Axonal transport of neurotrophins by visceral afferent and efferent neurons of the vagus nerve of the rat

The receptor-mediated axonal transport of [¹²⁵I]-labeled neurotrophins by afferent and efferent neurons of the vagus nerve was determined to predict the responsiveness of these neurons to neurotrophins in vivo. [¹²⁵I]-labeled neurotrophins were administered to the proximal stump of the transected cervical vagus nerve of adult rats. Vagal afferent neurons retrogradely transported [¹²⁵I]neurotrophin-3 (NT-3), [¹²⁵I]nerve growth factor (NGF), and [¹²⁵I]neurotrophin-4 (NT-4) to perikarya in the ipsilateral nodose ganglion, and transganglionically transported [¹²⁵I]NT-3, [¹²⁵I]NGF, and [¹²⁵I]NT-4 to the central terminal field, the nucleus tractus solitarius (NTS). Vagal afferent neurons showed minimal accumulation of [¹²⁵I]brain-derived neurotrophic factor (BDNF). In contrast, efferent (parasympathetic and motor) neurons located in the dorsal motor nucleus of the vagus and nucleus ambiguus retrogradely transported [¹²⁵I]BDNF, [¹²⁵I]NT-3, and [¹²⁵I]NT-4, but not [¹²⁵I]NGF. The receptor specificity of neurotrophin transport was examined by applying [¹²⁵I]-labeled neurotrophins with an excess of unlabeled neurotrophins. The retrograde transport of [¹²⁵I]NT-3 to the nodose ganglion was reduced by NT-3 and by NGF, and the transport of [¹²⁵I]NGF was reduced only by NGF, whereas the transport of [¹²⁵I]NT-4 was significantly reduced by each of the neurotrophins. The competition profiles for the transport of NT-3 and NGF are consistent with the presence of TrkA and TrkC and the absence of TrkB in the nodose ganglion, whereas the profile for NT-4 suggests a p75 receptor-mediated transport mechanism. The transport profiles of neurotrophins by efferent vagal neurons in the dorsal motor nucleus of the vagus and nucleus ambiguus are consistent with the presence of TrkB and TrkC, but not TrkA, in these nuclei. These observations describe the unique receptor-mediated axonal transport of neurotrophins in adult vagal afferent and efferent neurons and thus serve as a template to discern the role of specific neurotrophins in the functions of these visceral sensory and motor neurons in vivo. J. Comp. Neurol. 393:102–117, 1998. Published 1998 Wiley-Liss, Inc.

1 This article is a US government work and, as such, is in the public domain in the United States of America.

     

Learning deficits induced by sleep deprivation and recovery are not associated with altered [(3)H]muscimol and [(3)H]flunitrazepam binding

Several studies have shown that sleep deprivation produces deficits in learning tasks, but mechanisms underlying these effects remain unclear. Other lines of evidence indicate an involvement of brain GABA systems in cognitive processes. Here, we investigated the possibility that alterations in GABA(A) or benzodiazepine (BDZ) receptor binding might underlie avoidance deficits induced by sleep deprivation. Rats were deprived of sleep for 96 h using the platform method and then trained in a step-through inhibitory avoidance task, or allowed to recover sleep for 24 h before training (sleep rebound group). Thirty minutes after training, animals were given a retention test. Both sleep-deprived and sleep-recovered animals showed a significant impairment in avoidance responding compared to cage controls, and the sleep-deprived group performed significant worse than the sleep-recovered group. A separate group of animals was sacrificed either immediately after 96 h of sleep deprivation or after 96 h of sleep deprivation followed by 24 h of sleep recovery. [(3)H]muscimol and [(3)H]flunitrazepam binding were examined by quantitative autoradiography in 42 brain regions, including areas involved in cognitive processes. No significant differences among groups were found in any brain region, except for a reduction in [(3)H]flunitrazepam binding in the frontal cortex of sleep-recovered animals. These results confirm the deleterious effects of sleep loss on inhibitory avoidance learning, but suggest that such deficits cannot be attributed to altered GABA(A) or BDZ binding in brain.     

Corticotropin-releasing factor binding sites undergo axonal transport in the rat vagus nerve

Corticotropin-releasing factor (CRF) binding sites were found to be present in the rat vagus nerve and underwent axonal transport. Binding sites accumulated on both sides of ligatures placed on the nerve and at similar rates following ligation of right or left cervical vagal trunks of either male or female rats. CRF binding sites also accumulated proximal and distal to ligatures on subdiaphragmatic vagal trunks. Binding was specific, reversible and inhibited by the CRF receptor antagonist α-helical-CRF(9-41). [(125) l]Tyr(0) -ovine-CRF binding to rat vagus nerve was not guanine nucleotide-sensitive. CRF and cholecystokinin binding sites were transported at a similar rate in the cervical vagus, although turnover of CRF binding sites was more rapid. No differences in CRF binding site transport were observed between Zucker rats of lean or obese genotype.     

Axonal transport of α-bungarotoxin binding sites in rat sciatic nerve

[¹²⁵I]α-Bungarotoxin (α-BuTX) binding sites accumulate both proximal and distal to a ligature positioned around the sciatic nerve of rats. [¹²⁵I]α-BuTX binding sites, localized using quantitative receptor autoradiography, were found to accumulate at nerve ligatures at a relatively constant rate which suggests that they undergo both anterograde and retrograde axonal transport. [¹²⁵I]α-BuTX binding to sections of ligated sciatic nerve was saturable with apparent dissociation constants of 0.97 nM proximal and 0.53 nM distal to the ligature.d-Tubocurarine, nicotine, decamethonium and atropine displaced [¹²⁵I]α-BuTX from sciatic nerve sections with affinities comparable to those previously reported for the toxin binding component of rat brain. These data indicate that [¹²⁵I]α-BuTX binding sites pharmacologically similar to those of rat brain are transported in sciatic nerve. Axonally transported toxin binding sites may correspond to those previously localized to the plasma membrane of peripheral nerve axons and on the terminals of motor neurons.     

Anatomical distribution of sodium-dependent [(3)H]naloxone binding sites in rat brain

The sulfhydryl alkylating reagent N-ethylmaleimide (NEM) blocks opioid receptor binding and receptor/G-protein coupling. Sodium partially restores [(3)H]naloxone binding after inhibition by NEM to reveal sodium-dependent [(3)H]naloxone sites, defined as binding in the presence of 50-100 mM NaCl after treatment of membranes or sections with 750 microM NEM. In the present study, receptor autoradiography of [(3)H]naloxone binding in control and NEM-treated tissue was used to examine the anatomical distribution of sodium-dependent [(3)H]naloxone sites in rat brain. In brain membranes, the pharmacology of sodium-dependent [(3)H]naloxone sites was consistent with that of mu opioid receptors. Relatively high IC(50) values for agonists and lack of effect of Gpp(NH)p on DAMGO displacement of [(3)H]naloxone binding in NEM-treated membranes indicated that the sodium-dependent sites were low affinity sites, presumably uncoupled from G-proteins. Autoradiograms revealed that NEM treatment dramatically reduced [(3)H]naloxone binding in all brain regions. However, [(3)H]naloxone binding was increased in specific regions in NEM-treated sections in the presence of sodium, including bed nucleus of the stria terminalis, interpeduncular nucleus, periaqueductal gray, parabrachial nucleus, locus coeruleus, and commissural nucleus tractus solitarius. Sodium-dependent [(3)H]naloxone binding sites were not found in other areas that exhibited [(3)H]naloxone binding in control tissue, including the striatum and thalamus. These studies revealed the presence of a subpopulation of [(3)H]naloxone binding sites which are sodium-dependent and have a unique regional distribution in the rat brain.     

Axonal transport of [3H]proteins in a noradrenergic system of the rat brain.

Proteins synthesized from [3H]leucine injected into the rat nucleus locus coeruleus (LC) were transported through the hypothalamus in 5 successive waves at rates of 72-192, 24-48, 13-20, 3-4 and 1.4-2.9 mm/day (waves I through V, respectively). Waves I through IV began axoplasmic transport in the LC within the first few hours after [3H]protein synthesis began in the LC. Wave V was delayed in onset for 1.7-3.7 days and was also probably transported through the contralateral hypothalamus. Wave IV was not transported within the LC-hypothalamic axons ascending through the dorsal noradrenergic bundle since its transport was not blocked by 6-OHDA lesions in this bundle, as was transport of the other 4 waves. Unilateral dorsal bundle lesions caused a well defined caudal backup of dopamine-beta-hydroxylase immunofluorescence and a fall in dopamine-beta-hydroxylase activity in the ipsilateral frontal cortex and hypothalamus of 55% and 9%, respectively. Bilateral lesions caused only a significantly further reduction in hypothalamic levels indicating crossed innervation of the hypothalamus by the LC of 27%. Waves I and II have been classified as rapid transport and contained 33% of the transported [3H]protein. Wave V was slowly transported and contained 51% of the transported [3H]protein, while wave III was intermediate in rate and contained 16% of transported [3H]proteins.     

Guanine based purines inhibit [(3)H]glutamate and [(3)H]AMPA binding at postsynaptic densities from cerebral cortex of rats

Extracellular guanine-based purines (GBPs) have been implicated in neuroprotective effects against glutamate toxicity by modulating the glutamatergic system through mechanisms without the involvement of G proteins. Accordingly, GBPs have been shown to inhibit the binding of glutamate and its analogs in different brain membrane preparations. However, brain membrane preparations used for these studies are comprised of both post- and pre-neuronal and glial synaptic components. In this study we investigated the ability of GBPs to displaced glutamate and AMPA binding at postsynaptic densities (PSDs). PSDs are markedly prominent in glutamatergic synapses and retains the native apposition of membrane components and post synaptic receptors. The PSD fraction was prepared from cerebral cortex of Wistar rats and it was characterized as PSDs by electron microscopy and by an enrichment of PSD-95, a protein marker of PSDs (90% of immunodetection). Moreover, we detected an enrichment of glutamate receptors subunits that including NR1 subunit of NMDA receptors and GluR1 subunit of AMPA receptors. GppNp (poor hydrolyzable GTP analog) and GMP displaced 40 and 36% of glutamate binding, respectively, and guanosine only 23%. AMPA binding was not affected by guanosine and was inhibited 21 and 25% by GppNp and GMP, respectively. Hence, this study demonstrates that guanine based purines inhibited glutamate and AMPA binding at postsynaptic membrane preparations, contributing for a better understanding of the mechanisms by which GBPs antagonize glutamatergic neurotoxicicity, e.g. the possible involvement of glutamatergic postsynaptic receptors in their neuroprotective roles.     

Axonal transport of [H]proteins in a noradrenergic system of the rat brain

Proteins synthesized from [³H]leucine injected into the rat nucleus locus coeruleus (LC) were transported through the hypothalamus in 5 successive waves at rates of 72–192, 24–48, 13–20, 3–4 and 1.4–2.9 mm/day (waves I through V, respectively). Waves I through IV began axoplasmic transport in the LC within the first few hours after [³H]protein synthesis began in the LC. Wave V was delayed in onset for 1.7–3.7 days and was also probably transported through the contralateral hypothalamus. Wave IV was not transported within the LC-hypothalamic axons ascending through the dorsal noradrenergic bundle since its transport was not blocked by 6-OHDA lesions in this bundle, as was transport of the other 4 waves. Unilateral dorsal bundle lesions caused a well defined caudal backup of dopamine-β-hydroxylase immunofluorescence and a fall in dopamine-β-hydroxylase activity in the ipsilateral frontal cortex and hypothalamus of 55% and 9%, respectively. Bilateral lesions caused only a significantly further reduction in hypothalamic levels indicating crossed innervation of the hypothalamus by the LC of 27%. Waves I and II have been classified as rapid transport and contained 33% of the transported [³H]protein. Wave V was slowly transported and contained 51% of the transported [³H]protein, while wave III was intermediate in rate and contained 16% of transported [³H]proteins.     

Axonal transport of [H]fucosyl glycoproteins in noradrenergic neurons in the rat brain

The transport of [³H]fucosyl glycoproteins has been investigated in the noradrenergic pathway from the locus coeruleus (LC) to the hypothalamus in the rat brain by use of local injection ofl-[6-³H]fucose into the LC. Two discrete waves of³H-glycoprotein pass through the hypothalamus in a caudorostral direction traveling at 4 mm/h (96 mm/day) and 2 mm/h (48 mm/day) respectively. Both waves appear to originate from the LC at approximately 2 h after the injection of [³H]fucose, a time when incorporation into glycoprotein has not yet peaked in the LC. Lesioning the LC-hypothalamic pathway with 6-OHDA, but not carrier solution, blocks both waves of³H-glycoprotein demonstrating that transport occurs exclusively in the catecholamine axons within the pathway. DEAE-Sephadex chromatography combined with 6-OHDA lesions demonstrates clear differences in the character of³H-glycoproteins in the two waves undergoing transport. The relative lack of labelling of hypothalamic glycoproteins in the interval before and after these waves suggests that relatively few rapidly transported glycoproteins contribute to the non-terminal axon membrane and are probably primarily transported to the nerve terminal. No evidence for slow axoplasmic transport of [³H]fucosyl glycoproteins is found.     

Axonal transport and metabolism of [3H]fucose- and [35S]-sulfate-labeled macromolecules in the rat visual system.

The axonal transport of labeled macromolecules in retinal ganglion cells of rats was investigated from 1 to 20 days following intraocular injection of [3H]fucose and [35S]sulfate. Maximal incorporation of [3H]fucose into acid insoluble material in the retina was at 8 h, followed by a biphasic decline. Transported [3H]fucose (98% as glycoprotein) was in the optic nerve at 1 h, the optic tract and lateral geniculate body by 2 h, and the superior colliculus by 3 h after injection, indicating a rate of transport of approximately 200 mm/day. Radioactivity continued to accumulate in the superior colliculus for at least 8 h and began to decline rapidly by 24 h. Between 3 and 6 days levels rose again in both optic tract and superior colliculus before starting a gradual decline, indicating that a wave of rapidly transported material was delayed in leaving the retina. When proteins in the superior colliculus were fractionated by gel electrophoresis, the composition of the two fucosylated protein transport phases could be partially resolved. Radioactivity in individual gel peaks represented primarily in the first phase decayed with an average half-life of one day, althouth that in one prominent protein of molecular weight 280,000 turned over with a half-life of the order of 12 h. Radioactive peaks primarily in the second phase decayed with an average half-life of more than a week. Incorporation of [35S]sulfate into acid insoluble material in the retina was maximal at 1-2 h, after which there was a rapid loss of label. The appearance of [35S]sulfate in the optic tract, lateral geniculate body and superior colliculus preceded by a short time that of the [3H]fucose; indicating a shorter retinal processing time for this label. The total transported [35S]sulfate in the superior colliculus peaked by 4-8 h and had fallen by 65% at one day; no prominent second wave of transport was observed as was the case for [3H]fucose. Acid insoluble [35S]sulfate in the superior colliculus was equally divided between glycopeptides and glycosaminoglycans at all times examined, indicating that these macromolecules are transported at the same rate. [35S]Sulfate incorporated into various proteins fractionated by gel electrophoresis had heterogeneous turnover rates, the average being around 12 h. Radioactivity in one group of proteins, of molecular weight around 90,000, decayed with a half-life of only a few hours.     

Axonal transport and metabolism of [H]fucose- and [S]- Sulfate-labeled macromolecules in the rat visual system

The axonal transport of labeled macromolecules in retinal ganglion cells of rats was investigated from 1 to 20 days following intraocular injection of [³H]fucose and [³⁵S]sulfate. Maximal incorporation of [³H]fucose into acid insoluble material in the retina was at 8 h, followed by a biphasic decline. Transported [³H]fucose (98% as glycoprotein) was in the optic nerve at 1 h, the optic tract and lateral geniculate body by 2 h, and the superior colliculus by 3 h after injection, indicating a rate of transport of approximately 200 mm/day. Radioactivity continued to accumulate in the superior colliculus for at least 8 h and began to decline rapidly by 24 h. Between 3 and 6 days levels rose again in both optic tract and superior colliculus before starting a gradual decline, indicating that a wave of rapidly transported material was delayed in leaving the retina. When proteins in the superior colliculus were fractionated by gel electrophoresis, the composition of the two fucosylated protein transport phases could be partially resolved. Radioactivity in individual gel peaks represented primarily in the first phase decayed with an average half-life of one day, although that in one prominent protein of molecular weight 280,000 turned over with a half-life of the order of 12 h. Radioactive peaks primarily in the second phase decayed with an average half-life of more than a week.Incorporation of [³⁵S]sulfate into acid insoluble material in the retina was maximal at 1–2 h, after which there was a rapid loss of label. The appearance of [³⁵S]sulfate in the optic tract, loateral geniculate body and superior colliculus preceded by a short time that of the [³H]fucose; indicating a shorter retinal processing time for this label. The total transported [³⁵S]sulfate in the superior colliculus peaked by 4–8 h and had fallen by 65% at one day; no prominent second wave of transport was observed as was the case for [³H]fucose. Acid insoluble [³⁵S]sulfate in the superior colliculus was equally divided between glycopeptides and glycosaminoglycans at all times examined, indicating that these macromolecules are transported at the same rate. [³⁵S]Sulfate incorporated into various proteins fractionated by gel electrophoresis had heterogeneous turnover rates, the average being around 12 h. Radioactivity in one group of proteins, of molecular weight around 90,000, decayed with a half-life of only a few hours.     

Axonal transport of glycerophospholipids in regenerating sciatic nerve of the rat during aging

The effect of age upon the axoplasmic transport of glycerophospholipids has been studied using as a model the regenerating sciatic nerve of young (2-month-old), young adult (6-month-old), middle-aged (16-month-old), and aged (20-month-old) male rats. The right sciatic nerve was crushed 0.5 mm down the incisura ischiadica. Four and nine days after the lesion, a mixture of [2-3H] glycerol and [methyl-14C] choline was bilaterally injected into the spinal cord, at a level of the L4-L5 vertebrae. The animals were killed 18 hr after the isotope injection. Proximal and distal portions of crushed nerve and of contralateral sham-operated ones were dissected and consecutive 5-mm segments were subjected to lipid extraction and analysis. The findings of the present study are summarized as follows: (1) The accumulation of labeled lipid material axonally transported four days after nerve injury was mainly located at the crush site in young, young adult, middle-aged, and aged rats. The accumulation of both 3H-glycerolipids and 14C-choline phospholipids in postcrush segments was markedly higher for young and young adult than for aged rats, four and nine days after crush; (2) the average rate of axonal regeneration, determined between days 4 and 9 following crush injury was 3.6 and 4.2 mm/day for 2-month-old and 6-month-old rats, respectively; it decreased to the value of 2.5 mm/day for 16-20-month-old rats.     

Effect of neonatal capsaicin treatment on cholecystokinin-(CCK8) satiety and axonal transport of CCK binding sites in the rat vagus nerve

Cholecystokinin (CCK) binding sites which accumulate at ligatures placed on the rat vagus nerve may mediate the satiety actions of CCK. Treatment of neonatal rats with capsaicin attenuated the satiety effect of injected CCK in adult life. Capsaicin pretreatment also reduced, but did not eliminate, the accumulation of CCK binding sites proximal and distal to ligatures on either cervical trunk. A similar effect was observed following ligation of subdiaphragmatic vagal trunks. The CCK receptor antagonists, MK-329 and L-365,260, inhibited binding to capsaicin- and vehicle-treated nerves to a similar degree. Densities of CCK binding sites in the nucleus tractus solitarius and area postrema were also markedly affected by neonatal capsaicin treatment.     

Axonal transport and distribution of endogenous calcitonin gene-related peptide in rat peripheral nerve

Calcitonin gene-related peptide (CGRP) has been found in both sensory and motor neurons. It has been suggested that CGRP is transported from neuron cell bodies to their terminals, where it may act as an anterograde trophic factor. However, it is not known how fast CGRP is transported or whether CGRP found in the innervated target organ indeed originated in neural tissues. We have quantified endogenous CGRP in the rat peripheral nerve by a newly developed enzyme immunoassay. The CGRP immunoreactive material obtained from neural tissues coincided with synthetic rat CGRP in fractional distributions separated by gel filtration. After ligation of the sciatic nerve, tissue CGRP accumulated in the segment central to the ligature. The rate of anterograde transport of CGRP was about 1 mm/hr in both sensory and motor fibers. In the sciatic nerve, only a small fraction of CGRP measured was found to originate from the motor nerve fibers. This may be due in part to the disproportionately large number of sensory fibers in the sciatic nerve and in part to the possible presence of CGRP in sympathetic nerve fibers. The CGRP content in the dorsal root fibers was significantly lower than that in the peripheral processes of the sensory neurons. The CGRP content in the hind leg muscle was much higher than that expected from the amount of CGRP per nerve fiber in the sciatic nerve. Most CGRP in muscle disappeared following denervation. It is concluded that CGRP highly concentrated in nerve terminals is supplied by axonal transport from the neuron cell bodies.     

Differential effects of endocannabinoids on [(3)H]-GABA uptake in the rat globus pallidus

In the globus pallidus, cannabinoid CB(1) receptors are localized pre-synaptically on GABAergic neurons. We assessed the influence of the endocannabinoids, anandamide, 2-arachidonoyl-glycerol (2-AG) and noladin ether, on the uptake of [(3)H]-GABA in pallidal slices from rat. Both 2-AG and noladin ether increased [(3)H]-GABA uptake (by 40.8 +/- 8.0% and 38.4 +/- 12.5%). The effect of 2-AG was blocked by the cannabinoid CB(1) receptor antagonist AM 251. In contrast, neither anandamide nor the agonist WIN 55,212-2 had an effect on [(3)H]-GABA uptake. Different roles might be played by different endocannabinoids, both physiologically and in basal ganglia disorders, such as Parkinson's disease.     

Axonal transport of norepinephrine and choline acetyltransferase in regenerating sciatic nerve of the rat.

Axonal transport of norepinephrine (NE) and choline acetyltransferase (ChAc) in the distal portion to the transected and sutured region in the rat sciatic nerve was investigated together with correlations of the contractility of the reinnervated gastrocnemius-plantaris muscle. In the fluorescence histochemical study, the initial delay and the rate of regeneration of noradrenergic fibers were found to be approximately 2 days and 4 mm/day, respectively. In the regenerating nerves, the accumulation of ChAc was about one-half that of normal at 4 weeks and returned to normal at 12 weeks. The amplitude of muscle contraction induced by stimulation of the sciatic nerves was one-half that of normal at 4 weeks and recovered to normal at 12 weeks. The accumulation of NE was still one-fourth that of normal at 12 weeks, and catecholamine fluorescence in the sural artery was still weak, fine, and scattered, compared to that in the artery innervated by the normal sciatic nerve. These results indicate that the function of cholinergic motor fibers correlates with the accumulation of ChAc and that the recovery in amount of transportable NE in the sciatic nerve is much slower than that of the cholinergic component.     

Withdrawal from butorphanol dependence alters binding of [3H]phorbol dibutyrate to protein kinase C, but not of [3H]forskolin to adenylate cyclase

The time course of autoradiographic binding of major second messengers in the rat brain was studied at 2, 7, and 24 h after withdrawal from butorphanol infusion. [3H]Forskolin and [3H]phorbol 12,13-dibutyrate (PDBu) were used to label adenylate cyclase and protein kinase C (PKC), respectively. Rats were rendered dependent by intracerebroventricular infusion of butorphanol (26 nmol microliter-1 h-1) via osmotic minipumps for 3 days. Withdrawal was initiated by abrupt cessation of the butorphanol infusion. The levels of [3H]forskolin binding were not changed at any time or in any brain area, except for an increase following 7 h of withdrawal in the brainstem only. The levels of [3H]PDBu binding were significantly increased (13-47%) in multiple areas of the rat brain following 7 h of withdrawal from butorphanol infusion. These findings suggest that the phosphoinositide cycle system is more susceptible to alteration during butorphanol dependence than is the adenylate cyclase system in the rat brain.     

Release of [3H]L-glutamine acid (L-Glu) and [3H]D-aspartic acid (D-Asp) in the area of nucleus tractus solitarius in vivo produced by stimulation of the vagus nerve

We have investigated the effects of bilateral electrical stimulation of the vagus nerves in anesthetized, paralyzed rats on the release of exogenously administered [3H]L-glutamic acid ([3H]L-Glu) or [3H]D-aspartic acid ([3H]D-Asp) from the intermediate portion of the nucleus tractus solitarius (NTS). Electrical stimulation of afferent fibers with the frequency, pulse, duration, and intensity required to activate C-fibers, elicited hypotension and bradycardia. Such stimuli induced the release of [3H]L-Glu, or its stable analogue [3H]D-Asp, from the NTS into perfusate collected through push-pull cannulae. The release of radioactive materials, calculated as a percent of increase in radioactivity above the prestimulation level, was for [3H]L-Glu 114.4 +/- 25.1% (n = 20) during bilateral vagal stimulation, and 45.6 +/- 11.3% (n = 9) (P less than 0.001) during unilateral stimulation. The release of [3H]D-Asp induced by bilateral vagal stimulation was 100.4 +/- 31.9%. The release, which was anatomically specific and restricted to the NTS, was directly related to stimulus (and hence reflex) intensity. Overflow of the inert substances [14C]urea OR [14C]sucrose, co-administered with the [3H] amino acids, did not increase at the same time. Local depolarization of the cells in the NTS by K+ (53mM) increased the overflow of [3H]L-Glu, as well as [14C]urea, and was able to induce the release of [3H]L-Glu when electrical stimulation failed to have an effect. The results are consistent with the hypothesis that L-Glu is a neurotransmitter of neurons in the NTS mediating vasodepressor response from vagal afferents, including those from systemic baroreceptors.