many firing during the expiratory-inspiratory transition is consistent with the hypothesis that this nucleus plays a key role in respiratory rhythm generation in the adult rat.

Respiratory activity of the rat posterior cricoarytenoid muscle.

An anatomic and electrophysiological study of the rat posterior cricoarytenoid (PCA) muscle is described. The intramuscular nerve distribution of the PCA branch of the recurrent laryngeal nerve was demonstrated by a modified Sihler's stain. The nerve to the PCA was found to terminate in superior and inferior branches with a distribution that appeared to be confined to the PCA muscle. Electromyography (EMG) recordings of PCA muscle activity in anesthetized rats were obtained under stereotaxic control together with measurement of phrenic nerve discharge. A total of 151 recordings were made in 7 PCA muscles from 4 rats. Phasic inspiratory activity with a waveform similar to that of phrenic nerve discharge was found in 134 recordings, while a biphasic pattern with both inspiratory and post-inspiratory peaks was recorded from random sites within the PCA muscle on 17 occasions. The PCA EMG activity commenced 24.6 +/- 2.2 milliseconds (p < .0001) before phrenic nerve discharge. The results are in accord with findings of earlier studies that show that PCA muscle activity commences prior to inspiratory airflow and diaphragmatic muscle activity. The data suggest that PCA and diaphragm motoneurons share common or similar medullary pre-motoneurons. The earlier onset of PCA muscle activity may indicate a role for medullary pre-inspiratory neurons in initiating PCA activity.

Bötzinger neurons project towards bulbospinal neurons in the rostral ventrolateral medulla of the rat.

Sympathetic nerve activity often fluctuates with the respiratory cycle, but the central neurons that impose this respiratory modulation have not been conclusively identified. In the present study, we used intracellular recording and dye-filling to identify expiratory neurons in the Bötzinger complex. Our aim was to see if Bötzinger neurons project towards putative cardiovascular neurons in the rostral ventrolateral medulla. In the first series of experiments, histochemistry and immunohistochemistry were used to reveal the labelled Bötzinger neurons and neurons immunoreactive for tyrosine hydroxylase. Two out of four Bötzinger neurons had axon varicosities that were closely apposed to tyrosine hydroxylase-immunoreactive neurons with cell bodies located within 0.6 mm caudal to the facial nucleus (three and five close appositions, respectively). In a second series of studies, rats were injected with cholera toxin B into the intermediolateral cell column of the spinal cord 4-7 days before the electrophysiological recording. Eight of the fourteen labelled Bötzinger neurons had a direct projection towards cholera toxin B-labelled neurons in the rostral ventrolateral medulla. Close appositions were found on both somata and proximal dendrites (5 +/- 2 close appositions/neuron, n = 8). The present study supports the idea that a direct projection from Bötzinger neurons to presympathetic neurons in the rostral medulla plays a role in the respiratory modulation of sympathetic nerve activity.

Phosphate-activated glutaminase immunoreactivity in brainstem respiratory neurons.

The aim of this study was to determine if immunoreactivity for phosphate activated glutaminase (PAG), an enzyme involved in the biosynthesis of glutamate and a putative marker for neurons that use glutamate as a neurotransmitter, is present within respiratory neurons in the ventrolateral medulla oblongata. Intracellular recordings were obtained from neurons in the ventrolateral medulla of adult anaesthetised Sprague-Dawley rats. Neurons with a respiratory-related modulation of their membrane potential were filled with Neurobiotin (Vector, CA). After histochemical processing, sections of brainstem were examined by fluorescence and light microscopy. Some PAG immunoreactivity was found in all of the four types of respiratory neurons examined. PAG immunoreactivity was graded as strong or weak. (1) Of six inspiratory neurons in the rostral ventral respiratory group five were strongly PAG immunoreactive and one was weakly PAG immunoreactive. (2) Of six expiratory neurons in the caudal ventral respiratory group five were strongly PAG immunoreactive while one was weak. (3) Seven motoneurons in the nucleus ambiguous were all strongly PAG immunoreactive. (4) Five neurons in the Bötzinger area were examined. Four were weakly PAG immunoreactive while one contained strong PAG immunoreactivity. These data demonstrate a heterogeneity of PAG immunoreactivity amongst brainstem respiratory neurons.

Thyrotropin-releasing hormone immunoreactive boutons form close appositions with medullary expiratory neurons in the rat.

The aim of the present study was to assess the size of the input from TRH immunoreactive varicosities to medullary respiratory neurons in the Bötzinger complex and caudal ventral respiratory group. Neurobiotin was intracellularly injected into seven neurons in the Bötzinger complex, between 0.4 and 0.9 mm caudal to the facial nucleus. Five of the seven Bötzinger neurons had extensive local axonal projections, with bouton-like varicosities clustered predominantly between their somata and the nucleus ambiguus. Seven neurons in the caudal ventral respiratory group, located between 1.6 and 2.4 mm caudal to the facial nucleus, were also labelled. All but one caudal respiratory neurons had no, or very few, medullary collaterals. TRH immunoreactive fibres were seen in many medullary nuclei, including the ventral reticular formation. Bötzinger neurons were closely apposed by an average of 29 +/- 8 TRH immunoreactive boutons/neuron (mean +/- S.D., n = 7). In contrast, caudal ventral respiratory group neurons were apposed by only 5 +/- 3 TRH immunoreactive boutons/neuron (n = 7). Bötzinger neurons form many intramedullary and bulbospinal inhibitory connections with premotoneurons and motoneurons that are important in the timing, amplitude and shape, of respiratory activity. Our findings suggest a role for endogenous TRH-containing neurons in modulating the activity of inhibitory Bötzinger neurons and neurons in the caudal ventral respiratory group. The significance of the apparent difference in size of this input remains to be determined.

Bulbospinal sympatho-excitatory neurons in the rat caudal raphe.

OBJECTIVES: To explore the rat caudal raphe nuclei for neurons that respond to activation of baroreceptor nerves and that have a spinal axon, and to compare the behavioural properties of barosensitive bulbospinal neurons in the rat caudal raphe with the properties of barosensitive bulbospinal neurons in the rostral ventrolateral medulla. DESIGN: Extracellular unit recordings were obtained from an area extending up to 1.0 mm caudally from the caudal edge of the facial nucleus. Two sites were explored: the rostral ventrolateral medulla and the midline. MATERIALS AND METHODS: Single-unit recordings were made in anaesthetized (75 mg/kg chloral hydrate and 30 mg/kg sodium pentobarbitone then 3-6 mg intravenously as required), immobilized (2 mg pancuronium as required) Sprague-Dawley rats. Central respiratory drive was recorded from phrenic nerve discharge. The barosensitivity of single units was assessed by R-wave triggered histograms and by histograms of their responses to aortic nerve stimulation or to intravenous injection of phenylephrine. Nociceptors were activated by a brief pinch of the tail. RESULTS: Eleven spontaneously active units in the midline that were inhibited by baroreceptor stimulation and had a spinal axon were studied. Respiratory modulation was present and was predominantly inspiratory. Barosensitive neurons in the rostral ventrolateral medulla were activated by nociceptive inputs; midline barosensitive neurons were not. CONCLUSIONS: The behavioural characteristics of midline neurons differ from those of the bulbospinal barosensitive neurons in the rostral ventrolateral medulla, indicating that raphe spinal neurons have different sets of afferent inputs and may subserve to a distinct physiological role. The present paper is the first report of bulbospinal neurons in the rat caudal raphe that are inhibited by activation of arterial baroreceptors.

Thyrotropin-releasing hormone inputs are preferentially directed towards respiratory motoneurons in rat nucleus ambiguus.

In the present study, we assessed the extent of the thyrotropin-releasing hormone (TRH) input to motoneurons in the ambigual, facial, and hypoglossal nuclei of the rat using a combination of intracellular recording, dye filling, and immunohistochemistry. Twelve motoneurons in the rostral nucleus ambiguus were labelled by intracellular injection in vivo of Neurobiotin (Vector). Seven out of 12 ambigual motoneurons displayed rhythmic fluctuations of their membrane potential in phase with phrenic nerve discharge, whereas the other five had no modulations of any kind. Seven facial motoneurons and seven hypoglossal motoneurons were also filled with Neurobiotin. All three motor nuclei contained TRH-immunoreactive varicosities, with the largest numbers found in the nucleus ambiguus. Close appositions were seen between TRH-immunoreactive boutons and every labelled motoneuron. Respiratory-related motoneurons in the nucleus ambiguus received the largest number of TRH appositions with 74 +/- 38 appositions/neuron (mean +/- S.D.; n = 7). In contrast, nonrespiratory ambigual motoneurons received significantly fewer TRH appositions (11 +/- 5; n = 5; P < 0.05; Mann-Whitney U test). Facial motoneurons received about the same number of TRH appositions as nonrespiratory ambigual motoneurons, with 13 +/- 4 (n = 7). Hypoglossal motoneurons received the fewest appositions from TRH-containing boutons, with 8 +/- 2 (n = 7). There were no differences in the TRH inputs to respiratory and nonrespiratory motoneurons in the facial and hypoglossal nuclei. These results demonstrate that, among motoneurons in the medulla, respiratory motoneurons in the rostral nucleus ambiguus are preferentially innervated by the TRH-immunoreactive boutons.

Thyrotropin-releasing hormone-immunoreactive varicosities synapse on rat phrenic motoneurons.

The relationship between retrogradely labelled or intracellularly filled phrenic motoneurons and varicosities containing thyrotropin-releasing hormone immunoreactivity was investigated in rats by light and electron microscopy. Phrenic motoneurons were identified via retrograde tracing from the diaphragm with cholera toxin B subunit, which was followed by immunocytochemistry to visualise retrogradely labelled motoneurons and thyrotropin-releasing hormone-immunoreactive nerve fibres in their vicinity. At the light microscopic level, varicose thyrotropin-releasing hormone-immunoreactive nerve fibres were distributed sparsely in the phrenic motor nucleus, with some axons surrounding retrogradely labelled motoneurons. In separate intracellular experiments, four phrenic motoneurons identified by antidromic activation from the C5 phrenic nerve root were subsequently filled with Neurobiotin, and nerve fibres that contained thyrotropin-releasing hormone immunoreactivity were identified by immunocytochemistry. The numbers and locations of thyrotropin-releasing hormone-immunoreactive varicosities that were closely appeared to the intracellularly labelled motoneurons were mapped using a camera lucida technique. Close appositions by thyrotropin-releasing hormone-immunoreactive varicosities were seen on somata as well as on proximal and distal dendrites. The closely apposed varicosities were usually present in tight clusters, which were formed by single varicose axons. However, the distribution was nonuniform, in that some dendrites did not receive any close appositions. Ultrastructural analysis of random ultrathin sections through retrogradely labelled neurons showed that varicosities with thyrotropin-releasing hormone immunoreactivity made 1.8% of all synapses and direct contacts on somata and 2.3% of synapses and contacts with dendrites of the retrogradely labelled phrenic motoneurons. The results of these experiments suggest that thyrotropin-releasing hormone-immunoreactive varicosities provide similar numbers of inputs to both the somata and dendrites of phrenic motoneurons. These thyrotropin-releasing hormone-containing inputs seen via light and electron microscopy could modulate the excitability of phrenic motoneurons.

Central inspiration increases barosensitivity of neurons in rat rostral ventrolateral medulla.

Barosensitive neurons in the rostral ventrolateral medulla (RVLM) often have a respiratory-related modulation of their activity. However, the extent of the interaction between baroreceptor and respiratory inputs is controversial. The main aim of the present study was to determine the effect of central respiratory drive (CRD) on the barosensitivity of RVLM neurons. Extracellular recordings were obtained from 68 barosensitive neurons in the RVLM of anesthetized, paralyzed, and bilaterally vagotomized Sprague-Dawley rats. Examination of phrenic-triggered histograms revealed five activity patterns among barosensitive neurons: inspiratory depression (type I, n = 20), early inspiratory activation (type II, n = 14), postinspiratory activation (type III, n = 18), expiratory depression (type IV, n = 5) and no modulation (type V, n = 11). In most neurons (types I and III and 56% of type II) inhibition produced by aortic nerve stimulation was greater in inspiration than in expiration. Cardiac-related modulation, as an index of natural phasic baroreceptor activation, was also greater in inspiration than expiration in type III neurons. The results demonstrate that CRD modulates the baroreflex at the level of the RVLM.

Close appositions between tyrosine hydroxylase immunoreactive boutons and respiratory neurons in the rat ventrolateral medulla.

The extent of the adrenergic input to respiratory neurons in the ventrolateral medulla oblongata of rats was assessed by using a combination of intracellular recording, dye filling, and immunohistochemistry. Twenty-two neurons that displayed a pronounced respiration-related modulation of their membrane potential, and could not be antidromically activated by electrical stimulation of the superior laryngeal, vagus, or facial nerves, were labelled by intracellular injection of biocytin. Three types of respiration-related neurons were labelled: small neurons located in the Bötzinger complex between 0.5 and 1.0 mm caudal to the facial nucleus; medium-sized neurons located in the ventral respiratory group 1.0 to 2.0 mm caudal to the facial nucleus; and large motoneurons located within the nucleus ambiguus 0.5 to 2.0 mm caudal to the facial nucleus. Small Bötzinger neurons [length = 22 +/- 5 microns, width = 13 +/- 3 microns, area = 222 +/- 79 microns2; (mean +/- SD, n = 5)] had membrane potentials of -15 to -27 mV during the recording period. Four of five of these cells had profuse axonal terminations between 50 microns caudal and 450 microns rostral to their somata, suggesting that they may form part of local networks responsible for generating respiratory activity. Medium-sized ventral respiratory group neurons (length = 26 +/- 5 microns, width = 18 +/- 4 microns, area = 377 +/- 141 microns2; n = 5) were found in the vicinity of the nucleus ambiguus dorsal to the lateral reticular nucleus. Three of five of these neurons had an axon that crossed the midline and travelled caudally. One axon had a collateral with varicosities close to its soma. The somata of motoneurons (length = 29 +/- 6 microns, width = 21 +/- 4 microns, area = 485 +/- 142 microns2; n = 12) were located within the nucleus ambiguus, and had axons that could be traced to exist points from the medulla. Tyrosine hydroxylase immunoreactive cells and their terminal fibres within the medulla were localised by immunocytochemistry. Small Bötzinger neurons received the largest number of close appositions from tyrosine hydroxylase immunoreactive boutons (13 +/- 2 appositions/neuron; n = 5). Medium-sized ventral respiratory group neurons received fewer appositions (8 +/- 4 appositions/neuron; n = 5). Most motoneurons (n = 10) received few appositions from tyrosine hydroxylase immunoreactive boutons, while two received none. The average number was 3 +/- 3 appositions/neuron (n = 12).(ABSTRACT TRUNCATED AT 400 WORDS)