Role of Barrington's nucleus in micturition

Barrington's nucleus is a central component of the micturition circuit. This nucleus projects axons to the sacral parasympathetic nucleus, where preganglionic neurons innervating the urinary bladder are located. To clarify the functional role of this nucleus, the firing properties of Barrington's neurons that project axons to the spinal cord were examined. Based on these studies, a model begins to emerge that places Barrington's nucleus in the micturition pathway that is involved in increasing bladder pressure rapidly and strongly, while also maintaining high bladder pressure. In addition, Barrington's neurons are suggested to have another role, that is, increasing the probability of micturition contraction by activating a spinal excitatory pathway or disinhibiting a spinal inhibitory mechanism. In contrast to the excitatory role of Barrington's nucleus, this nucleus does not seem to trigger bladder relaxation.     

Axon branching of medullary expiratory neurons in the lumbar and the sacral spinal cord of the cat

Intraspinal axon collaterals of expiratory (E) neurons in the caudal nucleus retroambigualis extending their ascending spinal axons to the lower lumbar (L6-L7) and the sacral (S1-S3) segments were investigated in anesthetized cats. To search for axon collaterals of single E neurons in the lumbar segments, the spinal gray matter was microstimulated from the dorsal to the ventral sites at 100 μm intervals with an intensity of 150–250 μA at 1 mm intervals rostrocaudally along the spinal cord, and effective stimulating sites of antidromic activation in axon collaterals were systematically mapped. In addition, the detailed trajectory of collaterals in the upper lumbar (L1-L3), the middle lumbar (L4-L5), and the sacral (S1-S3) spinal cord was examined by microstimulation at a matrix of points 100–200 μm apart with a maximum stimulus intensity of 50 μA. The trajectory of axon collaterals was reconstructed on the basis of the location of low-threshold foci and the latency of antidromic spikes. Virtually all E neurons examined had 1–7 collaterals at widely separated segments of the lumbar cord. Many axon collaterals were found in the upper lumbar spinal cord as compared to the middle and the lower lumbar spinal cord. The locations of axon collaterals in the upper lumbar spinal cord overlapped with those of abdominal motoneurons. Axon collaterals in the sacral gray matter were found in 3 of 9 E neurons. Axon collateral were found within the nucleus of Onuf, in the region dorsal to the nucleus of Onuf, and in the intermediate region. The functional significance of the divergent distribution of multiple axon collaterals of single E neurons in different spinal levels of the lumbar and the sacral spinal cord is discussed in relation to the respiratory function of E neurons and other spinal motor activities.     

Properties of Barrington's neurones in cats: units that fire inversely with micturition contraction

Barrington's nucleus contains neurones that decrease their firing during micturition contraction, as well as neurones that increase their firing during this phase. These neurones are commonly termed inverse neurones and direct neurones, respectively. The aims of the present study were to determine whether inverse neurones send descending axons to the spinal cord and to clarify how these neurones regulate bladder contractility. Forty-five single neurones were recorded from the dorsolateral pontine tegmentum. Spinal-projecting neurones were identified by antidromic stimulation of the spinal cord. More than half of inverse neurones were located outside Barrington's nucleus. Only three were spinal-projecting neurones. The results were in marked contrast with direct neurones that we studied previously: the majority of them were located within Barrington's nucleus, and 56% were spinal-projecting neurones. The firing frequency of inverse neurones ranged between 6 and 37 Hz during the relaxation phase of the micturition contraction-relaxation rhythm. The firing of all neurones began to decrease within 8 s after the onset of micturition contraction. During micturition contraction, neurones displayed little firing, being virtually silent (n = 29) or displayed weak tonic firing (3-11 Hz; n = 16). All neurones began to increase their firing within 8 seconds after the onset of bladder relaxation. These results suggest that inverse neurones do not trigger bladder contraction or relaxation, despite the finding that a few of them send descending axons to the spinal cord. One possible role of the inverse neurone is to regulate firing activities of direct neurones in Barrington's nucleus.     

大鼠Barrington核内脊髓投射神经元直接接受腰骶髓传入的电镜研究

将结合生物素的葡聚糖胺 (BDA)注射到大鼠腰骶髓后 ,在电镜下观察脑桥Barrington核内腰骶髓投射神经元与来自腰骶髓传入投射纤维间的突触联系。与先前的研究相一致 ,注射BDA到腰 6和骶 1节段后 ,光镜下可见Barrington核内出现大量顺行标记的神经末梢和一定数量的逆行标记细胞。电镜下发现标记的轴突末梢和标记的树突之间存在直接的突触连接。结果表明 ,Barrington核直接接受腰骶髓的传入投射 ,提示大鼠脑桥排尿反射的脊髓内上行投射通路中可能存在一条直接通路。     

Collaterals of primate spinothalamic tract neurons to the periaqueductal gray

Collateral projections are an important feature of the organization of ascending projections from the spinal cord to the brain. Primate spinothalamic tract (STT) neurons with collaterals to the periaqueductal gray (PAG) were studied by means of a fluorescent double-labeling method. Granular Blue and rhodamine-labeled latex microspheres were placed in the ventral posterior lateral (VPL) nucleus of the thalamus and the periaqueductal gray, respectively. Single and double labeled neurons were studied in the upper cervical cord, cervical enlargement, thoracic cord, lumbar enlargement, and sacral segments. The laminar distribution of double labeled neurons was similar to that of spinomesencephalic tract (SMT) neurons. Most double labeled (STT-SMT) neurons were located in contralateral laminae I, V, VII, and X. Relatively more lamina I STT-SMT neurons were found in the cervical enlargement and more lamina V STT-SMT neurons in the lumbar enlargement. The density of STT-SMT neurons in the upper cervical segments and cervical enlargement was almost equal. The density of STT-SMT neurons in the lumbar enlargement was 40% of that in the cervical enlargement. The thoracic and sacral segments had the lowest density of STT-SMT neurons, about 10% of that in the cervical enlargement. STT-SMT neurons constituted 14.7% of SMT neurons and 6% of STT neurons in the cervical enlargement and 15.3% of SMT neurons and 2.9% of STT neurons in the lumbar enlargement. The branch points of eight STT-SMT axons were studied electrophysiologically. The average percentage of conduction time spent in the parent axon was more than 85% for an antidromic action potential from the VPL nucleus and 91% from the PAG. Branch points of STT-SMT axons were calculated to be 9-13 mm caudal to the PAG, in the pons or rostral medulla.     

Distribution and origin of corticotropin-releasing factor-immunoreactive axons in the female rat lumbosacral spinal cord.

Corticotropin-releasing factor (CRF) is a neuropeptide traditionally known for its hormonal role in the hypothalamic/pituitary/adrenal stress axis. However, CRF has been reported in axons in sites that may be considered outside of the direct stress axis, e.g., in axons in the lumbosacral spinal cord associated with the micturition response. Whether any of these CRF-immunoreactive axons interacts with uterine-related preganglionic autonomic neurons or projection neurons in the lumbosacral spinal cord is unknown. Thus, immunohistochemistry and retrograde tracing were employed to determine the presence, distribution, and origin of CRF-immunoreactive axons in the L6/S1 spinal cord of the female rat and to ascertain whether these axons are associated with uterine-related neurons. CRF-immunoreactive axons were present in the dorsal horn, medial and lateral collateral pathways, dorsal intermediate gray, laminae VlI and X, and sacral parasympathetic nucleus of the spinal cord. Nitric oxide-synthesizing, i.e., NADPH-d-positive neurons and pseudorabies virus labeled uterine-related neurons were in the sacral parasympathetic nucleus and were closely apposed by CRF-immunoreactive axons. Injection of retrograde tracers (fluorogold or fast blue) into the L6/S1 spinal cord labeled neurons in the hypothalamic paraventricular nucleus and pontine Barrington's nucleus, and some of these neurons were immunoreactive for CRF. This study demonstrates that CRF-immunoreactive axons are present in the L6/S1 spinal cord of the female rat in areas associated with sensory and autonomic processing. Some of these axons originate from the paraventricular nucleus and Barrington's nucleus and are adjacent to uterine-related neurons. These results indicate that CRF may influence neural activity related to the female reproductive system.     

Distribution of neurons in the lateral pontine tegmentum projecting to thoracic, lumbar and sacral spinal segments in the cat

The course of descending fibers projecting to the spinal cord and the arrangement of their parent cells located in various nuclei of the dorso-lateral pontine tegmentum were studied using the horseradish peroxidase (HRP) retrograde axonal transport technique. Retrogradely labeled neurons were found in the locus coeruleus (LC), subcoeruleus (SC), K?lliker-Fuse nucleus (KF) and in the lateral parabrachial nucleus (LPB) after HRP injections into various spinal segments. Neurons innervating the thoracic spinal cord were found to be arranged in the ventral portion of the LC and in the entire SC; their axons descended ipsilaterally. Neurons with descending axons to lumbar segments were seen mainly in the ventral portion of the LC and in the medial portion of SC. Most of their axons were also seen to descend ipsilaterally. Neurons projecting to sacral segments occurred in the entire LC and in the medial portion of the SC. Large part of descending fibers crossed the midline at the level of (or near) the termination site. Neurons of all portions of the KF and LPB projected to the thoracic spinal cord only ipsilaterally, while many descending fibers innervating the sacral segments crossed the midline.     

Central representation of bladder and colon revealed by dual transsynaptic tracing in the rat: substrates for pelvic visceral coordination

The neurocircuitry underlying regulation of bladder and distal colon function by Barrington's nucleus (the pontine micturition centre) was investigated in rats by identifying neurons which were transsynaptically labelled from these viscera, with pseudorabies virus (PRV) or genetically modified forms of PRV [PRV-beta-galactosidase (PRV-beta-Gal) and PRV-green fluorescent protein (PRV-GFP)]. PRV injection into the bladder or the colon of separate rats suggested an overlap in the distribution of bladder- and colon-related neurons in Barrington's nucleus, as well as a topographical arrangement whereby dorsal neurons were bladder-related and ventral neurons were colon-related. In rats injected with PRV-beta-Gal into one viscera and PRV-GFP into another, neurons in the major pelvic ganglion and lumbosacral spinal cord were primarily single-labelled at relatively early survival times. With longer survival times many double-labelled neurons (>70%) appeared in Barrington's nucleus, suggesting that individual Barrington's nucleus neurons are synaptically linked to preganglionic parasympathetic neurons which independently innervate the colon or the bladder. In addition to these dual-labelled neurons, Barrington's nucleus neurons which were single-labelled from either viscera were observed and these exhibited a viscerotopic organization consistent with the single-labelling studies. Together, these findings suggest the existence of three neuronal populations in Barrington's nucleus, one which is synaptically linked to both the bladder and the colon and the other two populations which are specifically linked to either viscera. These anatomical substrates may underlie the central coordination of bladder and colon function and play a role in disorders characterized by a coexistence of bladder and colonic symptoms.     

Sleep-waking discharge of basal forebrain projection neurons in cats

We have previously described a population of neurons in the magnocellular basal forebrain which have selectively elevated discharge rates during slow-wave sleep compared to waking; we postulate that these sleep-active neurons are a component of a basal forebrain sleep-promoting system. The purpose of the present experiment was to determine if sleep-active neurons contribute axons to recently described basal forebrain projection pathways. In cats prepared for chronic single unit and EEG-sleep recordings, stimulating electrodes were placed in the mesencephalic reticular formation, and the external capsule and anterior cingulate bundle, fiber bundles known to contain axons of basal forebrain projection neurons. Fifty-nine neurons were antidromically driven; differences in antidromic response latencies were related to sleep-waking discharge profiles. Of the cells with short antidromic latencies (less than 5 msec), the majority (9 of 12) had high discharge rates during waking and low rates during slow-wave sleep. Cells with long antidromic latencies had either very low discharge rates (less than 1 spike/sec) across all states, or had elevated discharge rates in slow-wave sleep. Sleep-active neurons were antidromically driven from external capsule (n = 9), anterior cingulate bundle (n = 9), or mesencephalic reticular formation (n = 5). Projection sleep-active neurons were recorded in the substantia innominata, ventral to the globus pallidus and medial to the central nucleus of the amygdala. Our study found that identified basal forebrain projection neurons in cats exhibit a variety of sleep-waking discharge patterns and conduction velocities. Sleep-active neurons were found to have slowly conducting axons, and to be a source of both ascending and descending projections.     

Barrington's nucleus in the guinea pig (Cavia porcellus): location in relation to noradrenergic cell groups and connections to the lumbosacral spinal cord

Micturition is largely controlled by Barrington's nucleus in the dorsolateral tegmentum of the pons. This nucleus coordinates simultaneous bladder contraction and external urethral sphincter relaxation, by means of a specific pattern of projections to the lumbosacral spinal cord. The most widely used small animal model in neurourological research is the rat. However, urodynamic studies suggest that, in sharp comparison to rat, guinea pig micturition is very similar to human micturition. Therefore, the present study, using retrograde and anterograde tracing and double immunofluorescence, was designed to investigate the location of Barrington's nucleus in the guinea pig, to identify Barrington's nucleus projections to the spinal cord and to clarify the relationship of Barrington's nucleus to pontine noradrenergic cell groups. Results show that Barrington's nucleus is located in the dorsolateral pons, projects to the intermediolateral and intermediomedial cell groups of the lumbosacral spinal cord and is clearly distinct from the pontine noradrenergic cell groups. These results show that the neuroanatomical circuitry in the spinal cord and brainstem that controls micturition in the guinea pig is similar to that in rat. This means that the differences between rat and guinea pig micturition on a behavioral level are not the result of different neuroanatomical connections in these parts of the central nervous system. These results provide a neuroanatomical basis for further neurourological studies in guinea pig.     

Reticulospinal vasomotor neurons of the rat rostral ventrolateral medulla: relationship to sympathetic nerve activity and the C1 adrenergic cell group

Neurons projecting from the rostral ventrolateral medulla (RVL) to the spinal cord were antidromically identified in rats anesthetized with urethane, paralyzed, and ventilated. The sites of lowest antidromic threshold were concentrated in the intermediolateral nucleus (IML). Their axonal conduction velocities were distributed bimodally, with the mean of the rapidly conducting fibers (greater than 1 m/sec) being 3.1 +/- 0.1 m/sec (n = 105), and of the slower axons being 0.8 +/- 0.03 m/sec (n = 25). Single-shock electrical stimulation of RVL elicited 2 bursts of excitation in splanchnic sympathetic nerve activity (SNA), which resulted from activation of 2 descending pathways with conduction velocities comparable to those of antidromically excited RVL-spinal neurons. The probability of discharge of RVL-spinal cells was synchronized both with the cardiac-related bursts in SNA with functional baroreceptor reflexes and with the free-running 2-6 Hz bursts in SNA following baroreceptor afferent denervation. On the average, their spontaneous discharges occurred 67 +/- 2 msec (n = 31) prior to the peak of the spontaneous bursts in splanchnic SNA. This time corresponded to the latency to the peak of the early excitatory potential in splanchnic SNA following electrical stimulation of RVL. Baroreceptor reflex activation inhibited RVL-spinal neurons. The recording sites of RVL-spinal vasomotor neurons were consistently located within 100 micron of cell bodies (C1 neurons) immunoreactive for the adrenaline-synthesizing enzyme phenylethanolamine N-methyltransferase (PNMT). Ultrastructural analysis of the lateral funiculus of the cervical and thoracic spinal cord demonstrated PNMT immunoreactivity within myelinated (0.6-2.1 micron diameter) and unmyelinated (0.1-0.8 micron diameter) axons. Estimated conduction velocities of these fibers were comparable to the antidromic conduction velocities of the rapidly and slowly conducting populations of RVL-spinal vasomotor neurons. We conclude that in rat, the discharge of RVL-spinal vasomotor neurons strongly influences SNA: the baroreceptor-mediated inhibition of these neurons is reflected in the cardiac locking of SNA, while, in the absence of baroreceptor input, the synchronous discharge of RVL-spinal neurons maintains a free-running 2-6 Hz bursting pattern in SNA. RVL-spinal neurons are located within, and may be elements of, the C1 adrenergic cell group, and they provide a sympathoexcitatory drive to neurons in the IML over rapidly and slowly conducting pathways that correspond to myelinated and unmyelinated spinal axons containing PNMT.     

Variations in conduction velocity and excitability following single and multiple impulses of visual callosal axons in the rabbit

The present study examined the conduction properties of 75 visual callosal axons of the awake rabbit. These axons were studied by measuring latency to antidromic activation of cell bodies following midline callosal and/or contralateral cortical stimulation. Seventy-three of 75 neurons (axon conduction velocities = 0.3 to 12.9 m/sec) demonstrated decreases in antidromic latency and threshold to a test stimulus which followed an antecedent conditioning stimulus at appropriate intervals. Control experiments indicated that (i) the latency and threshold variations resulted from prior impulse conduction along the axon, and (ii) the latency decrease reflected an increase in conduction velocity along the main axon trunk. The maximum magnitude of the latency decrease for different axons ranged from 3 to 22% of control values, and the duration ranged from 18 to 169 msec. The duration of the latency decrease was greater for slowly conducting axons than for fast conducting axons. Latency increases to an antidromic test stimulus occurred for up to several minutes following a train of antidromic conditioning pulses. Antidromic latency and threshold shifts were also observed in somatosensory callosal axons and in some corticotectal axons.     

Central motor conduction times using transcranial stimulation and F wave latencies

The purpose of this study was to determine central motor conduction times between the motor cortex and the C8 spinal cord level (MC/C8), the motor cortex and S1 cord level (MC/S1), and C8 and S1 (C8/S1). We stimulated 29 normal subjects with a transcranial high voltage (300-500 V), short duration spike waveform (time constant 50 mu sec) and recorded over the hypothenar or calf muscles. F wave and M response latencies were used to estimate peripheral conduction time. The mean MC/C8 conduction time was 4.4 +/- 0.6 msec. Out of 26 subjects tested, 14 had discernable responses in the calf. The mean MC/S1 conduction time was 13.1 +/- 2.5 msec, and the mean estimated C8/S1 conduction time was 8.5 +/- 2.3 msec. This technique accurately measures conduction time from the motor cortex to the cervical anterior horn cells but is less reliable for monitoring conduction to the sacral cord. These values will facilitate future studies in patients with suspected lesions of the descending motor pathways.     

Firing of micturition center neurons in the rat mesopontine tegmentum during urinary bladder contraction

Micturition is controlled by a network of brainstem neurons involving the Barrington's nucleus. To depict clearly the brainstem system for micturition control, the present study was designed to record single neuronal activity in the mesopontine tegmentum including the Barrington's nucleus, and to observe its precise timing in relation to bladder contraction recorded simultaneously. About 1/5 of neurons encountered had firing modulated in relation to bladder contraction. Three types of neurons were distinguished; those which fired only prior to the start of contraction (type E1), those whose firing started shortly prior to and was maintained during contraction (type E2), and those whose firing was strongly suppressed during contraction (type I). Type E2 neurons were most frequently observed in the Barrington's nucleus and its close vicinity, while the neurons of the other two types were scattered widely in the mesopontine tegmentum. The results show clearly that direct neural signals to induce bladder contraction may arise from the Barrington's nucleus, and that the nucleus may receive regulatory inputs from wide areas of the mesopontine tegmentum. In addition, the present study clarified that the noradrenergic and cholinergic neurons, which are located in nuclei adjoining the Barrington's nucleus and function to control sleep/wakefulness, may not be concerned in controlling micturition directly.     

Neurons in the lateral sacral cord of the cat project to periaqueductal grey, but not to thalamus

Previous work of our laboratory has shown that neurons in the lateral sacral cord in cat project heavily to the periaqueductal grey (PAG), in all likelihood conveying information from bladder and genital organs. In humans this information usually does not reach consciousness, which raises the question of whether the lateral sacral cell group projects to the thalamus. After wheatgerm agglutinin-horseradish peroxidase (WGA-HRP) injections into the sacral cord, anterogradely labelled fibers were found in the thalamus, specifically in the ventral anterior and ventral lateral nuclei, the medial and intralaminar nuclei, the lateral ventrobasal complex/ventroposterior lateral nucleus, and the nucleus centre median, lateral to the fasciculus retroflexus. Much denser projections were found to the central parts of the PAG, mainly to its dorsolateral and ventrolateral parts at caudal levels and lateral parts at intermediate levels. In a subsequent retrograde tracing study, injections were made in those parts of the thalamus that received sacral fibers, as found in the anterograde study. Labelled neurons were observed in the sacral cord, but not in the lateral sacral cell group. In contrast, a small control injection in the caudal PAG resulted in many labelled neurons in the lateral sacral cord. These results suggest that afferent information regarding micturition and sexual behaviour is relayed to the PAG, rather than to the thalamus.     

Tuberoinfundibular neurons in the basomedial hypothalamus of the rat: electrophysiological evidence for axon collaterals to hypothalamic and extrahypothalamic areas.

In pentobarbital or urethane anesthetized rats, the activity of 889 mediobasal hypothalamic neurons was studied for evidence of a response to median eminence stimulation. Evidence of antidromic invasion, which indicated a projection to the median eminence, identified 134 cells (15%) as 'tuberoinfundibular' neurons. Antidromic spike latencies ranged from 0.5 to 14.0 msec (4.3 +/- 2.9 S.D.); conduction velocities were under 1.0 m/sec and were generally slower for tuberoinfundibular neurons located closest to the ventral surface of the hypothalamus. Certain tuberoinfundibular neurons followed paired median eminence shocks at frequencies up to 500 Hz; an increase in both the threshold and the latency for the second antidromic spike was observed with interstimulus intervals under 4 msec. Only 38% of tuberoinfundibular neurons were spontaneously active; 24 of 29 spontaneously active neurons displayed evidence of recurrent inhibition with durations up to 150 msec and at latencies which approximated that of the antidromic spike but which did not depend upon antidromic invasion. Similar responses were observed from 33 spontaneously active non-tuberoinfundibular neurons. Evidence of orthodromic excitation in response to median eminence shocks was observed from 22 other medial hypothalamic neurons. Latencies for excitation ranged from 1.5 to 9.0 msec (mean 4.5 +/- 2.1 S.D.). Simultaneous antidromic invasion from other hypothalamic and extrahypothalamic sites was observed from 8 tuberoinfundibular neurons. These sites included the anterior hypothalamic area (2 cells), the preoptic area (3 cells) and the thalamic nucleus medialis dorsalis (3 cells). These results indicate the presence of axon collaterals within the tuberoinfundibular system; some appear to terminate locally within the hypothalamus, while others extend rostrally and dorsally into extrahypothalamic areas. These connections may provide pathways for extrahypothalamic distribution of peptides which regulate adenohypophyseal secretion, and suggest that these peptides may subserve alternate regulatory roles within the central nervous system.     

A new indirect method for measuring spinal conduction velocity in man

A non-invasive, indirect method for measuring spinal cord mixed afferent-efferent conduction is described. The method is based upon eliciting late reflex responses labelled R1 and R2 from voluntarily contracting thenar and tibialis anterior muscles by preferentially stimulating median and common peroneal sensory nerve fibres. The mean onset latencies of R1 measured 27.5 msec and 30.6 msec recorded from hand and leg muscles respectively. R2 mean onset latencies measured 46.0 msec and 65.1 msec respectively. R1 has characteristics similar to an H-reflex. R2 is a long-loop reflex of unknown pathway assumed to involve similar circuits and rostral turn around points when elicited by both arm and leg stimulation. Mean spinal cord conduction time between the seventh cervical and fifth lumbar spinous processes, is given by (formula; see text) It measured 7.95 msec and the calculated mean conduction velocity was 57.9 +/- 5.7 m/sec.     

Ultrastructure of endomorphin-1 immunoreactivity in the rat dorsal pontine tegmentum: evidence for preferential targeting of peptidergic neurons in Barrington's nucleus rather than catecholaminergic neurons in the peri-locus coeruleus

Endomorphins are opioid tetrapeptides that have high affinity and selectivity for mu-opioid receptors (muORs). Light microscopic studies have shown that endomorphin-1 (EM-1) -containing fibers are distributed within the brainstem dorsal pontine tegmentum. Here, immunoelectron microscopy was conducted in the rat brainstem to identify potential cellular interactions between EM-1 and tyrosine hydroxylase (TH) -labeled cellular profiles in the locus coeruleus (LC) and peri-LC, an area known to contain extensive noradrenergic dendrites of LC neurons. Furthermore, sections through the rostral dorsal pons, from colchicine-treated rats, were processed for EM-1 and corticotropin releasing factor (CRF), a neuropeptide known to be present in neurons of Barrington's nucleus. EM-1 immunoreactivity was identified in unmyelinated axons, axon terminals, and occasionally in cellular profiles resembling glial processes. Within axon terminals, peroxidase labeling for EM-1 was enriched in large dense core vesicles. In sections processed for EM-1 and TH, approximately 10% of EM-1-containing axon terminals (n=269) targeted dendrites that exhibited immunogold-silver labeling for TH. In contrast, approximately 30% of EM-1-labeled axon terminals analyzed (n = 180) targeted CRF-containing somata and dendrites in Barrington's nucleus. Taken together, these data indicate that the modulation of nociceptive and autonomic function as well as stress and arousal responses attributed to EM-1 in the central nervous system may arise, in part, from direct actions on catecholaminergic neurons in the peri-LC. However, the increased frequency with which EM-1 axon terminals form synapses with CRF-containing profiles in Barrington's nucleus suggests a novel role for EM-1 in the modulation of functions associated with Barrington's nucleus neurons such as micturition control and pelvic visceral function.     

Evidence for divergent projections to the brain noradrenergic system and the spinal parasympathetic system from Barrington's nucleus

The present study was designed to determine whether Barrington's nucleus, which lies ventromedial to the locus coeruleus (LC) and projects to the sacral parasympathetic nucleus, is a source of afferent projections to the LC. Restricted injections of the anterograde tracer, biocytin, into Barrington's nucleus labeled varicose fibers that extended from the injection site into the LC. Consistent with this, injections of the retrograde tracers, wheatgerm agglutinin conjugated to horseradish peroxidase coupled to gold particles (WGA-Au-HRP) or fluorescein-conjugated latex beads, into the LC labeled numerous (approximately 10%) Barrington's neurons that were also retrogradely labeled by Fluoro-Gold (FG) injections in the spinal cord. Retrograde tracing from the LC combined with corticotropin-releasing hormone (CRH) immunohistochemistry revealed that at least one third of the retrogradely labeled neurons in Barrington's nucleus were CRH-immunoreactive (CRH-IR). Finally, in triple labeling studies, CRH-Barrington's neurons were consistently observed that were retrogradely labeled from both the LC and spinal cord. These findings implicate Barrington's nucleus as an LC afferent and a source of CRH-IR fibers in the LC. Additionally, the results suggest that some Barrington's neurons diverge to innervate both the spinal cord and the LC. This divergent innervation may serve to coregulate the sacral parasympathetic nervous system and brain noradrenergic system, thus providing a mechanism for coordinating pelvic visceral functions with forebrain activity.