0306-4522/92$5.00+ 0.00 Pergamon Press plc Q 199I IBRO
NeuroscienceVol. 46, No. 2, pp. 455-463, I992 Printed in Great Britain
FINE STRUCTURE OF GABA-LABELED AXONAL ENDINGS IN THE INFERIOR COLLICULUS OF THE CAT: IMMUNOCYTOCHEMISTRY ON DEPLASTICIZED ULTRATHIN SECTIONS D.L.
OLIVER*
andG.E.
BECKIUS
Department of Anatomy and Center for Neurological Sciences, The University of Connecticut Health Center, Farmington, CT 06030, U.S.A. Abstract-Antisera to GABA conjugates and postembedding techniques were used to identify GABA-containing axonal endings at the electron microscopic level in the inferior colliculus. Over 90% of the GABA-labeled axonal endings bad a similar mo~hoIogy. They contained pleomorphic synaptic vesicles and made s~metrical synapses. The exceptional endings contained round vesicles and made s~rnet~~~ synaptic contacts or had pleomorphic vesicles with asymmetrical contacts. The majority of GABA-labeled axonal endings synapsed on dendrites; however, a few labeled axosomatic synapses were also found. Potential sources for these GABAergic synapses are neurons intrinsic to the inferior colliculus or from the dorsal nucleus of the lateral lemniscus. These findings suggest a basic similarity for most GABAergic endings in the inferior colliculus despite their possible origin from different cell types.
Inhibitory neurotransmitters, such as GABA, may play a significant role in shaping the neural responses in the inferior colliculus. Nelson and Erulkar” and othersg*” showed that inhibitory and excitatory postsynaptic potentials were common components of intracellular neural responses in the inferior colliculus. GABA may mediate some of these inhibitory potentials. Glutamate decarboxylase (GAD) has been shown to be present in the inferior colliculus in biochemical assays3 and in immun~ytochemi~al studies. ‘2*‘4~21~24*26~27~36 Moreover, at least one ascending projection to the inferior colliculus from the dorsal nucleus of the lateral lemniscus may use GABA as a transmitter.‘v32.33 Despite the potential importance of inhibitory synapses in the inferior colliculus, they have received little study at the electron microscopic level. Thus, the goal of this study was to identify synapses in the inferior colliculus where GABA may act as the neurotransmitter. A modification of the postembedding immunocytochemical procedure was developed for use with the electron microscope. This technique allowed the unobscured fine structure of synapses to be related to the presence of a particular transmitter. Since GABAergic axonal endings may arise from more than one source, this approach helped to identify morphologically distinct GABAergic endings.
iments.‘7~20*3Z The two additional animals were deeply anesthetized with sodium pentobarbital(40 mg/kg) and perfused through the aorta with mixed aldehyde fixatives similar to those used previously.“~20~32A washout solution (50 ml) contained 0.12 M phosphate buffer, pH 7.4, 2% sucrose, 0.05% lidocaine, and 0.004% CaCI,. The fixatives contained the same buffer and calcium salts and consisted of a dilute first solution (500 ml of 1% paraformaldehyde, 1% glutaraldehyde) and a second, more concentrated fixative (1000 ml of 2% paraformaldehyde and 3% glutaraldehyde). Histology
Sections, 7t.KlOO~m thick, were cut from the inferior colliculus in the transverse plane with a Vibratome. Every fifth section was processed for electron microscopy and the rest for light microscopy. The sections for electron microscopy were rinsed in phosphate buffer (0.12 M, pH 7.4) containing 6% sucrose, postfixed in 1% buffered osmium tetroxide, block-stained in 2% uranyi acetate, dehydrated in acetone, and embedded in Durcupan ACM (Pofyscience). The remaining light microscopic sections were either stained with Cresyl Violet or reacted with GABA antisera as free-floating sections.39,4’ Binding of the GABA antibody was revealed in the light microscopic sections with biotinylated anti-rabbit IgG coupled to an avidin-peroxidase complex (Vectastain, Vector Labs). Polyclonal rabbit antisera were used. One antiserum, donated by Dr Robert Wenthold (NIH), was affinitypurified antiserum against bovine serum albumin~utaraldehyde conjugates of GABA.399A’A second GABA antiserum (Pel-Freez. P4MV72~1) was raised against glutaraldeh~de conjugates of bovine serum alb;min, hemoglobin, and poly-lysine used in successive boosters3’ The cross-reactivity of both antibodies has been well characterized.“~39~4’ To prepare samples for colloidal gold immunostaining and electron microscope analysis, serial lOO-nm-thick sections were cut and collected in pairs with a wire loop on Formvar-coated glass slides. This technique was similar to that used routinely for electron microscopic autoradiography. *vi7The day before immunos~i~ng, the sections were transferred to grids by etching the Formvar coating with a dilute solution of hydrofluoric acid, floating off the Formvar
EXPERIMENTAL PROCEDURES The experimental material was obtained from six normal, adult cats, four of which were part of previous exper*To whom correspondence should be addressed. DNLL, dorsal nucleus of the lateral lemniscus; GAD, glutamate decarboxylase.
Abbreuiatims:
455
456
D. L. OLIVERand G. E.
film and section onto a water bath, and picking up the sections on grids. For each antibody dilution, a minimum of three sequential pairs of serial sections on three separate grids was used. The first pair of “normal” sections, collected on 200-mesh thin bar grids, was stained with 4% many1 acetate and 0.1% lead citrate.” The second pair of “immunoreactive” sections was collected on 300-mesh nickel grids for immunostaining. The third pair of “control” sections (300-mesh nickel grids) was handled like the second pair but was not exposed to the primary antibody. Colloidal gold immunostaining on deplasticized thin sections followed the method of Mar and Wight.” Thin sections on grids were deplasticized by immersion for 10 min in 45% sodium hydroxide in methanoP and were rehydrated in methanolic solutions. The grids were then floated on droplets of 0.12 M phosphate buffer, pH 7.4. This buffer was used in all aqueous solutions for immunohistochemistry and as a holding step. After blocking in 10% normal goat serum, the grids-were transferred without rinsing to primary antibodv at 4°C at dilutions of 1:200. 1: 500. and 1: 1000and incubated overnight (17-19 h). Control sections were kept in normal goat serum during this time. Following buffer rinses, the immunoreactive and control grids were exposed to a I : IO dilution of colloidal gold anti-rabbit IgG (E-Y Labs) for 4 h and postfixed successively in 2% glutaraldehyde, 1% osmium tetroxide, and 2% uranyl acetate solutions. Finally, the grids were dehydrated in ethanol, dipped in 2% Epon in absolute ethanol, polymerized overnight at 60°C on edge,‘” and counterstained with uranyl acetate and lead citrate. Analysis
Ultrathin sections processed for immunocytochemistry were compared to control and normal sections on a JEOL 1OOCXelectron microscope. The location of the sections relative to the grid bars was drawn, then the sections were photographed at x 160&2200. Sample grid squares were chosen at random or because a cell body with a visible nucleolus was present. Immunolabeled endings were at least x 5-10 above background, but the absolute labeling was not routinely counted. All labeled endings within a grid square were photographed at x 10,000. After the labeled endings were identified on the low magnification pictures, these same axonal endings were located in the adjacent normal sections and photographed at x lO,OOO-16,000. Most of the unlabeled endings were identified in the same micrographs as the labeled endings. RESULTS Immunoreactivity for GABA was studied in electron microscopic samples from the pars centralis in the central nucleus of the inferior colliculus. This region was rich in immunoreactive profiles at the light microscopic level (Refs 21 and 24; Oliver et al., in preparation). The electron microscopic samples were
BECKIUS
adjacent to areas in light microscopic sections that contained immunoreactive axons, puncta, and cell bodies. Comparisons of immunoreactive, control, and normal sections
After immunoreaction with antibodies to GABA, some axonal endings and myelinated axons were heavily labeled compared to control sections. Serial sections were used to compare the immunoreaction to control conditions in the same axonal endings. Figure 1 illustrates such an ending in three sections spaced 300 nm apart. The sections in panels A and B were prepared identically except that B was exposed to Pel-Freez GABA antibody. In the control section. the colloidal gold was distributed evenly throughout all tissue compartments. In contrast, the section exposed to antibody (Fig. 1B) showed a heavy accumulation of colloidal gold over the axonal ending (PL) and over several myelinated axons (LA). In the adjacent serial ultrathin section, the fine structure of the labeled axonal endings could be examined. Ultrastructural detail in the immunoreactive sections was lost because of the deplasticizing and the heavy gold labeling. In contrast, the morphology of the synaptic vesicles and synaptic contacts was intact in the adjacent normal sections. The section pictured in Fig. 1C showed that the labeled axonal ending (Fig. IB) contained pleomorphic synaptic vesicles. Most GABA-labeled synaptic vesicles
endings contained pleomorphic
Tables 1 and 2 describe the vesicular and synaptic morphology for the 272 endings in the present sample. The labeled axonal endings represented over half of the sample, and most of the labeled endings (97%) contained pleomorphic synaptic vesicles (Table 1). As shown in Fig. 2, the endings with pleomorphic vesicles were well labeled with the Wenthold antisera as compared to an ending with round vesicles (LR) and a dendrite in the same section. Labeling of these latter structures was at control levels. Only four labeled endings in the sample contained round synaptic vesicles (Table 2); these endings made symmetrical synaptic contacts in the two cases where the synaptic density was visible.
Table 1. Axonal endings with pleomorphic vesicles: GABA-labeled versus unlabeled Labeled 88-06-38 Labeled 80-07-68 Total labeled Unlabeled 88-06-38 Unlabeled 88-07-68 Total unlabeled
No. 50 91 141
A/D 4 0 4
A/S
1
S/D 17 30 47
II 9 20
0 I I
0 0 0
4 2 6
1 0
N
6 7
U/D 9 14 23
U/S 3 I 10
15 34 49
0 0 0
3 I 4
1 0 1
3 5 8
SIS
I
No., number of endings counted; A/D, asymmetric contacts on dendrite; A/S, asymmetric contacts on soma; S/D, symmetric contacts on dendrite; S/S, symmetric contacts on soma; U/D, unclear contacts on dendrites; U/S, unclear contacts on soma; N, no contact visible
GABA-labeled axonal endings in the inferior colliculus Table 2. Axonal endings with round vesicles:GABA-labeled versus unlab&d*
Unlabeled 88-06-38 Unlabeled 88-07-68 Total unlabeled
44 62 106
14 23 37
0 0 0
1 1 2
0 0 0
13 12 25
2
I 3
14 25 39
*See Table 1 for abbreviations.
Several types of endings were usually unlabeled. The largest group of endings in the sample (39%) were unlabeled and contained round vesicles. Two sizes of round vesicle were evident in micrographs.
Endings with larger round vesicles (Figs 24) were easily distinguished from endings with smaller round vesicles (RR’) in micrographs where both were present (Fig. 3). In most cases, these endings were
Abbreviations used in rhe jigures
A CB D F
unlabeled axon cell body dendrite flattened/pleomorphic synaptic vesicles
LR LA PL R
large, round synaptic vesicles labeled myelinated axon pleomorphic synaptic vesicles small round synaptic vesicle
Fig. 1. Control, immunoreactive, and normal sections spaced 3OOmnapart through the same axonal ending (PL). (A) The control section shows that gold particles are distributed randomly in a deplasticixed section. (B) After exposure to GABA antibody (1: 500, Pel-Free& colloidal gold label accumulatesover the ending and two myelinated axons (LA). The cell body (CB) may be lightly labeled. Other axons and glial processes between the labeled ending and labeled axons display far fewer gold particles. (C) In the normal section, pleomorphic synaptic vesiclesare visible.The ending forms a symmetricalsynapticcontact (arrowheads)on the somatic spine pictured in C. Case nos 88-07-68-39,-41, -42. Magnifications:(A), (B) x 25,380;(C) x 32,806.Scale bars = 0.5pm.
458
D L. OLIVERand G. E. BECKWS
Fig. 2. Axodendritie synapses. (A) In a normal section, three endings (PL) with pleomorphic synaptic vesicles make s~rnmetricalsynapses (arrowheads) on a dendrite (D). A fourth ending contains large, round vesicles (LR). (8) Immunoreactive section, 2OOnmfrom section in A, shows the same endings after exposure to GABA antib~y (I : 500,Wenthold), The pieiomorphic synaptic vesicle endings are labeled. while the large. round synaptic vesicle ending and dendrite are not. Case nos 88-06-38-96,-94. Magnifications:(A) x 35,062,(B) x 28,952.Scale bars = 0.5 urn. equally unlabeled. However, one ending (R’) in Fig. 3 dispiayed more gold binding than the large and smaller round vesicle endings. This level of signal was somewhat higher than surrounding structures, but it was less than the labeled endings with pleomorphic vesicles (Fig. 3). Besides endings with round vesicles, unlabeled endings with pleomo~hic vesicles were observed (7% of the sample). One of the 20 such endings is shown in Fig. 4. It is designated F to distinguish it from the labeled ending in the micrograph. Labeled endirrgs us~af~ymade s?~~rnetri~R~ spaptir contacts on dendrites Tables 1 and 2 show the types of synaptic contacts
made by labeled and unlabeled endings. Of the 141 labeled endings that contained p~eomorphic vesicles, 38% made symmetrical contacts while 58% had no contact or the contact was unclear. Figure 2 illustrates such a synaptic contact on a dendrite (arrowheads). Only a few (about 4%) labeled endings made asymmetrical synapses, and each of these endings contained pleomorphic synaptic vesicles. In contrast, unlabeled endings with round vesicles made asymmetrical synapses in all but two instances where the contact could be identified. Roughly 50% of the GABA-labeled endings were presynaptic to dendrites (Tables 1 and 2). Axosomatic synapses were also found but were less frequent. Only seven labeled endings with pleomo~hic
GABA-labeled axonal endings in the inferior colliculus
Fig. 3. Labeling density. (A) After i~unoreactian with Pel-Freez GABA antibody, two axonal endings (PL) are heavily labeled.Other endings(LR, R) are unlabeled.All endings in the micrographs are outlined with triangles. One axonal ending (R’) and the dendrite (D) may be labeled above background. (B) In the normal section 300nm away, the heaviest labeled endings contain pleomorphic vesicles (PL). Other endings contain either round (R) or large round (LR) vesicles Case nos 88-07-68-79,-76. Ma~ifications: (A) x 27,166;(B) 33,276.Scale bars = 0.5 pm.
459
460
D. L. OLWERand G. E. BECKIUS
Fig. 4. Axosomatic endings. (A) After immunoreaction, only one (PL) of three axosomatic endings pictured are labeled. A nearby axon (LA) is heavily labeled, while other endings {F, LR) and axons (A) are unlabeled.(B)The labeledending contains pleomorphicvesicles(PL) and makes a symmetricalsynapse on the dendrite (arrowhead). One unlabeled ending contains pleomorphic vesicles (F), and the other contains large round vesicles(LR) and makes an axosomaticsynapse (arrowhead),Samesections as Fig. 2. Magnifications:(A) x 28,670,(B) x 42,I 12. Scale bars = 0.5 pm.
GABA-labeled axonai endings in the inferior colliculus
461
colliculus of the rat. Most of their data concerned the identification of GAD-labeled cell types in the inferior cofliculus. However, they did report GADlabeled terminals containing pleomorphic or flattened synaptic vesicles that made symmetrical synapses. DISCUSSION They distinguished different types of labeled endings by the size and shape of the ending, not by the This study is the first to use postembedding techsynaptic vesicles or synaptic contacts. Thus, they did niques to examine the fine structure of GABA-connot report any exceptional types of GAD-labeled taining axonal endings in the inferior colliculus. Most synaptic endings. GABA-labeled axonal endings contained pleomorAll GABAergic endings are not identical in other phic synaptic vesicles. When the synaptic contact systems (e.g. Refs 6 and 38). Different GABAergic could be identified, GABA-labeled endings made symmetrical synapses. However, there were excep- synapses have been distinguished by vesicle size, the presence of dense-core vesicles, and by vesicle density. tions. Small numbers of endings contained round vesicles and made symmetrical synaptic contacts or Quantitative analysis of immunolabeled endings in the inferior colliculus will be carried out in the near had pleomorphie vesicles with asymmetrical contacts. Most GABA-labeled axonal endings synapsed on future since several types of endings with pleomorphic vesicles have been identified previously.” dendrites; however, a number of labeled axosomatic Not all of the endings with pleomorphic or flatsynapses were also found. These findings suggest a basic similarity for most GABAergic endings in the tened vesicles were immunolabeled in the inferior inferior colliculus despite the likelihood that they colliculus in the present study and in the previous arise from different cell types. study in the rat.27 It is possible that these unlabeled endings use a different transmitter such as glycine. Comparison of data from pre-embedding and postBoth the projections from the ipsilateral lateral suembedding studies perior oIive7*28,30 and the ventral nucleus of the lateral Pre-embedding t~h~ques at the electron microlemniscus428 have been implicated as glycinergic. Moreover, the presynaptic terminals opposite glycinscopic level have been the mainstay of previous ergic synapses have been reported to contain either immunocytochemical studies in the inferior collicu1us2’ and elsewhere in the auditory system.2~13*16.29,41 flattened or pleomorphic synaptic vesicles.5,a In some The strong immunocytochemical reactions carried cases, GABA-containing terminals may appose out before embedding tend to obscure the fine strucglycine receptors at central synapses3’ ture of the axonaf ending. Postembedding methods, Do GABAergic endings in the inferior coltic~lus come as used here and in other systems (e.g. Refs 34 and 35), preserve the fine structure. In the present study, from deferent sources?
vesicles made axosomatic synapses with symmetrical contacts (Table 1). Figures 1 and 4 (PL) illustrate this type of axosomatic, symmetrical synapse.
the use of deplasticized sections allowed for maximal immunoreactivity. In addition, it allowed the fine structure to be analysed on adjacent normal sections that were completely undisturbed by histochemical reactions. Are there d@erent types of GABAergic synapses in the inferior col~jcu~~s? Although most GABAergic endings in the inferior cofliculus were similar, there could be variations of GABAergic endings. A small number of labeled endings in this sample did not have either pleomorphic synaptic vesicles or symmetrical synaptic junctions. These exceptional endings were found in the tissue studied primarily with the Wenthold antibody. Since both the background and signal were lower with that antiserum, the exceptional axonal endings may have been easier to detect. Cross-reactivity of both antisera with other possible transmitters or transmitter precursors appear to be low (see Experimental Procedures). It is likely that these exceptional endings represent genuine immunolabeling for GABA and could represent axons from other GABAergic cell types (see below). Similar to the present results, Roberts and Ribakz7 reported observations of GAD fabeling in the inferior
There may be intrinsic and extrinsic sources for GABAergic synapses in the inferior colliculus. Within the inferior colliculus, there may be several cell types that use GABA as a transmitter. Two major cell classes, disc-shaped and stelfate, are identified in Golgi preparations,23 and both types emit axonal collaterals within the inferior colliculus.22 Both GABAergic disc-shaped and stellate cells have been identified.2’,24*27Even though multiple GABAergic cell types may exist within the inferior colliculus, the majority of GABAergic synapses in the inferior colliculus still have a similar morphology. Outside of the inferior colliculus, the dorsal nucleus of the lateral lemniscus (DNLL) is another potential source for GABAergic projections.’ Previous studies25+32 show that most of the endings from the DNLL have pleomorphic vesicles and make symmetrical synapses. These endings from DNLL are similar to the majority of the GABA-labeled endings in the present study and account for one-third of such endings in the central nucleus of the inferior colliculus. This is uniformly true for the projection from the contralateral DNLL to the inferior colliculus. However, projections from the ipsilateral DNLL to the colhcufus are heterogeneous. Some of the ipsilateral projections from DNLL may not use GABA as a
462
D. L. OLIVER and
transmitter since one-third of those endings contain round synaptic vesicles. In other studies, axonal endings from the dorsal cochlear nucleus’7,‘9 and ventral cochlear nucleus2’ are identified in the inferior colliculus at the electron microscopic level and contain round vesicles and make asymmetric synapses. This type of ending was never GABA-immunoreactive in the present study. Thus, the majority of the extrinsic GABA-labeled axons are likely to arise only from the DNLL and have a morphology similar to the intrinsic GABA-labeled axons. Postsynaptic targets
Both disc-shaped and stellate classes of cells are richly innervated by presumed GABAergic synapses
G. E. BECKIUS
on their somata and dendrites. Both classes of cells have projections ascending to the medial geniculate body and receive symmetrical, axosomatic synapses from endings with pleomorphic vesicles.‘8 Those synapses are the same type labeled with GABA antisera in the present study. Many of the axosomatic GABAergic endings may be intrinsic in origin since most of the endings from DNLL terminate primarily on dendrites.32 Ackno~iedgements-This
research was sponsored by the Deafness Research Foundation and by NIH grant LXOO189. Thanks arc due to Dr Robert Wenthold for the GABA antiserum and to Drs R. Saint Marie and I. Schwartz for helpful comments on the manuscript.
1. Adams J. C. and Mugnaini E. (1984) Dorsal nucleus of the lateral lemniscus: a nucleus of GABAergic projection neurons. Brain Res. Bull. 13, 585-590.
2. Adams J. C. and Mugnaini E. (1987) Patterns of glutamate decarboxylase immunostaining in the feline cochlear nuclear complex studied with silver enhancement and electron microscopy. J. camp. Neural. 262, 375-101. 3. Adams J. C. and Wenthold R. J. (1979) Distribution of putative amino acid transmitters. choline acetyltransferase and glutamate decarboxylase in the inferior colliculus. Neuroscience 4, I947-1951. 4. Adams J. C. and W~nthold R. J. (1987) Immunostaining of ascending auditory pathways with glycine antiserum. Assoc. Res. ~ioiaryngol. Abs/r. 10, 63. 5. Altschuler R. A., Betz H., Parakkal M. H., Reeks K. A. and Wenthold R. J. (1986) Identification of glycinergic synapses in the cochlear nucleus through immunocytochemical localization of the postsynaptic receptor. Bruin Res. 369,316-320. 6. De Zeeuw C. I., Holstege J. C., Ruigrok T. J. H. and Voogd J. (1989) Ultrastructural study of the GABAergic. cerebellar, and mesodiencephalic innervation of the cat medial accessory olive: anterograde tracing combined with immunocytochemistry. 1. camp. Neuroi. 284, 12-35. 7. Hutson K. A. (1988) Connections of the auditory midbrain: efferent projections of the dorsal nucleus of the lateral lemniscus, the nucleus sagulum, and the origins of the GABAergic commissure of Probst. Doctoral Dissertation, Florida State University, Tallahassee, FL. 8. Kooriwa B. M. (1973) A reliable. standardized method for ultrastructural electron microscopic radioautography. His>ochemistry 3?, l--l?. 9. Kuwada S., Yin T. C. T., Haberly L. B. and Wickesberg R. E. (1980) Binaural interaction in the cat inferior colliculus: ohvsioloev and anatomv. In Phvsioloeical and Behavioral Studies in HearinpI (eds Vander Brink G. and Bilsen F. A.), bp: 40128. Delft UniGersity Press, belft. 10, Mar H. and Wight T. N. (1988) Colloidal gold immunostaining on deplasticized ultra-thin sections. J. Hjsr~Jchenf. Cytochem. 34, 1387-1395. il. Moiseff A. (1985) Intracellular recordings from owl inferior colliculus. Sot. Neurosci. Abstr. 11, 735. 12. Moore J. K. and Moore R. Y. (1987) Glutamic acid decarboxylase-like immunoreactivity in brainstem auditory nuclei of the rat. J. camp. Neural. 260, 157--174. 13. Mugnaini E. (1985) GABA neurons in the superficial layers of the rat dorsal cochlear nucleus: light and electron microscopic i&mu&ytochemistry. J. camp. Neural. 235,. 61-8 I. 14. Mupnaini E. and Oertel W. H. f1985) An atlas of the distribution of GABAergic neurons and terminals in the rat CNS as revealed by GAD immuno~st~h~mistry. In handbook o~Chemica1 ~euro~atomy, Vol. 4: GABA and ~europeptide.~ in the CNS, Part 1 (eds Bjorklund A. and Hokfelt T.), pp. 436605. Elsevier, New York. Nelson P. G. and Erulkar S. D. (1963) Synaptic mechanisms of excitation and inhibition in the central auditory pathway. J. Neurophysiol. 26, 908-923. Oberdorfer M. D., Parakkal M. H., Altschuler R. A. and Wenthold R. J. (1988) Ultrastructural localization of GABA-immunoreactive terminals in the anteroventral cochlear nucleus of the guinea pig. Hearing Res. 33, 229-238. Oliver D. L. (1984) Dorsal cochlear nucleus projections to the inferior colliculus in the cat: a light and electron microscopic study. J. camp. Neural. 224, 155-172. Oliver D. L. (1984) Neuron types in the central nucleus of the inferior colliculus that project to the medial geniculate body. Neuroscience 11, 409424. ^ _ .. . 19. Oliver D. L. (1985) Quantitative analyses of axonal endings in the central nucleus ot the mterlor COIIICUIUS ana distribution of SH-labeling after injections in the dorsal cochlear nucleus. J. camp. Neural. 237, 343-359. 20. Oliver D. L. (1987) Projections of the inferior colliculus from the anteroventral cochlear nucleus in the cat: possible substrates for binaural interactions. J. camp. Neural. 264, 2446. 21. Oliver D. L. and Beckius G. (1989) GABA immunocytochemistry and speculation on the inhibitory circuits in the inferior colliculus of the cat. Assoc. Res. ~to~aryngo~. Abstr. 12, 91. 22. Oliver D. L., Kuwada S., Yin T. C. T., Haberly L. B. and Henkel C. K. (1991) Dendritic and axonal mo~hology of HRP-injected neurons in the inferior colliculus of the cat. J. camp. Neural. 303, 75-100. 23. Oliver D. L. and Morest D. K. (1984) The central nucleus of the inferior colliculus in the cat. J. romp. Neural. 222, 237-264. 24. Oliver D. L., Nuding S. C. and Beckius G. (1988) Multiple cell types have GABA immunoreactivity in the inferior colliculus of the cat. Sot. Neurosci. Abstr. 14, 490.
GABA-labeled axonal endings in the inferior colliculus
463
25. Oliver D. L. and Shneiderman A. (1989) An EM study of the dorsal nucleus of the lateral lemniscus: inhibitory, commissural, synaptic connections between ascending auditory pathways. J. Neurosci. 9, 967-982. 26. Roberts R. C. and Ribak C. E. (1987) GABAergic neurons and axonal terminals in the brainstem auditory nuclei of the gerbil. J. camp.Neural. 258, 267-280. 27. Roberts R. C. and Ribak C. E. (1987) An electron microscopic study of GABAergic neurons and terminals in the central nucleus of the inferior colliculus of the rat. J. Neurocytol. 16, 3333345. 28. Saint Marie R. L. and Baker R. A. (1990) Neurotransmitter-specific uptake and retrograde transport of [‘Hlglycine from the inferior colliculus by ipsilateral projections of the superior olivary complex and nuclei of the lateral lemniscus. Brain Rex 524, 244-253. 29. Saint Marie R. L., Morest D. K. and Brandon C. J. (1989) The form and distribution of GABAergic synapses on the principal cell types of the ventral cochlear nucleus of the cat. Hearing Res. 42, 97-l 12. 30. Saint Marie R. L., Ostapoff E. M. and Merest D. K. (1989) Glycine-immunoreactive projection of the cat lateral superior olive: possible role in midbrain ear dominance. J. camp. Neuroi. 279, 382-396. 31. Seguela P., Geffard M., Buijs R. M. and Le MoaI M. (1984) Antibodies against (gamma)-aminobutyric acid: specificity studies and immunocytochemical results. Proc. nutn. Acad. Sci. U.S.A. 81, 3888-3892. 32. Shneiderman A. and Oliver D. L. (1989) An EM autoradiographic study of the projections from the dorsal nucleus of the lateral lemniscus. A possible source of inhibitory inputs to the inferior colliculus. J. camp. Neural. 286, 28847.
33. Shneiderman A., Oliver D. L. and Henkel C. K. (1988) Connections of the dorsal nucleus of the lateral lemniscus: an inhibitory parallel pathway in the ascending auditory system. J. camp. Neural. 276, 188-208. 34. Somogyi P. and Hodgson A. J. (1985) Antiserum to (gamma)-aminobutyric acid. III. Demonstration of GABA in Go@-impregnated neurons and in conventional electron microscopic sections of cat striate cortex. J. Histochem. Cytochem. 33, 249-257. 3.5. Somogyi P. and Soltesz I.
36. 37. 38. 39. 40. 41.
(I 986) Immunogold demonstration of GABA in synaptic terminals of intraceIiularly recorded, horseradish peroxidase-tilled basket cells and clutch cells in the cat’s visual cortex. ~euroscje~ee 19, 1051-1065. Thompson G. C., Cortez A. M. and Lam D. M, (1985) Localization of GABA immunoreactivity in the auditory brainstem of guinea pigs. Brain Rex 339, 119-122. Triller A., Cluzeaud F. and Kom H. (1987) Gamma-aminobutyric acid-containing terminals can be apposed to glycine receptors at central synapses. J. Cell Biol. 104, 947.-956. Van Der Want J. J. L. and Nunes Cardozo J. J. (1988) GABA immunoelectron microscopy study of the nucleus of the optic tract in the rabbit. J. camp. Neural. 271, 229-242. Wenthold R. J., Huie D., Altschuler R. A. and Reeks K. A. (1987) Glycine immunoreactivity localized in the cochlear nucleus and superior olivary complex. Neuroscience 22, 897-912. Wenthold R. J., Parakkal M. H., Oberdorfer M. D. and Altschuler R. A. (1988) Glycine receptor immunoreactivity in the ventral cochlear nucleus of the guinea pig. J. camp. Neural. 276, 423-435. Wenthold R. J., Zempel J. M., Parakkal M. H., Reeks K. A. and Altschuler R. A. (1986) Immunocytochemical localization of GABA in the cochlear nucleus of the guinea pig. Brain Res. 380, 7-18. (Accepted 28 June 1991)
NeuroscienceVol. 46, No. 2, pp. 455-463, I992 Printed in Great Britain
FINE STRUCTURE OF GABA-LABELED AXONAL ENDINGS IN THE INFERIOR COLLICULUS OF THE CAT: IMMUNOCYTOCHEMISTRY ON DEPLASTICIZED ULTRATHIN SECTIONS D.L.
OLIVER*
andG.E.
BECKIUS
Department of Anatomy and Center for Neurological Sciences, The University of Connecticut Health Center, Farmington, CT 06030, U.S.A. Abstract-Antisera to GABA conjugates and postembedding techniques were used to identify GABA-containing axonal endings at the electron microscopic level in the inferior colliculus. Over 90% of the GABA-labeled axonal endings bad a similar mo~hoIogy. They contained pleomorphic synaptic vesicles and made s~metrical synapses. The exceptional endings contained round vesicles and made s~rnet~~~ synaptic contacts or had pleomorphic vesicles with asymmetrical contacts. The majority of GABA-labeled axonal endings synapsed on dendrites; however, a few labeled axosomatic synapses were also found. Potential sources for these GABAergic synapses are neurons intrinsic to the inferior colliculus or from the dorsal nucleus of the lateral lemniscus. These findings suggest a basic similarity for most GABAergic endings in the inferior colliculus despite their possible origin from different cell types.
Inhibitory neurotransmitters, such as GABA, may play a significant role in shaping the neural responses in the inferior colliculus. Nelson and Erulkar” and othersg*” showed that inhibitory and excitatory postsynaptic potentials were common components of intracellular neural responses in the inferior colliculus. GABA may mediate some of these inhibitory potentials. Glutamate decarboxylase (GAD) has been shown to be present in the inferior colliculus in biochemical assays3 and in immun~ytochemi~al studies. ‘2*‘4~21~24*26~27~36 Moreover, at least one ascending projection to the inferior colliculus from the dorsal nucleus of the lateral lemniscus may use GABA as a transmitter.‘v32.33 Despite the potential importance of inhibitory synapses in the inferior colliculus, they have received little study at the electron microscopic level. Thus, the goal of this study was to identify synapses in the inferior colliculus where GABA may act as the neurotransmitter. A modification of the postembedding immunocytochemical procedure was developed for use with the electron microscope. This technique allowed the unobscured fine structure of synapses to be related to the presence of a particular transmitter. Since GABAergic axonal endings may arise from more than one source, this approach helped to identify morphologically distinct GABAergic endings.
iments.‘7~20*3Z The two additional animals were deeply anesthetized with sodium pentobarbital(40 mg/kg) and perfused through the aorta with mixed aldehyde fixatives similar to those used previously.“~20~32A washout solution (50 ml) contained 0.12 M phosphate buffer, pH 7.4, 2% sucrose, 0.05% lidocaine, and 0.004% CaCI,. The fixatives contained the same buffer and calcium salts and consisted of a dilute first solution (500 ml of 1% paraformaldehyde, 1% glutaraldehyde) and a second, more concentrated fixative (1000 ml of 2% paraformaldehyde and 3% glutaraldehyde). Histology
Sections, 7t.KlOO~m thick, were cut from the inferior colliculus in the transverse plane with a Vibratome. Every fifth section was processed for electron microscopy and the rest for light microscopy. The sections for electron microscopy were rinsed in phosphate buffer (0.12 M, pH 7.4) containing 6% sucrose, postfixed in 1% buffered osmium tetroxide, block-stained in 2% uranyi acetate, dehydrated in acetone, and embedded in Durcupan ACM (Pofyscience). The remaining light microscopic sections were either stained with Cresyl Violet or reacted with GABA antisera as free-floating sections.39,4’ Binding of the GABA antibody was revealed in the light microscopic sections with biotinylated anti-rabbit IgG coupled to an avidin-peroxidase complex (Vectastain, Vector Labs). Polyclonal rabbit antisera were used. One antiserum, donated by Dr Robert Wenthold (NIH), was affinitypurified antiserum against bovine serum albumin~utaraldehyde conjugates of GABA.399A’A second GABA antiserum (Pel-Freez. P4MV72~1) was raised against glutaraldeh~de conjugates of bovine serum alb;min, hemoglobin, and poly-lysine used in successive boosters3’ The cross-reactivity of both antibodies has been well characterized.“~39~4’ To prepare samples for colloidal gold immunostaining and electron microscope analysis, serial lOO-nm-thick sections were cut and collected in pairs with a wire loop on Formvar-coated glass slides. This technique was similar to that used routinely for electron microscopic autoradiography. *vi7The day before immunos~i~ng, the sections were transferred to grids by etching the Formvar coating with a dilute solution of hydrofluoric acid, floating off the Formvar
EXPERIMENTAL PROCEDURES The experimental material was obtained from six normal, adult cats, four of which were part of previous exper*To whom correspondence should be addressed. DNLL, dorsal nucleus of the lateral lemniscus; GAD, glutamate decarboxylase.
Abbreuiatims:
455
456
D. L. OLIVERand G. E.
film and section onto a water bath, and picking up the sections on grids. For each antibody dilution, a minimum of three sequential pairs of serial sections on three separate grids was used. The first pair of “normal” sections, collected on 200-mesh thin bar grids, was stained with 4% many1 acetate and 0.1% lead citrate.” The second pair of “immunoreactive” sections was collected on 300-mesh nickel grids for immunostaining. The third pair of “control” sections (300-mesh nickel grids) was handled like the second pair but was not exposed to the primary antibody. Colloidal gold immunostaining on deplasticized thin sections followed the method of Mar and Wight.” Thin sections on grids were deplasticized by immersion for 10 min in 45% sodium hydroxide in methanoP and were rehydrated in methanolic solutions. The grids were then floated on droplets of 0.12 M phosphate buffer, pH 7.4. This buffer was used in all aqueous solutions for immunohistochemistry and as a holding step. After blocking in 10% normal goat serum, the grids-were transferred without rinsing to primary antibodv at 4°C at dilutions of 1:200. 1: 500. and 1: 1000and incubated overnight (17-19 h). Control sections were kept in normal goat serum during this time. Following buffer rinses, the immunoreactive and control grids were exposed to a I : IO dilution of colloidal gold anti-rabbit IgG (E-Y Labs) for 4 h and postfixed successively in 2% glutaraldehyde, 1% osmium tetroxide, and 2% uranyl acetate solutions. Finally, the grids were dehydrated in ethanol, dipped in 2% Epon in absolute ethanol, polymerized overnight at 60°C on edge,‘” and counterstained with uranyl acetate and lead citrate. Analysis
Ultrathin sections processed for immunocytochemistry were compared to control and normal sections on a JEOL 1OOCXelectron microscope. The location of the sections relative to the grid bars was drawn, then the sections were photographed at x 160&2200. Sample grid squares were chosen at random or because a cell body with a visible nucleolus was present. Immunolabeled endings were at least x 5-10 above background, but the absolute labeling was not routinely counted. All labeled endings within a grid square were photographed at x 10,000. After the labeled endings were identified on the low magnification pictures, these same axonal endings were located in the adjacent normal sections and photographed at x lO,OOO-16,000. Most of the unlabeled endings were identified in the same micrographs as the labeled endings. RESULTS Immunoreactivity for GABA was studied in electron microscopic samples from the pars centralis in the central nucleus of the inferior colliculus. This region was rich in immunoreactive profiles at the light microscopic level (Refs 21 and 24; Oliver et al., in preparation). The electron microscopic samples were
BECKIUS
adjacent to areas in light microscopic sections that contained immunoreactive axons, puncta, and cell bodies. Comparisons of immunoreactive, control, and normal sections
After immunoreaction with antibodies to GABA, some axonal endings and myelinated axons were heavily labeled compared to control sections. Serial sections were used to compare the immunoreaction to control conditions in the same axonal endings. Figure 1 illustrates such an ending in three sections spaced 300 nm apart. The sections in panels A and B were prepared identically except that B was exposed to Pel-Freez GABA antibody. In the control section. the colloidal gold was distributed evenly throughout all tissue compartments. In contrast, the section exposed to antibody (Fig. 1B) showed a heavy accumulation of colloidal gold over the axonal ending (PL) and over several myelinated axons (LA). In the adjacent serial ultrathin section, the fine structure of the labeled axonal endings could be examined. Ultrastructural detail in the immunoreactive sections was lost because of the deplasticizing and the heavy gold labeling. In contrast, the morphology of the synaptic vesicles and synaptic contacts was intact in the adjacent normal sections. The section pictured in Fig. 1C showed that the labeled axonal ending (Fig. IB) contained pleomorphic synaptic vesicles. Most GABA-labeled synaptic vesicles
endings contained pleomorphic
Tables 1 and 2 describe the vesicular and synaptic morphology for the 272 endings in the present sample. The labeled axonal endings represented over half of the sample, and most of the labeled endings (97%) contained pleomorphic synaptic vesicles (Table 1). As shown in Fig. 2, the endings with pleomorphic vesicles were well labeled with the Wenthold antisera as compared to an ending with round vesicles (LR) and a dendrite in the same section. Labeling of these latter structures was at control levels. Only four labeled endings in the sample contained round synaptic vesicles (Table 2); these endings made symmetrical synaptic contacts in the two cases where the synaptic density was visible.
Table 1. Axonal endings with pleomorphic vesicles: GABA-labeled versus unlabeled Labeled 88-06-38 Labeled 80-07-68 Total labeled Unlabeled 88-06-38 Unlabeled 88-07-68 Total unlabeled
No. 50 91 141
A/D 4 0 4
A/S
1
S/D 17 30 47
II 9 20
0 I I
0 0 0
4 2 6
1 0
N
6 7
U/D 9 14 23
U/S 3 I 10
15 34 49
0 0 0
3 I 4
1 0 1
3 5 8
SIS
I
No., number of endings counted; A/D, asymmetric contacts on dendrite; A/S, asymmetric contacts on soma; S/D, symmetric contacts on dendrite; S/S, symmetric contacts on soma; U/D, unclear contacts on dendrites; U/S, unclear contacts on soma; N, no contact visible
GABA-labeled axonal endings in the inferior colliculus Table 2. Axonal endings with round vesicles:GABA-labeled versus unlab&d*
Unlabeled 88-06-38 Unlabeled 88-07-68 Total unlabeled
44 62 106
14 23 37
0 0 0
1 1 2
0 0 0
13 12 25
2
I 3
14 25 39
*See Table 1 for abbreviations.
Several types of endings were usually unlabeled. The largest group of endings in the sample (39%) were unlabeled and contained round vesicles. Two sizes of round vesicle were evident in micrographs.
Endings with larger round vesicles (Figs 24) were easily distinguished from endings with smaller round vesicles (RR’) in micrographs where both were present (Fig. 3). In most cases, these endings were
Abbreviations used in rhe jigures
A CB D F
unlabeled axon cell body dendrite flattened/pleomorphic synaptic vesicles
LR LA PL R
large, round synaptic vesicles labeled myelinated axon pleomorphic synaptic vesicles small round synaptic vesicle
Fig. 1. Control, immunoreactive, and normal sections spaced 3OOmnapart through the same axonal ending (PL). (A) The control section shows that gold particles are distributed randomly in a deplasticixed section. (B) After exposure to GABA antibody (1: 500, Pel-Free& colloidal gold label accumulatesover the ending and two myelinated axons (LA). The cell body (CB) may be lightly labeled. Other axons and glial processes between the labeled ending and labeled axons display far fewer gold particles. (C) In the normal section, pleomorphic synaptic vesiclesare visible.The ending forms a symmetricalsynapticcontact (arrowheads)on the somatic spine pictured in C. Case nos 88-07-68-39,-41, -42. Magnifications:(A), (B) x 25,380;(C) x 32,806.Scale bars = 0.5pm.
458
D L. OLIVERand G. E. BECKWS
Fig. 2. Axodendritie synapses. (A) In a normal section, three endings (PL) with pleomorphic synaptic vesicles make s~rnmetricalsynapses (arrowheads) on a dendrite (D). A fourth ending contains large, round vesicles (LR). (8) Immunoreactive section, 2OOnmfrom section in A, shows the same endings after exposure to GABA antib~y (I : 500,Wenthold), The pieiomorphic synaptic vesicle endings are labeled. while the large. round synaptic vesicle ending and dendrite are not. Case nos 88-06-38-96,-94. Magnifications:(A) x 35,062,(B) x 28,952.Scale bars = 0.5 urn. equally unlabeled. However, one ending (R’) in Fig. 3 dispiayed more gold binding than the large and smaller round vesicle endings. This level of signal was somewhat higher than surrounding structures, but it was less than the labeled endings with pleomorphic vesicles (Fig. 3). Besides endings with round vesicles, unlabeled endings with pleomo~hic vesicles were observed (7% of the sample). One of the 20 such endings is shown in Fig. 4. It is designated F to distinguish it from the labeled ending in the micrograph. Labeled endirrgs us~af~ymade s?~~rnetri~R~ spaptir contacts on dendrites Tables 1 and 2 show the types of synaptic contacts
made by labeled and unlabeled endings. Of the 141 labeled endings that contained p~eomorphic vesicles, 38% made symmetrical contacts while 58% had no contact or the contact was unclear. Figure 2 illustrates such a synaptic contact on a dendrite (arrowheads). Only a few (about 4%) labeled endings made asymmetrical synapses, and each of these endings contained pleomorphic synaptic vesicles. In contrast, unlabeled endings with round vesicles made asymmetrical synapses in all but two instances where the contact could be identified. Roughly 50% of the GABA-labeled endings were presynaptic to dendrites (Tables 1 and 2). Axosomatic synapses were also found but were less frequent. Only seven labeled endings with pleomo~hic
GABA-labeled axonal endings in the inferior colliculus
Fig. 3. Labeling density. (A) After i~unoreactian with Pel-Freez GABA antibody, two axonal endings (PL) are heavily labeled.Other endings(LR, R) are unlabeled.All endings in the micrographs are outlined with triangles. One axonal ending (R’) and the dendrite (D) may be labeled above background. (B) In the normal section 300nm away, the heaviest labeled endings contain pleomorphic vesicles (PL). Other endings contain either round (R) or large round (LR) vesicles Case nos 88-07-68-79,-76. Ma~ifications: (A) x 27,166;(B) 33,276.Scale bars = 0.5 pm.
459
460
D. L. OLWERand G. E. BECKIUS
Fig. 4. Axosomatic endings. (A) After immunoreaction, only one (PL) of three axosomatic endings pictured are labeled. A nearby axon (LA) is heavily labeled, while other endings {F, LR) and axons (A) are unlabeled.(B)The labeledending contains pleomorphicvesicles(PL) and makes a symmetricalsynapse on the dendrite (arrowhead). One unlabeled ending contains pleomorphic vesicles (F), and the other contains large round vesicles(LR) and makes an axosomaticsynapse (arrowhead),Samesections as Fig. 2. Magnifications:(A) x 28,670,(B) x 42,I 12. Scale bars = 0.5 pm.
GABA-labeled axonai endings in the inferior colliculus
461
colliculus of the rat. Most of their data concerned the identification of GAD-labeled cell types in the inferior cofliculus. However, they did report GADlabeled terminals containing pleomorphic or flattened synaptic vesicles that made symmetrical synapses. DISCUSSION They distinguished different types of labeled endings by the size and shape of the ending, not by the This study is the first to use postembedding techsynaptic vesicles or synaptic contacts. Thus, they did niques to examine the fine structure of GABA-connot report any exceptional types of GAD-labeled taining axonal endings in the inferior colliculus. Most synaptic endings. GABA-labeled axonal endings contained pleomorAll GABAergic endings are not identical in other phic synaptic vesicles. When the synaptic contact systems (e.g. Refs 6 and 38). Different GABAergic could be identified, GABA-labeled endings made symmetrical synapses. However, there were excep- synapses have been distinguished by vesicle size, the presence of dense-core vesicles, and by vesicle density. tions. Small numbers of endings contained round vesicles and made symmetrical synaptic contacts or Quantitative analysis of immunolabeled endings in the inferior colliculus will be carried out in the near had pleomorphie vesicles with asymmetrical contacts. Most GABA-labeled axonal endings synapsed on future since several types of endings with pleomorphic vesicles have been identified previously.” dendrites; however, a number of labeled axosomatic Not all of the endings with pleomorphic or flatsynapses were also found. These findings suggest a basic similarity for most GABAergic endings in the tened vesicles were immunolabeled in the inferior inferior colliculus despite the likelihood that they colliculus in the present study and in the previous arise from different cell types. study in the rat.27 It is possible that these unlabeled endings use a different transmitter such as glycine. Comparison of data from pre-embedding and postBoth the projections from the ipsilateral lateral suembedding studies perior oIive7*28,30 and the ventral nucleus of the lateral Pre-embedding t~h~ques at the electron microlemniscus428 have been implicated as glycinergic. Moreover, the presynaptic terminals opposite glycinscopic level have been the mainstay of previous ergic synapses have been reported to contain either immunocytochemical studies in the inferior collicu1us2’ and elsewhere in the auditory system.2~13*16.29,41 flattened or pleomorphic synaptic vesicles.5,a In some The strong immunocytochemical reactions carried cases, GABA-containing terminals may appose out before embedding tend to obscure the fine strucglycine receptors at central synapses3’ ture of the axonaf ending. Postembedding methods, Do GABAergic endings in the inferior coltic~lus come as used here and in other systems (e.g. Refs 34 and 35), preserve the fine structure. In the present study, from deferent sources?
vesicles made axosomatic synapses with symmetrical contacts (Table 1). Figures 1 and 4 (PL) illustrate this type of axosomatic, symmetrical synapse.
the use of deplasticized sections allowed for maximal immunoreactivity. In addition, it allowed the fine structure to be analysed on adjacent normal sections that were completely undisturbed by histochemical reactions. Are there d@erent types of GABAergic synapses in the inferior col~jcu~~s? Although most GABAergic endings in the inferior cofliculus were similar, there could be variations of GABAergic endings. A small number of labeled endings in this sample did not have either pleomorphic synaptic vesicles or symmetrical synaptic junctions. These exceptional endings were found in the tissue studied primarily with the Wenthold antibody. Since both the background and signal were lower with that antiserum, the exceptional axonal endings may have been easier to detect. Cross-reactivity of both antisera with other possible transmitters or transmitter precursors appear to be low (see Experimental Procedures). It is likely that these exceptional endings represent genuine immunolabeling for GABA and could represent axons from other GABAergic cell types (see below). Similar to the present results, Roberts and Ribakz7 reported observations of GAD fabeling in the inferior
There may be intrinsic and extrinsic sources for GABAergic synapses in the inferior colliculus. Within the inferior colliculus, there may be several cell types that use GABA as a transmitter. Two major cell classes, disc-shaped and stelfate, are identified in Golgi preparations,23 and both types emit axonal collaterals within the inferior colliculus.22 Both GABAergic disc-shaped and stellate cells have been identified.2’,24*27Even though multiple GABAergic cell types may exist within the inferior colliculus, the majority of GABAergic synapses in the inferior colliculus still have a similar morphology. Outside of the inferior colliculus, the dorsal nucleus of the lateral lemniscus (DNLL) is another potential source for GABAergic projections.’ Previous studies25+32 show that most of the endings from the DNLL have pleomorphic vesicles and make symmetrical synapses. These endings from DNLL are similar to the majority of the GABA-labeled endings in the present study and account for one-third of such endings in the central nucleus of the inferior colliculus. This is uniformly true for the projection from the contralateral DNLL to the inferior colliculus. However, projections from the ipsilateral DNLL to the colhcufus are heterogeneous. Some of the ipsilateral projections from DNLL may not use GABA as a
462
D. L. OLIVER and
transmitter since one-third of those endings contain round synaptic vesicles. In other studies, axonal endings from the dorsal cochlear nucleus’7,‘9 and ventral cochlear nucleus2’ are identified in the inferior colliculus at the electron microscopic level and contain round vesicles and make asymmetric synapses. This type of ending was never GABA-immunoreactive in the present study. Thus, the majority of the extrinsic GABA-labeled axons are likely to arise only from the DNLL and have a morphology similar to the intrinsic GABA-labeled axons. Postsynaptic targets
Both disc-shaped and stellate classes of cells are richly innervated by presumed GABAergic synapses
G. E. BECKIUS
on their somata and dendrites. Both classes of cells have projections ascending to the medial geniculate body and receive symmetrical, axosomatic synapses from endings with pleomorphic vesicles.‘8 Those synapses are the same type labeled with GABA antisera in the present study. Many of the axosomatic GABAergic endings may be intrinsic in origin since most of the endings from DNLL terminate primarily on dendrites.32 Ackno~iedgements-This
research was sponsored by the Deafness Research Foundation and by NIH grant LXOO189. Thanks arc due to Dr Robert Wenthold for the GABA antiserum and to Drs R. Saint Marie and I. Schwartz for helpful comments on the manuscript.
1. Adams J. C. and Mugnaini E. (1984) Dorsal nucleus of the lateral lemniscus: a nucleus of GABAergic projection neurons. Brain Res. Bull. 13, 585-590.
2. Adams J. C. and Mugnaini E. (1987) Patterns of glutamate decarboxylase immunostaining in the feline cochlear nuclear complex studied with silver enhancement and electron microscopy. J. camp. Neural. 262, 375-101. 3. Adams J. C. and Wenthold R. J. (1979) Distribution of putative amino acid transmitters. choline acetyltransferase and glutamate decarboxylase in the inferior colliculus. Neuroscience 4, I947-1951. 4. Adams J. C. and W~nthold R. J. (1987) Immunostaining of ascending auditory pathways with glycine antiserum. Assoc. Res. ~ioiaryngol. Abs/r. 10, 63. 5. Altschuler R. A., Betz H., Parakkal M. H., Reeks K. A. and Wenthold R. J. (1986) Identification of glycinergic synapses in the cochlear nucleus through immunocytochemical localization of the postsynaptic receptor. Bruin Res. 369,316-320. 6. De Zeeuw C. I., Holstege J. C., Ruigrok T. J. H. and Voogd J. (1989) Ultrastructural study of the GABAergic. cerebellar, and mesodiencephalic innervation of the cat medial accessory olive: anterograde tracing combined with immunocytochemistry. 1. camp. Neuroi. 284, 12-35. 7. Hutson K. A. (1988) Connections of the auditory midbrain: efferent projections of the dorsal nucleus of the lateral lemniscus, the nucleus sagulum, and the origins of the GABAergic commissure of Probst. Doctoral Dissertation, Florida State University, Tallahassee, FL. 8. Kooriwa B. M. (1973) A reliable. standardized method for ultrastructural electron microscopic radioautography. His>ochemistry 3?, l--l?. 9. Kuwada S., Yin T. C. T., Haberly L. B. and Wickesberg R. E. (1980) Binaural interaction in the cat inferior colliculus: ohvsioloev and anatomv. In Phvsioloeical and Behavioral Studies in HearinpI (eds Vander Brink G. and Bilsen F. A.), bp: 40128. Delft UniGersity Press, belft. 10, Mar H. and Wight T. N. (1988) Colloidal gold immunostaining on deplasticized ultra-thin sections. J. Hjsr~Jchenf. Cytochem. 34, 1387-1395. il. Moiseff A. (1985) Intracellular recordings from owl inferior colliculus. Sot. Neurosci. Abstr. 11, 735. 12. Moore J. K. and Moore R. Y. (1987) Glutamic acid decarboxylase-like immunoreactivity in brainstem auditory nuclei of the rat. J. camp. Neural. 260, 157--174. 13. Mugnaini E. (1985) GABA neurons in the superficial layers of the rat dorsal cochlear nucleus: light and electron microscopic i&mu&ytochemistry. J. camp. Neural. 235,. 61-8 I. 14. Mupnaini E. and Oertel W. H. f1985) An atlas of the distribution of GABAergic neurons and terminals in the rat CNS as revealed by GAD immuno~st~h~mistry. In handbook o~Chemica1 ~euro~atomy, Vol. 4: GABA and ~europeptide.~ in the CNS, Part 1 (eds Bjorklund A. and Hokfelt T.), pp. 436605. Elsevier, New York. Nelson P. G. and Erulkar S. D. (1963) Synaptic mechanisms of excitation and inhibition in the central auditory pathway. J. Neurophysiol. 26, 908-923. Oberdorfer M. D., Parakkal M. H., Altschuler R. A. and Wenthold R. J. (1988) Ultrastructural localization of GABA-immunoreactive terminals in the anteroventral cochlear nucleus of the guinea pig. Hearing Res. 33, 229-238. Oliver D. L. (1984) Dorsal cochlear nucleus projections to the inferior colliculus in the cat: a light and electron microscopic study. J. camp. Neural. 224, 155-172. Oliver D. L. (1984) Neuron types in the central nucleus of the inferior colliculus that project to the medial geniculate body. Neuroscience 11, 409424. ^ _ .. . 19. Oliver D. L. (1985) Quantitative analyses of axonal endings in the central nucleus ot the mterlor COIIICUIUS ana distribution of SH-labeling after injections in the dorsal cochlear nucleus. J. camp. Neural. 237, 343-359. 20. Oliver D. L. (1987) Projections of the inferior colliculus from the anteroventral cochlear nucleus in the cat: possible substrates for binaural interactions. J. camp. Neural. 264, 2446. 21. Oliver D. L. and Beckius G. (1989) GABA immunocytochemistry and speculation on the inhibitory circuits in the inferior colliculus of the cat. Assoc. Res. ~to~aryngo~. Abstr. 12, 91. 22. Oliver D. L., Kuwada S., Yin T. C. T., Haberly L. B. and Henkel C. K. (1991) Dendritic and axonal mo~hology of HRP-injected neurons in the inferior colliculus of the cat. J. camp. Neural. 303, 75-100. 23. Oliver D. L. and Morest D. K. (1984) The central nucleus of the inferior colliculus in the cat. J. romp. Neural. 222, 237-264. 24. Oliver D. L., Nuding S. C. and Beckius G. (1988) Multiple cell types have GABA immunoreactivity in the inferior colliculus of the cat. Sot. Neurosci. Abstr. 14, 490.
GABA-labeled axonal endings in the inferior colliculus
463
25. Oliver D. L. and Shneiderman A. (1989) An EM study of the dorsal nucleus of the lateral lemniscus: inhibitory, commissural, synaptic connections between ascending auditory pathways. J. Neurosci. 9, 967-982. 26. Roberts R. C. and Ribak C. E. (1987) GABAergic neurons and axonal terminals in the brainstem auditory nuclei of the gerbil. J. camp.Neural. 258, 267-280. 27. Roberts R. C. and Ribak C. E. (1987) An electron microscopic study of GABAergic neurons and terminals in the central nucleus of the inferior colliculus of the rat. J. Neurocytol. 16, 3333345. 28. Saint Marie R. L. and Baker R. A. (1990) Neurotransmitter-specific uptake and retrograde transport of [‘Hlglycine from the inferior colliculus by ipsilateral projections of the superior olivary complex and nuclei of the lateral lemniscus. Brain Rex 524, 244-253. 29. Saint Marie R. L., Morest D. K. and Brandon C. J. (1989) The form and distribution of GABAergic synapses on the principal cell types of the ventral cochlear nucleus of the cat. Hearing Res. 42, 97-l 12. 30. Saint Marie R. L., Ostapoff E. M. and Merest D. K. (1989) Glycine-immunoreactive projection of the cat lateral superior olive: possible role in midbrain ear dominance. J. camp. Neuroi. 279, 382-396. 31. Seguela P., Geffard M., Buijs R. M. and Le MoaI M. (1984) Antibodies against (gamma)-aminobutyric acid: specificity studies and immunocytochemical results. Proc. nutn. Acad. Sci. U.S.A. 81, 3888-3892. 32. Shneiderman A. and Oliver D. L. (1989) An EM autoradiographic study of the projections from the dorsal nucleus of the lateral lemniscus. A possible source of inhibitory inputs to the inferior colliculus. J. camp. Neural. 286, 28847.
33. Shneiderman A., Oliver D. L. and Henkel C. K. (1988) Connections of the dorsal nucleus of the lateral lemniscus: an inhibitory parallel pathway in the ascending auditory system. J. camp. Neural. 276, 188-208. 34. Somogyi P. and Hodgson A. J. (1985) Antiserum to (gamma)-aminobutyric acid. III. Demonstration of GABA in Go@-impregnated neurons and in conventional electron microscopic sections of cat striate cortex. J. Histochem. Cytochem. 33, 249-257. 3.5. Somogyi P. and Soltesz I.
36. 37. 38. 39. 40. 41.
(I 986) Immunogold demonstration of GABA in synaptic terminals of intraceIiularly recorded, horseradish peroxidase-tilled basket cells and clutch cells in the cat’s visual cortex. ~euroscje~ee 19, 1051-1065. Thompson G. C., Cortez A. M. and Lam D. M, (1985) Localization of GABA immunoreactivity in the auditory brainstem of guinea pigs. Brain Rex 339, 119-122. Triller A., Cluzeaud F. and Kom H. (1987) Gamma-aminobutyric acid-containing terminals can be apposed to glycine receptors at central synapses. J. Cell Biol. 104, 947.-956. Van Der Want J. J. L. and Nunes Cardozo J. J. (1988) GABA immunoelectron microscopy study of the nucleus of the optic tract in the rabbit. J. camp. Neural. 271, 229-242. Wenthold R. J., Huie D., Altschuler R. A. and Reeks K. A. (1987) Glycine immunoreactivity localized in the cochlear nucleus and superior olivary complex. Neuroscience 22, 897-912. Wenthold R. J., Parakkal M. H., Oberdorfer M. D. and Altschuler R. A. (1988) Glycine receptor immunoreactivity in the ventral cochlear nucleus of the guinea pig. J. camp. Neural. 276, 423-435. Wenthold R. J., Zempel J. M., Parakkal M. H., Reeks K. A. and Altschuler R. A. (1986) Immunocytochemical localization of GABA in the cochlear nucleus of the guinea pig. Brain Res. 380, 7-18. (Accepted 28 June 1991)
