Heterogeneity of GABAA receptor in goldfish retina

Effects of picrotoxin on light adapted frog electroretinogram are not due entirely to its action in proximal retina

In order to evaluate the site of action of picrotoxin (antagonist of ionotropic GABA receptors) on the electroretinographic (ERG) b- and d-waves, in this study we compared its effects on the intensity-response function of the ERG waves in intact light adapted frog eyecup preparations with its effects in eyecups, where the activity of proximal neurons was blocked by 1 mMN-methyl-d-aspartate (MNDA). Picrotoxin markedly enhanced the b- and d-wave amplitude and slowed the time course of the responses at all stimulus intensities in the intact eyecups. Perfusion with NMDA alone caused significant enhancement of the b-wave amplitude and diminution of the d-wave amplitude without altering their time course in the entire intensity range. When picrotoxin was applied in combination with NMDA, an enhancement of the b-wave amplitude and slowing of its time course were observed at all stimulus intensities. The increase of the b-wave amplitude was significantly higher than that seen in NMDA group. Combined application of picrotoxin and NMDA caused an enhancement of the d-wave amplitude at the lower stimulus intensities and its diminution at the higher ones, while the d-wave time course was delayed over the entire intensity range. The results obtained indicate that a part of picrotoxin effects on the amplitude and time course of the photopic ERG b- and d-waves are due to its action in the distal frog retina.

Lateral interactions in the outer retina

Lateral interactions in the outer retina, particularly negative feedback from horizontal cells to cones and direct feed-forward input from horizontal cells to bipolar cells, play a number of important roles in early visual processing, such as generating center-surround receptive fields that enhance spatial discrimination. These circuits may also contribute to post-receptoral light adaptation and the generation of color opponency. In this review, we examine the contributions of horizontal cell feedback and feed-forward pathways to early visual processing. We begin by reviewing the properties of bipolar cell receptive fields, especially with respect to modulation of the bipolar receptive field surround by the ambient light level and to the contribution of horizontal cells to the surround. We then review evidence for and against three proposed mechanisms for negative feedback from horizontal cells to cones: 1) GABA release by horizontal cells, 2) ephaptic modulation of the cone pedicle membrane potential generated by currents flowing through hemigap junctions in horizontal cell dendrites, and 3) modulation of cone calcium currents (I(Ca)) by changes in synaptic cleft proton levels. We also consider evidence for the presence of direct horizontal cell feed-forward input to bipolar cells and discuss a possible role for GABA at this synapse. We summarize proposed functions of horizontal cell feedback and feed-forward pathways. Finally, we examine the mechanisms and functions of two other forms of lateral interaction in the outer retina: negative feedback from horizontal cells to rods and positive feedback from horizontal cells to cones.

Intraocular injection of muscimol induces illusory motion reversal in goldfish

Induced activation of the gamma-aminobutyric acid(A) (GABA(A)) receptor in the retina of goldfish caused the fish to rotate in the opposite direction to that of the spinning pattern during an optomotor response (OMR) measurement. Muscimol, a GABA(A) receptor agonist, modified OMR in a concentration-dependent manner. The GABA(B) receptor agonist baclofen and GABA(C) receptor agonist CACA did not affect OMR. The observed modifications in OMR included decreased anterograde rotation (0.01~0.03 microM), coexistence of retrograde rotation and decreased anterograde rotation (0.1~30 microM) and only retrograde rotation (100 microM~1 mM). In contrast, the GABA(A) receptor antagonist bicuculline blocked muscimolinduced retrograde rotation. Based on these results, we inferred that the coding inducing retrograde movement of the goldfish retina is essentially associated with the GABA(A) receptor-related visual pathway. Furthermore, from our novel approach using observations of goldfish behavior the induced discrete snapshot duration was approximately 573 ms when the fish were under the influence of muscimol.

Organization of the torus longitudinalis in the rainbow trout (Oncorhynchus mykiss): an immunohistochemical study of the GABAergic system and a DiI tract-tracing study

The torus longitudinalis (TL) is a tectum-associated structure of actinopterygian fishes. The organization of the TL of rainbow trout was studied with Nissl staining, Golgi methods, immunocytochemistry with antibodies to gamma-aminobutyric acid (GABA), glutamic acid decarboxylase (GAD), and the GABA(A) receptor subunits delta and beta2/beta 3, and with tract tracing methods. Two types of neuron were characterized: medium-sized GABAergic neurons and small GABA-negative granule cells. GABA(A) receptor subunit delta-like immunoreactivity delineated two different TL regions, ventrolateral and central. Small GABAergic cells were also observed in marginal and periventricular strata of the optic tectum. These results indicate the presence of local GABAergic inhibitory circuits in the TL system. For tract-tracing, a lipophilic dye (DiI) was applied to the TL and to presumed toropetal nuclei or toral targets. Toropetal neurons were observed in the optic tectum, in pretectal (central, intermediate, and paracommissural) nuclei, in the subvalvular nucleus, and associated with the pretectocerebellar tract. Torofugal fibers were numerous in the stratum marginale of the optic tectum. Toropetal pretectal nuclei also project to the cerebellum, and a few TL cells project to the cerebellar corpus. The pyramidal cells of the trout tectum were also studied by Golgi methods and local DiI labeling. The connections of trout TL revealed here were more similar to those recently reported in carp and holocentrids (Ito et al. [2003] J. Comp. Neurol. 457:202-211; Xue et al. [2003] J. Comp. Neurol. 462:194-212), than to those reported in earlier studies. However, important differences in organization of toropetal nuclei were noted between salmonids and these other teleosts.

Functional GABA(B) receptors are expressed at the cone photoreceptor terminals in bullfrog retina

GABA(B) receptors at the cone terminals in bullfrog retina were characterized by immunocytochemical and whole-cell patch clamp techniques in retinal slice preparations. Somata, axons and synaptic terminals (pedicles) of cones were both GABA(B) receptor (GABA(B)R) 1 and GABA(B)R2 immunoreactive. Physiologically, barium/calcium currents of cones to voltage steps were significantly reduced in size when GABA was puffed to cone terminals in the presence of picrotoxin that is supposed to block both GABA(A) and GABA(C) receptors. Similar reduction in barium currents was obtained with puff application of baclofen to cone terminals. These results suggest the presence of functional GABA(B) receptors at the bullfrog cone terminals. Suppression of barium currents of cones by baclofen was dose-dependent. Moreover, barium currents of cones were potentiated by background illumination, as compared with those recorded in the dark. 6,7-Dinitroquinoxaline-2,3-dione, an antagonist of non-NMDA receptors that hyperpolarizes horizontal cells and reduces GABA release from these cells, and saclofen, a GABA(B) receptor antagonist, both potentiated barium currents of cones in the dark, thereby mimicking the effects of background illumination. It is suggested that changes in calcium influx into the cone synaptic terminals due to activation of GABA(B) receptors may provide a negative feedback mechanism for regulating signal transmission between cones and second-order neurons in the retina by modifying the amount of glutamate released from the cones.

Characterization of receptors for glutamate and GABA in retinal neurons

Glutamate and gamma-aminobutyric acid (GABA) are major excitatory and inhibitory neurotransmitters in the vertebrate retina, "a genuine neural center" (Ramón y Cajal, 1964, Recollections of My Life, C.E. Horne (Translater) MIT Press, Cambridge, MA). Photoreceptors, generating visual signals, and bipolar cells, mediating signal transfer from photoreceptors to ganglion cells, both release glutamate, which induces and/or changes the activity of the post-synaptic neurons (horizontal and bipolar cells for photoreceptors; amacrine and ganglion cells for bipolar cells). Horizontal and amacrine cells, which mediate lateral interaction in the outer and inner retina respectively, use GABA as a principal neurotransmitter. In recent years, glutamate receptors and GABA receptors in the retina have been extensively studied, using multi-disciplinary approaches. In this article some important advances in this field are reviewed, with special reference to retinal information processing. Photoreceptors possess metabotropic glutamate receptors and several subtypes of GABA receptors. Most horizontal cells express AMPA receptors, which may be predominantly assembled from flop slice variants. In addition, these cells also express GABAA and GABAC receptors. Signal transfer from photoreceptors to bipolar cells is rather complicated. Whereas AMPA/KA receptors mediate transmission for OFF type bipolar cells, several subtypes of glutamate receptors, both ionotropic and metabotropic, are involved in the generation of light responses of ON type bipolar cells. GABAA and GABAC receptors with distinct kinetics are differentially expressed on dendrites and axon terminals of both ON and OFF bipolar cells, mediating inhibition from horizontal cells and amacrine cells. Amacrine cells possess ionotropic glutamate receptors, whereas ganglion cells express both ionotropic and metabotropic glutamate receptors. GABAA receptors exist in amacrine and ganglion cells. Physiological data further suggest that GABAC receptors may be involved in the activity of these neurons. Moreover, responses of these retinal third order neurons are modulated by GABAB receptors, and in ganglion cells there exist several subtypes of GABAB receptors. A variety of glutamate receptor and GABA receptor subtypes found in the retina perform distinct functions, thus providing a wide range of neural integration and versatility of synaptic transmission. Perspectives in this research field are presented.

Postsynaptic localization of ?-aminobutyric acid transporters and receptors in the outer plexiform layer of the goldfish retina: An ultrastructural study

The gamma-aminobutyric acid (GABA)-ergic system in the outer plexiform layer (OPL) of the goldfish retina was studied via light and electron immunohistochemistry. The subcellular distributions of immunoreactivity (-IR) of plasma membrane GABA transporters GAT2 and GAT3, the alpha1 and alpha3 subunits of the ionotropic GABA(A) receptor, and the rho1 subunit of the ionotropic GABA(C) receptor were determined. The localization of the GAT2-IR and GAT3-IR to horizontal cell dendrites at the base of the cone synaptic complex was the main characteristic at the ultrastructural level. Very rarely, GAT2-IR and GAT3-IR were found in horizontal cell dendrites innervating rod spherules. alpha1-IR and alpha3-IR were seen in wide bands in the OPL, whereas rho1-IR appeared as a narrow band in the OPL. Most alpha1-IR was intracellular in rod and cone terminals. Membrane-associated alpha1-IR was observed in cone pedicles but not in rod spherules; postsynaptic elements were also labeled. alpha3-IR was concentrated in the lateral elements of horizontal cell dendrites in cone pedicles. In contrast, rho1-IR was found mainly on the spinules of the horizontal cell dendrites in cone pedicles. In addition, in another type of cone pedicle, rho1-IR was found at the position of OFF-bipolar cell dendrites. alpha3-IR and rho1-IR were rarely found in horizontal cell dendrites innervating rods. We suggest that two GABAergic pathways exist in the outer retina- first, a GABAergic positive loop with GABA receptors mainly on the horizontal cell dendrites and spinules and, second, a GABAergic feedback pathway involving GABA receptors on cone pedicles and GABA transporters on horizontal cells and that this pathway presumably modulates feedback strength from horizontal cells to cones.

Benzodiazepine and kainate receptor binding sites in the RCS rat retina

BACKGROUND

The effect of age and photoreceptor degeneration on the kainate subtype of glutamate receptors and on the benzodiazepine-sensitive gamma-aminobutyric acid-A receptors (GABA(A)) in normal and RCS (Royal College of Surgeons) rats were investigated.

METHODS

[(3)H]Kainate and [(3)H]flunitrazepam were used as radioligands for kainate and GABA(A)/benzodiazepine()receptors, respectively, using the quantitative receptor autoradiography technique.

RESULTS

In both normal and RCS rat retina we observed that [(3)Eta]flunitrazepam and [(3)Eta]kainate binding levels were several times higher in inner plexiform layer (IPL) than in outer plexiform layer (OPL) at all four ages studied (P17, P35, P60 and P180). Age-related changes in receptor binding were observed in normal rat retina: [(3)Eta]flunitrazepam binding showed a significant decrease of 25% between P17 and P60 in IPL,and [(3)Eta]kainate binding showed significant decreases between P17 and P35 in both synaptic layers (71% in IPL and 63% in OPL). Degeneration-related changes in benzodiazepine and kainate receptor binding were observed in RCS rat retina. In IPL, [(3)Eta]flunitrazepam and [(3)Eta]kainate binding levels were higher than in normal retina at P35 (by 24% and 86%, respectively). In OPL, [(3)Eta]flunitrazepam binding was higher in RCS than in normal retina on P35 (74%) and also on P60 (62%).

CONCLUSIONS

The results indicate that postnatal changes occur in kainate and benzodiazepine receptor binding sites in OPL and IPL of the rat retina up to 6 months of age. The data also suggest that the receptor binding changes observed in the RCS retina could be a consequence of the primary photoreceptor degeneration.

GABA-Mediated inhibition between amacrine cells in the goldfish retina

Retinal amacrine cells have abundant dendro-dendritic synapses between neighboring amacrine cells. Therefore an amacrine cell has both presynaptic and postsynaptic aspects. To understand these synaptic interactions in the amacrine cell, we recorded from amacrine cells in the goldfish retinal slice preparation with perforated- and whole cell-patch clamp techniques. As the presynaptic element, 19% of the cells recorded (15 of 78 cells) showed spontaneous tetrodotoxin (TTX)-sensitive action potentials. As the postsynaptic element, all amacrine cells (n = 9) were found to have GABA-evoked responses and, under perforated patch clamp, 50 microM GABA hyperpolarized amacrine cells by activating a Cl(-) conductance. Bicuculline-sensitive spontaneous postsynaptic currents, carried by Cl(-), were observed in 82% of the cells (64 of 78 cells). Since the source of GABA in the inner plexiform layer is amacrine cells alone, these events are likely to be inhibitory postsynaptic currents (IPSCs) caused by GABA spontaneously released from neighboring amacrine cells. IPSCs were classified into three groups. Large amplitude IPSCs were suppressed by TTX (1 microM), indicating that presynaptic action potentials triggered GABA release. Medium amplitude IPSCs were also TTX sensitive. Small amplitude IPSCs were TTX insensitive (miniature IPSCs; n = 26). All of spike-induced, medium amplitude, and miniature IPSCs were Ca(2+) dependent and blocked by Co(2+). Blocking of glutamatergic inputs by DL-2-amino-phosphonoheptanoate (AP7; 30 microM) and 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX; 2 microM) had almost no effect on spontaneous GABA release from presynaptic amacrine cells. We suggest that these dendro-dendrotic inhibitory networks contribute to receptive field spatiotemporal properties.

Subcellular localization and complements of GABA(A) and GABA(C) receptors on bullfrog retinal bipolar cells

gamma-Aminobutyric acid (GABA) receptors on retinal bipolar cells (BCs) are highly relevant to spatial and temporal integration of visual signals in the outer and inner retina. In the present work, subcellular localization and complements of GABA(A) and GABA(C) receptors on BCs were investigated by whole cell recordings and local drug application via multi-barreled puff pipettes in the bullfrog retinal slice preparation. Four types of the BCs (types 1-4) were identified morphologically by injection of Lucifer yellow. According to the ramification levels of the axon terminals and the responses of these cells to glutamate (or kainate) applied at their dendrites, types 1 and 2 of BCs were supposed to be OFF type, whereas types 3 and 4 of BCs might be ON type. Bicuculline (BIC), a GABA(A) receptor antagonist, and imidazole-4-acetic acid (I4AA), a GABA(C) receptor antagonist, were used to distinguish GABA receptor-mediated responses. In all BCs tested, not only the axon terminals but also the dendrites showed high GABA sensitivity mediated by both GABA(A) and GABA(C) receptors. Subcellular localization and complements of GABA(A) and GABA(C) receptors at the dendrites and axon terminals were highly related to the dichotomy of OFF and ON BCs. In the case of OFF BCs, GABA(A) receptors were rather evenly distributed at the dendrites and axon terminals, but GABA(C) receptors were predominantly expressed at the axon terminals. Moreover, the relative contribution of GABA(C) receptors to the axon terminals was prevalent over that of GABA(A) receptors, while the situation was reversed at the dendrites. In the case of ON BCs, GABA(A) and GABA(C) receptors both preferred to be expressed at the axon terminals; relative contributions of these two GABA receptor subtypes to both the sites were comparable, while GABA(C) receptors were much less expressed than GABA(A) receptors. GABA(A), but not GABA(C) receptors, were expressed clusteringly at axons of a population of BCs. In a minority of BCs, I4AA suppressed the GABA(C) responses at the dendrites, but not at the axon terminal, implying that the GABA(C) receptors at these two sites may be heterogeneous. Taken together, these results suggest that GABA(A) and GABA(C) receptors may play different roles in the outer and inner retina and the differential complements of the two receptors on OFF and ON BCs may be closely related to physiological functions of these cells.

Subunit composition and pharmacological characterization of gamma-aminobutyric acid type A receptors in frog pituitary melanotrophs

The frog pars intermedia is composed of a single population of endocrine cells directly innervated by gamma-aminobutyric acid (GABA)ergic nerve terminals. We have previously shown that GABA, acting through GABA(A) receptors, modulates both the electrical and secretory activities of frog pituitary melanotrophs. The aim of the present study was to take advantage of the frog melanotroph model to determine the relationship between the subunit composition and the pharmacological properties of native GABA(A) receptors. Immunohistochemical labeling revealed that in situ and in cell culture, frog melanotrophs were intensely stained with alpha2-, alpha3-, gamma2-, and gamma3-subunit antisera and weakly stained with a gamma1-subunit antiserum. Melanotrophs were also immunolabeled with a monoclonal antibody to the beta2/beta3-subunit. In contrast, frog melanotrophs were not immunoreactive for the alpha1-, alpha5-, and alpha6-isoforms. The effects of allosteric modulators of the GABA(A) receptor on GABA-activated chloride current were tested using the patch-clamp technique. Among the ligands acting at the benzodiazepine-binding site, clonazepam (EC50, 5 x 10(-9) M), diazepam (EC50, 10(-8) M), zolpidem (EC50, 3 x 10(-8) M), and beta-carboline-3-carboxylic acid methyl ester (EC50, 10(-6) M) were found to potentiate the whole cell GABA-evoked current in a dose-dependent manner. Methyl-6,7-dimethoxy-4-ethyl-beta-carboline-3-carboxylate (IC50, 3 x 10(-5) M) inhibited the current, whereas Ro15-4513 had no effect. Among the ligands acting at other modulatory sites, etomidate (EC50, 2 x 10(-6) M) enhanced the GABA-evoked current, whereas 4'-chlorodiazepam (IC50, 4 x 10(-7) M), ZnCl2 (IC50, >5 x 10(-5) M), and furosemide (IC50, >3 x 10(-4) M) depressed the response to GABA. PK 11195 did not affect the GABA-evoked current or its inhibition by 4'-chlorodiazepam. The results indicate that the native GABA(A) receptors in frog melanotrophs are formed by combinations of alpha2-, alpha3-, beta2/3-, gamma1-, gamma2-, and gamma3-subunits. The data also demonstrate that clonazepam is the most potent, and zolpidem is the most efficient positive modulator of the native receptors. Among the inhibitors, 4'-chlorodiazepam is the most potent, whereas ZnCl2 is the most efficient negative modulator of the GABA(A) receptors. The present study provides the first correlation between subunit composition and the functional properties of native GABA(A) receptors in nontumoral endocrine cells.

Differences in the retinal GABA system among control, spastic mutant and retinal degeneration mutant mice

Immunocytochemical methods were used to compare the GABA system in control mice and two mutant strains: spastic which has reduced glycine receptors and retinal degeneration mutant in which the photoreceptors degenerate and reportedly have increased GABA and GAD levels. We found that the spastic mutant retina had reduced GABA-immunoreactivity (IR) in the proximal retina, reduced staining for GAD-1440 in the OPL, and reduced GABAA receptor staining in the OPL, compared to control. The retinal degeneration mutant retinas had enhanced GABA-IR throughout the retina, particularly in Müller cells, bipolar cells and IPL, and enhancement of GABAA receptor staining in the OPL, compared to control. The distributions of GABA-IR, GAD-1440-IR and GABAA receptor-IR in retinas of spastic mutant mice that also expressed the retinal degeneration phenotype resembled those found in retinas of mice that expressed only the retinal degeneration phenotype rather than those that expressed only the spastic mutation. No differences were observed among the conditions for GAD-65, GAD-67 or GABA-T. Our results with the spastic and retinal degeneration mutant mice demonstrate that attenuation in the glycinergic system and photoreceptor degeneration, respectively, is accompanied by alterations in different aspects of the GABA system, giving impetus for caution in the interpretation of experiments involving genetic manipulation of complex phenotypes.

The autoradiographic perspective of central benzodiazepine receptors; a short review

1. We reviewed studies performed to characterize central benzodiazepine binding sites. 2. An overview of the different radioligands used to characterize BZ1 and BZ2 binding sites and a mapping of these central benzodiazepine sites are described. 3. Saturation studies carried out by autoradiogram quantification also are reviewed. 4. The specific use of the autoradiographic technique to carry out studies on ontogeny, development, and phylogeny is discussed, as well as studies performed using this technique on some diseases and experimental conditions, such as drug treatments or chemical and mechanical lesions.

Immunocytochemical localization of the GABA(C) receptor rho subunits in the cat, goldfish, and chicken retina

Polyclonal antibodies against the N-terminus of the rat rho1 subunit were used to study the distribution of gamma-aminobutyric acid C (GABA(C)) receptors in the cat, goldfish, and chicken retina. Strong punctate immunoreactivity was present in the inner plexiform layer (IPL) of all three species. The punctate labelling suggests a clustering of the GABA(C) receptors at synaptic sites. Weak label was also found in the outer plexiform layer (OPL) and over the cell bodies of bipolar cells. Double immunostaining of vertical sections with an antibody against protein kinase C (PKC) showed the punctate immunofluorescence to colocalize with bipolar cell axon terminals. In the goldfish retina, the axon terminals of Mb1 bipolar cells were enclosed by rho-immunoreactive puncta. In the chicken retina, several distinct strata within the IPL showed a high density of rho-immunoreactive puncta. The results suggest a high degree of sequence homology between the rho subunits of different vertebrate species, and they show that the retinal localization of GABA(C) receptors is similar across different species.

Light adaptation affects synaptic vesicle density but not the distribution of GABAA receptors in goldfish photoreceptor terminals

GABA is a likely feedback transmitter from H1 horizontal cells to cone photoreceptors in fish retinas. Spinules arise from H1 cell dendrites in light-adapted retinas, are correlated with responses attributed to feedback, and have been proposed to be the GABA release sites. We used mAb 62-3G1, an antibody against the beta 2/beta 3 subunits of the GABAA receptor complex, to visualize GABAA receptor immunoreactivity (GABAr-IR) in photoreceptors as a function of light and dark adaptation at the electron microscopical level. Regardless of adaptation, GABAr-IR was restricted to the synaptic terminals of all cones and most rods; synaptic vesicular membrane and plasma membrane, exhibited GABAr-IR. Contrary to expectations, the density of GABAr-IR was least on the plasma membrane within the invagination, regardless of the presence or absence of spinules. Dense GABAr-IR was observed on the lateral surface of cone pedicles, on cone processes proximal to the invagination, and on presumed telodendria from nearby cones. There was no difference in GABAr-IR of rod plasma membranes within or outside of the invagination or with adaptation. The only novel effect of adaptation was in regards to the density of synaptic vesicles. Cones showed a 29% increase in vesicle density with dark adaptation, whereas rods showed a 17% decrease. We conclude that all goldfish photoreceptors will be GABA-sensitive and that the sensitivity is distributed over the surface of the synaptic terminal rather than localized to within the invagination. The role of spinules in GABA release remains to be determined, but we conclude that spinules are not related to the GABA sensitivity of goldfish photoreceptors.

Distribution of immunoreactivity for the beta 2 and beta 3 subunits of the GABAA receptor in the mammalian spinal cord

The localization of GABAA receptors in cat and rat spinal cord was analyzed using two monoclonal antibodies specific for an epitope shared by the beta 2 and beta 3 subunits of the receptor. beta 2/beta 3-subunit immunoreactivity was the most intense in inner lamina II, lamina III, and lamina X, and it was the least intense in lamina IX. In laminae I-III, generally, the staining had a rather diffuse appearance, but the surfaces of small cell bodies in these laminae were outlined clearly by discrete labeling, as were many cell bodies and dendrites in deeper laminae. Rhizotomy experiments and ultrastructural observations indicated that beta 2/beta 3-subunit immunoreactivity in the dorsal horn was largely localized in intrinsic neuropil elements rather than in the terminals of primary afferent fibers, even though labeling overlapped with the terminal fields of different types of primary afferents and was also detected on the membranes of dorsal root ganglion neurons. With few exceptions (most notably, a highly immunoreactive group of dorsolaterally located cells in the cat lumbar ventral horn), motoneurons expressed low levels of beta 2/beta 3-subunit immunoreactivity. Labeling of neuronal membranes was fairly continuous, but focal accumulations of beta 2/beta 3-subunit immunoreactivity were also detected using immunofluorescence. Focal "hot spots" correlated ultrastructurally with the presence of synaptic junctions. Dual-color immunofluorescence revealed that focal accumulations of beta 2/beta 3-subunit immunoreactivity were frequently apposed by glutamic acid decarboxylase (GAD)-immunoreactive terminals. However, the density of continuous-membrane beta 2/beta 3 immunolabeling and GAD terminal density were not correlated in many individual neurons. The results suggest the existence of "classical" (synaptic) and "nonclassical" (paracrine) actions mediated via spinal cord GABAA receptors. The study also revealed the relative paucity of beta 2/beta 3-subunit immunoreactivity postsynaptic to certain GABAergic terminals, particularly those presynaptic to motoneurons or primary afferent terminals.