* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Stimulation Within the Rostral Ventrolateral Medulla Can Evoke
Long-term depression wikipedia , lookup
Neural engineering wikipedia , lookup
Artificial general intelligence wikipedia , lookup
Neuroeconomics wikipedia , lookup
Environmental enrichment wikipedia , lookup
Action potential wikipedia , lookup
Signal transduction wikipedia , lookup
NMDA receptor wikipedia , lookup
Resting potential wikipedia , lookup
Apical dendrite wikipedia , lookup
Activity-dependent plasticity wikipedia , lookup
Metastability in the brain wikipedia , lookup
Synaptogenesis wikipedia , lookup
Mirror neuron wikipedia , lookup
End-plate potential wikipedia , lookup
Multielectrode array wikipedia , lookup
Axon guidance wikipedia , lookup
Transcranial direct-current stimulation wikipedia , lookup
Neural coding wikipedia , lookup
Development of the nervous system wikipedia , lookup
Endocannabinoid system wikipedia , lookup
Neural oscillation wikipedia , lookup
Biological neuron model wikipedia , lookup
Neurotransmitter wikipedia , lookup
Nonsynaptic plasticity wikipedia , lookup
Caridoid escape reaction wikipedia , lookup
Electrophysiology wikipedia , lookup
Central pattern generator wikipedia , lookup
Circumventricular organs wikipedia , lookup
Clinical neurochemistry wikipedia , lookup
Neuroanatomy wikipedia , lookup
Single-unit recording wikipedia , lookup
Premovement neuronal activity wikipedia , lookup
Spike-and-wave wikipedia , lookup
Neurostimulation wikipedia , lookup
Stimulus (physiology) wikipedia , lookup
Nervous system network models wikipedia , lookup
Optogenetics wikipedia , lookup
Molecular neuroscience wikipedia , lookup
Feature detection (nervous system) wikipedia , lookup
Chemical synapse wikipedia , lookup
Pre-Bötzinger complex wikipedia , lookup
Neuropsychopharmacology wikipedia , lookup
Synaptic gating wikipedia , lookup
Stimulation Within the Rostral Ventrolateral Medulla Can Evoke Monosynaptic GABAergic IPSPs in Sympathetic Preganglionic Neurons In Vitro SUSAN A. DEUCHARS, K. MICHAEL SPYER, AND MICHAEL P. GILBEY Royal Free Hospital School of Medicine, London NW3 2PF, United Kingdom INTRODUCTION The rostral ventrolateral medulla (RVLM) contains neurons that are involved in the maintenance of sympathetic tone and blood pressure (see Guyenet 1990; Ross et al. 1984) and that have direct projections onto sympathetic preganglionic neurons (SPNs) in the spinal cord (Zagon and Smith 1993). Recent experiments using a neonatal rat brain stem-spinal cord preparation have shown that stimulation of the RVLM can elicit excitatory postsynaptic potentials (EPSPs) in SPNs that are mediated by an excitatory amino acid acting on both non-N-methyl-D-aspartate (NMDA) and NMDA receptors (Deuchars et al. 1995a). At least a proportion of this excitatory response was mediated by a monosynaptic input. During such experiments, on several occasions when the EPSPs were blocked by non-NMDA and NMDA receptor antagonists, underlying inhibitory postsynaptic potentials (IPSPs) were observed after RVLM stimulation. Such observations are of particular interest because the RVLM is regarded as an area involved in providing a purely excitatory input onto SPNs and the existence of a descending inhibitory input onto SPNs has yet to be shown electrophysiologically. Thus stimulation of the RVLM has been shown to elicit increases in blood pressure and sympathetic activity (Guertzenstein and Silver 1974; Morrison et al. 1988) whereas lesions of this region, or applications of inhibitory neurotransmitters, decreased blood pressure, heart rate, and sympathetic postganglionic activity (Granata et al. 1985; Yardley et al. 1989). A recent study revealed a number of RVLM neurons retrogradely labelled after injections into the intermediolateral cell column (IML) that were immunoreactive for g-aminobutyric acid (GABA) (Miura et al. 1994). Thus some GABAergic RVLM neurons may innervate SPNs in the thoracic spinal cord. Therefore, using a neonatal rat brain stem-spinal cord preparation, it was decided to study in detail the IPSPs evoked in SPNs after stimulation within the RVLM. The possibility that these IPSPs were due to activation of a long descending monosynaptic inhibitory pathway onto SPNs was of particular interest because the existence of such a pathway has yet to shown. Preliminary reports of this study have been published (Deuchars et al. 1994). METHODS Anesthesia was induced in neonatal rats (2–5 days) with isoflurane. Rats then were placed on ice to maintain anesthesia by hypothermia as described previously (Deuchars et al. 1995b). Decerebration was performed rapidly and the brain stem-spinal cords were isolated (see Deuchars et al. 1995b) and pinned in a recording chamber, twisting the spinal cord so that the ventral 0022-3077/97 $5.00 Copyright q 1997 The American Physiological Society / 9k0b$$ja24 J-022-6 08-13-97 17:58:39 neupa LP-Neurophys 229 Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 17, 2017 Deuchars, Susan A., K. Michael Spyer, and Michael P. Gilbey. Stimulation within the rostral ventrolateral medulla can evoke monosynaptic GABAergic IPSPs in sympathetic preganglionic neurons in vitro. J. Neurophysiol. 77: 229–235, 1997. The inhibitory responses of identified sympathetic preganglionic neurons (SPNs) to stimulation within the rostral ventrolateral medulla (RVLM) were studied to determine their nature and pharmacology. Whole cell patch-clamp recordings were made from 36 SPNs in the upper thoracic segments of the spinal cord in a neonatal rat brain stem-spinal cord preparation. Neurons were identified as SPNs on the basis of their antidromic activation after stimulation of the ipsilateral segmental ventral root and their morphology and location in the intermediolateral cell column and intercalated nucleus. In all SPNs, electrical stimulation of the RVLM evoked fast excitatory postsynaptic potentials (EPSPs) that were mediated by non-N-methyl-D-aspartate (NMDA) and NMDA receptors. These excitatory responses were the most prominent response in control artificial cerebrospinal fluid and have been studied previously. In 22 of the SPNs, RVLM stimulation also elicited fast inhibitory postsynaptic potentials (IPSPs), which increased in amplitude as the membrane was depolarized. Five of these neurons were not studied further as they responded occasionally with IPSPs that had highly variable onset latencies indicating the involvement of a polysynaptic pathway. In the remaining SPNs (n Å 17), the evoked IPSPs persisted in the presence of the excitatory amino acid antagonists 6-cyano-7-nitroquinoxaline-2,3,-dione and D,L-2-amino-5phosphonopentanoic acid. In eight of these SPNs, it was necessary to block the EPSPs to reveal the IPSPs. In the 7 SPNs tested, the onset latencies of the IPSPs were not significantly different from the onset latencies of the fast EPSPs. The low sweep-to-sweep fluctuations in onset latency of individual IPSPs (absolute average deviation: 0.4 ms) indicated that the IPSPs were elicited by activation of a monosynaptic pathway. The amplitudes of the IPSPs decreased in amplitude as the membrane was hyperpolarized and reversed in polarity at 070.3 { 1.7 mV (mean { SD), which was close to the equilibrium potential for chloride ions. In addition, in seven SPNs, bath applications of 5 mM bicuculline, a g-aminobuturic acid-A (GABAA ) antagonist, abolished or reduced the evoked IPSPs. Five SPNs also were studied that displayed ongoing IPSPs. The amplitudes of these IPSPs increased with membrane depolarization and were blocked by bath applications of 5 mM bicuculline, suggesting that they also were mediated by activation of GABAA receptors. These results demonstrate the existence of a bulbospinal GABAergic pathway impinging directly onto SPNs. This pathway may be tonically active in the neonatal rat brain stem-spinal cord preparation. 230 S. A. DEUCHARS, K. M. SPYER, AND M. P. GILBEY surface of the brain stem and the lateral surface of the thoracic spinal cord were uppermost (Deuchars et al. 1995a) (see Fig. 1A). The preparation was perfused at a rate of 5 ml/min with artificial cerebrospinal fluid (ACSF) composed of (in mM) 128 NaCl, 3 KCl, 0.5 NaH2PO4rH2O, 1.5 CaCl2r2H2O, 1 MgSO4 , 23.5 NaHCO3 , 30 glucose, and 2 mannitol equilibrated with 95% O2 – 5% CO2 and maintained at 26–277C. A suction electrode was placed on an upper thoracic ventral root and used to record ongoing activity to determine the viability of the preparation and stimulate axons within the root for the antidromic activation of SPNs. Ongoing ventral root activity displayed rhythmic respiratory activity that was characterized by bursts of activity during inspiration while the ongoing activity had a peak in the power spectral analysis, which occurred in the 2- to 6-Hz range, similar to that observed in sympathetic nerves recorded in vivo (see Deuchars et al. 1995b). A monopolar stimulating electrode was placed in the RVLM. The RVLM was stimulated with single or twin pulses using an isolated stimulator (Digitimer DS2; 1 ms pulse width; 6 ms apart, 50–100 mA) (see Deuchars et al. 1995b). RVLM stimulation elicited a ‘‘fast’’ response occurring Ç20–60 ms after stimulation and triggered a premature inspiratory burst that reset the respiratory rhythm (see Deuchars et al. 1995a). Using these parameters to position the electrodes, all stimulating sites were found to be located within the RVLM. If the electrode was moved 500 mm either rostral or medial to the RVLM, stimulation of these areas using the same parameters did not evoke a response in the ventral root. Returning the electrode to the RVLM and stimulating there restored responses in the ventral root. Whole cell patch-clamp recordings were made from SPNs using a blind patch technique where rupture of the neuronal membrane was attempted when the seal resistance had reached a value of ¢1 GV (see Deuchars et al. 1995a). The electrodes were filled with (in mM) 145 Kgluconate, 2 MgCls2 , 5 N-2-hydroxyethylpiperazine-N *-2-ethanesulfonic acid, 1.1 ethylene glycol-bis( b-aminoethyl ether)-N,N,N *,N *-tetraacetic acid, 0.1 CaCl2 , and 5 K2ATP 5 (pH 7.2; osmolality 310 mOsmol/kg H2O). Using this intracellular solution together with the extracellular solution, junction potentials averaging 1.9 { 0.6 mV (mean { SD) were measured for six electrodes. This value has not been subtracted from the values / 9k0b$$ja24 J-022-6 given in the text. The solution also contained 0.5–1% biocytin for labelling neurons and identifying them as SPNs (see Fig. 1C). Recordings were made using an Axoclamp 2A amplifier in bridge balance mode. During the search for a neuron, current pulses of 50–1250 pA (10- to 30-ms pulse width) were applied at a rate of 2 Hz. To determine the input resistance of a neuron, currents of 10–50 pA (200- to 350-ms pulses width) were applied and the resulting voltage deflections measured. All drugs were dissolved in ACSF at known concentrations and applied to the perfusing medium and as such were subject to a dead space of Ç30 ml due to the volume of the bath and perfusion tube. The following drugs were applied: the non-NMDA receptor antagonist, 6-cyano-7-nitroquinoxaline-2,3,-dione (CNQX); the NMDA receptor antagonist, D,L-2-amino-5-phosphonopentanoic acid (AP-5; Research Biochemicals International) and the GABAA receptor antagonist bicuculline methochloride (Sigma). All drug concentrations given are the final concentrations in the perfusing medium. Data analysis Membrane potentials, current injections, whole nerve recordings and trigger pulses were stored via an interface (Instrutech; 11 kHz/ channel sampling rate) on video tape for analysis. All data analysis was carried out using an IBM-compatible microcomputer (interface and software supplied by Cambridge Electronic Design UK Ltd). Using the Nernst equation and knowing the intracellular and extracellular concentrations of chloride, an equilibrium potential of 088 mV for this ion was calculated. The equilibrium potential for potassium ( 0101 mV) was calculated also in this way. The fluctuations in onset latency of the IPSPs for each SPN were determined by measuring the onset latencies of 20 single sweep IPSPs. The mean amount by which these values varied from the mean onset latency then was calculated for each neuron (average absolute deviation). The differences in onset latency of the control EPSPs and the IPSPs (in CNQX and AP-5) were determined by statistical comparison of the onset latencies of 20 single sweep EPSPs and IPSPs using the Mann-Whitney U test. The reversal potentials of IPSPs were calculated by plotting changes in IPSP amplitude against holding potential and the linear regression for each neuron 08-13-97 17:58:39 neupa LP-Neurophys Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 17, 2017 FIG . 1. Experimental setup and recording and stimulating sites. A: line drawing of brain stemspinal cord preparation shows positioning of electrodes in rostral ventrolateral medulla (RVLM), spinal cord, and on ventral root. B: photomicrograph of a section of brain stem with a lesion in RVLM indicating stimulation site. Scale bar Å 0.2 mm. C: photomicrograph of a section of thoracic spinal cord showing a filled SPN with cell body in intermediolateral cell column (IML) and axon coursing ventrally to exit in ventral root. Scale bar Å 100 mm. MEDULLARY INHIBITION OF SYMPATHETIC NEURONS 231 was determined. The effects of bicuculline on the amplitudes of IPSPs were determined using the Wilcoxon matched pairs test (on untransformed data). For the ongoing IPSPs, measurements of the amplitudes of 20 consecutive ongoing IPSPs in control and bicuculline solutions were taken to obtain a mean amplitude and the Wilcoxon matched pairs test was carried out on the untransformed data. All values are given as means { SE unless stated. Histology SPNs were filled with biocytin during the recording period, and at the end of the experiment, 80-mm sections of thoracic spinal cord were cut and processed for visualization of the SPNs (see Deuchars et al. 1995a) (Fig. 1C). In addition, the sites in the medulla were lesioned at the end of each experiment and 100-mm sections of medulla cut and stained with neutral red or thioneine then viewed under a microscope to locate the sites of stimulation (see Fig. 1B). Presented in this paper are data from 36 neurons, which were located at depths of 47–329 mm below the lateral surface of the spinal cord. These were identified as SPNs on the basis of their antidromic activation after stimulation of the ipsilateral segmental ventral root and their morphology and location in the spinal cord. Antidromically evoked action potentials were characterized by a constant latency of response and by an inflection on the rising phase of the action potential that became more prominent as the membrane was hyperpolarized. Such inflections were due to the initial segment-somatodendritic (IS-SD) break. If the membrane was further hyperpolarized, the SD spike failed leaving just the long-lasting IS spike (see Deuchars et al. 1995a). All neurons were located in the IML (see Fig. 1C for an example) or intercalated nucleus. The mean resting membrane potential for these neurons was 042 { 0.9 mV and the input resistance measured for nine SPNs was 289 { 45 MV. Action potentials displayed overshoots and were similar in amplitude and duration to that observed previously (Deuchars et al 1995a). The recordings obtained from the neurons used in this study were stable throughout the recording period of °2 h. These electrophysiological and morphological properties of the neurons were equivalent to those reported previously (Deuchars et al. 1995a). Responses of SPNs to stimulation of the RVLM Electrical stimulation of the RVLM elicited excitatory responses in all SPNs tested (n Å 36). These responses had similar characteristics to those reported by Deuchars et al. (1995a) in that, in all SPNs tested, RVLM stimulation elicited fast EPSPs that displayed conventional voltage relationships and were sensitive to applications of the non-NMDA receptor antagonist CNQX; and in a proportion of neurons, slow EPSPs were evoked also that decreased in amplitude with membrane hyperpolarization and were reduced in amplitude by applications of AP-5. In 22 of these SPNs (61% of neurons tested), RVLM stimulation also elicited IPSPs that merited further analysis. Of these 22 SPNs, 5 responded with only occasional IPSPs of variable latency, and in two such cases, IPSPs were abolished by applications of the excitatory amino acid antago- / 9k0b$$ja24 J-022-6 FIG . 2. Single sweeps of the inhibitory responses of 4 sympathetic preganglion neurons (SPNs; membrane potential of 040 mV) to RVLM stimulation. Ai: in this neuron, RVLM stimulation elicited excitatory postsynaptic potentials (EPSPs) that were dominant event at all potentials and, at this potential, reached threshold for firing thus masking inhibitory postsynaptic potential (IPSP). Scale bars Å 20 mV and 50 ms. Aii: EPSP was blocked by applications of 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) to reveal underlying IPSPs, which occurred at similar onset latencies, and this onset latency was not significantly different from latency to onset of fast CNQXsensitive EPSP. Scale bars Å 5 mV and 50 ms. Bi: this is an example of a SPN where RVLM stimulation induced an EPSP/IPSP complex at a membrane potential of 040 mV. Bii: in presence of CNQX and D,L-2amino-5-phosphonopentanoic acid (AP-5), IPSP persisted, and single sweeps of IPSP show low degree of fluctuation of onset latency. In addition, onset latency of IPSP was not significantly different from that of EPSP. Inflections on the rising phases of the IPSPs are marked by arrows. Scale bars Å 3 mV and 50 ms. C: in this SPN, responses at 040 mV were predominantly inhibitory, and 3 sweeps show low degree of fluctuation in sweep-to-sweep onset latency of IPSP. Scale bars Å 5 mV and 50 ms. Di: this neuron responded to RVLM stimulation with an IPSP that fluctuated in onset latency from sweep to sweep. In the 2nd trace, RVLM stimulation failed to elicit an IPSP, and only ongoing synaptic potentials are observed on the trace. Dii: applications of CNQX and AP-5 blocked evoked IPSPs but not ongoing IPSP illustrated. Thus evoked IPSP in this case was considered polysynaptic and was not studied further. Scale bars Å 3 mV and 50 ms. nists CNQX and AP-5 (see Fig. 2D). Such observations indicate that these responses were elicited via activation of a polysynaptic pathway. Consequently these neurons were not included in the remainder of the study. It was, however, interesting to note the difference in the variabilities of the latencies of the IPSPs evoked over presumed polysynaptic pathways and those short latency IPSPs elicited in the SPNs 08-13-97 17:58:39 neupa LP-Neurophys Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 17, 2017 RESULTS 232 S. A. DEUCHARS, K. M. SPYER, AND M. P. GILBEY that formed the bulk of the study (see Fig. 2 and below for values of absolute average deviation). In addition, RVLM stimulation did not always elicit IPSPs in these five neurons, in contrast to the other 17 SPNs where short constant latency IPSPs were elicited after every RVLM stimulation (see Fig. 2). The sites at which IPSPs were elicited were all located in the RVLM (see Fig. 1B for example), and there was no variation in the sites at which IPSPs could or could not be elicited. Indeed, on several occasions it was possible to record one SPN in which an IPSP was evoked and one SPN which was not inhibited by stimulation at the same RVLM site. Characteristics of the fast IPSPs / 9k0b$$ja24 J-022-6 FIG . 3. EPSPs and IPSPs evoked by RVLM stimulation. A: in this SPN, EPSPs were elicited that increased in amplitude as membrane was hyperpolarized. At 040 mV, EPSP reached threshold for firing; this trace shows averaged action potentials that occur at different points on the EPSPs. B: in presence of CNQX and AP-5, stimulating in same area with same parameters revealed an IPSP that decreased in amplitude as membrane was hyperpolarized. Calculated reversal potential using linear regression for this neuron was 065.8 mV, although from raw data, it can be seen that IPSP is not present at 060 mV. All EPSPs and IPSPs shown are averages of 10– 15 sweeps. Inflections on rising phases of the IPSPs are marked by arrows. C: graph of changes in IPSP amplitude against holding potential for 5 SPNs. Reversal potentials are calculated using linear regression, and mean reversal potential was 070.3 { 1.7 mV. the fast EPSPs in control medium. In seven SPNs, the single sweep onset latencies of EPSPs and IPSPs (10–20 sweeps of each) were measured for each SPN, and it was shown that the onset latencies for the EPSPs and the IPSPs were not significantly different in all SPNs tested (smallest value of P ú 0.18). The amplitude of the IPSPs was measured at 040 mV, and the mean amplitude was 6.8 { 0.9 mV (n Å 17). As the membrane was hyperpolarized, the amplitudes of the IPSPs decreased, and in a number of SPNs (n Å 5), the reversal potentials for the IPSPs were calculated as 070.3 { 1.7 mV (Fig. 3, B and C). This value is more depolarized than the calculated equilibrium potential for chloride ions ( 088 mV) and may be due to a number of factors. The junction potentials measured for the intracellular and extracellular solutions used in this study were small (1.9 { 0.6 mV, mean { SD, n Å 6) and were not taken into 08-13-97 17:58:39 neupa LP-Neurophys Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 17, 2017 In control ACSF, RVLM stimulation elicited an EPSP that was the dominant event in 15 of 17 SPNs, even at potentials of 050 mV or less. Moreover in 8 of the 17 SPNs where RVLM stimulation elicited both EPSPs and IPSPs, the inhibitory effect of RVLM stimulation was masked completely by the EPSP because at depolarized potentials ( 050 mV or less), the EPSP reached the threshold potential for firing (Fig. 2A). In these eight neurons, applications of CNQX (10 mM) and AP-5 (50 mM) revealed RVLM-evoked IPSPs that persisted throughout the applications of these excitatory amino acid antagonists (Fig. 2A). Seven of the 17 SPNs displayed multiphasic EPSP/IPSP complexes at potentials of 050 mV or less (Fig. 2B). In the other two SPNs, the IPSP was the dominant event at 040 mV (Fig. 2C). In both of these groups, the IPSPs persisted in the presence of CNQX and AP-5. IPSPs elicited by RVLM stimulation were characterized by inflections on the rising phases (arrows in Figs. 2 and 3). These are likely to be due to activation of a population of neurons whose axons have a range of conduction velocities—similar notches also were observed on the rising phases of RVLM-evoked EPSPs (see Deuchars et al. 1995a). The onset latencies of all IPSPs (n Å 17) were calculated at a holding potential of 040 mV in the presence of CNQX and AP-5 because this unmasked the actual onset of the IPSP. The mean onset latency was 31.4 { 2.3 ms (n Å 17). Taking the conduction distance as the distance from the stimulating electrode to the recording electrode, an approximation of conduction velocities was calculated as 0.39 { 0.02 m/s (n Å 17), which falls within the range for C fibers as observed previously (see Deuchars et al 1995a). For a number of SPNs (n Å 14), the sweep-to-sweep fluctuations in IPSP onset latencies were evaluated by measuring single sweep onset latencies for each neuron and determining the amount by which these deviate from the mean latency for the neuron. This calculation provides a value for the average absolute deviation of latency. The mean of the average absolute deviations for all SPNs tested was 0.40 { 0.04 ms (n Å 14). This value could be compared with a value of 6.3 { 1.0 ms (n Å 4) for the absolute average deviation of those IPSPs elicited by activation of presumed polysynaptic pathways as described in the above section. In addition, the onset latencies of some of the IPSPs in the presence of CNQX and AP-5 were compared with those of MEDULLARY INHIBITION OF SYMPATHETIC NEURONS 233 was 3.1 mV at a holding potential of 040 mV and decreased as the membrane was hyperpolarized (see Fig. 5A). In all three SPNs tested, these IPSPs persisted in the presence of CNQX and AP-5. For three neurons, the mean reversal potential of the IPSPs was calculated as 069.2 mV. For two of these SPNs, the effects of applications of 5 mM bicuculline were tested and found to decrease the amplitude of the ongoing IPSPs to 13.3% of the control amplitudes (Fig. 5B). Recovery from this antagonism was observed in one neuron (not shown). DISCUSSION Evidence of a monosynaptic inhibitory pathway and its chemical nature In this preparation, IPSPs that persisted in the presence of the excitatory amino acid antagonists, CNQX and AP-5, could be elicited in SPNs by stimulation of the RVLM. FIG . 4. Bicuculline antagonism of evoked IPSPs. A: this trace shows an average of EPSPs evoked in this SPN by stimulation of RVLM at a membrane potential of 070 mV and sensitivity of the response to excitatory amino acid antagonists CNQX (15 mM) and AP5 (50 mM). B: these traces in presence of CNQX and AP-5 reveal an IPSP at depolarized potential of 040 mV. This IPSP was antagonised by bath application of g-aminobuturic acid-A receptor antagonist bicuculline (5 mM). C: graph of changes in IPSP amplitude against holding potential for this neuron. Calculated reversal potential was 072 mV. account in the measurements of the reversal potentials. This small value may contribute in part to the depolarized values measured. In addition, it may be that the chloride channels in SPNs are slightly permeable to bicarbonate ions (FatimaShad and Barry 1993); this leads to a slightly more depolarized reversal potential. Other receptors also may be activated that contribute slightly to the postsynaptic potentials at more hyperpolarized potentials. These data, however, do suggest a role for chloride channels in mediating these responses, rather than potassium channels because the calculated equilibrium potential for potassium was 0101 mV. Effects of application of bicuculline In seven SPNs, bath applications of the GABAA antagonist bicuculline (5 mM) caused a significant decrease in the amplitude of the IPSP to 17.2 { 5.8% of the control response (i.e., 8.3 { 1.8 mV to 1.9 { 0.8 mV at a holding potential of 040 mV, P õ 0.05; see Fig. 4). Partial recovery of the IPSP was observed after returning to control medium. Ongoing IPSPs in the brain stem-spinal cord preparation In five SPNs, ongoing IPSPs were observed in the absence of RVLM stimulation. The average amplitude of these IPSPs / 9k0b$$ja24 J-022-6 FIG . 5. Ongoing IPSPs that are sensitive to bicuculline. A: this neuron (and 4 other neurons) exhibited ongoing IPSPs that were voltage sensitive. Thus hyperpolarization of neuron from 030 to 040 mV decreased amplitude of IPSPs, whereas at 070 mV, IPSPs were reversed in polarity. B: Application of 5 mM bicuculline antagonised these IPSPs. 08-13-97 17:58:39 neupa LP-Neurophys Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 17, 2017 These observations indicate that stimulation of the RVLM elicits IPSPs in SPNs and that at least in part, these IPSPs are mediated by activation of a monosynaptic bulbospinal inhibitory pathway. The IPSPs were reduced by applications of bicuculline, indicating that GABAA receptors are involved in mediating these responses. These results therefore reveal the presence of a monosynaptic bulbospinal inhibitory pathway onto SPNs from the RVLM or sites more rostral. 234 S. A. DEUCHARS, K. M. SPYER, AND M. P. GILBEY / 9k0b$$ja24 J-022-6 nal cord transverse slice preparations. In these preparations, ongoing IPSPs or those evoked by focal stimulation near the recorded neurons were mediated by activation of glycinergic receptors (Dun and Mo 1989; Krupp and Feltz 1993; Spanswick et al. 1994), because strychnine, not bicuculline, blocked these events. The presence of GABAA receptors affecting SPNs in the spinal cord was illustrated by GABAinduced hyperpolarizing currents (Krupp and Feltz 1993). Therefore the GABAergic neurons were remote from the SPNs because the pathway was not maintained in the transverse slice. GABAergic interneurons commonly are found within laminae I–III of the spinal cord (Todd and Sullivan 1990), while a recent study, which used wheat germ agglutinin to transsynaptically label neurons that impinge onto SPNs, showed that the majority of interneurons innervating SPNs were located in laminae IV, V, and VII, close to the labelled SPNs (Cabot et al. 1994). These results therefore suggest that spinal interneurons directly innervating SPNs are not GABAergic. Interestingly, in our preparation, which maintains a high degree of the supraspinal inputs onto SPNs, the ongoing IPSPs observed also were mediated by activation of GABA receptors in contrast to the glycine receptormediated responses demonstrated in the spinal cord slice preparation (Dun and Mo 1989; Krupp and Feltz 1993; Spanswick et al. 1994). This suggests that the ongoing inhibitory activity in this preparation is generated by supraspinal GABAergic neurons, which are lost in the transverse slice. Origin of the GABAergic neurons The results presented in this study show the effects of electrical stimulation of the RVLM on the membrane potentials of SPNs. It remains possible, however, that the observed responses were due either to activation of cell bodies within the RVLM and/or fibers of passage coursing through the RVLM. In our previous study concerning EPSPs evoked by RVLM stimulation, chemical activation of neurons within the RVLM was achieved by microinjections of glutamate. In the present study, however, IPSPs were observed in the presence of bath applied CNQX and AP-5, which would block the effects of glutamate on the RVLM. In two initial experiments, the effects of microinjections of GABA into the RVLM on the ongoing IPSPs were examined and revealed a tendency to decrease the frequency of IPSPs (Deuchars, unpublished observations). However, further studies are necessary to fully determine the contribution of neurons in the RVLM to ongoing SPN activity and possible contributions of neurons in other regions cannot be ruled out. In addition, even if neurons in the RVLM did contribute to the ongoing inhibitory activity in SPNs, they may not be the source of the evoked inhibitory input onto SPNs. The fact that GABA-containing neurons within the RVLM are known to project to the IML of the upper thoracic segments of spinal cord (Miura et al. 1994) suggests that at least some of the inhibitory effects may be due to activation of these neurons in the RVLM. It is not possible to rule out a contribution from neurons in other supraspinal areas whose axons pass through the RVLM and descend to input directly onto SPNs. One area in particular, the A5 region of the pons, is known to contain neurons that directly innervate SPNs and are involved in the control of sympathetic activity. Stimula- 08-13-97 17:58:39 neupa LP-Neurophys Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 17, 2017 These observations indicate that these bulbospinal inhibitory influences on SPNs are not mediated by polysynaptic pathways using inotropic excitatory amino acid receptors. However, this does not rule out a role for other bulbospinal pathways of different chemistry whose axons descend and synapse onto inhibitory interneurons. If such were the case, the neurotransmitter must be activating another fast ionic channel-coupled receptor [e.g., ATP (Evans et al. 1992)] because it was observed that the single sweep onset latencies of the IPSPs recorded in the presence of CNQX and AP-5 were not significantly different from the onset latencies of the control EPSPs. This suggests that if a polysynaptic pathway is involved, its kinetics must be faster than those of excitatory amino-acid receptor mediated events to compensate for the additional synapses. It may be that the bulbospinal neurons involved in mediating the IPSPs have faster conduction velocities than the neurons mediating the fast EPSPs. However, this is an unappealing extrapolation as there is little evidence for myelination in the neonatal rat at this age (Davison and Dobbing 1966). The strongest argument for a monosynaptic descending inhibitory pathway onto SPNs is the degree of fluctuation in the sweep to sweep latencies of each IPSP. The mean absolute average deviations of the onset latencies calculated from single sweep IPSPs from 14 SPNs was 0.4 ms; this means that, on average, the onset latency fluctuated by õ1 ms from sweep to sweep. This is a low degree of fluctuation for the length of pathway ( ú1 cm) in a preparation maintained at 267C (see Deuchars et al. 1995a). This is clearly different from the five SPNs where stimulation of the RVLM-elicited IPSPs, which had very variable onset latencies (average absolute deviation of 6.3 ms) and were blocked by applications of CNQX and AP5, indicating the involvement of neurons in a polysynaptic pathway that use excitatory amino acids. Thus the electrophysiological evidence points to a monosynaptic inhibitory bulbospinal pathway innervating SPNs. The possibility that some components of the IPSP (especially the later components on the decay phase of the IPSPs) are due to activation of polysynaptic pathways cannot be ruled out. The shape of the decay phases of the IPSPs were not affected by CNQX and AP-5 unlike the five SPNs that exhibited purely polysynaptic IPSPs and were abolished by these antagonists. Thus any remaining polysynaptic pathway must use receptors other than inotropic excitatory amino acid receptors. It was observed that both the IPSPs evoked by stimulation of the RVLM and the ongoing IPSPs observed had reversal potentials close to the equilibrium potential for chloride and the amplitudes of the IPSPs were decreased by applications of the GABAA receptor antagonist, bicuculline (5 mM). These two observations indicate that the IPSPs were mediated by activation of GABAA receptors. The monosynaptic nature of the pathway, and its neurochemistry, is consistent with a recent anatomic study that showed that a proportion of neurons retrogradely labelled from injections into the intermediolateral cell column were also immunoreactive for GABA (Miura et al. 1994). In addition, in the adult rat, Bogan and associates (1989) reported the presence of thinly myelinated GABA-LIR axons in the lateral funiculus, which is the location of the descending axons from the brain stem. These results are also in keeping with other electrophysiological studies using rat spi- MEDULLARY INHIBITION OF SYMPATHETIC NEURONS We thank Professor S. F. Morrison, who was involved in some of the initial experiments, and Dr. J. Deuchars for histological advice and comments on the manuscript. This work was supported by the British Heart Foundation and the Wellcome Trust. Address reprint requests to S. A. Deuchars. Received 16 January 1996; accepted in final form 4 September 1996. REFERENCES BOGAN, N., MENNONE, A., AND CABOT, J. B. Light microscopic and ultrastructural localization of GABA-like immunoreactive input to retrogradely labelled sympathetic preganglionic neurons. Brain Res. 505: 257– 270, 1989. CABOT, J. B., ALESSI, V., CARROLL, J., AND LIGORIO, M. Spinal cord lamina V and lamina VII interneuronal projections to sympathetic preganglionic neurons. J. Comp. Neurol. 347: 515–530, 1994. DAVISON, A. N. AND DOBBING, J. Myelination as a vulnerable period in brain development. Br. Med. Bull. 22: 40–44, 1966. DEUCHARS, S. A., MORRISON, S. F., AND GILBEY, M. P. Medullary-evoked EPSPs in neonatal rat sympathetic preganglionic neurons in vitro. J. Physiol. Lond. 487: 453–463, 1995a. DEUCHARS, S. A., MORRISON, S. F., SPYER, K. M., AND GILBEY, M. P. Medullary stimulation can evoke IPSPs in sympathetic preganglionic neurons in the brainstem-spinal cord preparation. Soc. Neurosci. Abstr. 20 Suppl. 1&2: 559.7, 1994. DEUCHARS, S. A., SPYER, K. M., BROOKS, P. A., AND GILBEY, M. P. A / 9k0b$$ja24 J-022-6 study of sympathetic preganglionic neuronal activity in a neonatal rat brainstem-spinal cord preparation. J. Auton. Nerv. Syst. 52: 51–63, 1995b. DUN, N. J. AND MO, N. Inhibitory postsynaptic potentials in neonatal rat sympathetic preganglionic neurons in vitro. J. Physiol. Lond. 410: 267– 281, 1989. EVANS, R. J., DERK ACH, V., AND SURPRENANT, A. ATP mediates fast synaptic transmission in mammalian neurons. Nature Lond. 357: 503–505, 1992. FATIMA-SHAD, K. AND BARRY, P. H. Anion permiability in GABA- and glycine-gated channels of mammalian hippocampal neurons. Proc. Royal Soc. Lond. Ser. B Biol. Sci. 253: 69–73, 1993. GRANATA, A. R., RUGGIERO, D. A., PARK, D. H., JOH, T. H., AND REIS, D. J. Brain stem area with C1 epinephrine neurons mediates baroreflex vasodepressor responses. Am. J. Physiol. 248 (Heart Circ. Physiol. 17): H547–H567, 1985. GUERTZENSTEIN, P. G. AND SILVER, A. Fall in blood pressure produced from discrete regions of the ventral surface of the medulla by glycine and lesions. J. Physiol. Lond. 242: 489–503, 1974. GUYENET, P. G. Role of ventral medulla oblongata in blood pressure regulation. In: Central Regulation of Autonomic Function, edited by A. D. Loewy and K. M. Spyer. New York: Oxford Univ. Press, 1990, p. 145– 167. HUANGFU, D., HWANG, L. J., RILEY, T. A., AND GUYENET, P. G. Splanchnic nerve response to a5-area stimulation in rats. Am. J. Physiol. 263 (Regulatory Integrative Comp. Physiol. 32): R437–R446, 1992. KRUPP, J. AND FELTZ, P. Synaptic-induced and agonist-induced chloride currents in neonatal rat sympathetic preganglionic neurons in vitro. J. Physiol. Lond. 471: 729–748, 1993. LI, Y.-W., GIEROBA, Z. J., MC ALLEN, R. M., AND BLESSING, W. W. Neurons in the rabbit caudal ventrolateral medulla inhibit bulbospinal barosensitive neurons in rostral medulla. Am. J. Physiol. 261 (Regulatory Integrative Comp. Physiol. 30): R44-R51, 1991. LI, Y.-W. AND DAMPNEY, R. A. L. Expression of fos-like protein in brain following sustained hypertension and hypotension in conscious rabbits. Neuroscience 61: 613–634, 1994. MIURA, M., TAK AYAMA, K., AND OK ADA, J. Distribution of glutamateand GABA-immunoreactive neurons projecting to the cardioacceleratory center of the intermediolateral nucleus of the thoracic cord of SHR and WKY rats—a double-labelling study. Brain Res. 638: 139–150, 1994. MORRISON, S. F., MILNER, T. A., AND REIS, D. J. Reticulospinal vasomotor neurons of the rat rostral ventrolateral medulla: relationship to sympathetic nerve activity and the C1 adrenergic cell group. J. Neurosci. 8: 1286–1301, 1988. ROSS, C. A., RUGGIERO, D. A., JOH, T. H., PARK, D. H., AND REIS, D. J. Rostral ventrolateral medulla: selective projections to the thoracic autonomic cell column from the region containing C1 adrenaline neurons. J. Comp. Neurol. 228: 168–185, 1984. SPANSWICK, D., PICKERING, A. E., GIBSON, I. C., AND LOGAN, S. D. Inhibition of sympathetic preganglionic neurons by spinal glycinergic interneurons. Neuroscience 62: 205–216, 1994. TODD, A. J. AND SULLIVAN, A. C. Light microscope study of the coexistence of GABA-like and glycine-like immunoreactivites in the spinal cord of the rat. J. Comp. Neurol. 296: 496–505, 1990. YARDLEY, C. P., STEIN, R. D., AND WEAVER, L. C. Tonic influences from the rostral medulla affect sympathetic nerves differentially. Am. J. Physiol. 256 (Regulatory Integrative Comp. Physiol. 25): R323–R331, 1989. ZAGON, A. AND SMITH, A. D. Monosynaptic projections from the rostral ventrolateral medulla-oblongata to identified sympathetic preganglionic neurons. Neuroscience 54: 729–743, 1993. 08-13-97 17:58:39 neupa LP-Neurophys Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 17, 2017 tion of this region causes a complex pattern of cardiovascular and sympathetic responses. Thus an overall modest decrease in blood pressure and heart rate was accompanied by an increase in splanchnic and renal sympathetic nerve activity and a decrease in lumbar sympathetic nerve activity (Huangfu et al. 1992). The axons of these neurons descend through the RVLM and then into the lateral funiculus to synapse onto SPNs. Thus these neurons also may contribute to the inhibitory control of SPNs. The RVLM has been considered a purely sympathoexcitatory area (see INTRODUCTION ). However, in support of a sympathoinhibitory role for some neurons within the RVLM, Li et al. (1991) observed that in the rabbit, raising arterial pressure excited ú10% of RVLM neurons tested. Moreover, Li and Dampney (1994) located fos-positive neurons within the RVLM of rabbit after intravenous infusions of phenylephrine to increase blood pressure. Thus the suggestion that stimulating the RVLM can inhibit as well as excite SPNs is not without other indirect support. In conclusion, this study has shown that SPNs can be inhibited by activation of a bulbospinal GABAergic pathway that synapses directly onto SPNs. The inhibitory effects are mediated by activation of GABAA receptors. This is the first electrophysiological evidence of such a pathway onto SPNs. 235