Brain Research, 516 (1990) 201-207
201
Elsevier BRES 15449
Some characteristics of baclofen-evoked responses of primary afferents in frog spinal cord A.L. Padjen and G.M. Mitsoglou Department of Pharmacology and Therapeutics, McGill University, Montreal, Quebec (Canada) (Accepted 10 October 1989)
Key words: 7-Aminobutyric acid; Baclofen; Frog spinal cord; Ba; 4-Aminopyridine; Primary afferent terminal; Baclofen antagonist
Baclofen has been shown to be a selective agonist for a subclass of GABA receptors (GABAB) in many regions of the vertebrate nervous system. On the intraspinal terminals of dorsal roots (DRT), it evokes a pure hyperpolarizing response. We have previously shown that the response of DRT to GABA and some of its analogs (e.g. kojic amine) in isolated frog spinal cord is dual in nature, consisting of a bicuculline-sensitive depolarizing component and a bicuculline-resistant hyperpolarizing component. Under the working hypothesis that the hyperpolarizing component of the GABA-evoked response is mediated by the activation of GABAB receptors, we have examined, using the sucrose gap technique, some characteristics of the response of DRT to baclofen. We have found that this response is stereospecific (L-baclofen being about 100 times more potent than D-baclofen), dependent on [K]o (response amplitude inversely related to [K]o), blocked by barium (0.5 mM causing a reduction of the response amplitude to 37% of control), and is not significantly affected by 4-aminopyridine, nor by inorganic calcium channel blockers (manganese, cobalt, cadmium). Some proposed GABA B antagonists (6-aminovaleric acid, 6-aminolaevulinic acid, phaclofen) are also rather ineffective at blocking it. These results are therefore consistent with the notion that the baclofen-evoked response of DRT is mediated by an increase in conductance to potassium ions. INTRODUCTION G A B A is considered to be the most widely distributed inhibitory n e u r o t r a n s m i t t e r in the v e r t e b r a t e CNS 27. A t the first sensory synapse in the spinal cord, it r e p o r t e d l y plays a role in presynaptic inhibition 18'29. Recently, convincing d a t a have surfaced in support of G A B A r e c e p t o r multiplicity in the v e r t e b r a t e nervous system 7. Two sites have b e e n established by binding studies and designated by B o w e r y and colleagues 6 as G A B A A and G A B A B. Presently, one of the m a j o r distinctions between these two classes of receptors is that the former is sensitive to the antagonistic action of bicuculline while the latter is bicuculline-resistant and selectively activated by baclofen, a chlorophenyl derivative of G A B A 7. Baclofen is also capable of depressing neurotransmission across the first sensory synapse by interacting at a presynaptic site 9'44. Consequently, it is possible that G A B A can p r o d u c e presynaptic inhibition through m o r e than one mechanism. A s previously d e m o n s t r a t e d in this l a b o r a t o r y , t h e responses to G A B A and o t h e r neutral amino acids (e.g. kojic amine, taurine) on the dorsal root terminals ( D R T ) of isolated frog spinal cord are dual in nature, consisting of b o t h a depolarizing c o m p o n e n t and a hyperpolarizing
c o m p o n e n t . T h e latter b e c o m e s m o r e evident at low agonist concentrations or in the presence of bicuculline 5' 35,37,39 Baclofen, on the o t h e r hand, evokes a pure hyperpolarizing response at this site. U n d e r the working hypothesis that all the h y p e r p o l a r izing effects are mediated by the same receptor ( G A B A u ) , we chose to further characterize the responses of D R T to baclofen because of its a p p a r e n t selectivity. This was done in two ways. Firstly, we e x a m i n e d the ionic r e q u i r e m e n t s to see if they were consistent with those of G A B A u - e v o k e d responses f o u n d elsewhere in the vertebrate nervous system. Secondly, we tested the capacity of p r o p o s e d G A B A A antagonists to block these responses. Preliminary results have b e e n c o m m u n i c a t e d 3°.
MATERIALS AND METHODS Experiments were done on isolated hegnisected spinal cord from leopard frog, Rana pipiens (both northern and southern varieties), kept in aquaria at 4 °C for periods of 1 week to several months. Details of the methods were described previously2'42. Briefly, frogs were cooled to an anesthetic state in crushed ice, and decapitated before a dorsal or ventral laminectomy was performed. The spinal cord with its attached roots was carefully removed to a dissecting dish, sagittally hemisected, and then placed in the central compartment (bath volume = 0.15 ml) of a sucrose gap chamber. The attached dorsal roots (9th and/or 10th segment) were led out by way
Correspondence: A.L. Padjen, Department of Pharmacology and Therapeutics, McGill University, 3655 Drummond Street, Montreal, Quebec H3G 1Y6, Canada. 0006-8993/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
202 of slits through the sucrose compartment into two separate pools of normal Ringer solution. All slits, as well as a closely fitted cover, were coated with White Petrolatum (USP) to provide a leak-proof separation of compartments. The central and sucrose compartments were continuously perfused with unoxygenated Ringer solution (flow rate 1.5-2.0 ml/min) and 244 mM sucrose (0.5-1.0 ml/min) respectively. Temperature (10-12 °C) was controlled by means of a Peltier device. Normal Ringer solution (pH 7.3) contained the following (in mM): sodium chloride, 115; potassium chloride, 2; calcium chloride, 2; HEPES (pH 7.3), 5; D-glucose, 10. Tetrodotoxin (0.2/tM) was present in all solutions superfusing the spinal cord to minimize indirect drugevoked responses. Drugs were applied by interchange of solutions superfusing the central compartment via a set of stopcocks. Drug solutions were made by diluting appropriate amounts of concentrated stock solution in normal Ringer solution and pH adjusted if necessary. All drugs were purchased from Sigma, St. Louis, MO, except for kojic amine (gift from Merck Frosst Laboratories, Kirkland, P.Q., Canada) and baclofen (gift from Ciba-Geigy, Basel, Switzerland). Electrical activity was recorded by monitoring the potential difference between two silver-silver chloride pellet electrodes making contact with the solution in the central (ground) and root compartments via salt bridges (4% agar Ringer). Permanent records were obtained from a Gould (Brush 2400) rectilinear pen recorder. Signals were filtered by a low-pass filter (0.5 s time constant) without distortion of drug responses. The observed potential changes represent an attenuated version of the average change in the membrane potential of the intramedullary part of a population of primary afferents. Under our recording conditions, an upward deflection of the pen represents depolarization. Results are expressed as the mean + S.E.M. wherever possible. RESULTS
Differential sensitivity of D R T to G A B A and baclofen Sucrose gap recordings from D R T (Figs. 2-6) reveal that baclofen consistently evokes a concentration-dependent hyperpolarizing response, unlike its chemical analogue G A B A . This hyperpolarization has a characteristically slow rise (time to peak about 4 rain) and is very slow to return to baseline, especially when it is brought about by higher concentrations of baclofen (up to 1.5 h at 0.1 mM). Responses to baclofen, at concentrations greater than 0.1 mM, often never fully recover, even after several hours of wash out. It is very likely that the cause of this slow recovery is the lack of an effective uptake system for the removal of baclofen 6. In addition, we found that D R T from different segments of the spinal cord are differentially sensitive to baclofen and G A B A (see Table I below). Baclofen-evoked responses have greater amplitudes on D R T of segment IX (DRT9) than
on D R T of segment X (DRT10) for the same agonist concentration, with the reverse being true for G A B A .
Stereospecificity of baclofen-evoked responses Unlike G A B A , baclofen exists in either the dextrorotatory or levorotatory structural configurations. The stereospecificity of the baclofen-evoked response of D R T is evident after comparing the ECs0 values from the concentration-response curves for the two stereoisomers (Fig. 1). The L-isomer, with an ECso of 0.01 mM, is about 100 times more potent than the D-form (see Fig. 3 for samples of typical recordings). The maximum amplitude of the hyperpolarizing response evoked by L-baclofen occurs at a concentration of 0.1 mM.
Sensitivity to changes in [K+]o The responses of D R T to several neutral amino acids (analogs of G A B A ) can exhibit a hyperpolarizing component under certain conditions, such as low agonist concentrations (i.e. lower than 0.1 mM) or in the presence of bicuculline 5'39. This is particularly obvious with some analogs, such as kojic amine (Fig. 2). However, even when a duality is not evident under bicuculline blockade, lowering [K]o can bring out such a component, as exemplified by the G A B A and taurineevoked responses in Fig. 2 (see also Fig. 3). Hyperpolarizing responses already visible in the presence of bicuculline (i.e. kojic amine and baclofen), are increased when [K]o is reduced (see also Figs. 3 and 5). The potentiation of the baclofen-evoked response depends upon the extent of [K]o reduction: in 0.5 mM [K]o, the response amplitude increases to 120% (N = 2) of control (Fig. 3), while in 0.1 mM [K]o, it approximately doubles (215% + 25%; N = 6; Figs. 2 and 5). Parallel increases in the slope of the responses (rise phase) are also evident,
1.6
o--o L-boclofen ~E
~
.--.
T
0.8
°
04 ~
1 / o
o.0 0.001
TABLE I
[ o[
D b o l co f e n
/ [ 0010
I O, 1 O0
I 10 0 0
Concentrotion (mM)
Response amplitude (m V)
DRT9 DRT10
O.5 mM GA BA
O.1 mM L-baclofen
1.86+0.10 (N=31) 2.93+0.19 (N=33)
-1.24+0.10 (N=8) -0.94+0.11 (N=8)
Fig. 1. Concentration-response curves of baclofen stereoisomers on dorsal root terminals of the 9th segment (DRT9). Note that L-baclofen is more potent than D-baclofen in hyperpolarizing DRT (mean _+ S.E.M. wherever possible, N = 1-8; the difference between the curves is significant at the 1% level, using a paired Student's t-test, for 0.1 mM agonist concentration only).
203 Control
0.02 mM BIC
A
(3ABA TAU 1 0.5 0.1K/BIC
1
KJA 0,1
GLU 0.5
L--BAC 0.1
3.5K
KJAO.2
GABAO.5 TAU1
5
Fig. 2. Effect of low [K]o on neutral amino acid-evoked responses in the presence of bicuculline. Abbreviations: BIC, bicucuUine; TAU, taurine; KJA, kojic amine; L-BAC,L-baclofen;0.1K, 0.1 mM potassium. On this and other figures: drug application is marked by a horizontal bar; depolarization upwards; calibrations: mV (vertical), min (horizontal). Bicuculline depresses depolarizing responses (see text); in this case, the kojic amine-evoked depolarization (not shown) was changed to a hyperpolarizing response. Note that decreasing [K]o brings out the hyperpolarizing components of responses to GABA and taurine and increases the hyperpolarizing responses to kojic amine and L-baclofen.
possibly reflecting a change in the driving force as a result of a new ionic gradient. In contrast, raising the level of [K]o depresses the baclofen-evoked response in a reversible manner (to 15% of control in 4 m M [K]o; N = 2; Fig. 3).
Effect of barium Barium has been shown to block a variety of K + currents T M . In our experiments, two concentrationdependent effects indicative of this channel-blocking ability are depolarization of the resting membrane potential of D R T and blockade (and/or reversal) of the K-evoked depolarizing responses (Figs. 4, 5) 31'32'39 . T h e responses to all the amino acids tested (i.e. G A B A , taurine, kojic amine, baclofen, glutamate) are depressed to varying degrees (Fig. 4) with baclofen exhibiting the
Ba 0.5
D.L,-BAC 0.2 0.5 m M BQ
Fig. 4. Effect of barium (0.5 mM) on the responses of DRT to amino acids and increased potassium. Abbreviations: GLU, glutamate; Ba, barium (arrow indicates start of perfusion). Note that the hyperpolarizing response to D,L-baclofen and the depolarizing response to increased [K]o (3.5 mM) are depressed more than the other responses by 0.5 mM barium, and thatthe taurine-evoked response becomes multiphasic. greatest sensitivity (amplitude reduced to 27% of control in the example illustrated; mean 37% _+ 7% of control, N = 4). The taurine-evoked response in particular exhibits a considerable depression with the emergence of a hyperpolarizing component sl,a2. Further increasing the barium concentration in the presence of low [K+]o (reflected in a much greater baseline depolarization) 32 causes a total blockade of the baclofen-evoked response, while the response to G A B A is only partially depressed (Fig. 5).
Effect of 4-aminopyridine (4-AP) 4-AP is another well-known blocker of potassium
Control Control
GABA~ 0.5
L-BAC 0.2
-- ~ D
GA-BA 0.5
K4"~
D.L.- BAC 0.1
Wosh
_z_ _
0.1K/5Bo m
Fig. 3. Effect of changes in [Klo on the responses of DRT to L-baclofen and GABA. Note that the amplitude of baclofen-evoked responses is inversely proportional to [K]o, and that L-baclofen is more potent than D-baclofen in hyperpolarizing DRT (see also Fig. 1).
Fig. 5. Effect of barium (5 mM) in low [K]o on the responses of DRT to amino acids and K. Note that the D,L-baclofen-evoked response is completely blocked and the K-evoked response is reversed while the response to GABA is partially depressed.
204 conductance 14'5°'51. On frog DRT, this agent proved to be ineffective at depressing the baclofen-evoked response in concentrations of 2-5 mM (N = 4; Fig. 6). It did, however, cause a small depolarization of the resting membrane potential (1.43 + 0.50 mV at 5 mM, N = 4), and it increased the amplitude of the GABA-evoked responses (160% of control + 10%, N --- 4) and the K-evoked response (not shown; see Padjen and Smith, 1981). It is interesting that the taurine-evoked response is depressed by 4-AP (42% of control + 7%, N = 4; not shown), in agreement with recent findings in hippocampal slices 53. In preliminary experiments, tetraethylammonium, another blocker of certain potassium currents 14'5°, only attenuated (by less than 40%) the baclofen-evoked response of D R T at a concentration of 10 mM.
Effect of inorganic calcium channel blockers There were 3 main reasons for testing the effects of agents known to block calcium conductance: (i) if baclofen-evoked responses are directly mediated by an increase in K conductance, then the channels involved may be calcium-dependent4; (ii) the responses of D R T to baclofen may be due to a transient inhibition of the chronic release of a depolarizing substance onto DRT, even in the presence of tetrodotoxin; (iii) baclofen effects have been explained as being due to a direct inactivation of a voltage-dependent calcium current 15'16. The 3 inorganic calcium channel blockers applied were manganese (2 mM; N = 2), cobalt (2-10 raM; N = 6), and cadmium (0.5 mM; N = 2). None of these agents had any significant effect on the baclofen-evoked responses. Both manganese and cobalt did however produce a depolarization of D R T and augmented the G A B A evoked responses while cadmium caused a small hyperpolarization of DRT, as previously reported 41.
DRT Control
.j
_ GABA 0.5
-~
4-A%
L-BAG01
5mM 4-AP
2
Effects of putative GABA B antagonists We tested several agents proposed to antagonize G A B A B effects. Delta-aminovaleric acid 33 (1 raM) produced a dual effect on the resting membrane potential and reduced the responses of D R T to 0.5 mM G A B A (54% of control + 4%, N = 4) but had little effect at reducing the amplitude of the baclofen-evoked response (i.e. < 20% depression, N = 2). 6-Aminolaevulinic acid 8 (1 mM) caused only a modest (i.e. < 30%, N = 2) attenuation of the baclofen-evoked response. Phaclofen has recently been suggested to be another possible baclofen antagonist 11'26. Its use in the rat hippocampus allowed the tentative identification of a synaptic inhibitory pathway mediated by a GABAB receptor 17. Our preliminary data on the responses of D R T to baclofen show only a modest antagonistic effect (i.e. < 30% depression) at high concentrations of phaclofen (i.e. > 0.5 mM). In conclusion, our studies so far have not identified any substance capable of antagonizing the baclofenevoked responses of D R T with a potency or selectivity useful for pharmacological exploration. DISCUSSION On the basis of a number of physiological and binding studies, a separate site of G A B A action has been described (GABAB 6'7) on presynaptic terminals in both the peripheral and central nervous system of vertebrates. Although CNS terminals are rarely accessible for direct analysis, sucrose gap recording on dorsal roots permits the measurement of the polarization of a whole population of DRT. Using this technique, we have identified hyperpolarizing responses of D R T to G A B A , taurine and analogs 5'3°'39 that have many attributes of a GABAB site, in agreement with some binding studies on the distribution of G A B A B sites in the spinal cord 46. The present study describes some characteristics of these responses, particularly those evoked by baclofen as the most potent and selective agonist at this site. The results demonstrate that the baclofen-evoked responses of D R T are: (i) tetrodotoxin and manganese resistant, and therefore very likely evoked directly on D R T membranes; (ii) stereospecific; (iii) insensitive to bicuculline, and largely unaffected by several putative G A B A B antagonists; (iv) dependent on [K]o; (v) blocked by barium; (vi) unaffected by 4-AP or inorganic calcium channel blockers.
5
\
Fig. 6. Effect of 4-AP on the responses of DRT to L-baclofen and GABA. Note the lack of effect of 4-AP on the amplitude of the response evoked by L-baclofen, while that evoked by GABA is increased.
Comparison with baclofen-evoked responses of other cells Baclofen has a hyperpolarizing effect on amphibian motoneurons (unpublished observation). In addition, baclofen has been shown to evoke hyperpolarizing responses in other cells of the vertebrate CNS, particu-
205 larly hippocampal pyramidal cells. The characteristics of these responses appear to be very similar to those of baclofen-evoked responses in our preparation. For example, the ratio of the EC50 for the levorotatory versus the dextrorotatory isomers of baclofen on DRT (i.e. approximately 100:1) is close to that obtained in hippocampus 34. Similarly, the observed dependence of our baclofen-evoked responses on [K]o and their sensitivity to barium ions is in agreement with intracellular studies of hippocampal pyramidal cells22-24"34. The observed effects of such manipulations are unlikely to be due solely to the associated changes in the membrane polarization level since these would then be contrary to what is expected of hyperpolarizing responses. It has been reported 22-24'34 that baclofen-evoked responses are voltage-sensitive in the depolarizing direction but this factor could not totally account for the depressant action of barium. In contrast to one report on hippocampal neurons 23, we did not find the baclofen-evoked responses to be sensitive to 4-AP, even at relatively higher concentrations. The possibility therefore exists that different potassium conductances are involved. Baclofen has been also reported to reduce a calcium current in cultured D R G without any action on potassium conductance and/or resting membrane potential 1°' 13,15,16,47,48 This type of baclofen action seems to be distinct from the one described here (see later). Our experimental data are thus consistent with the hypothesis that one type of baclofen-evoked response in vertebrate CNS reflects the direct activation of a potassium conductance. This response appears to originate predominantly on thin-diameter fibers 10'37"39'46. The differential distribution of these fibers 52 may then explain the segmental differences in baclofen potency demonstrated in this study. In addition to baclofen and GABA, other agents, such as kojic amine and taurine, have a similar hyperpolarizing action. The lack of a potent and specific antagonist of the hyperpolarizing component (see below) does not allow any possible distinct pharmacology of it to be done and we could assume, for the moment, that all of these agents might act at the same site to produce the observed hyperpolarization. It is of interest that the depolarizing component of the responses of DRT to taurine and kojic amine (but not GABA) is sensitive to 6-aminomethyl3-methyl-4H-1,2,4-benzothiadiazine-l,l-dioxide4°, thus indicating that these compounds activate a site which G A B A does not.
Physiological significance The possible physiological role of baclofen-evoked responses, including those on the presynaptic terminals of primary afferents, derives from the assumption that
baclofen is an analog of endogenous G A B A at the G A B A B receptors. By virtue of its electrophysiological effect, baclofen could thus be assumed to inhibit the release of a transmitter from D R T in the frog spinal cord 9"21, in general agreement with its main pharmacological mechanism of action 7'45. This may then explain the clinical efficacy of baclofen in the treatment of spinal spasticity, which occurs at an intrathecal concentration range 43 that is similar to that used in this study. Synaptically evoked hyperpolarization of presynaptic terminals (PAH) has been associated with presynaptic inhibition in invertebrate preparations 3'2°'25, but has not been identified in the vertebrate spinal cord, where the depolarization of primary afferents (PAD) is normally associated with a presynaptic inhibitory mechanism 36. Other evidence indicates that PAH in invertebrates may be due to the depression of a tonic depolarization of DRT 36. Even though this would suggest that PAH is not a synaptic entity in vertebrate spinal cord, it is possible that both PAD and PAH occur simultaneously on DRT but that the more dominant depolarizing response masks the presence of PAH. Although the depolarizing responses to both G A B A and PAD could be blocked by bicuculline or picrotoxin 2, the associated convulsions would prevent any possible recognition of PAH. The dual nature of the GABA-evoked responses of DRT 5,39 suggests that G A B A could be the physiological mediator of both PAD and (possibly) PAH via interaction at G A B A A and GABAB receptors, respectively. Although these two receptors seem to be different, the activation of both could lead to presynaptic inhibition by a similar mechanism, i.e. a shunt of current that keeps the membrane potential from substantial depolarization 18'38 thus preventing the opening of voltage-dependent calcium channels. Direct participation of a calciumdependent potassium conductance in the baclofenevoked (GABAB) responses is unlikely because of the ineffectiveness of inorganic calcium channel blockers, as shown in this study (also ref. 22). An additional mode of action of baclofen, i.e. a direct depression of calcium fluxes at the primary afferent terminals, has been suggested on the basis of observations at the somata of the s a m e cells 1°,15,16,47. At DRT, however, this mechanism is not amenable to direct analysis, and the present study is therefore not able to address this issue. It is however entirely possible that more than one mechanism is involved in presynaptic inhibition mediated via G A B A a receptors. The issue of whether the hyperpolarizing responses are physiologically important in the primary afferents of vertebrates awaits two developments: (i) identification of a synaptic event that would be its physiological equivalent, i.e. the PAH; (ii) discovery of potent competitive
206 antagonists of G A B A B receptors that produce concur-
sponses (e.g. baclofen). Such antagonists would also provide i n d e p e n d e n t evidence as to whether or not the
In conclusion, our study suggests that the baclofenevoked hyperpolarizing response of D R T is mediated by potassium fluxes and that this may provide a mechanism for presynaptic inhibition mediated by G A B A B receptors. These responses may therefore serve as a model
hyperpolarizing responses elicited by other neutral amino acids (e.g. G A B A , taurine, kojic amine) 3°'39 are medi-
system for future studies of G A B A B pharmacology in the vertebrate CNS.
rent depression of presynaptic inhibition (or P A H ) and responses evoked by agonists of hyperpolarizing re-
ated by the same receptor. Unfortunately, the present results, as well as similar studies on isolated rat spinal cord 19, suggest that the discovery of a potent and specific antagonist is still in the future. REFERENCES 1 Arhem, P., Effects of rubidium, caesium, strontium, barium, and lanthanum on ionic currents in myelinated nerve fibers from Xenopus laevis, Acta Physiol. Scand., 108 (1980) 7-16. 2 Barker, J.L., Nicoll, R.A, and Padjen, A.L., Studies on convulsants in the isolated frog spinal cord. I. Antagonism of amino acid responses, J. Physiol. (Lond.), 245 (1975) 521-536. 3 Baxter, D.A. and Bittner, G.D., Intracellular recordings from crustacean motor axons during presynaptic inhibition, Brain Research, 223 (1981) 422-428. 4 Blaxter, T.J., Carlen, P.L., Davies, M.E and Kujtan, P.W., Gamma-aminobutyric acid hyperpolarizes rat ~hippocampal pyramidal cells through a calcium-dependent potassium conductance, J. Physiol, (Lond.), 373 (1986) 181-194. 5 Bourne, G.W. and Padjen, A.L., Kojic amine: a GABA agonist with dual action, Soc. Neurosci. Abstr., 6 (1980) 148. 6 Bowery, N.G., Hill, D,R. and Hudson, A.L., Characteristics of GABAB receptor binding sites on rat whole brain synaptic membranes, Br. J. PharmacoL, 73 (1983) 191-206. 7 Bowery, N.G., Price, G.W., Hudson, A.L., Hill, D.R., Wilkin, G.P. and Turnbull, M.J., GABA receptor multiplicity: visualization of different receptor types in the mammalian CNS, Neuropharmacology, 23 (1984) 219-231. 8 Brennan, M.J.W. and Cantrill, R.C., b-Aminolaevulinic acid is a potent agonist for GABA autoreceptors, Nature (Lond.), 280 (1979) 514-515. 9 Davidoff, R.A. and Sears, E.S., The effects of lioresal on synaptic activity in the isolated spinal cord, Neurology, 24 (1974) 957-963. 10 Desarmenien, M., Feltz, P., Occhipinti, G., Santangelo, E and Schlichter, R., Coexistence of GABAA and GABAB receptors on A-delta and C primary afferents, Br. J. Pharmacol., 81 (1984) 327-333. 11 Dickenson, H.W., Allan, R.D., Ong, J. and Johnston, G.A.R., GABAB receptor antagonist and GABAA receptor agonist properties of a ~,-aminovalericacid derivative, Z-5-aminopent2-enoic acid, Neurosci. Lett., 86 (1988) 351-355. 12 Dodd, J. and Horn, J.P., Muscarinic inhibition of sympathetic C neurones in the bullfrog, Br. J. Pharmacol., 74 (1981) 579-585. 13 Dolphin, A.C. and Scott, R.H., Inhibition of calcium currents in cultured rat dorsal root ganglion neurones by (-)-baclofen, Br. J. Pharmacol., 88 (1986) 213-220. 14 Dubois, J.M., Potassium currents in the frog node of Ranvier, Progr. Biophys. Mol. BioL, 42 (1983) 120. 15 Dunlap, K. and Fischbach, G.D., Neurotransmitters decrease the calcium component of sensory neurone action potentials, Nature (Lond.), 276 (1978) 837-839. 16 Dunlap, K. and Fischbach, G.D., Neurotransmitters decrease the calcium conductance activated by depolarization of embryonic chick sensory neurones, J. Physiol. (Lond.), 317 (1981) 519-553. 17 Dutar, P. and Nicoll, R.A., A physiological role for GABAB receptors in the central nervous system, Nature (Lond.), 332
Acknowledgements. This work was supported by the Medical Research Council of Canada. G.M.M. held a studentship from the Fonds de la Recherche en Sant6 du QuEbec. The authors wish to thank Alain SEvigny for some technical assistance.
(1988) 156-158. 18 Eccles, J.C., Presynaptic inhibition. In The Physiology of Synapses, Springer, Berlin, 1964, pp, 220-238. 19 Evans, R.H., Pharmacology of amino acid receptors on vertebrate primary afferent nerve fibres, Gen. Pharmacol., 17 (1986) 5-11. 20 Fuchs, P.A., Intracellular analysis of presynaptic inhibition in the crayfish claw opener, Soc. Neurosci. Abstr., 3 (1977) 176. 21 Fukuda, H., Kudo, Y. and Ono, H., Effects of beta-(parachlorophenyl)-GABA (baclofen) on spinal synaptic activity, Eur. J. Pharmacol., 44 (1987) 17-24. 22 G/ihwiler, B.H. and Brown, D.A., GABAa-receptor-activated K+ current in voltage-clamped CA3 pyramidal cells in hippocampal cultures, Proc. Natl. Acad. Sci. U.S.A., 82 (1985) 1558-1562. 23 Inoue, M., Matsuo, T. and Ogata, N., Baclofen activates voltage-dependent and 4-aminopyridine sensitive K+ conductance in guinea-pig hippocampal pyramidal cells maintained in vitro, Br. J. Pharmacol., 84 (1985) 833-841. 24 Inoue, M., Matsuo, T. and Ogata, N,, Possible involvement of K+-conductance in the action of GABA in the guinea-pig hippocampus, Br. J. PharmacoL, 86 (1985) 515-524. 25 Kawai, N. and Niwa, A., Hyperpolarization of the excitatory nerve terminals by inhibitory nerve stimulation in lobster, Brain Research, 137 (1977) 365-368. 26 Kerr, D.I.B., Ong, J., Prager, R.H., Gynther, B.D. and Curtis, D.R., Phaclofen: a peripheral and central baclofen antagonist, Brain Research, 405 (1987) 150-154. 27 Krnjevic, K., Chemical nature of synaptic transmission in vertebrates, Physiol. Rev., 54 (1974) 418-540. 28 Krnjevic, K., Pumain, R. and Renaud, L., Effects of barium and tetraethylammonium on cortical neurones, J. Physiol. (Lond.), 215 (1971) 223-245. 29 Levy, R.A., The role of GABA in primary afferent depolarization, Progr. Neurobiol., 9 (1977) 211-267. 30 Mitsoglou, G.M. and Padjen, A.L., Hyperpolarizing responses to GABA and analogs on primary afferent fibres, Soc. Neurosci. Abstr., 11 (1985) 282. 31 Mitsoglou, G.M. and Padjen, A.L., Contributionof [K+]o to the neutral amino acid responses on primary afferents, Proc. Can. Fed. Biol. Soc., 28 (1985) 169. 32 Mitsoglou, G.M. and Padjen, A.L., Differential effect of barium on responses of frog primary afferents to neutral amino acids, submitted. 33 Muhyaddin, M., Roberts, P.J. and Woodruff, G.N., Presynaptic gamma-aminobutyric acid receptors in the rat anococcygeus muscle and their antagonism by 5-aminovaleric acid, Br. J. Pharmacol., 77 (1982) 163-168. 34 Newberry, N.R. and Nicoll, R.A, Comparison of the action of baclofen with GABA on rat hippocampal pyramidal cells in vitro, J. Physiol. (Lond.), 360 (1985) 161-185. 35 Nicoll, R.A., The interaction of porphyrin precursors with GABA receptors in the isolated frog spinal cord, Life Sci., 19 (1976) 521-526.
207 36 Nicoll, R.A. and Alger, B.E., Presynaptic inhibition: transmitter and ionic mechanisms, Int. Rev. Neurobiol., 21 (1979) 217-258. 37 Padjen, A.L. and Hashiguchi, T., Intraceilular analysis of GABA-evoked responses of frog primary afferent axons, Soc. Neurosci. Abstr., 8 (1982) 577. 38 Padjen, A.L. and Hashiguchi, T., Primary afferent depolarization in frog spinal cord is associated with an increase in membrane conductance, Can. J. Physiol. Pharmacol., 61 (1983) 626-631. 39 Padjen, A.L., Hashiguchi, T., Mitsoglou, G.M. and Bourne, G.W., Dual effect of GABA and kojic amine on primary afferents: electrophysiological evidence for the activation of GABAn receptors, 1988. 40 Padjen, A.L., Mitsoglou, G.M. and Hassessian, H., Further evidence in support of taurine as a mediator of synaptic transmission in the frog spinal cord, Brain Research, 488 (1989) 288-296. 41 Padjen, A.L. and Smith, P.A., Possible role of divalent cations in amino acid responses of frog spinal cord. In EV. De Feudis and E Mandel (Eds.), Amino Acid Neurotransmitters, Raven Press, New York, 1981, pp. 271-280. 42 Padjen, A.L. and Smith, P.A., The role of the electrogenic sodium pump in the glutamate afterhyperpolarization of frog spinal cord, J. Physiol. (Lond.), 336 (1983) 433-451. 43 Penn, R.D., Savoy, S.M., Corcos, D., Latash, M., Gottlieb, G., Parke, B. and Kroin, J.S., Intrathecal baclofen for severe spinal spasticity, N. Engl. J. Med., 320 (1989) 1517-1521. 44 Pierau, E-K., Matheson, G.K. and Wuster, R.D., Presynaptic action of fl-(4-chlorophenyl)-GABA, Exp. Neurol., 48 (1975) 343-351. 45 Potashner, S.J., Baclofen: effects on amino acid release and
46
47
48
49
50 51 52 53
metabolism in slices of guinea pig cerebral cortex, J. Neurochem., 32 (1979) 103-109. Price, G.W., Wilkin, G.E, Turnbull, M.J. and Bowery, N.G., Are baclofen-sensitive GABAB receptors present on primary afferent terminals of the spinal cord?, Nature (Lond.), 307 (1984) 71-74. Robertson, B. and Taylor, R., Effects of gamma-aminobutyric acid and (-)-baclofen on calcium and potassium currents in cat dorsal root ganglion neurones in vitro, Br. J. Pharmacol., 89 (1986) 661-672. Schlichter, R., Demeneix, B.A., Desarmenien, M., Desaulles, E., Loeffler, J.P. and Feltz, P., Properties of the GABA receptors located on spinal primary afferent neurones and hypophyseal neuroendocrine cells of the rat, Neurosci. Lett., 47 (1984) 257-263. 8tanden, N.B. and Stanfield, ER., A potential- and timedependent blockade of inward rectification in frog skeletal muscle fibres by barium and strontium ions, J. Physiol. (Lond.), 280 (1978) 169-191. Thompson, S.H., Three pharmacologically distinct potassium channels in molluskan neurones, J. Physiol. (Lond.), 265 (1977) 465-488. Ulbricht, W. and Wagner, H.H., Block of potassium channels of the nodal membrane by 4-aminopyridine and its partial removal on depolarization, Pfliigers Arch., 367 (1976) 77-87. Wilhelm, G.B. and CoggeshaU, R.E., An electron microscopic analysis of the dorsal root in the frog, J. Comp. Neurol., 196 (1981) 421-429. Zeise, M., Taurine on hippocampal slices: comparison to GABA and glycine, and antagonism by 4-aminopyridine, Prog. Clin. Biol. Res., 179 (1985) 281-287.
201
Elsevier BRES 15449
Some characteristics of baclofen-evoked responses of primary afferents in frog spinal cord A.L. Padjen and G.M. Mitsoglou Department of Pharmacology and Therapeutics, McGill University, Montreal, Quebec (Canada) (Accepted 10 October 1989)
Key words: 7-Aminobutyric acid; Baclofen; Frog spinal cord; Ba; 4-Aminopyridine; Primary afferent terminal; Baclofen antagonist
Baclofen has been shown to be a selective agonist for a subclass of GABA receptors (GABAB) in many regions of the vertebrate nervous system. On the intraspinal terminals of dorsal roots (DRT), it evokes a pure hyperpolarizing response. We have previously shown that the response of DRT to GABA and some of its analogs (e.g. kojic amine) in isolated frog spinal cord is dual in nature, consisting of a bicuculline-sensitive depolarizing component and a bicuculline-resistant hyperpolarizing component. Under the working hypothesis that the hyperpolarizing component of the GABA-evoked response is mediated by the activation of GABAB receptors, we have examined, using the sucrose gap technique, some characteristics of the response of DRT to baclofen. We have found that this response is stereospecific (L-baclofen being about 100 times more potent than D-baclofen), dependent on [K]o (response amplitude inversely related to [K]o), blocked by barium (0.5 mM causing a reduction of the response amplitude to 37% of control), and is not significantly affected by 4-aminopyridine, nor by inorganic calcium channel blockers (manganese, cobalt, cadmium). Some proposed GABA B antagonists (6-aminovaleric acid, 6-aminolaevulinic acid, phaclofen) are also rather ineffective at blocking it. These results are therefore consistent with the notion that the baclofen-evoked response of DRT is mediated by an increase in conductance to potassium ions. INTRODUCTION G A B A is considered to be the most widely distributed inhibitory n e u r o t r a n s m i t t e r in the v e r t e b r a t e CNS 27. A t the first sensory synapse in the spinal cord, it r e p o r t e d l y plays a role in presynaptic inhibition 18'29. Recently, convincing d a t a have surfaced in support of G A B A r e c e p t o r multiplicity in the v e r t e b r a t e nervous system 7. Two sites have b e e n established by binding studies and designated by B o w e r y and colleagues 6 as G A B A A and G A B A B. Presently, one of the m a j o r distinctions between these two classes of receptors is that the former is sensitive to the antagonistic action of bicuculline while the latter is bicuculline-resistant and selectively activated by baclofen, a chlorophenyl derivative of G A B A 7. Baclofen is also capable of depressing neurotransmission across the first sensory synapse by interacting at a presynaptic site 9'44. Consequently, it is possible that G A B A can p r o d u c e presynaptic inhibition through m o r e than one mechanism. A s previously d e m o n s t r a t e d in this l a b o r a t o r y , t h e responses to G A B A and o t h e r neutral amino acids (e.g. kojic amine, taurine) on the dorsal root terminals ( D R T ) of isolated frog spinal cord are dual in nature, consisting of b o t h a depolarizing c o m p o n e n t and a hyperpolarizing
c o m p o n e n t . T h e latter b e c o m e s m o r e evident at low agonist concentrations or in the presence of bicuculline 5' 35,37,39 Baclofen, on the o t h e r hand, evokes a pure hyperpolarizing response at this site. U n d e r the working hypothesis that all the h y p e r p o l a r izing effects are mediated by the same receptor ( G A B A u ) , we chose to further characterize the responses of D R T to baclofen because of its a p p a r e n t selectivity. This was done in two ways. Firstly, we e x a m i n e d the ionic r e q u i r e m e n t s to see if they were consistent with those of G A B A u - e v o k e d responses f o u n d elsewhere in the vertebrate nervous system. Secondly, we tested the capacity of p r o p o s e d G A B A A antagonists to block these responses. Preliminary results have b e e n c o m m u n i c a t e d 3°.
MATERIALS AND METHODS Experiments were done on isolated hegnisected spinal cord from leopard frog, Rana pipiens (both northern and southern varieties), kept in aquaria at 4 °C for periods of 1 week to several months. Details of the methods were described previously2'42. Briefly, frogs were cooled to an anesthetic state in crushed ice, and decapitated before a dorsal or ventral laminectomy was performed. The spinal cord with its attached roots was carefully removed to a dissecting dish, sagittally hemisected, and then placed in the central compartment (bath volume = 0.15 ml) of a sucrose gap chamber. The attached dorsal roots (9th and/or 10th segment) were led out by way
Correspondence: A.L. Padjen, Department of Pharmacology and Therapeutics, McGill University, 3655 Drummond Street, Montreal, Quebec H3G 1Y6, Canada. 0006-8993/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
202 of slits through the sucrose compartment into two separate pools of normal Ringer solution. All slits, as well as a closely fitted cover, were coated with White Petrolatum (USP) to provide a leak-proof separation of compartments. The central and sucrose compartments were continuously perfused with unoxygenated Ringer solution (flow rate 1.5-2.0 ml/min) and 244 mM sucrose (0.5-1.0 ml/min) respectively. Temperature (10-12 °C) was controlled by means of a Peltier device. Normal Ringer solution (pH 7.3) contained the following (in mM): sodium chloride, 115; potassium chloride, 2; calcium chloride, 2; HEPES (pH 7.3), 5; D-glucose, 10. Tetrodotoxin (0.2/tM) was present in all solutions superfusing the spinal cord to minimize indirect drugevoked responses. Drugs were applied by interchange of solutions superfusing the central compartment via a set of stopcocks. Drug solutions were made by diluting appropriate amounts of concentrated stock solution in normal Ringer solution and pH adjusted if necessary. All drugs were purchased from Sigma, St. Louis, MO, except for kojic amine (gift from Merck Frosst Laboratories, Kirkland, P.Q., Canada) and baclofen (gift from Ciba-Geigy, Basel, Switzerland). Electrical activity was recorded by monitoring the potential difference between two silver-silver chloride pellet electrodes making contact with the solution in the central (ground) and root compartments via salt bridges (4% agar Ringer). Permanent records were obtained from a Gould (Brush 2400) rectilinear pen recorder. Signals were filtered by a low-pass filter (0.5 s time constant) without distortion of drug responses. The observed potential changes represent an attenuated version of the average change in the membrane potential of the intramedullary part of a population of primary afferents. Under our recording conditions, an upward deflection of the pen represents depolarization. Results are expressed as the mean + S.E.M. wherever possible. RESULTS
Differential sensitivity of D R T to G A B A and baclofen Sucrose gap recordings from D R T (Figs. 2-6) reveal that baclofen consistently evokes a concentration-dependent hyperpolarizing response, unlike its chemical analogue G A B A . This hyperpolarization has a characteristically slow rise (time to peak about 4 rain) and is very slow to return to baseline, especially when it is brought about by higher concentrations of baclofen (up to 1.5 h at 0.1 mM). Responses to baclofen, at concentrations greater than 0.1 mM, often never fully recover, even after several hours of wash out. It is very likely that the cause of this slow recovery is the lack of an effective uptake system for the removal of baclofen 6. In addition, we found that D R T from different segments of the spinal cord are differentially sensitive to baclofen and G A B A (see Table I below). Baclofen-evoked responses have greater amplitudes on D R T of segment IX (DRT9) than
on D R T of segment X (DRT10) for the same agonist concentration, with the reverse being true for G A B A .
Stereospecificity of baclofen-evoked responses Unlike G A B A , baclofen exists in either the dextrorotatory or levorotatory structural configurations. The stereospecificity of the baclofen-evoked response of D R T is evident after comparing the ECs0 values from the concentration-response curves for the two stereoisomers (Fig. 1). The L-isomer, with an ECso of 0.01 mM, is about 100 times more potent than the D-form (see Fig. 3 for samples of typical recordings). The maximum amplitude of the hyperpolarizing response evoked by L-baclofen occurs at a concentration of 0.1 mM.
Sensitivity to changes in [K+]o The responses of D R T to several neutral amino acids (analogs of G A B A ) can exhibit a hyperpolarizing component under certain conditions, such as low agonist concentrations (i.e. lower than 0.1 mM) or in the presence of bicuculline 5'39. This is particularly obvious with some analogs, such as kojic amine (Fig. 2). However, even when a duality is not evident under bicuculline blockade, lowering [K]o can bring out such a component, as exemplified by the G A B A and taurineevoked responses in Fig. 2 (see also Fig. 3). Hyperpolarizing responses already visible in the presence of bicuculline (i.e. kojic amine and baclofen), are increased when [K]o is reduced (see also Figs. 3 and 5). The potentiation of the baclofen-evoked response depends upon the extent of [K]o reduction: in 0.5 mM [K]o, the response amplitude increases to 120% (N = 2) of control (Fig. 3), while in 0.1 mM [K]o, it approximately doubles (215% + 25%; N = 6; Figs. 2 and 5). Parallel increases in the slope of the responses (rise phase) are also evident,
1.6
o--o L-boclofen ~E
~
.--.
T
0.8
°
04 ~
1 / o
o.0 0.001
TABLE I
[ o[
D b o l co f e n
/ [ 0010
I O, 1 O0
I 10 0 0
Concentrotion (mM)
Response amplitude (m V)
DRT9 DRT10
O.5 mM GA BA
O.1 mM L-baclofen
1.86+0.10 (N=31) 2.93+0.19 (N=33)
-1.24+0.10 (N=8) -0.94+0.11 (N=8)
Fig. 1. Concentration-response curves of baclofen stereoisomers on dorsal root terminals of the 9th segment (DRT9). Note that L-baclofen is more potent than D-baclofen in hyperpolarizing DRT (mean _+ S.E.M. wherever possible, N = 1-8; the difference between the curves is significant at the 1% level, using a paired Student's t-test, for 0.1 mM agonist concentration only).
203 Control
0.02 mM BIC
A
(3ABA TAU 1 0.5 0.1K/BIC
1
KJA 0,1
GLU 0.5
L--BAC 0.1
3.5K
KJAO.2
GABAO.5 TAU1
5
Fig. 2. Effect of low [K]o on neutral amino acid-evoked responses in the presence of bicuculline. Abbreviations: BIC, bicucuUine; TAU, taurine; KJA, kojic amine; L-BAC,L-baclofen;0.1K, 0.1 mM potassium. On this and other figures: drug application is marked by a horizontal bar; depolarization upwards; calibrations: mV (vertical), min (horizontal). Bicuculline depresses depolarizing responses (see text); in this case, the kojic amine-evoked depolarization (not shown) was changed to a hyperpolarizing response. Note that decreasing [K]o brings out the hyperpolarizing components of responses to GABA and taurine and increases the hyperpolarizing responses to kojic amine and L-baclofen.
possibly reflecting a change in the driving force as a result of a new ionic gradient. In contrast, raising the level of [K]o depresses the baclofen-evoked response in a reversible manner (to 15% of control in 4 m M [K]o; N = 2; Fig. 3).
Effect of barium Barium has been shown to block a variety of K + currents T M . In our experiments, two concentrationdependent effects indicative of this channel-blocking ability are depolarization of the resting membrane potential of D R T and blockade (and/or reversal) of the K-evoked depolarizing responses (Figs. 4, 5) 31'32'39 . T h e responses to all the amino acids tested (i.e. G A B A , taurine, kojic amine, baclofen, glutamate) are depressed to varying degrees (Fig. 4) with baclofen exhibiting the
Ba 0.5
D.L,-BAC 0.2 0.5 m M BQ
Fig. 4. Effect of barium (0.5 mM) on the responses of DRT to amino acids and increased potassium. Abbreviations: GLU, glutamate; Ba, barium (arrow indicates start of perfusion). Note that the hyperpolarizing response to D,L-baclofen and the depolarizing response to increased [K]o (3.5 mM) are depressed more than the other responses by 0.5 mM barium, and thatthe taurine-evoked response becomes multiphasic. greatest sensitivity (amplitude reduced to 27% of control in the example illustrated; mean 37% _+ 7% of control, N = 4). The taurine-evoked response in particular exhibits a considerable depression with the emergence of a hyperpolarizing component sl,a2. Further increasing the barium concentration in the presence of low [K+]o (reflected in a much greater baseline depolarization) 32 causes a total blockade of the baclofen-evoked response, while the response to G A B A is only partially depressed (Fig. 5).
Effect of 4-aminopyridine (4-AP) 4-AP is another well-known blocker of potassium
Control Control
GABA~ 0.5
L-BAC 0.2
-- ~ D
GA-BA 0.5
K4"~
D.L.- BAC 0.1
Wosh
_z_ _
0.1K/5Bo m
Fig. 3. Effect of changes in [Klo on the responses of DRT to L-baclofen and GABA. Note that the amplitude of baclofen-evoked responses is inversely proportional to [K]o, and that L-baclofen is more potent than D-baclofen in hyperpolarizing DRT (see also Fig. 1).
Fig. 5. Effect of barium (5 mM) in low [K]o on the responses of DRT to amino acids and K. Note that the D,L-baclofen-evoked response is completely blocked and the K-evoked response is reversed while the response to GABA is partially depressed.
204 conductance 14'5°'51. On frog DRT, this agent proved to be ineffective at depressing the baclofen-evoked response in concentrations of 2-5 mM (N = 4; Fig. 6). It did, however, cause a small depolarization of the resting membrane potential (1.43 + 0.50 mV at 5 mM, N = 4), and it increased the amplitude of the GABA-evoked responses (160% of control + 10%, N --- 4) and the K-evoked response (not shown; see Padjen and Smith, 1981). It is interesting that the taurine-evoked response is depressed by 4-AP (42% of control + 7%, N = 4; not shown), in agreement with recent findings in hippocampal slices 53. In preliminary experiments, tetraethylammonium, another blocker of certain potassium currents 14'5°, only attenuated (by less than 40%) the baclofen-evoked response of D R T at a concentration of 10 mM.
Effect of inorganic calcium channel blockers There were 3 main reasons for testing the effects of agents known to block calcium conductance: (i) if baclofen-evoked responses are directly mediated by an increase in K conductance, then the channels involved may be calcium-dependent4; (ii) the responses of D R T to baclofen may be due to a transient inhibition of the chronic release of a depolarizing substance onto DRT, even in the presence of tetrodotoxin; (iii) baclofen effects have been explained as being due to a direct inactivation of a voltage-dependent calcium current 15'16. The 3 inorganic calcium channel blockers applied were manganese (2 mM; N = 2), cobalt (2-10 raM; N = 6), and cadmium (0.5 mM; N = 2). None of these agents had any significant effect on the baclofen-evoked responses. Both manganese and cobalt did however produce a depolarization of D R T and augmented the G A B A evoked responses while cadmium caused a small hyperpolarization of DRT, as previously reported 41.
DRT Control
.j
_ GABA 0.5
-~
4-A%
L-BAG01
5mM 4-AP
2
Effects of putative GABA B antagonists We tested several agents proposed to antagonize G A B A B effects. Delta-aminovaleric acid 33 (1 raM) produced a dual effect on the resting membrane potential and reduced the responses of D R T to 0.5 mM G A B A (54% of control + 4%, N = 4) but had little effect at reducing the amplitude of the baclofen-evoked response (i.e. < 20% depression, N = 2). 6-Aminolaevulinic acid 8 (1 mM) caused only a modest (i.e. < 30%, N = 2) attenuation of the baclofen-evoked response. Phaclofen has recently been suggested to be another possible baclofen antagonist 11'26. Its use in the rat hippocampus allowed the tentative identification of a synaptic inhibitory pathway mediated by a GABAB receptor 17. Our preliminary data on the responses of D R T to baclofen show only a modest antagonistic effect (i.e. < 30% depression) at high concentrations of phaclofen (i.e. > 0.5 mM). In conclusion, our studies so far have not identified any substance capable of antagonizing the baclofenevoked responses of D R T with a potency or selectivity useful for pharmacological exploration. DISCUSSION On the basis of a number of physiological and binding studies, a separate site of G A B A action has been described (GABAB 6'7) on presynaptic terminals in both the peripheral and central nervous system of vertebrates. Although CNS terminals are rarely accessible for direct analysis, sucrose gap recording on dorsal roots permits the measurement of the polarization of a whole population of DRT. Using this technique, we have identified hyperpolarizing responses of D R T to G A B A , taurine and analogs 5'3°'39 that have many attributes of a GABAB site, in agreement with some binding studies on the distribution of G A B A B sites in the spinal cord 46. The present study describes some characteristics of these responses, particularly those evoked by baclofen as the most potent and selective agonist at this site. The results demonstrate that the baclofen-evoked responses of D R T are: (i) tetrodotoxin and manganese resistant, and therefore very likely evoked directly on D R T membranes; (ii) stereospecific; (iii) insensitive to bicuculline, and largely unaffected by several putative G A B A B antagonists; (iv) dependent on [K]o; (v) blocked by barium; (vi) unaffected by 4-AP or inorganic calcium channel blockers.
5
\
Fig. 6. Effect of 4-AP on the responses of DRT to L-baclofen and GABA. Note the lack of effect of 4-AP on the amplitude of the response evoked by L-baclofen, while that evoked by GABA is increased.
Comparison with baclofen-evoked responses of other cells Baclofen has a hyperpolarizing effect on amphibian motoneurons (unpublished observation). In addition, baclofen has been shown to evoke hyperpolarizing responses in other cells of the vertebrate CNS, particu-
205 larly hippocampal pyramidal cells. The characteristics of these responses appear to be very similar to those of baclofen-evoked responses in our preparation. For example, the ratio of the EC50 for the levorotatory versus the dextrorotatory isomers of baclofen on DRT (i.e. approximately 100:1) is close to that obtained in hippocampus 34. Similarly, the observed dependence of our baclofen-evoked responses on [K]o and their sensitivity to barium ions is in agreement with intracellular studies of hippocampal pyramidal cells22-24"34. The observed effects of such manipulations are unlikely to be due solely to the associated changes in the membrane polarization level since these would then be contrary to what is expected of hyperpolarizing responses. It has been reported 22-24'34 that baclofen-evoked responses are voltage-sensitive in the depolarizing direction but this factor could not totally account for the depressant action of barium. In contrast to one report on hippocampal neurons 23, we did not find the baclofen-evoked responses to be sensitive to 4-AP, even at relatively higher concentrations. The possibility therefore exists that different potassium conductances are involved. Baclofen has been also reported to reduce a calcium current in cultured D R G without any action on potassium conductance and/or resting membrane potential 1°' 13,15,16,47,48 This type of baclofen action seems to be distinct from the one described here (see later). Our experimental data are thus consistent with the hypothesis that one type of baclofen-evoked response in vertebrate CNS reflects the direct activation of a potassium conductance. This response appears to originate predominantly on thin-diameter fibers 10'37"39'46. The differential distribution of these fibers 52 may then explain the segmental differences in baclofen potency demonstrated in this study. In addition to baclofen and GABA, other agents, such as kojic amine and taurine, have a similar hyperpolarizing action. The lack of a potent and specific antagonist of the hyperpolarizing component (see below) does not allow any possible distinct pharmacology of it to be done and we could assume, for the moment, that all of these agents might act at the same site to produce the observed hyperpolarization. It is of interest that the depolarizing component of the responses of DRT to taurine and kojic amine (but not GABA) is sensitive to 6-aminomethyl3-methyl-4H-1,2,4-benzothiadiazine-l,l-dioxide4°, thus indicating that these compounds activate a site which G A B A does not.
Physiological significance The possible physiological role of baclofen-evoked responses, including those on the presynaptic terminals of primary afferents, derives from the assumption that
baclofen is an analog of endogenous G A B A at the G A B A B receptors. By virtue of its electrophysiological effect, baclofen could thus be assumed to inhibit the release of a transmitter from D R T in the frog spinal cord 9"21, in general agreement with its main pharmacological mechanism of action 7'45. This may then explain the clinical efficacy of baclofen in the treatment of spinal spasticity, which occurs at an intrathecal concentration range 43 that is similar to that used in this study. Synaptically evoked hyperpolarization of presynaptic terminals (PAH) has been associated with presynaptic inhibition in invertebrate preparations 3'2°'25, but has not been identified in the vertebrate spinal cord, where the depolarization of primary afferents (PAD) is normally associated with a presynaptic inhibitory mechanism 36. Other evidence indicates that PAH in invertebrates may be due to the depression of a tonic depolarization of DRT 36. Even though this would suggest that PAH is not a synaptic entity in vertebrate spinal cord, it is possible that both PAD and PAH occur simultaneously on DRT but that the more dominant depolarizing response masks the presence of PAH. Although the depolarizing responses to both G A B A and PAD could be blocked by bicuculline or picrotoxin 2, the associated convulsions would prevent any possible recognition of PAH. The dual nature of the GABA-evoked responses of DRT 5,39 suggests that G A B A could be the physiological mediator of both PAD and (possibly) PAH via interaction at G A B A A and GABAB receptors, respectively. Although these two receptors seem to be different, the activation of both could lead to presynaptic inhibition by a similar mechanism, i.e. a shunt of current that keeps the membrane potential from substantial depolarization 18'38 thus preventing the opening of voltage-dependent calcium channels. Direct participation of a calciumdependent potassium conductance in the baclofenevoked (GABAB) responses is unlikely because of the ineffectiveness of inorganic calcium channel blockers, as shown in this study (also ref. 22). An additional mode of action of baclofen, i.e. a direct depression of calcium fluxes at the primary afferent terminals, has been suggested on the basis of observations at the somata of the s a m e cells 1°,15,16,47. At DRT, however, this mechanism is not amenable to direct analysis, and the present study is therefore not able to address this issue. It is however entirely possible that more than one mechanism is involved in presynaptic inhibition mediated via G A B A a receptors. The issue of whether the hyperpolarizing responses are physiologically important in the primary afferents of vertebrates awaits two developments: (i) identification of a synaptic event that would be its physiological equivalent, i.e. the PAH; (ii) discovery of potent competitive
206 antagonists of G A B A B receptors that produce concur-
sponses (e.g. baclofen). Such antagonists would also provide i n d e p e n d e n t evidence as to whether or not the
In conclusion, our study suggests that the baclofenevoked hyperpolarizing response of D R T is mediated by potassium fluxes and that this may provide a mechanism for presynaptic inhibition mediated by G A B A B receptors. These responses may therefore serve as a model
hyperpolarizing responses elicited by other neutral amino acids (e.g. G A B A , taurine, kojic amine) 3°'39 are medi-
system for future studies of G A B A B pharmacology in the vertebrate CNS.
rent depression of presynaptic inhibition (or P A H ) and responses evoked by agonists of hyperpolarizing re-
ated by the same receptor. Unfortunately, the present results, as well as similar studies on isolated rat spinal cord 19, suggest that the discovery of a potent and specific antagonist is still in the future. REFERENCES 1 Arhem, P., Effects of rubidium, caesium, strontium, barium, and lanthanum on ionic currents in myelinated nerve fibers from Xenopus laevis, Acta Physiol. Scand., 108 (1980) 7-16. 2 Barker, J.L., Nicoll, R.A, and Padjen, A.L., Studies on convulsants in the isolated frog spinal cord. I. Antagonism of amino acid responses, J. Physiol. (Lond.), 245 (1975) 521-536. 3 Baxter, D.A. and Bittner, G.D., Intracellular recordings from crustacean motor axons during presynaptic inhibition, Brain Research, 223 (1981) 422-428. 4 Blaxter, T.J., Carlen, P.L., Davies, M.E and Kujtan, P.W., Gamma-aminobutyric acid hyperpolarizes rat ~hippocampal pyramidal cells through a calcium-dependent potassium conductance, J. Physiol, (Lond.), 373 (1986) 181-194. 5 Bourne, G.W. and Padjen, A.L., Kojic amine: a GABA agonist with dual action, Soc. Neurosci. Abstr., 6 (1980) 148. 6 Bowery, N.G., Hill, D,R. and Hudson, A.L., Characteristics of GABAB receptor binding sites on rat whole brain synaptic membranes, Br. J. PharmacoL, 73 (1983) 191-206. 7 Bowery, N.G., Price, G.W., Hudson, A.L., Hill, D.R., Wilkin, G.P. and Turnbull, M.J., GABA receptor multiplicity: visualization of different receptor types in the mammalian CNS, Neuropharmacology, 23 (1984) 219-231. 8 Brennan, M.J.W. and Cantrill, R.C., b-Aminolaevulinic acid is a potent agonist for GABA autoreceptors, Nature (Lond.), 280 (1979) 514-515. 9 Davidoff, R.A. and Sears, E.S., The effects of lioresal on synaptic activity in the isolated spinal cord, Neurology, 24 (1974) 957-963. 10 Desarmenien, M., Feltz, P., Occhipinti, G., Santangelo, E and Schlichter, R., Coexistence of GABAA and GABAB receptors on A-delta and C primary afferents, Br. J. Pharmacol., 81 (1984) 327-333. 11 Dickenson, H.W., Allan, R.D., Ong, J. and Johnston, G.A.R., GABAB receptor antagonist and GABAA receptor agonist properties of a ~,-aminovalericacid derivative, Z-5-aminopent2-enoic acid, Neurosci. Lett., 86 (1988) 351-355. 12 Dodd, J. and Horn, J.P., Muscarinic inhibition of sympathetic C neurones in the bullfrog, Br. J. Pharmacol., 74 (1981) 579-585. 13 Dolphin, A.C. and Scott, R.H., Inhibition of calcium currents in cultured rat dorsal root ganglion neurones by (-)-baclofen, Br. J. Pharmacol., 88 (1986) 213-220. 14 Dubois, J.M., Potassium currents in the frog node of Ranvier, Progr. Biophys. Mol. BioL, 42 (1983) 120. 15 Dunlap, K. and Fischbach, G.D., Neurotransmitters decrease the calcium component of sensory neurone action potentials, Nature (Lond.), 276 (1978) 837-839. 16 Dunlap, K. and Fischbach, G.D., Neurotransmitters decrease the calcium conductance activated by depolarization of embryonic chick sensory neurones, J. Physiol. (Lond.), 317 (1981) 519-553. 17 Dutar, P. and Nicoll, R.A., A physiological role for GABAB receptors in the central nervous system, Nature (Lond.), 332
Acknowledgements. This work was supported by the Medical Research Council of Canada. G.M.M. held a studentship from the Fonds de la Recherche en Sant6 du QuEbec. The authors wish to thank Alain SEvigny for some technical assistance.
(1988) 156-158. 18 Eccles, J.C., Presynaptic inhibition. In The Physiology of Synapses, Springer, Berlin, 1964, pp, 220-238. 19 Evans, R.H., Pharmacology of amino acid receptors on vertebrate primary afferent nerve fibres, Gen. Pharmacol., 17 (1986) 5-11. 20 Fuchs, P.A., Intracellular analysis of presynaptic inhibition in the crayfish claw opener, Soc. Neurosci. Abstr., 3 (1977) 176. 21 Fukuda, H., Kudo, Y. and Ono, H., Effects of beta-(parachlorophenyl)-GABA (baclofen) on spinal synaptic activity, Eur. J. Pharmacol., 44 (1987) 17-24. 22 G/ihwiler, B.H. and Brown, D.A., GABAa-receptor-activated K+ current in voltage-clamped CA3 pyramidal cells in hippocampal cultures, Proc. Natl. Acad. Sci. U.S.A., 82 (1985) 1558-1562. 23 Inoue, M., Matsuo, T. and Ogata, N., Baclofen activates voltage-dependent and 4-aminopyridine sensitive K+ conductance in guinea-pig hippocampal pyramidal cells maintained in vitro, Br. J. Pharmacol., 84 (1985) 833-841. 24 Inoue, M., Matsuo, T. and Ogata, N,, Possible involvement of K+-conductance in the action of GABA in the guinea-pig hippocampus, Br. J. PharmacoL, 86 (1985) 515-524. 25 Kawai, N. and Niwa, A., Hyperpolarization of the excitatory nerve terminals by inhibitory nerve stimulation in lobster, Brain Research, 137 (1977) 365-368. 26 Kerr, D.I.B., Ong, J., Prager, R.H., Gynther, B.D. and Curtis, D.R., Phaclofen: a peripheral and central baclofen antagonist, Brain Research, 405 (1987) 150-154. 27 Krnjevic, K., Chemical nature of synaptic transmission in vertebrates, Physiol. Rev., 54 (1974) 418-540. 28 Krnjevic, K., Pumain, R. and Renaud, L., Effects of barium and tetraethylammonium on cortical neurones, J. Physiol. (Lond.), 215 (1971) 223-245. 29 Levy, R.A., The role of GABA in primary afferent depolarization, Progr. Neurobiol., 9 (1977) 211-267. 30 Mitsoglou, G.M. and Padjen, A.L., Hyperpolarizing responses to GABA and analogs on primary afferent fibres, Soc. Neurosci. Abstr., 11 (1985) 282. 31 Mitsoglou, G.M. and Padjen, A.L., Contributionof [K+]o to the neutral amino acid responses on primary afferents, Proc. Can. Fed. Biol. Soc., 28 (1985) 169. 32 Mitsoglou, G.M. and Padjen, A.L., Differential effect of barium on responses of frog primary afferents to neutral amino acids, submitted. 33 Muhyaddin, M., Roberts, P.J. and Woodruff, G.N., Presynaptic gamma-aminobutyric acid receptors in the rat anococcygeus muscle and their antagonism by 5-aminovaleric acid, Br. J. Pharmacol., 77 (1982) 163-168. 34 Newberry, N.R. and Nicoll, R.A, Comparison of the action of baclofen with GABA on rat hippocampal pyramidal cells in vitro, J. Physiol. (Lond.), 360 (1985) 161-185. 35 Nicoll, R.A., The interaction of porphyrin precursors with GABA receptors in the isolated frog spinal cord, Life Sci., 19 (1976) 521-526.
207 36 Nicoll, R.A. and Alger, B.E., Presynaptic inhibition: transmitter and ionic mechanisms, Int. Rev. Neurobiol., 21 (1979) 217-258. 37 Padjen, A.L. and Hashiguchi, T., Intraceilular analysis of GABA-evoked responses of frog primary afferent axons, Soc. Neurosci. Abstr., 8 (1982) 577. 38 Padjen, A.L. and Hashiguchi, T., Primary afferent depolarization in frog spinal cord is associated with an increase in membrane conductance, Can. J. Physiol. Pharmacol., 61 (1983) 626-631. 39 Padjen, A.L., Hashiguchi, T., Mitsoglou, G.M. and Bourne, G.W., Dual effect of GABA and kojic amine on primary afferents: electrophysiological evidence for the activation of GABAn receptors, 1988. 40 Padjen, A.L., Mitsoglou, G.M. and Hassessian, H., Further evidence in support of taurine as a mediator of synaptic transmission in the frog spinal cord, Brain Research, 488 (1989) 288-296. 41 Padjen, A.L. and Smith, P.A., Possible role of divalent cations in amino acid responses of frog spinal cord. In EV. De Feudis and E Mandel (Eds.), Amino Acid Neurotransmitters, Raven Press, New York, 1981, pp. 271-280. 42 Padjen, A.L. and Smith, P.A., The role of the electrogenic sodium pump in the glutamate afterhyperpolarization of frog spinal cord, J. Physiol. (Lond.), 336 (1983) 433-451. 43 Penn, R.D., Savoy, S.M., Corcos, D., Latash, M., Gottlieb, G., Parke, B. and Kroin, J.S., Intrathecal baclofen for severe spinal spasticity, N. Engl. J. Med., 320 (1989) 1517-1521. 44 Pierau, E-K., Matheson, G.K. and Wuster, R.D., Presynaptic action of fl-(4-chlorophenyl)-GABA, Exp. Neurol., 48 (1975) 343-351. 45 Potashner, S.J., Baclofen: effects on amino acid release and
46
47
48
49
50 51 52 53
metabolism in slices of guinea pig cerebral cortex, J. Neurochem., 32 (1979) 103-109. Price, G.W., Wilkin, G.E, Turnbull, M.J. and Bowery, N.G., Are baclofen-sensitive GABAB receptors present on primary afferent terminals of the spinal cord?, Nature (Lond.), 307 (1984) 71-74. Robertson, B. and Taylor, R., Effects of gamma-aminobutyric acid and (-)-baclofen on calcium and potassium currents in cat dorsal root ganglion neurones in vitro, Br. J. Pharmacol., 89 (1986) 661-672. Schlichter, R., Demeneix, B.A., Desarmenien, M., Desaulles, E., Loeffler, J.P. and Feltz, P., Properties of the GABA receptors located on spinal primary afferent neurones and hypophyseal neuroendocrine cells of the rat, Neurosci. Lett., 47 (1984) 257-263. 8tanden, N.B. and Stanfield, ER., A potential- and timedependent blockade of inward rectification in frog skeletal muscle fibres by barium and strontium ions, J. Physiol. (Lond.), 280 (1978) 169-191. Thompson, S.H., Three pharmacologically distinct potassium channels in molluskan neurones, J. Physiol. (Lond.), 265 (1977) 465-488. Ulbricht, W. and Wagner, H.H., Block of potassium channels of the nodal membrane by 4-aminopyridine and its partial removal on depolarization, Pfliigers Arch., 367 (1976) 77-87. Wilhelm, G.B. and CoggeshaU, R.E., An electron microscopic analysis of the dorsal root in the frog, J. Comp. Neurol., 196 (1981) 421-429. Zeise, M., Taurine on hippocampal slices: comparison to GABA and glycine, and antagonism by 4-aminopyridine, Prog. Clin. Biol. Res., 179 (1985) 281-287.
