Glutamate and GABA-mediated synaptic currents in neurons of the rat dorsal motor nucleus of the vagus R. A. TRAVAGLI, R. A. GILLIS, C. D. ROSSITER, AND S. VICINI FIDIA Georgetown Institute for the Neurosciences and Department of Pharmacology, Georgetown University School of Medicine, Washington, DC 20007

TRAVAGLI, R. A., R. A. GILLIS, C. D. ROSSITER, AND S. VICINI. Glutamate and GABA-mediated synaptic currents in neurons of the rat dorsal motor nucleus of the vagus. Am. J. Physiol. 260 (Gastrointest. Liver Physiol. 23): G531-G536, 1991.-We report the presence of excitatory and inhibitory spontaneous and evoked synaptic currents in the dorsal motor nucleus of the vagus (DMV) in the rat upon vagal and perivagal stimulation. Whole cell current-clamp recordings from anatomically identified DMV neurons in rat brain stem slices show that these neurons are capable of sustained slow-frequency action potential firing probably because of the presence of a pacemaker current. Spontaneously occurring, tetrodotoxin-resistant miniature inhibitory and excitatory synaptic potentials were observed. Stimulation of the vagus mostly induced antidromic action potentials in DMV neurons. However, careful positioning of the stimulating electrode in the tissue surrounding the recording neuron, and sometimes in the vagus itself, was capable of evoking orthodromic-evoked mixed inhibitoryexcitatory postsynaptic potentials, and eventually, action potentials. Whole cell voltage-clamp recordings of the synaptic currents corresponding to these synaptic potentials in the presence of pharmacological antagonists of the neurotransmitters y-aminobutyric acid (GABA), glutamate, and glycine receptor subtypes indicate that the inhibitory synaptic currents are mediated by GABA-activated Cl- channels, while the excitatory synaptic currents are due to activation of ionotropic glutamate receptors of the N-methyl-D-aspartic acid (NMDA) and nonNMDA subtypes. brain stem slices; spontaneous and evoked synaptic currents; whole cell clamp; excitatory amino acid receptors; y-aminobutyric acid receptors

DATA from previous in vivo studies of the dorsal motor

nucleus of the vagus (DMV) indicate that these neurons are the major source of vagal fibers innervating the esophagus and stomach, and an important role of this nucleus in control of esophageal-gastric function is well documented (5). Many of the major neural inputs to the DMV have been elucidated (5, ZZ), but the neurotransmitters released by these inputs have not been determined. Attempts have been made to reveal these neurotransmitters using immunohistochemical techniques (10) and application of putative neurotransmitters to 0193-1857/91

$1.50

Copyright

0 1991

DMV neurons in vivo (10, 13, 21). The in vitro slice preparation of the brain stem offers a better opportunity for elucidating these neurotransmitters. In the slice preparation, synaptic connections between tractus solitarius afferents and DMV neurons have been demonstrated (l), and data obtained clearly indicate that the entire reflex arc from the afferent fibers to the efferent DMV neurons can be preserved (1, 30). Studies of neurotransmitters in the DMV slice preparation have been limited to assessingthe electrophysiological effects of putative neurotransmitters on DMV neurons. The neurotransmitters include acetylcholine (11, 29), thyrotropin-releasing hormone (23), cholecystokinin (19)) somatostatin (17), and substance P (18). Observing changes in the electrophysiological behavior of DMV neurons upon application of a putative neurotransmitter in a slice preparation does not provide the necessary evidence for concluding that a particular substance is the neurotransmitter that permits communication between afferent inputs and DMV neurons. This evidence could be obtained by evaluating the effects of specific antagonists to putative neurotransmitters on the occurrence of spontaneous and evoked synaptic activity recorded intracellularly from DMV motoneurons by means of the whole cell recording technique in thin brain slices. The purpose of the present study was to obtain this type of evidence using the recently devised brain slice whole cell recording technique devised by Edwards et al. (2). METHODS

Brain stem slices preparation. Young Sprague Dawley rats 3-4 wk of age were decapitated after ether anesthesia. The brain stem was quickly removed, and the cerebellum was detached and placed in cold oxygenated buffer. Four 200~pm coronal slices from the medulla rostra1 to the obex and four 200-pm coronal slices caudal to the obex were prepared with an Oxford vibratome and incubated at 37°C in oxygenated buffer. Drugs, bath, and patch pipette solutions. Physiological rat Ringer was composed of 120 mM NaCl, 3.1 mM KCl, 26 mM NaHC03, 1.25 mM K2HP04, 5 mM dextrose, 2 mM MgCl, 2 mM CaC12, and 500 nM glycine. The the American

Physiological

Society

G531

Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 18, 2019.

G532

SYNAPTIC

CURRENTS

bathing medium was maintained at pH 7.4 with 95% 025% CO, bubbling. Whole cell recordings of visually identified DMV neurons were performed with patch pipettes filled with 145 mM potassium gluconate, 1 mM MgC12, 0.5 mM ethylene glycol-bis(P-aminoethyl ether)N,N,N’,N’-tetraacetic acid (EGTA), 2 mM ATP, 0.6 mM GTP, and 10 mM N-Z-hydroxyethylpiperazine-N’2-ethanesulfonic acid (HEPES)-KOH (pH 7.2). Cesium gluconate (145 mM) was used as a substitute for potassium gluconate for synaptic current recordings because cesium ions block potassium channels and improve the voltage clamp at depolarized potentials. Drugs were added by perfusion in the bathing medium (4 ml/min). Tetrodotoxin (TTX; Sigma, St.Louis, MO), 3-[ (R,S)-2carboxypiperazin-4-yl] -propyl-1-phosphonic acid (CPP; Tocris, Buckhurst Hill, UK), Z-(carboxy-3’-propyl)-3amino-6-paramethoxy-phenylpyridazinium bromide (SR 95531; a gift of Dr. J. M. Mienville, Sanofi Research, Montpellier, France), 2,3-dihydro-6-nitro-7-sulfamoylbenzo(f)quinoxaline (NBQX; a gift of Dr. T. Honor& Ferrosan Research, Soeborg, Denmark), and strychnine (Sigma) were dissolved in water and diluted in the bathing solution. 3a,21-Dihydroxy-5a-pregnan-ZO-one (allotetrahydrodeoxycorticosterone) (THDOC; a gift of Dr. R. Purdy, San Antonio, TX) was dissolved in dimethyl sulfoxide (0.1% maximal final dilution). For recordings of pure excitatory postsynaptic currents (EPSCs), picrotoxin (50 PM, Sigma) was routinely added to the bathing solution to prevent GABAergic currents. Electrophysiological recordings and stimulation. By means of the gigaseal patch-clamp method of Hamill et al. (7), we measured potentials and currents from DMV neurons. Current and voltage signals at the headstage of the patch-clamp amplifier (EPC-7, List, Darmstadt, FRG) were filtered at 1,500 Hz @-pole low-pass Bessel; Frequency Devices, Haverhill, MA) and continuously displayed on an oscilloscope. Stimulation was performed by using a concentric stimulating electrode (D. Kopf, Tujunga, CA). Synaptic currents were evoked by squarewave electric pulses (100-300 PA for 50 pus at 0.5 Hz). Data collection and analysis. Data were recorded with a VCR magnetic tape (VR IO-A, Instrutech, Elmont, NY) for off-line analysis (LSI 11/73 computer, Indec System, Sunny Vale, CA). The amplitude and the decay time constants of synaptically evoked currents were determined by exponential fit with the 11/73 system and an entirely automated least-squares procedure (a gift from the late Dr. Stephen M. Schuetze). This method makes use of a Simplex algorithm to fit the data to either a single or double exponential equation, yielding the peak amplitudes of the EPSC, the contribution to the peak amplitude of the slow component, and the decay time constants (28). Results are expressed as means t SE. RESULTS

Properties of Dh4V neurons in brain stem slices. Current-clamp whole cell recordings were performed from 34 vagal motoneurons. Eighteen cells were in slices taken caudal to the obex and 16 cells were in slices taken rostra1 to the obex. The cells were identified by their anatomical location at the edge of the hypoglossal nu-

IN

RAT

DMV

NEURONS

A

RP

C

-54

L--LL v-=-L 40

/

Control

200

200

mV pA

II

ms

D

B

TTX

I

40

pM

200 200

RP

-57

mV pA

ms

FIG. 1. A: current-clamp recording of repetitive action potentials firing from a DMV neuron in a brain stem slice. Each action potential is followed by a large afterhyperpolarization and a slow depolarizing current triggering the next spike. B: TTX (1 ,uM) perfusion blocks fast sodium spike but leaves unaffected the spontaneous membrane potential oscillation and smaller, probably calcium-mediated, spikes. C: evoked action potential firing by current injection in a silent DMV neuron from a resting potential of -54 mV. D: membrane responses to hyperpolarizing current pulses in a vagal motoneuron.

cleus and between the fourth ventricle and the vagus nerve below the nucleus tractus solitarius (nTS) area. At high magnification (x500), large compact cell bodies ZO30 pm in diameter were clearly distinguishable from the adjacent smaller nTS neurons and from the larger hypoglossal neurons located in the dark, heavily myelinated hypoglossal nucleus. Most DMV neurons exibited spontaneous “pacemaker-like” action potentials firing with a frequency ranging from 0.5 to 5 Hz (Fig. IA). Because of the spontaneous discharge and afterhyperpolarization (Fig. 1A ), it was difficult to accurately determine the resting membrane potential (RMP) of these neurons. However, in 10 of the cells studied, no spontaneous activity was present and the RMP was -53 t 5 mV. In these cells (10 of lo), small steady depolarization by current injection immediately triggered a typical spontaneous repetitive firing consisting of a fast spike of 7080 mV and slow (>20 ms) and huge afterhyperpolarization, followed by a slow repolarization until the next action potential. This repetitive firing was specific for neurons located in the DMV and differed from the electrical activity of adjacent hypoglossal or nTS neurons (unpublished observations). (Note: repetitive firing occurred in nTS neurons but with a pattern of discharge distinctly different from the pattern of discharge of DMV neurons). Bath perfusion with TTX (1 PM) abolished the fast component of the action potential (n = 3 neurons) but left unaffected the spontaneous membrane potential fluctuation (Fig. 1B). Depolarizing current pulses triggered action potentials similar in shape and duration (Fig. 1C) to the action potentials arising from spontaneous membrane potential fluctuation (Fig. 1A). Membrane input resistance was estimated from membrane voltage responses to hyperpolarizing current pulses (Fig. lo), and data derived from 18 cells indicated that input resistance averaged

Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 18, 2019.

SYNAPTIC

CURRENTS

555 t 23 MQ, a value significantly higher than previously reported for DMV neurons (see Ref. 30). This difference is probably due to the young age of the animals used and the minor damage occurring with the whole cell recording technique compared with the more extensive damage occurring with conventional intracellular recordings. No significant differences were observed between caudal and rostra1 slices containing DMV neurons in terms of the above-described basic electrophysiological properties. Evoked and spontaneous synaptic activity in DMV neurons. Antidromic activation of DMV neurons was achieved by stimulating the vagus nerve traversing laterally through the dorsal half of the brain stem slice, and an example of an all or none response appears in Fig. 2A. Orthodromically evoked excitatory postsynaptic potentials (EPSPs), with some EPSPs reaching threshold to trigger action potentials, were elicited by perivagal and, in few cases, vagal stimulation (Fig. 2B). Depolarizing the DMV neurons by steady injection of depolarizing current resulted in the occurence of a mixed excitatory-inhibitory postsynaptic potential (EPSP-IPSP) sequence upon perivagal stimulation (Fig. 2C). Whole cell voltage-clamp of the DMV neurons with an intracellular solution low in Cl- at -40 mV holding voltage (V,) revealed the occurrence of miniature spontaneous

IN

RAT

DMV

G533

NEURONS

Vh

A x

B

Ii

Control

0

SR

20

95531

Control

M THDOC

A IR~ -59

,uM

5 ,uM

50

pA

Ii 3. A: inhibitory synaptic current averages in a DMV neuron upon perivagal stimulation at a holding potential of 0 mV recorded using cesium gluconate as a major current carrier. B: bath perfusion with the GABAA antagonist SR 95531 (20 PM) completely abolished the synaptic current while strychnine (50 PM) failed to affect the synaptic current. C: perfusion with the neurosteroid THDOC (5 PM) potentiated these outward chloride currents. Each current in the figure is representative of the average of 5-1 .O currents . All drug effects were re versible after 10 min of continous perfusion with control solution (not shown). FIG.

25

mV

Vh

C

RP

-40

- 35

50 r 2. A: antidromically

evoked action potential in DMV neuron upon vagal stimulation. B: orthodromically evoked EPSP and action potential in a DMV neuron upon perivagal stimulation. C: orthodromic mixed EPSP-IPSP in a DMV neuron during steady current injection to hold the cell resting membrane potential at a depolarized level of -35 mV. Action potentials were often triggered after the IPSPs. D: at a holding potential of -40 mV spontaneous miniature EPSCs and spontaneus miniature IPSCs were observed as inward (downward) and outward (upward) currents, respectively, and were not abolished by the presence of 1 PM TTX. FIG.

inward EPSCs and miniature spontaneous outward inhibitory postsynaptic currents (IPSCs) (Fig. ZD). Bath perfusion with TTX (1 PM) did not abolish these spontaneous events (Fig. 2D), indicating that they were not due to spontaneous activation of single presynaptic inputs. In voltage-clamped DMV neurons, outward IPSCs were observed at depolarized potentials upon perivagal stimulation (Fig. 3), and at more negative holding potentials, inward EPSCs were noted (Fig. 4). In most cases, both excitatory and inhibitory responses were observed at a specific stimulation site in the area between the vagus and the recorded neuron. However, in some cases, readjustment of the position of the stimulation electrode resulted in large IPSCs without the occurrence of EPSCs, indicating that direct stimulation of a separate inhibitory afferent input to the DMV cells was being activated. Studies of whole cell voltage-clamp recordings of IPSCs. Spontaneously occurring (Fig. 20) and stimulationevoked (Fig. 3) outward IPSCs were recorded from vagal motoneurons. The average amplitude of evoked IPSCs at 0 mV Vh was 186 t 23 pA (n = 8, range 70-300 PA). The rise time of the IPSCs was between 1 and 3 ms and the IPSCs decayed with a fast component of 14 t 2 ms (n = 8, range 9-27 ms), followed by a slower component

Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 18, 2019.

G534

SYNAPTIC -

Vh

CURRENTS

RAT

DMV

1. Pharmacological characterization of EPSCs and IPSCs of DMV neurons in rat brain stem slices

Jr++---

Experimental Conditions

5 pM

NBQX

20

PA

mV) (20 pM) mV) (50 PM) mV) (10 PM) Excitatory

#LAM

Control

Vh

Amplitude,

Inhibitory Control (0 SR 95531 Control (0 Strychnine Control (0 THDOC

-

CPP

C

NEURONS

TABLE

70

Control

Vh

IN

(-70 mV) NBQX (5 PM) Control (-70 mV) CPP (20 PM) Control (+50 mV) NBQX (5 PM) Control (+50 mV) CPP (20 /.LM) +NBQX (5 ,uM)

+au

CPP

20

postsynaptic 137rt14 19t7 187k37 149-t-8 229t21 248t26 postsynaptic 106k7 16-1-2 71tlO 52t9 73k15 39t10 132t36 46t19 5&l

Time Constant ms

Slow Component, %

n

currents 14t4 11t2 14&l lOt3 16k6 40t8

29t11 18t10 21kll 30217 28k12 46k12

currents 6tl 47k20 5t1 5tl 26t4 51zkll 2624 lOt4 lOt2

12t5 100 16k7 0 25t14 72t18 7323 7k4 6t3

Values are means k SE of n neurons from different slices. In each DMV neuron, 5-20 synaptic currents were averaged before and after bath perfusion with drug.

,uM

\ To record only inward EPSCs without IPSCs from DMV neurons, experiments were performed in the presence of 50 ms picrotoxin (50 PM). The amplitude of evoked EPSCs recorded from DMV motoneurons averaged 133 t 39 pA FIG. 4. Average of 5-10 EPSCs in a DMV neuron upon perivagal (n = 17, range 40-200 PA) at Vh -70 mV. Inward EPSCs stimulation in the presence of picrotoxin (50 PM) and various antagonists of excitatory amino acid receptor subtypes. A: non-NMDA recephad a rise time