Ultrastructural evidence for GABA-mediated disinhibitory circuits in the spinal cord of the cat

Neuroscience Letters, 138 (1992) 183-187

183

Elsevier Scientific Publishers Ireland Ltd.

NSL 08555

Ultrastructural evidence for GABA-mediated disinhibitory circuits in the spinal cord of the cat H a n t a o Liu a, Ida J. Llewellyn-Smithd, Paul Pilowskyd and Allan I. Basbaum a-c Departments of "Anatomy, bphysiology and CKeck Centerfor Integrative Neurosciences, University of CaliJornia San Francisco, San Francisco, CA 94143 (USA), and dCentrefor Neuroscience and Department of Medicine, Bedford Park, S.A. (Australia) (Received 23 September 1991; Revised version received 2 December 1991; Accepted 10 January 1992)

Key words: GABA; Enkephalin; Analgesia; Dorsal horn The synaptic relationships between y-aminobutyric acid (GABA)-immunoreactive and enkephalin-immunoreactive profiles in the cat spinal cord were examined using combined pre-embedding immunoperoxidase and post-embedding immunogold electron microscopic immunocytochemistry. Although colchicine was not used, enkephalin-immunoreactive somata and dendrites were detected in regions associated with nociceptive transmission, including laminae I, II, V and X. In each of these laminae, many GABA-immunoreactive terminals were found presynaptic to enkephalinimmunoreactive cell bodies and dendrites. We propose that disinhibition of opioid-containing neurons may be a common feature of pain-related circuits in the cat spinal cord.

Intrathecal injection of opiates or opioid peptides produces a profound analgesia [25]. This is presumed to result from an action at spinal cord opioid receptors normally innervated by enkephalin and possibly dynorphinergic neurons [3]. Since enkephalin and other opioids hyperpolarize neurons [19] it has been proposed that the analgesic action of these agents results from a postsynaptic inhibition of nociresponsive neurons of the spinal cord. Consistent with this hypothesis, Ruda et al. [21] demonstrated that some enkephalin terminals are presynaptic to projection neurons of laminae I and V of the dorsal horn, two major loci containing nociresponsive neurons. Intrathecal administration of baclofen, an analogue of the inhibitory neurotransmitter y-aminobutyric acid (GABA) also produces analgesia [24], and baclofen, like opioids, profoundly inhibits the firing of spinal cord nociresponsive neurons [8]. What is not established is the extent to which GABAergic and opioidergic inhibitory systems interact in the spinal cord. Therefore, to evaluate the interaction between GABAergic and enkephalinergic antinociceptive circuitry in the spinal cord, we have initiated an electron microscopic double labelling study of the synaptic interactions made by enkephalin- and GABA-immunoreactive neurons. We report that in Correspondence: H. Liu, Department of Anatomy, UCSF, Box 0452, San Francisco, CA 94143, USA. Fax: ( 1) (415) 476-4845.

laminae I, IIo, V of the dorsal horn and in lamina X, another area of the spinal cord involved in nociceptive processing [10], GABA-immunoreactive terminals are commonly located presynaptic to enkephalin-immunoreactive neurons. These data provide support for the existence of a novel mechanism of antinociceptive control, namely a disinhibition of opioidergic mechanisms via a GABAergic inhibition of enkephalinergic neurons. Cats were deeply anesthetized with pentobarbital and perfused through the ascending aorta with a fixative solution containing 2.5% glutaraldehyde and 0.5% formaldehyde in a 0.1 M phosphate buffer, pH 7.4. The cervical spinal cord (C6-C~) was removed, divided transversely into half segments, and postfixed in the same solution for 3 h. Details of the pre-embedding immunocytochemical protocol have been described previously [12,14]. Briefly, 70-/tm-thick coronal vibratome sections were collected and immediately incubated in 50% ethanol (in distilled water) for 30 min, to improve antibody penetration. For immunocytochemical demonstration of enkephalin, the sections were blocked with 10% horse serum (NHS) and then incubated in a mouse monoclonal antiserum directed against methionine-enkephalin (Seralab) at a dilution of 1:125,000 for 72 h at room temperature. By titration, this dilution of antiserum gave the maximum number of immunoreactive fibers with the minimum level of non-specific background staining. After several washes, the sections were

184

Fig. 1. This electron micrograph illustrates a GABA-immunoreactiveterminal (g) that makes symmetrical synaptic contacts (arrowheads) with an enkephalin-immunoreactive cell body (Cb) in lamina I. Arrows point to the horseradish peroxidase immunoreaction product in the cell body. The GABA-immunoreactive terminal also contacts a small enkepha[in-immunoreactive dendrite (double arrowheads). An unlabelled axon terminal (a) that is presynaptic to an unlabelled dendrite makes contact with the GABA-immunoreactiveterminal and with the small enkephalin-immunoreactive dendrite. Bar - 0.5 #m.

incubated in biotinylated antimouse IgG (Sigma) for 24 h, followed by a 1:1500 dilution of ExtrAvidin-HRP (Sigma) for 24 h. The sections were washed 3 × 30 rain in Tris-phosphate-buffered saline after each incubation. A nickel-enhanced glucose oxidase protocol with diaminobenzidine was used to localize the H R P [13]. After several washes in phosphate buffer, the sections were osmicated (0.5% OsO4) for 1 h, en bloc stained in 2% aqueous uranyl acetate and then dehydrated and flat embedded in Durcupan. Areas containing either laminae l and II, lateral lamina V, or lamina X, were trimmed and mounted on blocks for sectioning. Ultrathin sections were collected onto nickel grids and then immunostained for G A B A [6]. Briefly, after several washes in phosphate buffer containing Triton X- 100, the sections were blocked in 3% normal goat serum and then incubated overnight in a rabbit anti-GABA antiserum (IncStar) diluted 1:1500. After washing, the sections were incubated for one hour in a 1:25 dilution of 10 nM colloidal gold-labelled goat antirabbit lgG (Jannsen Life Sciences; Ted Pella, Inc.). Sections were then stained with lead citrate. Adjacent control sections were incubated in primary antiserum preabsorbed with a G A B A - B S A conjugate (100 #g/ml [9]). This protocol eliminated staining.

Based on counts of gold particles in synaptic profiles from the same grid square, we identified two populations of terminals. One population contained from 24 to 33 gold particles per terminal; the second contained from 4 to 8 gold particles per terminal. We considered the former to be GABA-immunoreactive. As described previously, GABA-immunoreactive terminals were readily detected in all regions of the gray matter examined [4, 5, 7, 17, 18]. These generally contained clear, round and pleomorphic vesicles (Figs. 1 3). The gold particles overlay vesicles and mitochondria (Figs. 1 and 2). Although colchicine was not used, we detected many enkephalinimmunoreactive cell bodies and dendrites as well as a dense concentration of labelled terminals in laminae I and outer II of the superficial dorsal horn. The lateral part of lamina V contained a moderate number of labelled terminals and isolated immunoreactive cell bodies and dendrites, albeit fewer than in the superficial laminae. In lamina X, around the central canal, large numbers of immunoreactive terminals and a few labelled cell bodies were found. Since enkephalin-immunoreactive cell bodies and dendrites were present, it was possible to evaluate the relationship between GABA-immunoreactive terminals

185

Fig. 2. This electronmicrograph illustrates probable convergenceof a GABA-immunoreactiveterminal (g) and an unlabelled axon terminal (a) upon an enkephalin-immunoreactiveproximal dendrite (den) in lamina V. Arrows point to the horseradish peroxidase immunoreactionproduct in the dendrite. The GABA-immunoreactiveterminal makes an asymmetrical synaptic contact (arrowhead) with the dendrite. No synaptic specialization between the unlabelled terminal and the dendrite can be discerned. Bar = 0.5/.tm.

and the enkephalin neurons. In fact, we found numerous examples in which GABA-immunoreactive terminals were located presynaptic to enkephalin-immunoreactive cell bodies, and to proximal and distal dendrites (Figs. 1 3). The synaptic specialization was typically symmetrical (Figs. 1 and 3), however, an occasional asymmetric specialization was observed (Fig. 2). These synaptic arrangements were found in all of the regions examined. Often the postsynaptic enkephalin dendrite was also postsynaptic to unlabelled axon terminals (Figs. 1 and 3). In several cases, the latter resembled primary afferent terminals. Based on a variety of anatomical and biochemical studies it has been proposed that the analgesic action of intrathecally administered opioids results, at least in part, from a direct postsynaptic inhibition of nociresponsive projection neurons [3]. Thus enkephalin-immunoreactive terminals are located presynaptic to [21] and opioids hyperpolarize spinal dorsal horn nociresponsive projection neurons [23]. A presynaptic regulation of the release of transmitters from nociceptive primary af-

ferents has also been demonstrated [11]. Similar mechanisms have been reported for the antinociceptive effect of GABA, in particular by baclofen, the GABAB ligand. The presence of GABA B binding sites on small diameter primary afferents [20] and of immunoreactive GABA [4, 5, 7, 18], or glutamic acid decarboxylase terminals [1, 4] presynaptic to primary afferents and to dendrites, is of course, consistent with a both pre- and postsynaptic GABAergic control of nociceptive neurons [8]. On the other hand, since both GABA and opioids inhibit the output of nociresponsive projection neurons, it was surprising to find that GABA-immunoreactive terminals were commonly presynaptic to enkephalin cell bodies and dendrites, in several nocirecipient regions of the dorsal horn. This indicates that the enkephalin regulatory mechanism is under GABAergic control. Any condition which increases GABAergic tone would, therefore, have a dual effect, increased inhibition of the projection neurons and facilitation, or even excitation, of the projection neuron, via a disinhibition of the opioid regulation. The precise level at which the nociceptive

186

Fig. 3. This electron micrograph illustrates convergenceof a GABA-immunoreactiveterminal (g) and an unlabelled terminal (a) upon a distal enkephalin-immunoreactivedendrite in lamina Iio. The arrow points to the horseradish peroxidaseimmunoreactionproduct in the dendrite. Bar 0.5 pm.

threshold is set would, of course, depend on the tonic levels of activity in the opioid and GABAergic terminals. Since colchicine was not used in this study, the number of enkephalin-immunoreactive cells was small. It is, therefore, difficult to provide meaningful figures of the prevalence of these GABA-enkephalin relationships. Nevertheless we, examined a total of 100 examples in lamina I and outer II, in which a GABA-immunoreactive terminal contacted an enkephalin-immunoreactive profile (cell body, dendrite, or terminal). In 50 of these cases, the postsynaptic element was either a cell body or dendrite and in many of those examples, a synaptic contact was present. Based on this small sample we believe that this is, in fact, a common synaptic relationship. Other studies emphasized the interactions between opioid and GABAergic neurons. For example, Lovick and Wotstencroft [15, 16] reported that iontophoresis of the GABAA antagonist, bicuculline in the dorsal horn, blocked the increases in primary afferent terminal excitability produced by electrical stimulation in the medullary nucleus raphe magnus and that this effect was, itself blocked by iontophoresis of enkephalin. These studies indicate opioids can regulate the release of GABA from spinal interneuons. In vitro studies by Suzue and Jessel [22] similarly demonstrated that enkephalin inhibits dorsal root potentials, which are likely to be GABAmediated. Taken together with the present results, these

data suggest that the interactions between GABAergic and opioidergic neurons are not only common, but may be bidirectional. What is not known is the identity of the receptor through which GABA inhibits the enkephalin neuron. As described above, most studies point to GABAB mechanisms in the analgesic actions of GABA. If the GABAergic inhibition of the enkephalin neuron is mediated via the A-type receptor, it may be possible to pharmacologically dissociate the enkephalin neuron inhibition from the antinociceptive action of GABA. Specifically, a precise combination of baclofen and a GABA A antagonist may produce an analgesia more powerful than that produced by baclofen alone. Importantly, since intrathecal injection of either GABA or opioid agonists bypasses much of the circuitry through which the nociceptive projection neuron is regulated, the contribution of the GABAergic regulation of the enkephalin neuron would be missed. By contrast, analgesia produced by electrical stimulation or microinjection of drugs at supraspinal sites, for example, the periaqueductal grey [2], might provide a more 'natural' activation of the circuitry in the spinal cord. By appropriate use of antagonists at the level of the spinal cord it may be possible to identify the conditions under which these disinhibitory circuits come into play.

187

These studies were approved by the Flinders University Animal Ethics Review Subcommittee and was supported by grants from the NIH: NS14627 and 21445 and from the National Health and Medical Research Council of Australia, the National Heart Foundation of Australia and the National SIDS Council of Australia. We thank Adrian Wright and Margaret McLaren for their expert technical assistance. 1 Barber, R.P., Vaughn, J.E., Saito, K., McLaughlin, B.J. and Roberts, E., GABAergic terminals are presynaptic to primary afferent terminals in the substantia gelatinosa of the rat spinal cord, Brain Res., 141 (1978) 35 55. 2 Basbaum, A.I. and Fields, H.L., Endogenous pain control systems: Brainstem spinal pathways and endorphin circuitry, Annu. Rev. Neurosci., 7 (1984) 309-338. 3 Basbaum, A.I., Functional analysis of the cytochemistry of the spinal dorsal horn, Adv. Pain Res. Ther., 9 (1985) 149-175. 4 Basbaum, A.I., Glazer, E.J. and Oertel, W., Immunoreactive glutamic acid decarboxylase in the trigeminal nucleus caudalis of the cat: a light- and electron-microscopic analysis~ Somatosens. Res., 4 (1986) 77 94. 5 Carlton, S.M. and Hayes, E.S., Light microscopic and ultrastructural analysis of GABA-immunoreactive profiles in the monkey spinal cord, J. Comp. Neurol., 300 (1990) 162-182. 6 de Zeeuw, C.I., Holstege, J.C., Clakoen, F., Ruigrok, T.J.H. and Voogd, A., A new combination of WGA-HRP anterograde tracing and GABA immunocytochemistry applied to afferents of the cat inferior olive at the ultrastructural level, Brain Res., 447 (1988) 369375. 7 Difiglia, M., Aronin, N. and Leeman, S.E., Ultrastructural localization of immunoreactive neurotensin in the monkey superficial dorsal horn, J. Comp. Neurol., 225 (1984) 1-12. 8 Henry, J.L., Pharmacological studies on the prolonged depressant effects of baclofen on lumbar dorsal horn units in the cat, Neuroo pharmacology, 21 (1982) 1085-1093. 9 Hepler, J.R., Toomin, C.S., McCarth, K.D., Conti, F.~ Battaglia, G. and Rustioni, A., Characterization of antisera to glutamate and aspartate, J. Histochem. Cytochem., 36 (1988) 13 22. 10 Honda, C.H. and PerL E.R., Functional and morphological features of neurons in the midline region of the caudal spinal cord of the cat, Brain Res., 340 (1985) 285-295. 11 Jessell, T.M. and Iversen, L.L., Opiate analgesics inhibit substance P release from rat trigeminal nucleus, Nature, 268 (1977) 549-551.

12 LlewellynoSmith, I.J., Minson, J.B., Morilak, D.A., Oliver, J.R. and Chalmers, J.E, Neuropeptide Y-immunoreactive synapses in the intermediolateral cell column of rat and rabbit spinal cord, Neurosci. Lett., 108 (1990) 243 248. 13 Llewellyn-Smith, I.J., Minson, J.B. and Pilowsky, P.M., Retrograde tracing with cholera toxin B-gold or immunocytochemically-detected cholera toxin B in the central nervous system. In P.M. Conn (Ed.), Academic Press, New York, in press. 14 Llewellyn-Smith, I.J., Phend, K.D., Minson, J.B., Pilowsky, P.M. and Chalmers, J.P., Glutamate immunoreactive synapses on retrogradely labelled sympathetic preganglionic neurons in rat thoracic spinal cord, Brain Res., in press. 15 Lovick, T.A., Enkephalin blocks primary afferent depolarization of A~ tooth pulp afferents evoked by stimulation in nucleus raphe magnus in the decerebrate cat, Neurosci. Lett., 37 (1983) 273-278. 16 Lovick, T.A. and Wolstencroft, J.H., Inhibition from nucleus raphe magnus of tooth pulp responses in medial reticular neurones of the cat can be antagonized by bicuculline, Neurosci. Lett., 19 (1980) 325 330. 17 Magoul, R., Onteniente, B., Geffard, M. and Calas, A., Anatomical distribution and ultrastructural organization of the gabaergic system in the rat spinal cord. An immunocytochemical study using anti-GABA antibodies, Neuroscience, 20 (1987) 1001-1009. 18 Matthews, M.A., McDonald, G.K. and Hernandez, T.V., GABA distribution in a pain-modulating zone of trigeminal subnucleus interpolaris, Somatosens. Res., 5 (1988) 205 217. 19 Nicoll, R.A., Alger, B.E. and Jahr, C., Enkephalin blocks inhibitory pathways in the vertebrate CNS, Nature, 287 (1980) 22-25. 20 Price, G.W., Kelly, J.S. and Bowery, N.G., The location of GABAb receptor binding sites in mammalian spinal cord, Synapse, 1 (1987) 530 538. 21 Ruda, M.A., Coffield, J. and Dubner, R., Demonstration of postsynaptic opioid modulation of thalamic projection neurons by the combined techniques of retrograde horseradish peroxidase and enkephalin immunocytochemistry, J. Neurosci., 4 (1984) 2117-2132. 22 Suzue, T. and Jessell, T.M., Opiate analgesics and endorphins inhibit rat dorsal root potential in vitro, Neurosci. Lett., 16 (1980) 161-166. 23 Willcockson, W.S., Chung, J.M., Hori, Y., Lee, K.H. and Willis, W.D., Effects of iontophoretically released amino acids and amines on primate spinothalamic tract cells, J. Neurosci., 4 (1984) 732-740. 24 Wilson, P.R. and Yaksh, T.L., Baclofen is antinociceptive in the spinal intrathecal space of animals, Eur. J. Pharmacol., 51 (1978) 323-330. 25 Yaksh, T.L. and Rudy, T.A., Analgesia mediated by a direct spinal action of narcotics, Science, 192 (1976) 1357 1358.