D3n1qfjutz9qb9.cloudfront.net

Am J Physiol Gastrointest Liver Physiol280: G1106–G1113, 2001.
Mechanism of action of baclofen in rat dorsalmotor nucleus of the vagus K. N. BROWNING AND R. A. TRAVAGLIDivision of Gastroenterology and Department of Physiology, University ofMichigan Medical Center, Ann Arbor, Michigan 48109 Received 28 September 2000; accepted in final form 14 January 2001 Browning, K. N., and R. A. Travagli. Mechanism of
Functional in vivo studies have shown that subcuta- action of baclofen in rat dorsal motor nucleus of the vagus.
neous or intracerebroventricular application of the Am J Physiol Gastrointest Liver Physiol 280: G1106–G1113, GABAB receptor agonist baclofen increases gastric and 2001.— Using whole cell patch-clamp recordings, we inves- intestinal motility (3, 10) as well as initiating vagal tigated the effects of the GABAB receptor agonist baclofen in discharges similar to those obtained on stimulation of thin slices of rat brain stem containing identified gastric- or gastric secretion (13). These apparently contradictory intestinal-projecting dorsal motor nucleus of the vagus results can be explained if activation of GABA (DMV) neurons. Perfusion with baclofen (0.1–100 ␮M) in- duced a concentration-dependent outward current (EC tors involves separate subpopulations of DVC neurons.
␮M) in 54% of DMV neurons with no apparent differences One population would comprise excitatory DMV neu- between gastric- and intestinal-projecting neurons. The out- rons that do not have GABAB receptors on their mem- ward current was attenuated by pretreatment with the se- brane but rather receive an inhibitory input, most likely from NTS, which is inhibited by activation of B antagonists saclofen and 2-hydroxysaclofen, but not by the synaptic blocker TTX, indicating a direct effect GABAB receptors. The disinhibition of these DMV neu- at GABAB receptors on DMV neurons. Using the selective ion rons would result in an excitatory effect such as in- channel blockers barium, nifedipine, and apamin, we showed crease in gastric motility. Indeed, electrophysiological that the outward current was due to effects on potassium and studies have shown that baclofen acts directly on NTS calcium currents as well as calcium-dependent potassium neurons to produce a membrane hyperpolarization (7), currents. The calcium-mediated components of the outward as well as indirectly to inhibit synaptic transmission current were more prominent in intestinal-projecting neu- from vagal afferents (7, 23). Another DMV neuronal rons than in gastric-projecting neurons. These data indicate population, most likely participating in inhibitory con- that although baclofen inhibits both intestinal- and gastric- trol of gastric functions, would have GABA projecting neurons in the rat DMV, its mechanism of action differs among the neuronal subpopulations.
on the membrane but would not receive inhibitoryinputs containing GABAB receptors on the presynaptic brain stem; electrophysiology; gastrointestinal membrane. The inhibition of these DMV neurons re-sulting from activation of GABAB receptors would re-duce their inhibitory influence, causing an increase in THE DORSAL MOTOR NUCLEUS OF the vagus (DMV) contains gastric pressure. Indeed, Andrews et al. (3) hypothe- the neurons that provide the parasympathetic efferent sized that the observed atropine-insensitive increase in outflow to the subdiaphragmatic viscera (12, 22). Af- gastric pressure obtained by baclofen was probably ferent neurons innervating the gastrointestinal tract mediated via an action at a central site to reduce the project to, and terminate within, the DMV either di- tonic vagal drive to nonadrenergic noncholinergic rectly or via interneurons in the nucleus of the solitary (NANC) inhibitory neurons.
tract (NTS) (2, 25, 26, 30). DMV neurons are under The cellular mechanisms coupled to postsynaptic inhibitory control from GABAergic neurons of the NTS GABAB receptor activation are well documented (5, 11, (31, 32). The inhibitory actions of GABA on DMV 14, 17, 27, 34, 35). The mechanisms comprise activa- neurons are thought to be exerted primarily by fast tion of potassium- as well as inhibition of voltage- ionic currents mediated via GABAA receptor activation dependent calcium conductances and adenylate cy- (32). Although autoradiographic studies (1, 6, 20) have demonstrated that GABAB receptors also are located We (8, 9) have shown recently that the DMV is throughout the dorsal vagal complex (DVC), i.e., the composed of heterogeneous neuronal populations in DMV plus the NTS, to date there is no evidence to terms of both membrane as well as pharmacological suggest that functional GABAB receptors exist on DMV properties. These populations can be further distin- neuronal membranes.
guished based on their peripheral targets.
Address for reprint requests and other correspondence: R. A.
The costs of publication of this article were defrayed in part by the Travagli, Dept. of Internal Medicine, Division of Gastroenterology, payment of page charges. The article must therefore be hereby 3912 Taubman Center, 1500 East Medical Center Drive, Ann Arbor, marked ‘‘advertisement'' in accordance with 18 U.S.C. Section 1734 MI 48109-0362 (E-mail:[email protected]).
solely to indicate this fact.
0193-1857/01 $5.00 Copyright 2001 the American Physiological Society BACLOFEN IN RAT DMV The aims of this study were to investigate 1) whether Drug application. Drugs were applied to the bath via a series of manually operated valves. Baclofen (10 ␮M) was B receptors are functionally present on the DMV membrane, and if so, 2) what the mechanism of action applied to all neurons to ascertain whether or not it had any effect per se before the appropriate experiment was contin- B agonist baclofen is on the DMV mem- brane, and 3) whether there are differences in the ued. When voltage clamped at ⫺50 mV, a minimum peakcurrent of 20 pA had to be induced by baclofen (10 ␮M) to responses of DMV neurons identified as per their pe- classify the neuron as responsive. Drug effects were assessed with each neuron acting as its own control, i.e., the resultsobtained after administration of a receptor antagonist or channel blocker were compared with those obtained beforeadministration using the paired t-test. Drugs were applied in Retrograde tracing. The retrograde tracer DiI was applied concentrations shown previously to be effective. Results are to discrete gastrointestinal regions in 12-day-old Sprague- expressed as means ⫾ SE with significance defined as P ⬍ Dawley rat pups of either sex, as described previously (8).
0.05. To minimize sampling biases, only one experiment was Briefly, rats were anesthetized deeply with a 6% solution of conducted per each brain stem slice and only one experiment halothane in accordance with the guidelines of the Animal of each type was conducted per animal.
Care and Use Committee of the University of Michigan Chemicals and solutions. Krebs solution was composed of Medical Center (Ann Arbor, MI). The depth of anesthesia (in mM): 126 NaCl, 25 NaHCO3, 2.5 KCl, 1.2 MgCl2, 2.4 (abolition of the foot pinch withdrawal reflex) was monitored CaCl2, 1.2 NaH2PO4, and 11 dextrose, maintained at pH 7.4 before and during surgery. The abdominal area was shaved by bubbling with 95%O2-5% CO2. Intracellular solution con- and cleaned with alcohol before a laparotomy was performed.
sisted of (in mM): 128 potassium gluconate, 10 KCl, 0.3 Crystals of DiI were applied to either the major curvature of the gastric fundus or corpus, the antrum/pylorus, the duode- 2, 1 MgCl2, 1 HEPES, 1 EGTA, 2 ATP, and 0.25 GTP, adjusted to pH 7.35 with KOH. 1,1⬘-Dioctadecyl-3,3,3⬘,3⬘- num (the antimesenteric surface at the level of the hepatic tetramethylindocarbocyanine perchlorate [DiIC and pancreaticoduodenal arteries), or the minor curvature of was purchased from Molecular Probes (Eugene, OR). R(⫹) the cecum (at the level of the ileocecal junction). The appli- Baclofen and 2-hydroxysaclofen were purchased from RBI cation site was embedded in a fast-hardening epoxy resin (Natick, MA). Halothane and all other drugs and chemicals before the entire surgical area was washed with warm, ster- were purchased from Sigma Chemical (St. Louis, MO).
ile saline. The wound was closed with 4-0 sutures, and theanimal was allowed to recover for 10–15 days.
Electrophysiology. The method used for tissue slice prepa- ration was as described previously (32). Briefly, rats wereanesthetized deeply (halothane bubbled with air) before be- Baclofen induces outward current in subpopulation ing killed through severing of the major blood vessels in the of DMV neurons. The effects of the GABAB receptor chest. The brain stem was removed and placed in chilled agonist baclofen were assessed in 394 gastrointestinal- (4°C) oxygenated Krebs solution (see below for composition).
projecting neurons of the DMV (211 gastric projecting Using a vibratome, we cut six to eight coronal slices contain- and 187 intestinal projecting). Baclofen induced a con- ing the DMV. Slices were stored in oxygenated Krebs solu- centration-dependent (0.1–100 ␮M) outward current tion at 32°C for at least 1 h before use. A slice was then placed in 211 (i.e., 54%) of these neurons. No differences were on a custom-made perfusion chamber (vol 500 ␮l) and main- observed in the magnitude of the baclofen-induced cur- tained at 35°C by continual perfusion with warmed oxygen- rent among the gastric and intestinal groups; data ated Krebs solution at a rate of 2.5 ml/min.
Retrogradely labeled neurons were identified before elec- were thus pooled and provided an estimated EC50 of 3 trophysiological recording using a Nikon E600FS microscope ␮M (Fig. 1). The percentage of responsive neurons did fitted with DIC (Nomarski) optics and tetramethylrhodamine not differ between the gastric- and intestinal-project- isothiocyanate epifluorescent filters. The brief periods of il- ing neurons; in fact, 112 of 211 gastric-projecting neu- lumination required to detect the fluorescent neurons have rons (i.e., 53%) and 99 of 187 intestinal-projecting not been observed to cause any damage (16, 21). Once a neurons (i.e., 54%) responded to perfusion with 10 ␮M labeled neuron was identified, electrophysiological record- baclofen. Similarly, the distribution of the magnitude ings were made under bright-field illumination.
of the response to 10 ␮M baclofen did not differ be- Whole cell recordings were made only from retrogradely tween gastric- and intestinal-projecting neurons (Fig.
labeled gastrointestinal-projecting neurons using patch pi- pettes filled with potassium gluconate solution of resistance3–8 M⍀ (see below for composition) and a single electrode In six neurons (3 gastric and 3 intestinal), the effects voltage-clamp amplifier (Axopatch 1D, Axon Instruments, of baclofen (10 ␮M) were assessed before and after Foster City, CA). Data were filtered at 2 kHz, digitized via a exposure to GABAB receptor antagonists. Baclofen in- Digidata 1200C interface (Axon Instruments), and acquired duced a 51 ⫾ 4 pA outward current in control condi- and stored on an IBM personal computer using pClamp8 tions. After perfusion with the selective antagonists software (Axon Instruments). Only neurons having a series saclofen (300 ␮M, n ⫽ 2) or 2-hydroxysaclofen (100 ␮M, resistance (i.e., pipette ⫹ access resistance) ⬍15 M⍀ were n ⫽ 4), the baclofen-induced current was reduced to used. A neuron was accepted for recording if it met several 10 ⫾ 3 pA, i.e., 17% of control (P ⬍ 0.05). However, criteria, including a membrane stable at the holding poten- incubation with the synaptic blocker TTX (1 ␮M) had tial that returned to baseline after the action potential af-terhyperpolarization plus an action potential of at least 60 no effect on the baclofen-induced outward current. In mV amplitude. The membrane potential was manually cor- fact, baclofen induced a current of 45 ⫾ 8 and 46 ⫾ 8 pA rected for junction potential before starting each experiment.
in control conditions and in the presence of TTX, re- Data analysis was performed using pClamp8 software.
spectively (n ⫽ 6, 3 gastric and 3 intestinal; P ⬎ 0.05).
BACLOFEN IN RAT DMV Fig. 1. Baclofen induces a concentration-dependent outward current. A: representative traces from an intestinal-projecting neuron showing that baclofen produced a concentration-dependent outward current. The cell wasvoltage clamped at ⫺50 mV. The complete washout of the baclofen effect (i.e., recovery to the initial holdingcurrent) was obtained after 7–10 min (not shown). B: concentration-response curves for the baclofen-inducedoutward current expressed as %maximal response (Imax). The EC50 for the baclofen-response was 3 ␮M. Data forgastric- and intestinal-projecting neurons have been pooled (n ⫽ 7–34). Each neuron contributed at least 3 datapoints. C: relative % of the neurons tested with baclofen perfusion showed unimodal distribution of the currentresponse. No differences were observed between gastric- and intestinal-projecting neurons at any response value.
Baclofen effect involves several ionic conductances. was reduced from 76 ⫾ 8.9 pA under control conditions The effects of the nonselective potassium channel to 40 ⫾ 4.7 pA in the presence of nifedipine, i.e., 53 ⫾ blocker barium (2 mM) on the outward current induced 2.4% of control (P ⬍ 0.05; data not shown, however, see by baclofen (10 ␮M) were assessed in 14 neurons (7 Fig. 3). Significant differences were uncovered, how- gastric and 7 intestinal). The baclofen-induced current ever, in the magnitude of this reduction between gas- was reduced from 66 ⫾ 14 pA in control conditions to tric- and intestinal-projecting neurons. In gastric-pro- 25 ⫾ 5 pA in the presence of barium, i.e., 41 ⫾ 3.6% of jecting neurons, nifedipine reduced the baclofen- control (P ⬍ 0.05; data not shown, however, see Fig. 3).
induced current from 85 ⫾ 14 to 47 ⫾ 7 pA, i.e., 58 ⫾ No significant differences were observed in the actions 3% of control, whereas in intestinal-projecting neu- of barium between gastric- and intestinal-projecting rons, nifedipine reduced the baclofen-induced current neurons. In detail, barium reduced the baclofen-in- from 66 ⫾ 9 to 32 ⫾ 6 pA, i.e., 47 ⫾ 3% of control (P ⬍ duced current from 73 ⫾ 24 to 25 ⫾ 7 pA in gastric- projecting neurons and from 60 ⫾ 17 to 24 ⫾ 8 pA in The involvement of both potassium and calcium con- intestinal-projecting neurons, i.e., 41 ⫾ 5% and 41 ⫾ ductances in mediating the actions of baclofen was 6% of control in gastric- and intestinal-projecting neu- confirmed by assessing the reversal potential for the rons, respectively.
baclofen-induced outward current in the absence and The effects of the L-type calcium channel blocker presence of both channel blockers. The baclofen-in- nifedipine (3 ␮M) on the baclofen (10 ␮M)-induced duced current reversed at approximately ⫺95 mV (Fig.
outward current were assessed in 34 neurons (19 gas- 2). In the presence of barium, however, the reversal of tric and 15 intestinal). The baclofen-induced current the baclofen-induced current shifted to the more posi- BACLOFEN IN RAT DMV between gastric- and intestinal-projecting neuronswere observed in the total inhibition of the baclofencurrent by the combined antagonists.
Given that perfusion with the nonselective potas- sium channel blocker barium in combination with theselective calcium-L type channel antagonist nifedipinedid not block the baclofen-induced outward currentcompletely, in five neurons (3 gastric and 2 intestinal),we perfused the slice first with nifedipine alone andthen in combination with the nonselective calciumchannel blocker cadmium (200 ␮M). Perfusion withnifedipine reduced the baclofen-induced current from76 ⫾ 4 to 43 ⫾ 2 pA, i.e., a 44 ⫾ 3.9% reduction of thebaclofen-induced current; addition of cadmium to thenifedipine perfusate caused the baclofen-induced cur-rent to be reduced to 34 ⫾ 4 pA, i.e, an additional 11 ⫾ 4.5% (nifedipine vs. nifedipine ⫹ cadmium, P ⬍ 0.05).
These data then suggested that calcium currents otherthan the L-type may be involved in the baclofen-in-duced outward current.
Because there appeared to be some overlap between the inhibitory actions of the potassium channel blockerbarium and the calcium channel blocker nifedipine (i.e., 59 ⫾ 3.5% reduction by barium, 47 ⫾ 2.4% reduc-tion by nifedipine but an 81% to 88% reduction whencombined), we investigated the involvement of a calci-um-dependent potassium conductance using the small-conductance potassium (SK) channel blocker apamin.
The outward current induced by baclofen (10 ␮M) wasassessed before and after apamin (100 nM) in 19 neu- Fig. 2. Current-voltage relationship for the baclofen-induced out-ward current. Dorsal motor nucleus of the vagus (DMV) neurons rons (12 gastric, 7 intestinal). In the presence of were voltage clamped at ⫺50 mV and stepped to ⫺120 mV for 800 ms apamin, the baclofen-induced current was reduced every 5 s in ⫺10-mV increments. Each data point represents an from 107 ⫾ 15 to 63 ⫾ 9 pA, i.e., 62 ⫾ 3.5% of control average of 6–24 neurons. In control medium, the baclofen equilib- (P ⬍ 0.05; Fig. 4A). Differences were again apparent rium potential (Eb) was ⫺95 mV (A). After perfusion with the between gastric- and intestinal-projecting neurons. In potassium channel blocker barium (2 mM), the Eb shifted to lessnegative potentials (approximately ⫺55 mV) (B). Conversely, pre- detail, in gastric-projecting neurons, apamin reduced treatment with the L-type calcium channel antagonist nifedipine (3 the baclofen-induced current from 95 ⫾ 18 to 62 ⫾ 12 ␮M) shifted Eb closer to potassium equilibrium potential.
pA, i.e., 68 ⫾ 4% of control, whereas in intestinal-projecting neurons, apamin reduced the baclofen-in- tive value of ⫺55 mV (n ⫽ 6), i.e., further from potas-sium equilibrium potential (EK). In the presence ofnifedipine, however, the reversal of the baclofen-in-duced current shifted to the more negative value of ⫺110 mV (n ⫽ 6), i.e., closer to EK.
In eight cells (5 gastric and 3 intestinal), perfusion with barium reduced the baclofen-induced currentfrom 68 ⫾ 21 to 26 ⫾ 6 pA, i.e., a 55 ⫾ 3.0% reductionof the baclofen-induced current; addition of nifedipineto the barium perfusate caused the baclofen-inducedcurrent to be reduced from 26 ⫾ 6 to 9 ⫾ 3 pA, i.e., 12 ⫾2% of the initial control value (barium vs. barium ⫹nifedipine, P ⬍ 0.05). Likewise, in nine cells (5 gastricand 4 intestinal), perfusion with nifedipine reduced the Fig. 3. Baclofen-induced outward current is antagonized by a com-bination of nifedipine and barium. Representative trace showing baclofen-induced current from 110 ⫾ 28 to 55 ⫾ 14 pA, that in the voltage-clamp configuration [holding potential (HP) ⫽ i.e., a 51 ⫾ 3.7% reduction of the baclofen-induced ⫺50 mV], perfusion with baclofen (10 ␮M) induced an outward current; addition of barium to the nifedipine perfusate current in a DMV neuron. After pretreatment with a solution con- caused the baclofen-induced current to be reduced from taining the selective calcium channel L-type antagonist nifedipine (3 55 ⫾ 14 to 22 ⫾ 6 pA, i.e., 19 ⫾ 3.5% of the initial M) decreased the baclofen-induced current to ⬃50% of control, addition of the nonselective potassium channel blocker barium (Ba; control value (nifedipine vs. nifedipine ⫹ barium, P ⬍ 2 mM) to the nifedipine-containing solution further reduced the 0.05; Fig. 3). No statistically significant differences baclofen-induced outward current.
BACLOFEN IN RAT DMV in the presence of nifedipine plus apamin (n⫽ 5; P ⬎0.05; Fig. 4B). These data suggested that, unlike gas-tric-projecting neurons, the apamin-sensitive currentin intestinal-projecting neurons is activated by a com-bination of calcium currents rather than by the L-typecalcium current only. Data are summarized in Fig. 5.
To confirm the involvement of a calcium-dependent potassium conductance, the actions of baclofen to alterthe action potential afterhyperpolarization (previouslydemonstrated to involve the SK-channel; Refs. 8, 24,and 29) were assessed in 20 neurons (9 gastric, 11intestinal). Baclofen caused a 16 ⫾ 1.6% reduction inthe action potential afterhyperpolarization (25 ⫾ 1.0vs. 20 ⫾ 0.9 mV with baclofen; P ⬍ 0.05), a 42 ⫾ 4%reduction in the afterhyperpolarization rate of decay Fig. 4. Baclofen-induced outward current is antagonized differen- tially by a combination of apamin and nifedipine or nifedipine and ⫾ 10 vs. 62 ⫾ 4.7 ms with baclofen; P ⬍ 0.05), and apamin. Representative trace showing that in the voltage-clamp an 8 ⫾ 1.6% increase in the action potential duration configuration (HP ⫽ ⫺50 mV), perfusion with baclofen (10 ␮M) (2.61 ⫾ 0.14 vs. 2.82 ⫾ 0.15 ms with baclofen; P ⬍ 0.05) induced an outward current in a DMV neuron. A: after pretreatmentwith a solution containing the selective calcium-dependent potas-sium antagonist apamin (100 nM) decreased the baclofen-inducedcurrent to ⬃60% of control, addition of the selective L-type calciumchannel blocker nifedipine (3 ␮M) to the apamin-containing solutionfurther reduced the baclofen-induced outward current. B: in con- trast, in a gastric-projecting neuron, addition of apamin to thenifedipine-containing solution did not further reduce the baclofen-induced outward current.
duced current from 126 ⫾ 26 to 63 ⫾ 15 pA, i.e., 53 ⫾5% of control (P ⬍ 0.05).
Because extracellular barium is not an effective blocker of calcium-dependent potassium currents (15),we tested its effects on the baclofen-induced current in combination with apamin. In 11 cells (6 gastric and 5intestinal), perfusion with apamin decreased the ba-clofen-induced current from 88 ⫾ 13 to 48 ⫾ 5 pA, i.e,60 ⫾ 5.5% of control; the addition of barium to theapamin perfusate caused the baclofen-induced currentto be reduced to 24 ⫾ 3 pA, i.e., 30 ⫾ 2.5% of the initialcontrol value (apamin vs. apamin ⫹ barium, P ⬍ 0.05).
Likewise, in 10 cells (5 gastric and 5 intestinal), per-fusion with apamin decreased the baclofen-inducedcurrent from 136 ⫾ 21 to 79 ⫾ 15 pA, i.e., 58 ⫾ 2.8% ofcontrol; addition of nifedipine to the apamin perfusatecaused the baclofen-induced current to be reduced to57 ⫾ 12 pA, i.e., 40 ⫾ 2.4% of the initial control value(apamin vs. apamin ⫹ nifedipine, P ⬍ 0.05; Fig. 4A).
No statistically significant differences between gastric-and intestinal-projecting neurons were observed on thetotal inhibition of the baclofen current by the combinedantagonists. Interestingly, though, when apamin wasperfused in combination with, and after exposure to, Fig. 5. Summary of the effects of baclofen in the presence of ion nifedipine, a further 10 ⫾ 3.1% inhibition of the ba- channel antagonists. A: after incubation with barium (2 mM), nifed- clofen-induced current was observed in intestinal-pro- ipine (3 ␮M), or apamin (100 nM), the amplitude of the baclofen (10 jecting neurons. In detail, in intestinal-projecting neu- ␮M)-induced outward current was reduced in gastric- and intestinal-projecting DMV neurons. Note that nifedipine and apamin both rons, the baclofen-induced current was reduced from produced a larger inhibition of the baclofen-induced current in in- 50 ⫾ 5 pA under control conditions to 30 ⫾ 6 pA in the testinal compared with gastric neurons. B: in gastric-projecting neu- presence of nifedipine and 25 ⫾ 5 pA in the presence of rons, nifedipine (3 ␮M) reduced the current induced by 10 ␮M nifedipine plus apamin (n ⫽ 4; P ⬍ 0.05). In gastric- baclofen; addition of apamin (100 nM) had no additional effect.
projecting neurons, the baclofen-induced current was Conversely, in intestinal-projecting neurons, addition of apamin tothe nifedipine superfusate resulted in an additional decrease in the reduced from 61 ⫾ 8 pA under control conditions to baclofen-induced current. * P ⬍ 0.05, ** P ⬍ 0.05 compared with 34 ⫾ 2 pA in the presence of nifedipine and 34 ⫾ 3 pA relevant control.
BACLOFEN IN RAT DMV (data not shown). No differences were observed in the with the nonselective calcium channel blocker cad- effects of baclofen on action potentials from gastric vs.
mium. These data indicate that calcium channels other intestinal neurons.
than the L-type are also affected, although such chan-nels play a minor role in the overall baclofen response.
The present study seems to indicate differences be- tween gastric- and intestinal-projecting neurons with The present study provides the first direct evidence regard to the source of calcium necessary to activate that functional GABAB receptors are located on the the apamin-sensitive calcium-dependent potassium membrane of DMV neurons. In fact, the GABAB ago- current. In fact, the baclofen current obtained in the nist baclofen evoked a concentration-dependent out- presence of nifedipine compared with nifedipine in ward (inhibitory) current in a subpopulation of gastro- combination with a supramaximal concentration of intestinal-projecting DMV neurons via direct activation apamin (24) did not differ in gastric-projecting neu- of postsynaptic GABAB receptors. The baclofen-in- rons. When the same cocktail of antagonists was tested duced outward current was mediated by several ionic in intestinal-projecting neurons, however, a further conductances, namely a potassium conductance, a cal- 10% inhibition in the baclofen-induced current was cium conductance, and a calcium-dependent potassium observed after pretreatment with nifedipine and apamin. These data would suggest that the apamin- Although the proportion of neurons responding to sensitive current in gastric-projecting neurons is fully baclofen and the magnitude of response were similar in activated by calcium entry via L-type channels, gastric- and intestinal-projecting neurons, the action of whereas in intestinal neurons sources other than the baclofen, however, appeared to differ. In fact, in intes- L-type calcium channels also play a role. Indeed, a tinal-projecting neurons, the calcium-mediated compo- different complement of voltage-dependent calcium nents of the outward current appeared to exert a more currents has been shown to be present in rat vagal prominent role than in gastric-projecting neurons.
motoneurons (24, 28), although their selective localiza- Such differential effects provide further evidence for tion has not been investigated.
the nonuniformity of gastrointestinal-projecting DMV Although elucidating the sources of calcium neces- neurons (8, 9).
sary to activate calcium-dependent potassium currents In both gastric- and intestinal-projecting neurons, is beyond the scope of the present study, the observed ⬃60% of the baclofen-induced response was mediated differences between gastric- and intestinal-projectingvia activation of a barium-sensitive potassium conduc- neurons are a further indication of the distinct basic tance. Gastric- and intestinal-projecting neurons dif- characteristics of DMV neurons projecting to separate fered, however, in the proportion of the baclofen-in- areas of the gastrointestinal tract (8, 9).
duced response attributable to an effect on the inward Activation of central GABAB receptors results in L-type calcium conductance (42% vs. 53%, respectively; several gastrointestinal effects, such as increase in see nifedipine experiments), in the proportion of the gastric and intestinal motility (3, 10), increase in gas- response attributable to an effect on the apamin-sen- tric acid secretion (13), increase in lower esophageal sitive calcium-dependent potassium (SK) conductance sphincter (LES) pressure (4), and decrease of gluta- (32% vs. 48%, respectively; see apamin experiments), mate-induced LES relaxation (1). A similar dichotomy, and in the contribution of the L-type calcium conduc- i.e., both excitatory and inhibitory central effects of tance to the effect on the SK conductance (100% vs.
baclofen, has also been seen in a recent clinical study 84%, respectively; see nifedipine and apamin experi- (19). In their work on healthy volunteers, Lidums and colleagues (19) showed that baclofen decreased the rate Baclofen has been reported previously (14, 18, 35) to of transient LES relaxations, but at the same time, inhibit L-, N- and P/Q-type calcium channels, depend- increased basal LES pressure, probably via a vagally ing on the neuronal type investigated. For example, in mediated pathway.
rat hippocampus inhibitory neurons, baclofen inhibits To the best of our knowledge, no in vivo studies in L-, N- and P/Q-type calcium channels (18), whereas in animal models have been performed in which baclofen rat supraoptic nucleus neurons, baclofen inhibits only has been microinjected directly in the DVC while gas- N- and P/Q-type channels (14) and in cerebellar gran- trointestinal effects were monitored. However, con- ule cells, baclofen inhibits L-type channels (35). Al- vincing evidence (1, 3, 4, 10, 13) points toward the DVC though the effects of baclofen on calcium channels as the central site of action of baclofen on gastrointes- other than the L-type channel were not investigated in tinal function.
detail in the present study, it would appear that the In an in vivo study using a ferret model, Andrews major proportion of the baclofen-induced inhibition of and colleagues (3) administered baclofen subcutane- calcium channels in vagal motoneurons is mediated via ously and observed an increase in gastric pressure as inhibition of L-type calcium channels. Indeed, the com- well as an increase in the amplitude of rhythmic con- bination of nifedipine and barium almost completely tractions. Both effects were abolished by vagotomy (3).
abolished the baclofen-induced outward current. How- In the same study, Andrews et al. (3) reported that, in ever, after pretreatment with nifedipine alone, a small the presence of cholinergic and sympathetic blockade, proportion of the baclofen-induced current was further the actions of baclofen were restricted to an increase in inhibited by perfusion of nifedipine in combination gastric corpus pressure. Such actions can be explained BACLOFEN IN RAT DMV if one considers baclofen to have dual central effects: 1) Portions of this work have been presented previously in abstract an increase in vagal excitatory cholinergic drive to mediate the increase in rhythmic contractions and 2) adecrease in tonic vagal drive to NANC inhibitory neu- rons to mediate the increase in gastric pressure (3). At 1. Abrahams TP, Ekstrand J, Hyland NP, and Hornby PJ.
the cellular level, then, such an apparent contradiction Immunocytochemical localization of GABAB receptors in ferret can be resolved if one assumes that the increase in dorsal vagal complex and functional role in control of loweresophageal sphincter pressure (Abstract). Soc Neurosci Abstr 25: vagal cholinergic drive occurs as a result of the vagal disinhibition that follows blockade of GABAergic NTS 2. Altschuler SM, Bao X, Bieger D, Hopkins DA, and Miselis
neurons impinging on DMV. On the other hand, direct RR. Viscerotopic representation of the upper alimentary tract in
inhibition of DMV neurons by baclofen would result in the rat: sensory ganglia and nuclei of the solitary and spinaltrigeminal tracts. J Comp Neurol 283: 248–268, 1989.
the relief of inhibitory NANC drive to the stomach.
3. Andrews PL, Bingham S, and Wood KL. Modulation of the
Indeed, electrophysiological studies have shown that vagal drive to the intramural cholinergic and non-cholinergic baclofen acts directly on all NTS neurons to produce a neurones in the ferret stomach by baclofen. J Physiol (Lond) 388: membrane hyperpolarization (7), as well as indirectly 25–39, 1987.
4. Blackshaw LA, Smid SD, O'Donnel TA, and Dent J. GABA
to inhibit synaptic transmission from vagal afferents receptor-mediated effects on vagal pathways to the lower oesoph- (7, 23). Such actions would relieve the tonic inhibition ageal sphincter and heart. Br J Pharmacol 130: 279–288. 2000.
that the GABAergic NTS neurons exert over the DMV 5. Bowery NG. Metabotropic GABAB receptors. Neurotransmis-
and lead to an increased vagal motor output (31).
sions 15: 3–18, 1999.
Similarly, the baclofen-mediated increase in gastric 6. Bowery NG, Hudson AL, and Price GW. GABAA and GABAB
receptor site distribution in the rat central nervous system.
acid secretion (13) can be explained by an increase in Neuroscience 20: 365–383. 1987.
vagal activity obtained by disinhibition of NTS 7. Brooks PA, Glaum SR, Miller RJ, and Spyer KM. The
GABAergic neurons.
actions of baclofen on neurones and synaptic transmission in the In contrast, the present study has shown that acti- nucleus tractus solitarii of the rat in vitro. J Physiol (Lond) 457:115–129, 1992.
vation of GABAB receptors by baclofen is also capable 8. Browning KN, Renehan WE, and Travagli RA. Electrophys-
of inducing a direct outward current, or membrane iological and morphological heterogeneity of rat dorsal vagal hyperpolarization, in a subpopulation of DMV neurons.
neurones which project to specific areas of the gastrointestinal The direct inhibition by baclofen of DMV neurons, tract. J Physiol (Lond) 517: 521–532, 1999.
9. Browning KN and Travagli RA. Characterization of the in
which, in our hypothesis, control inhibitory NANC vitro effects of 5-hydroxytryptamine (5-HT) on identified neu- intramural neurons, provides a cellular substrate to rones of the rat dorsal motor nucleus of the vagus (DMV). Br J explain the atropine-insensitive increase in gastric cor- Pharmacol 128: 1307–1315, 1999.
pus pressure (3) and the excitatory effects on the LES 10. Fargeas MJ, Fioramonti J, and Bueno L. Central and pe-
ripheral action of GABA A and GABAB agonists on small intes- tine motility in rats. Eur J Pharmacol 150: 163–169, 1988.
In conclusion, we have shown that 1) GABAB recep- 11. Gage PW. Activation and modulation of neuronal K⫹ channels
tors are functionally present on the membrane of a by GABA. Trends Neurosci 15: 46–51, 1992.
large percentage of DMV neurons, 2) activation of 12. Gillis RA, Quest JA, Pagani FD, and Norman WP. Control
centers in the central nervous system for regulating gastrointes- GABAB receptors by baclofen inhibits several ionic tinal motility. In: Handbook of Physiology. The Gastrointestinal conductances, namely a potassium conductance, a cal- System. Motility and Circulation. Bethesda, MD: Am. Physiol.
cium conductance, and a calcium-dependent potassium Soc., 1989, sect. 6, vol. I, pt. 1, chapt. 17, p. 621–683.
conductance; and 3) the action of baclofen appeared to 13. Goto Y, Tache Y, Debas H, and Novin D. Gastric acid and
vagus nerve response to GABA agonist baclofen. Life Sci 36: differ between gastric- and intestinal-projecting neu- 2471–2475, 1985.
rons. In fact, in intestinal-projecting neurons, the cal- 14. Harayama N, Shibuya I, Tanaka K, Kabashima N, Ueta Y,
cium-mediated components of the outward current ap- and Yamashita H. Inhibition of N- and P/Q type calcium
peared to exert a more prominent role than in gastric- channels by postsynaptic GABAB receptor activation in rat su-praoptic neurones. J Physiol (Lond) 509: 371–383, 1998.
projecting neurons.
15. Hille B. Ionic Channels of Excitable Membranes. Sunderland,
Perspectives. Our data, would then provide an expla- MA: Sinauer, 1992.
nation at the cellular level of the apparently contradic- 16. Honig MG and Hume RI. DiI and DiO: versatile fluorescent
tory (i.e., excitatory and inhibitory) effects of baclofen dyes for neuronal labelling and pathway tracing. Trends Neuro- on vagal motor activity. In fact, the overall effect of sci 12: 333–341, 1989.
17. Kaupmann K, Huggel K, Heid J, Flor PJ, Bischoff S,
brain stem GABAB receptor activation seems to be Mickel SJ, McMaster G, Angst C, Bittiger H, Froestl W,
highly dependent on the neuronal circuit under study and Bettler B. Expression cloning of GABAB receptors uncovers
as well as the complement and activity of ionic conduc- similarity to metabotropic glutamate receptors. Nature 386: tances within the DMV gastrointestinal motoneurons.
239–246, 1997.
18. Lambert NA and Wilson WA. High-threshold Ca2⫹ currents in
Future studies combining in vivo and in vitro experi- rat hippocampal interneurones and their selective inhibition by ments aimed at separating these circuits will be re- activation of GABAB receptors. J Physiol (Lond) 492: 115–127, quired to resolve the neurochemical phenotype and the membrane properties of the neurons involved in the 19. Lidums I, Lehman A, Checklin H, Dent J, and Holloway
RH. Control of transient lower esophageal sphincter relaxation
effects of GABAB receptor activation in the dorsal va- and reflux by the GABAB agonist baclofen in normal subjects.
gal complex.
Gastroenterology 118: 7–13, 2000.
BACLOFEN IN RAT DMV 20. Mcdermott BJ, Ekstrand J, and Hornby PJ. Immunocyto-
28. Sah P. Different calcium channels are coupled to potassium
chemical GABAB receptor staining in the hindbrain of rodents channels with distinct physiological roles in vagal neurons. Proc (Abstract). Soc Neurosci Abstr 25: 940, 1999.
R Soc Lond B Biol Sci 260: 105–111, 1995.
21. Mendelowitz D and Kunze DL. Identification and dissociation
29. Sah P and McLachlan EM. Potassium currents contributing to
of cardiovascular neurons from the medulla for patch clamp action potential repolarization and the afterhyperpolarization in analysis. Neurosci Lett 132: 217–221, 1991.
rat vagal motoneurons. J Neurophysiol 68: 1834–1841, 1992.
22. Pagani FD, Norman WP, and Gillis RA. Medullary parasym-
30. Shapiro RE and Miselis RR. The central organization of the
pathetic projections innervated specific sites in the feline stom- vagus nerve innervating the stomach of the rat. J Comp Neurol ach. Gastroenterology 95: 277–288, 1988.
238: 473–488, 1985.
23. Page AJ and Blackshaw LA. GABA
31. Sivarao DV, Krowicki ZK, and Hornby PJ. Role of GABAA
B receptors inhibit mech- anosensitivity of primary afferent endings. J Neurosci 19: 8597– receptors in rat hindbrain nuclei controlling gastric motor func-tion. Neurogastroenterol Motil 10: 305–313, 1998.
32. Travagli RA, Gillis RA, Rossiter CD, and Vicini S. Gluta-
24. Pedarzani P, Kulik A, Muller M, Ballanyi K, and Stocker M.
mate and GABA-mediated synaptic currents in neurons of the Molecular determinants of Ca2⫹-dependent K⫹ channel function in rat dorsal motor nucleus of the vagus. Am J Physiol Gastrointest rat dorsal vagal neurones. J Physiol (Lond) 527: 283–290, 2000.
Liver Physiol 260: G531–G536, 1991.
25. Powley TL, Berthoud HR, Fox EA, and Laughton W. The
33. Travagli RA, Lewis MW, and Browning KN. Effects of ba-
dorsal vagal complex forms a sensory-motor lattice: the circuitry clofen on identified rat dorsal motor nucleus of the vagus (DMV) of gastrointestinal reflexes. In: Neuroanatomy and Physiology of neurons projecting to the stomach or intestine (Abstract). Soc Abdominal Vagal Afferents, edited by Ritter S, Ritter RC, and Neurosci Abstr 29: 380.3, 1999.
Barnes CD. Boca Raton, FL: CRC, 1992, p. 55–79.
34. Travagli RA, Ulivi M, and Wojcik WJ. Gamma-aminobutyric
26. Rinaman L, Card JP, Schwaber JS, and Miselis RR. Ultra-
acid-B receptors inhibit glutamate release from cerebellar gran- structural demonstration of a gastric monosynaptic vagal circuit ule cells: consequences of inhibiting cyclic AMP and calcium in the nucleus of the solitary tract in rat. J Neurosci 9: 1985– influx. J Pharmacol Exp Ther 258: 903–909, 1991.
35. Wojcik WJ, Travagli RA, Costa E, and Bertolino M. Ba-
27. Rusin KI and Moises HC. ␮-Opioid and GABAB receptors
clofen inhibits with high affinity an L-type-like voltage-depen- modulate different types of Ca2⫹ currents in rat nodose ganglion dent calcium channel in cerebellar granule cell cultures. Neuro- neurons. Neuroscience 85: 939–956, 1998.
pharmacology 29: 969–972, 1990.

Source: http://d3n1qfjutz9qb9.cloudfront.net/content/280/6/G1106.full-text.pdf