0306-4522/91 $3.00+ 0.00 Perpmon Prea plc
Neuroscience Vol. 43, No. 2/3, pp. 513-529, 1991 Printed in Great Britain
© 1991 IBRO
EVIDENCE FOR INTRINSIC EXPRESSION OF ENKEPHALIN-LIKE IMMUNOREACTIVITY A N D OPIOID BINDING SITES IN CAT SUPERIOR COLLICULUS D. M. BERSON,*t A. M. GRAYBIEL,~W. D. BOWL~* and U A. THOMPSON§ *Division of Biology and Medicine and §Department of Psychology, Brown University, Providence, RI 02912, U.S.A. ~:Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology, Cambridge, MA 02139, U.S.A. Almract--We have investigated the cellular localization of opioid peptides and binding sites in the cat's superior colliculus by testing the effects of retinal deafferentation and intracollicular excitotoxin lesions on patterns of enkephalin-like immunostaining and opiate receptor ligand binding. In normal cats, enkephalin-like immunoreactivity marks a thin tier in the most dorsal stratum griseum superficiale, small neurons of the stratum griseum superliciale, and patches of fibers in the intermediate and deeper gray layers. Eliminating crossed retinotectal afferents by contralaterul eye enucleation had little immediate effect on this pattern, although chronic eye enucleation from birth did reduce immqnoreactivity in the superficial layers. By contrast, fiber-sparing destruction of collicular neurons by the excitotoxins N-methyl-D-asparrate and ibotenic acid virtually eliminated enkephalin-like immunoreaetivity in the neuropil of the upper stratum griseum superficiale, presumably by kilting enkephalinergi¢ cells of the superficial layers. Such lesions did not eliminate the patches of enkephalin-like immunoreactivity in the deeper layers. In normal cats, opiate receptor ligand binding is dense in the stratum griseum superliciale, particularly in its upper tier, and moderately dense in the intermediate gray layer. Contralateral eye removal had no detectable effect on the binding pattern, but excitotoxin lesions of the colliculus dramatically reduced binding in both superficial and deep layers. Some ligand binding, including part of that in the upper stratum griseum superficiale, apparently survived such lesions. Similar effects were observed in the lateral genicuiate nucleus: enucleation produced no change in binding, whereas excitotoxin lesions greatly reduced specific opiate binding. We conclude that in the superficial collicular layers, both enkephalin-like opioid peptides and their membrane receptors are largely expressed by neurons of intrinsic collicular origin. The close correspondence between the location of these intrinsic opioid elements and the tier of retinal afferents terminating in the upper stratum griseum superficiale further suggests that opiatergic interneurons may modulate retinotectal transmission postsynaptically.
Opioid peptides and opiate receptors are of multiple types and are distributed widely in the brain, implying a diversity of functions. They arc found not only in structures linked to nociception and mood, but also in a wide array of other regions where their physiological significance is less certain. 2 ~ A case in point is the superior colliculus, a midbrain visual structure that participates in the control of saccadic eye movements. Autoradiographic studies have demonstrated very high concentrations of opiate binding sites in the superior colliculus in a variety of species (e.g. see Refs 26, 37, 54 and 60). Opioid receptor ligand binding is densest in the retinorecipi?To whom correspondence should be addressed. Abbreviations: A, layer A of dorsal lateral geniculate nucleus; AI, layer AI of dorsal lateral geniculate nucleus; DAGO, ISHHD-AIa2,N-Me-Phe4,Gly(ol)S]-enkephafin; DPDPE, [3l-l]-{D-Pen~-~kephalin; HRP, horseradish peroxidase; LGv, ventral lateral geniculate nucleus; MG, medial geniculate nucleus; PAG, periaqueductal gray; PAP, peroxidue-antiperoxidase; SGI, stratum griseum intermedial¢; SGS, stratum grise-m superticiale; SO, stratum opticum. 513
ent superficial layers of the colliculus, but significant binding is also found in the deeper muitisensory and premotor layers. In most species, immunohistochemical studies have revealed only scattered opioid peptide-positive neurons and fibers in the colliculus,21,2~'61 raising questions about the functional significance of the high opiate ligand binding densities. In the cat, however, substantial cnkephalin-like immunoreactivity has been observed in the stratum griseum superficiale (SGS) and in sets of regularly spaced patches in the deeper layers. 19 Neither the identity of opiatergic elements in the cat's colliculus nor their targets have been fully established, but available data raise some intriguing possibilities. In the superficial layers, enkephalin-like immunoreactivity is heavily concentrated in a thin sheet in the upper part of the SGS (the "upper S G S ' ) . This enkephalin-positive sheet closely matches the tier of densest retinal input) sa4 an input which apparently originates in a subset of contralateraily projecting retinotectal ganglion cells, s,'3,32 The correspondence is topographic as well as laminar, because both the retinal input and the enkephalin-like immuno-
514
D.M. BERSONet al.
staining in the upper SGS are weakest in the area centralis representation and exhibit a prominent gap in the representation of the blind spot. 7ajS'lg':4~ This correspondence raises the possibility that enkephalins or related peptides are contained within the terminals of certain contralaterally projecting retinocollicular afferents. There are precedents for the localization of neuropeptides in subsets of retinal ganglion cells, including some of those that project to the superior colliculus or optic tectum in rabbits, frogs and birds. W2"~'~9 A second potential extrinsic source for the enkephalin-like immunoreactivity is the crossed projection of the parabigeminal nucleus. This afferent system has a laminar and topographic distribution nearly identical to those of the dense contralateral retinal input to the upper SGS and the enkephalinlike immunostaining there. ~ Arguing against a localization within retinal or parabigeminal afferents, however, is the fact that enkephalin-like immunoreactivity has not as yet been observed in ganglion cells or in parabigeminal neurons in the cat. ~9"~N o r did an ultrastructural immunohistochemical study of the cat superior colliculus find any evidence for enkephalinlike immunostaining in presumed retinotectal afferent terminals. 46 An alternative is that the enkephalin-like immunoreactivity is largely of intrinsic collicular origin. There are substantial numbers of enkephalin-immunoreactive neuronal cell bodies in the superficial gray layer, especially in and near the upper SGS. j9"46Many of these appear to be horizontal cells, which extend their dendrites parallel to the coilicular surface for considerable distances. At the ultrastructural level, much of the enkephalin-like immunoreactivity of the upper SGS appears to be contained within dendritic profiles. 4~ in the deeper layers, the fibrous composition of the patches of enkephalin-like immunoreactivity suggest an origin in collicular afferents, many of which exhibit patchy distributions in the cat. ~:9 ~ However, they could arise, instead, from local intrinsic enkephalinergic neurons or tYom enkephalin-containing neurons of the superficial layers; interlaminar connections are known to extend from the superficial layers to the deeper strata. '~ There is equal uncertainty about the sources of opiate ligand binding in the superior colliculus. Retinotectal afferents represent one candidate source in the superficial layers. The close match between the patterns of enkephalin-like immunoreactivity and contralateral retinal input suggests that collicular opioids might act directly on retinal terminals. Indeed, binding of opiate ligands is heaviest in the tier of densest retinal terminations. ~5'~'~° Presynaptic opiate receptors have been localized to retinal afferents in the rat accessory optic nuclei) To distinguish a m o n g these possibilities, we studied the effects of destroying retinal afferents or intrinsic collicular neurons on the patterns of enkephalin-likc immunoreactivity and opiate receptor ligand binding in the superior colliculus.
EXPERIMENTAL PROCEDURES
The distribution of enkephalin-like immunoreactivity in the superior collieulus was studied by the peroxidase antiperoxidase (PAP) technique ~ in tissue from eight adult cats (Liberty Laboratories, Liberty Corner. N J), two with unilateral eye enucleations and six with excitotoxin lesions of the colliculus. Opiate binding site distributions were studied by in vitro autoradiography in tissue from four cats; one of these had undergone unilateral eye enucleation and two had received unilateral excitotoxin injections into the superior colliculus. Experimental lesion.~
All surgery was performed with standard antibiotic prophylaxis and sterile precaution under deep surgical anesthesia induced either with Nembutal (35 mg/kg, i,p.J or a combination of ketamine hydrochloride (30 mg/kg, i,p.) and acepromazine maleate (I mg/kg, i.m). One eye was removed in three cats. Two of these (COR- 1, CP-48) were enucleated as adult animals 14 days prior to being killed. The third animal (CP-46) was enucleated neonatally. For excitotoxin lesions, ibotenic acid (1% w/v; Sigma) and N-methyl-o-aspartate (5-6%; Sigma) were injected unilaterally into the superior colliculus of eight adult cats (COR-2: COR-4; CP-47; and CENK-I to CENK-5). The excitotoxins were dissolved in a vehicle of I% Tris base, 0.75% NaC1 and 0.17 N NaOH in 0.08 M phosphate buffer, pH 7.6. Toxin deposit sites were selected after recording multiunit visual responses through the tip of the injection device, which was either a glass mieropipette (< 100tzm o.d.) or an insulated 30 g needle attached to a 5/al Hamilton syringe. Deposits were made at two to nine separate sites in each colliculus. Total injected volumes ranged from 2.5 to 7.7 #1. In two of the cats used in the ligand-binding studies (COR-2 and COR-4), a single deposit of excitotoxin ( 1 2 p 1) was delivered by the same method into the A-laminae of the lateral geniculate nucleus on the side eontralateral to the collicular injection. Postsurgical survival times were: six days for CP-47. CENK-5 and COR-2: seven days for CENK-3; eight days for CENK-4; I2 days for CENK-I: 13 days for CENK-2; and nine weeks for COR-4. In order to confirm the integrity of afferent fibers in the excitotoxin-injected colliculi, retinal fibers were labeled in two animals (CP-47 and CENK-1) by injection of horseradish peroxidase (HRP; 50% w/v; Boehringer Mannheim Grade I or Sigma Type V1 in distilled water containing 2% dimethyl sulfoxide) into the vitreous chamber of the contralateral eye one day before being killed; in a third animal (CENK-2), bilateral intraocular HRP injections were made. The injections were carried out under deep ketamine-xylazine anesthesia (see above) by means of a I00 #1 Hamilton syringe inserted through a small slit in the sclera just behind the limbus. The insertion of the needle was monitored opbthalmoscopically to minimize damage to the retina. Volumes injected ranged from 20 to 25/~1. lmmunohistochemL~trv Protocols. Following an overdose of barbiturate, animals
were perfused through the heart with 0,9% NaC1 followed by fixative (4% paraformaldehyde and 5% sucrose, in 0.1 M phosphate buffer, pH 7.4 at 4"C or room temperature). Trimmed tissue blocks were immersed for up to 2 h in the same fixative, washed in 20% sucrose in phosphate buffer tbr up to 24 h and cut in the transverse plane on a sliding microtome at 30/am. Alternate series of sections were processed immunohistochemically, stained with Cresyl Violet, reacted for cytochrome oxidase6~ or. in cases of eyc injection, reacted for HRP by using tetra.methyl benzidine. 4" Enkephalin-like immunoreactivity was demonstrated with rabbit anti-[MetS]e'nkephalin antiserum. We used RI53H antiserum provided by Dr R. P. Elde except in one case (CENK-4) in which we used antiserum from Immuno Nuclear Corp ,~tions were taken free-floating through
Opioid peptides and binding sites in superior colliculns the following solutions in sequence: (I) 10% methyl alcohol and 3% hydrogen peroxide in 0.5 M "Iris buffer (pH 7.4) containing 0.9% NaCI ("Tris-saline"), 5 rain to suppress endogenous peroxidase staining; (2) 0.2% Triton X-100 in Tris-saline, 5 rain; (3) normal goat serum (1: 30), 30 rain; (4) anti-cnkephalin antiserum (1:600 or I:1000) in Tris-saline containing normal goat serum (1:100), normal cat serum (1:100) and Thimerasol (an antibacterial agent; 0.01%) at 4°C, 2-11 days; (5) goat anti-rabbit immunoglobulin G (1 : 30 to 1: 50; Antibodies Inc.), 30 rain at room temperature or overnight at 4°C; (6) PAP (1:30 to 1:50; StembergerMeyer), 30-60 rain at room temperature; (7) cobalt chloride (0.05%), 5--10rain; and (8) diaminobenzidine (0.05%) and hydrogen peroxide (0.0024°/,), 13-61 rain at room temperature. Steps in the sequence were separated by thorough washing in Tris--s~ne buffer. lmmunohistochemical controls. The specificity of the immunoreactivity observed in the cat superior collicuins with the RI53H antiserum has been described in a previous study./9 In the present experiments, controls for specificity of the immunostaining included (1) omission of the primary anti-enkepbalin antiserum or (2) substrate blocks in which the primary incubation was carried out in the lm:seneeof an excess conoentration (I0-3M) of synthetic [MetS]- or [LeuS]enkephalin(Sigma). Despite these attempts to ensure that the immunoreactivityobserved was specificto enkepbalin peptides, our methods do not permit us to distinguish between [MetS]- and [LeuS]enkephaiinas antigens, nor do they exclude possible reactivity of the antiserum with opiate precursor molecules or compounds antigenically related to these peptides. Although for conveniencewe have described our findings in terms of the enkephalins, we emphasize that our observations pertain specifically to immunoreactivity detected with the antisera used.
515
Research, Inc., St Catharines, Ontario). Density values were converted to units of /~Ci/g by reference to calibrated standards placed on each film. Optical density values within selected regions were converted to estimates of specific binding by subtracting values in matching sectors of the corresponding control sections. RESULTS
Immunoh is t ochemis try Effects o f eye enucleation. Short-term removal of the contralateral eye produced no obvious change in the pattern or the intensity of enkephalin-like immunostaining in the superior colliculus. This lack of effect, seen at a two-week survival period chosen to ensure complete degeneration of crossed retinotectal afferents, ss is illustrated in Fig. 1 for CP-48. As in normal animals,19'~ a nearly continuous sheet of dense immunoreactivity marked the upper SGS 0ayer Ill of Kanaseki and Sprague'u). Much of the labefing in this band was clearly punctate or fibrous in form, but well-stained neurons were also evident in this sublamina. The enkephafin-positive neurons were relatively small (c. 8-12/~m diameter), and many clearly had the morphology of horizontal cells (Fig. IC, D), as reported previously./9.~ The horizontally disposed dendrites of these neurons could often be traced for several hundred micrometers in and near the upper SGS. In the rostromedial part of the colliculus, the nearly continuous sheet of immunoIn vitro receptor autoradiography staining in the upper SGS was interrupted by a small Brains were removed from animals under profound Nemgap, approximately 300/~m wide, within which labelbutal anesthesia (40-60 mg/kg) within 5 rain of respiratory arrest, immersed in ice-cold normal saline for 5min, ing of both neurons and neuropil was minimal. This blocked, frozen in powdered dry ice, and cut at 24/~m on gap, which apparently corresponds topographically a cryostat. Sections were thaw-mounted onto subbed slides, to the collicular representation of the optic disc, is a desiccated under vacuum for at least 12 h at 4°C and stored consistent feature of enkephalin immunostaining in frozen at - 70°C.2e In each brain, opiate binding sites were radiolabeled the eat superior colliculns.19 Deeper in the SGS by incubating sections in [31-1]etorphine (42Ci/mm; (layers 112 and 113, Ref. 34), the density of immunoAmersham, Arlington Heights, IL), which has a high alfmity stained neurons was much lower, and staining of the for mu, delta, and kappa opiate receptors. In two cats, neuropil was relatively sparse. Immunostaining in the mu opiate binding sites were labeled with the highly mu-selective ligand [3H]-[v-Ala2,N-Me-Phe,4Gly(ol)S]- stratum opticum was also weak, but the intermediate enkephalin (DAGO; New England Nuclear, Boston, MA; gray layer was marked by scattered immunoreactive 44.7Ci/mm).2~ Delta opiate sites were labeled with the cells and, caudally, by a series of patches of relatively highly selective ligand [3H]-[D-Pen2.5]-enkephalin(DPDPE, strongly immunoreactive fibers as previously deNew England Nuclear, Boston, MA; 43 Ci/mm).¢ The incubation medium consisted of radioligand at a concen- scribed) 9 Apparently normal patterns of enkephalintration of 2 nM and 50 mM Tris buffer at pH 7.4 containing like immunoreactivity also persisted in the pretectal 0.3% bovine serum albumin. Sections were incubated for region and in the ventral lateral geniculate nucleus 90 rain at 25°C and then rinsed in four 20 s changes of ice despite the unilateral retinal deafferentation. cold 10 mM Tris buffer. Controls for non-specific binding The survival of the immunoreactivity in this case were prepared and incubated as above except that 10#M ruled out crossed retinal afferents as the exclusive unlabeled levallorphan was added to the incubation media. To determine the specificityof binding, selected sections and source of the enkephalin-like immunoreactivity in the matching controls were scraped off slides and radioactivity SGS. However, such survival left open the possibility was assessed by liquid scintillation spectrophotometry. that retinal afferents contribute a part of the immunoLevels of specific binding ranged from 60 to 71% for etorphine, 67-77% for DAGO, and 56-63% for DPDPE. reactivity, because in that case the enucleation would Slides were apposed to tritium-sensitive film (LKB have produced only a modest reduction in immunoUltrofilm) for 3.5-15 weeks before being developed (D-19, staining. Such subtle reductions would not have been Kodak). Selected sections were postlLxed following auto- detectable by comparisons with control brains, due to radiography with paraformaldehyde vapor under reduced pressure (2 h, 80°C) and counterstained with Cresyl Violet. the variable staining from brain to brain, but would Photographs were printed directly from the films. Quantitat- presumably have been observable through bilateral ive densitometry was carried out on selected autoradiograms comparisons. Retinal input to the upper SGS is using a Microcomputer Imaging Device (MCID, Imaging overwhelmingly crossed, 3,4.~5.ss so eye removal would
~'ig. I. ~ght-fi~d phor,~gapl~ 'illustrating c t ~ l i n - l i k e immunor0activity in a tramverse section through the superior c O l ~ l ~ of an adult :at (O',411) which had ~ o ~ e a ~eft m ~ e r a ] on~cleation 14 da~s l:~or to being killed. COUiculm i ~ l to h em~le~io~ is l~:)wn n A and C; corn~x)nJng ~ o g r a ; , ~ (made under identical conditions) for the coUiculm contralmeral to the e~ucJeatk,n are shown in B todD. hmnunostaiaing1~.t teyns.api~r g r o s ~ normal on both sides. Note extremely ~ i m m t m o r ~ t y "m~lls and n~uropfl o f the . t ~ r ~GS. Patches o f ~ g m the i.n!er~__late ggay layer, though prormnent more cau0talty, are.uarcuy vlm~e, at,uns u ~ a l u ~ , ~vel. ~.t mgnc t a~agnific,ation, staining of the upper SGS appears slightly denser contralateral to the enuc|eation (D) than tpsnaterauy tu)- t.Jense stamtng a pial surface of the superior colliculus is a non-specific edge artifact. Scale bar in D represents 1 mm for A and B and lOOgm for C and D.
7
©
Opioid peptidcs and binding sites in superior colliculus have been expected to reduce immunostaining more in the contralateral than in the ipeilateral upper SGS. Paradoxically, however, the immunoreactivity in the IIi tier of the SGS appeared to be slightly more intense on the contralateral side (Fig. I B, D) than on the ipsilateral side (Fig. l A, C). This difference was subtle enough to be barely detectable in the high-contrast photomicrographs of Fig. 1. Nonetheless, the discrepancy in staining intensity on the two sides was discernible in both the neurons and neuropil throughout the caudal two-thirde of the colliculus and was especially clear medially. It was observable both in the staining of individual cells and in the density of neuropil labeling in the upper SGS. Although these results make it highly unlikely that any of the collicular enkephalin is contained in retinal afferents, long-term interruption of these afferents can apparently change the pattern of distribution of
,.' '~'
517
collicular enkephalin-like immunoreactivity. Figure 2 illustrates enkephalin immunmtaining in the SGS of the neonataUy enucleated cat (CT-46) killed more than 10 years after eye removal. Contralateral to the eye enucleation (Fig. 2B), considerable immunoreactivity persisted in both the neuropil and neuronal somata of the SGS. However, in sharp contrast to the normal pattern, which persisted on the~psilateral side (Fig. 2A), immunostalning was markedly reduced in the upper SGS. In addition, there appeared to be a reduction both in the number of enkephalin-immunoreactive neurons and in the intensity of immunostaining in individual neurons of the contralateral SGS. Further work is needed to determine whether the reduction in immunostaining in this animal is attributable to the long duration of the deafl'erentat.ion or to the early age at which the enucleation was performed.
~•
Fig. 2. Photographs from a single transverse section illustrating enkephalin-like immtmoreactivity in the upper superficial gray layer of the superior collicttlus of an adult cat (CP-46) which had undergone a left unilateral enucleation within the first few weeks of life. (A) CoUiculus ipsilateral to enucleation (left). (B) Matching part of contralateral (right) colliculus. Note reduction in number of stained perikarya and in density of neuropil staining in the contralateral colliculus. The band of staining at the pial surface of the superior colliculus is a non-specific edge artifact. Scale bar ffi 100~m.
518
D.M. BERSONet al.
Effects of excitotoxin lesions. To determine whether the enkephalin-like immunoreactivity was of intrinsic collicular origin, we injected the superior colliculus in five cats with mixtures of the excitotoxins ibotenic acid and N-methyl-o-aspartate, excitatory amino acid analogs that locally destroy neurons but spare axons of passage, s°'s7 Figure 3 illustrates results from CENK-4, one of three cases in which we made multiple large deposits of excitotoxins in order to kill neurons in as much of the SGS as possible. Eight days after the deposit, the excitotoxin lesion was clearly visible in Nissl-stained sections (Fig. 3A; arrowheads). The lesion was largely confined to the superficial layers, but encroached slightly upon the intermediate gray layer (not illustrated) and spread across the midline into the medial edge of the contralateral (left) superior colliculus. Within the lesion, there was a nearly total loss of neuronal somata, disruption of normal cytoarchitecture, and an increased number of small presumably glial cells, Enkephalin-like immunoreaetivity was dramatically reduced in the superficial gray layer: the sheet of dense neuropil immunostaining in the upper tier of the SGS was virtually abolished, and there was a complete loss of enkephalin-positive cell bodies (Fig. 3B-D). Two sorts of faint residual immunostaining could be identified in some sections. First, the upper half of the SGS contained diffuse staining slightly above background intensity. In addition, a few well-stained immunoreactive puncta and fibers were scattered throughout the superficial gray layer (inset in Fig. 3D). Neither type of immunostaining exhibited the precise laminar distribution characteristic of the contralateral superior colliculus. No such immunostaining was visible in control sections in which the primary antiserum was omitted. Immunoreactivity in the deeper collicular layers, largely spared from the lesion, appeared normal. Discrete patches of fiber labeling persisted in the intermediate gray layer despite virtually complete neuronal cell loss in the overlying superficial layers (Fig. 3B). Comparable widespread loss of upper SGS immunoreactivity occurred in both of the other cases of large superficial excitotoxin injection (CENK-2 and CENK-5). Cell loss in all three of these cases of superficial injection extended to the deeper collieular layers and to varying degrees to neighboring nuclei, but the critical site of the lesion for producing the effect appeared to be the SGS itself. In a fourth cat (CENK-I), a small excitotoxin lesion apparently confined to the SGS produced a circumscribed zone of reduced immunostaining there, whereas in a fifth cat (CENK-3) a much larger deposit centered in the deeper tectal layers and sparing the SGS had no detectable effect on upper SGS immunostaining (not illustrated). The enkephalin-immunoreaetive patches in the deeper collicular layers survived excitotoxin lesions of the deeper layers (Fig. 4).
The excitotoxin-induced loss of enkephahn-like immunoreactivity in the SGS appeared attributable to destruction of collicular neurons rather than to unintended damage to afferent fibers, in addition to the surviving immunopositive fibers within the lesion, retinotectal axons were still present and capable of anterograde transport throughout the zone of induced neuronal death as shown by anterograde labeling after intraocular HRP injection (not illustrated). Finally, in sections stained for cytochrome oxidase, staining was remarkably normal within the lesions despite the loss of stained neurons, making non-specific tissue damage an unlikely explanation for the effect.
Ligand binding Our observations in untreated colliculi confirmed the pattern reported previously for opiate receptor binding in the cat's superior colliculus?° Very dense labeling marked the SGS. Ligand binding was particularly dense in the upper third of this layer, the tier of densest enkephalin-like immunostaining. Labeling was moderately dense in the intermediate gray layer. but did not exhibit the pronounced patchiness of the enkephalin4ike immunoreaetivity. Binding was weak in the optic layer and in the intermediate white and deeper layers. Effects of eye removal. If the opiate binding sites in the SGS were mainly on retinal terminals, eye enucleation should reduce collicular opiate binding, and because retinocollicular input is predominantly crossed, the effect should be especially dramatic in the contralateral colliculus. Instead, enucleation had no apparent effect on opiate receptor binding patterns. Figure 5 illustrates the distribution of labeling by thc non-selective opiate receptor ligand [3H]etorphine in case COR-1 14 days after right unilateral enucleation. The pattern was indistinguishable from that in normal animals. Labeling density in the superficial layers contralateral to the enucleation (left coUiculus) was within I% of that in the ipsilateral (right) colliculus. Ligand binding patterns also exhibited normal bilateral symmetry with the mu-selective ligand [3I-I]DAGO (and with the delta-selective ligand [3H]DPDPE, though the binding was extremely weak, as previously reportedt°). We examined other primary retinal targets tbr possible effects of the enucleation but found none with any of the ligands. Figure 5C and D shows that there was no obvious reduction in receptor binding in the layers of the dorsal lateral geniculate nucleus innervated by the enucleated eye (e.g. layer A of the left and layer AI of the right lateral geniculate nucleus). Labeling in the ventral division of the lateral geniculate nucleus and in the pretectal region also appeared unaffected by the eye removal. The labeling patterns in the accessory optic nuclei were of particular interest to us, because, in the rat. retinal afferents in this system possess presynaptic opiate receptors. ~ However, we found little if any binding in
520
D . M . BERSON et al.
..~ ~ ~ ~ ~ . ~
~. .~ ~ ,..~
~
Neuroscience Vol. 43, No. 2/3, pp. 513-529, 1991 Printed in Great Britain
© 1991 IBRO
EVIDENCE FOR INTRINSIC EXPRESSION OF ENKEPHALIN-LIKE IMMUNOREACTIVITY A N D OPIOID BINDING SITES IN CAT SUPERIOR COLLICULUS D. M. BERSON,*t A. M. GRAYBIEL,~W. D. BOWL~* and U A. THOMPSON§ *Division of Biology and Medicine and §Department of Psychology, Brown University, Providence, RI 02912, U.S.A. ~:Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology, Cambridge, MA 02139, U.S.A. Almract--We have investigated the cellular localization of opioid peptides and binding sites in the cat's superior colliculus by testing the effects of retinal deafferentation and intracollicular excitotoxin lesions on patterns of enkephalin-like immunostaining and opiate receptor ligand binding. In normal cats, enkephalin-like immunoreactivity marks a thin tier in the most dorsal stratum griseum superficiale, small neurons of the stratum griseum superliciale, and patches of fibers in the intermediate and deeper gray layers. Eliminating crossed retinotectal afferents by contralaterul eye enucleation had little immediate effect on this pattern, although chronic eye enucleation from birth did reduce immqnoreactivity in the superficial layers. By contrast, fiber-sparing destruction of collicular neurons by the excitotoxins N-methyl-D-asparrate and ibotenic acid virtually eliminated enkephalin-like immunoreaetivity in the neuropil of the upper stratum griseum superficiale, presumably by kilting enkephalinergi¢ cells of the superficial layers. Such lesions did not eliminate the patches of enkephalin-like immunoreactivity in the deeper layers. In normal cats, opiate receptor ligand binding is dense in the stratum griseum superliciale, particularly in its upper tier, and moderately dense in the intermediate gray layer. Contralateral eye removal had no detectable effect on the binding pattern, but excitotoxin lesions of the colliculus dramatically reduced binding in both superficial and deep layers. Some ligand binding, including part of that in the upper stratum griseum superficiale, apparently survived such lesions. Similar effects were observed in the lateral genicuiate nucleus: enucleation produced no change in binding, whereas excitotoxin lesions greatly reduced specific opiate binding. We conclude that in the superficial collicular layers, both enkephalin-like opioid peptides and their membrane receptors are largely expressed by neurons of intrinsic collicular origin. The close correspondence between the location of these intrinsic opioid elements and the tier of retinal afferents terminating in the upper stratum griseum superficiale further suggests that opiatergic interneurons may modulate retinotectal transmission postsynaptically.
Opioid peptides and opiate receptors are of multiple types and are distributed widely in the brain, implying a diversity of functions. They arc found not only in structures linked to nociception and mood, but also in a wide array of other regions where their physiological significance is less certain. 2 ~ A case in point is the superior colliculus, a midbrain visual structure that participates in the control of saccadic eye movements. Autoradiographic studies have demonstrated very high concentrations of opiate binding sites in the superior colliculus in a variety of species (e.g. see Refs 26, 37, 54 and 60). Opioid receptor ligand binding is densest in the retinorecipi?To whom correspondence should be addressed. Abbreviations: A, layer A of dorsal lateral geniculate nucleus; AI, layer AI of dorsal lateral geniculate nucleus; DAGO, ISHHD-AIa2,N-Me-Phe4,Gly(ol)S]-enkephafin; DPDPE, [3l-l]-{D-Pen~-~kephalin; HRP, horseradish peroxidase; LGv, ventral lateral geniculate nucleus; MG, medial geniculate nucleus; PAG, periaqueductal gray; PAP, peroxidue-antiperoxidase; SGI, stratum griseum intermedial¢; SGS, stratum grise-m superticiale; SO, stratum opticum. 513
ent superficial layers of the colliculus, but significant binding is also found in the deeper muitisensory and premotor layers. In most species, immunohistochemical studies have revealed only scattered opioid peptide-positive neurons and fibers in the colliculus,21,2~'61 raising questions about the functional significance of the high opiate ligand binding densities. In the cat, however, substantial cnkephalin-like immunoreactivity has been observed in the stratum griseum superficiale (SGS) and in sets of regularly spaced patches in the deeper layers. 19 Neither the identity of opiatergic elements in the cat's colliculus nor their targets have been fully established, but available data raise some intriguing possibilities. In the superficial layers, enkephalin-like immunoreactivity is heavily concentrated in a thin sheet in the upper part of the SGS (the "upper S G S ' ) . This enkephalin-positive sheet closely matches the tier of densest retinal input) sa4 an input which apparently originates in a subset of contralateraily projecting retinotectal ganglion cells, s,'3,32 The correspondence is topographic as well as laminar, because both the retinal input and the enkephalin-like immuno-
514
D.M. BERSONet al.
staining in the upper SGS are weakest in the area centralis representation and exhibit a prominent gap in the representation of the blind spot. 7ajS'lg':4~ This correspondence raises the possibility that enkephalins or related peptides are contained within the terminals of certain contralaterally projecting retinocollicular afferents. There are precedents for the localization of neuropeptides in subsets of retinal ganglion cells, including some of those that project to the superior colliculus or optic tectum in rabbits, frogs and birds. W2"~'~9 A second potential extrinsic source for the enkephalin-like immunoreactivity is the crossed projection of the parabigeminal nucleus. This afferent system has a laminar and topographic distribution nearly identical to those of the dense contralateral retinal input to the upper SGS and the enkephalinlike immunostaining there. ~ Arguing against a localization within retinal or parabigeminal afferents, however, is the fact that enkephalin-like immunoreactivity has not as yet been observed in ganglion cells or in parabigeminal neurons in the cat. ~9"~N o r did an ultrastructural immunohistochemical study of the cat superior colliculus find any evidence for enkephalinlike immunostaining in presumed retinotectal afferent terminals. 46 An alternative is that the enkephalin-like immunoreactivity is largely of intrinsic collicular origin. There are substantial numbers of enkephalin-immunoreactive neuronal cell bodies in the superficial gray layer, especially in and near the upper SGS. j9"46Many of these appear to be horizontal cells, which extend their dendrites parallel to the coilicular surface for considerable distances. At the ultrastructural level, much of the enkephalin-like immunoreactivity of the upper SGS appears to be contained within dendritic profiles. 4~ in the deeper layers, the fibrous composition of the patches of enkephalin-like immunoreactivity suggest an origin in collicular afferents, many of which exhibit patchy distributions in the cat. ~:9 ~ However, they could arise, instead, from local intrinsic enkephalinergic neurons or tYom enkephalin-containing neurons of the superficial layers; interlaminar connections are known to extend from the superficial layers to the deeper strata. '~ There is equal uncertainty about the sources of opiate ligand binding in the superior colliculus. Retinotectal afferents represent one candidate source in the superficial layers. The close match between the patterns of enkephalin-like immunoreactivity and contralateral retinal input suggests that collicular opioids might act directly on retinal terminals. Indeed, binding of opiate ligands is heaviest in the tier of densest retinal terminations. ~5'~'~° Presynaptic opiate receptors have been localized to retinal afferents in the rat accessory optic nuclei) To distinguish a m o n g these possibilities, we studied the effects of destroying retinal afferents or intrinsic collicular neurons on the patterns of enkephalin-likc immunoreactivity and opiate receptor ligand binding in the superior colliculus.
EXPERIMENTAL PROCEDURES
The distribution of enkephalin-like immunoreactivity in the superior collieulus was studied by the peroxidase antiperoxidase (PAP) technique ~ in tissue from eight adult cats (Liberty Laboratories, Liberty Corner. N J), two with unilateral eye enucleations and six with excitotoxin lesions of the colliculus. Opiate binding site distributions were studied by in vitro autoradiography in tissue from four cats; one of these had undergone unilateral eye enucleation and two had received unilateral excitotoxin injections into the superior colliculus. Experimental lesion.~
All surgery was performed with standard antibiotic prophylaxis and sterile precaution under deep surgical anesthesia induced either with Nembutal (35 mg/kg, i,p.J or a combination of ketamine hydrochloride (30 mg/kg, i,p.) and acepromazine maleate (I mg/kg, i.m). One eye was removed in three cats. Two of these (COR- 1, CP-48) were enucleated as adult animals 14 days prior to being killed. The third animal (CP-46) was enucleated neonatally. For excitotoxin lesions, ibotenic acid (1% w/v; Sigma) and N-methyl-o-aspartate (5-6%; Sigma) were injected unilaterally into the superior colliculus of eight adult cats (COR-2: COR-4; CP-47; and CENK-I to CENK-5). The excitotoxins were dissolved in a vehicle of I% Tris base, 0.75% NaC1 and 0.17 N NaOH in 0.08 M phosphate buffer, pH 7.6. Toxin deposit sites were selected after recording multiunit visual responses through the tip of the injection device, which was either a glass mieropipette (< 100tzm o.d.) or an insulated 30 g needle attached to a 5/al Hamilton syringe. Deposits were made at two to nine separate sites in each colliculus. Total injected volumes ranged from 2.5 to 7.7 #1. In two of the cats used in the ligand-binding studies (COR-2 and COR-4), a single deposit of excitotoxin ( 1 2 p 1) was delivered by the same method into the A-laminae of the lateral geniculate nucleus on the side eontralateral to the collicular injection. Postsurgical survival times were: six days for CP-47. CENK-5 and COR-2: seven days for CENK-3; eight days for CENK-4; I2 days for CENK-I: 13 days for CENK-2; and nine weeks for COR-4. In order to confirm the integrity of afferent fibers in the excitotoxin-injected colliculi, retinal fibers were labeled in two animals (CP-47 and CENK-1) by injection of horseradish peroxidase (HRP; 50% w/v; Boehringer Mannheim Grade I or Sigma Type V1 in distilled water containing 2% dimethyl sulfoxide) into the vitreous chamber of the contralateral eye one day before being killed; in a third animal (CENK-2), bilateral intraocular HRP injections were made. The injections were carried out under deep ketamine-xylazine anesthesia (see above) by means of a I00 #1 Hamilton syringe inserted through a small slit in the sclera just behind the limbus. The insertion of the needle was monitored opbthalmoscopically to minimize damage to the retina. Volumes injected ranged from 20 to 25/~1. lmmunohistochemL~trv Protocols. Following an overdose of barbiturate, animals
were perfused through the heart with 0,9% NaC1 followed by fixative (4% paraformaldehyde and 5% sucrose, in 0.1 M phosphate buffer, pH 7.4 at 4"C or room temperature). Trimmed tissue blocks were immersed for up to 2 h in the same fixative, washed in 20% sucrose in phosphate buffer tbr up to 24 h and cut in the transverse plane on a sliding microtome at 30/am. Alternate series of sections were processed immunohistochemically, stained with Cresyl Violet, reacted for cytochrome oxidase6~ or. in cases of eyc injection, reacted for HRP by using tetra.methyl benzidine. 4" Enkephalin-like immunoreactivity was demonstrated with rabbit anti-[MetS]e'nkephalin antiserum. We used RI53H antiserum provided by Dr R. P. Elde except in one case (CENK-4) in which we used antiserum from Immuno Nuclear Corp ,~tions were taken free-floating through
Opioid peptides and binding sites in superior colliculns the following solutions in sequence: (I) 10% methyl alcohol and 3% hydrogen peroxide in 0.5 M "Iris buffer (pH 7.4) containing 0.9% NaCI ("Tris-saline"), 5 rain to suppress endogenous peroxidase staining; (2) 0.2% Triton X-100 in Tris-saline, 5 rain; (3) normal goat serum (1: 30), 30 rain; (4) anti-cnkephalin antiserum (1:600 or I:1000) in Tris-saline containing normal goat serum (1:100), normal cat serum (1:100) and Thimerasol (an antibacterial agent; 0.01%) at 4°C, 2-11 days; (5) goat anti-rabbit immunoglobulin G (1 : 30 to 1: 50; Antibodies Inc.), 30 rain at room temperature or overnight at 4°C; (6) PAP (1:30 to 1:50; StembergerMeyer), 30-60 rain at room temperature; (7) cobalt chloride (0.05%), 5--10rain; and (8) diaminobenzidine (0.05%) and hydrogen peroxide (0.0024°/,), 13-61 rain at room temperature. Steps in the sequence were separated by thorough washing in Tris--s~ne buffer. lmmunohistochemical controls. The specificity of the immunoreactivity observed in the cat superior collicuins with the RI53H antiserum has been described in a previous study./9 In the present experiments, controls for specificity of the immunostaining included (1) omission of the primary anti-enkepbalin antiserum or (2) substrate blocks in which the primary incubation was carried out in the lm:seneeof an excess conoentration (I0-3M) of synthetic [MetS]- or [LeuS]enkephalin(Sigma). Despite these attempts to ensure that the immunoreactivityobserved was specificto enkepbalin peptides, our methods do not permit us to distinguish between [MetS]- and [LeuS]enkephaiinas antigens, nor do they exclude possible reactivity of the antiserum with opiate precursor molecules or compounds antigenically related to these peptides. Although for conveniencewe have described our findings in terms of the enkephalins, we emphasize that our observations pertain specifically to immunoreactivity detected with the antisera used.
515
Research, Inc., St Catharines, Ontario). Density values were converted to units of /~Ci/g by reference to calibrated standards placed on each film. Optical density values within selected regions were converted to estimates of specific binding by subtracting values in matching sectors of the corresponding control sections. RESULTS
Immunoh is t ochemis try Effects o f eye enucleation. Short-term removal of the contralateral eye produced no obvious change in the pattern or the intensity of enkephalin-like immunostaining in the superior colliculus. This lack of effect, seen at a two-week survival period chosen to ensure complete degeneration of crossed retinotectal afferents, ss is illustrated in Fig. 1 for CP-48. As in normal animals,19'~ a nearly continuous sheet of dense immunoreactivity marked the upper SGS 0ayer Ill of Kanaseki and Sprague'u). Much of the labefing in this band was clearly punctate or fibrous in form, but well-stained neurons were also evident in this sublamina. The enkephafin-positive neurons were relatively small (c. 8-12/~m diameter), and many clearly had the morphology of horizontal cells (Fig. IC, D), as reported previously./9.~ The horizontally disposed dendrites of these neurons could often be traced for several hundred micrometers in and near the upper SGS. In the rostromedial part of the colliculus, the nearly continuous sheet of immunoIn vitro receptor autoradiography staining in the upper SGS was interrupted by a small Brains were removed from animals under profound Nemgap, approximately 300/~m wide, within which labelbutal anesthesia (40-60 mg/kg) within 5 rain of respiratory arrest, immersed in ice-cold normal saline for 5min, ing of both neurons and neuropil was minimal. This blocked, frozen in powdered dry ice, and cut at 24/~m on gap, which apparently corresponds topographically a cryostat. Sections were thaw-mounted onto subbed slides, to the collicular representation of the optic disc, is a desiccated under vacuum for at least 12 h at 4°C and stored consistent feature of enkephalin immunostaining in frozen at - 70°C.2e In each brain, opiate binding sites were radiolabeled the eat superior colliculns.19 Deeper in the SGS by incubating sections in [31-1]etorphine (42Ci/mm; (layers 112 and 113, Ref. 34), the density of immunoAmersham, Arlington Heights, IL), which has a high alfmity stained neurons was much lower, and staining of the for mu, delta, and kappa opiate receptors. In two cats, neuropil was relatively sparse. Immunostaining in the mu opiate binding sites were labeled with the highly mu-selective ligand [3H]-[v-Ala2,N-Me-Phe,4Gly(ol)S]- stratum opticum was also weak, but the intermediate enkephalin (DAGO; New England Nuclear, Boston, MA; gray layer was marked by scattered immunoreactive 44.7Ci/mm).2~ Delta opiate sites were labeled with the cells and, caudally, by a series of patches of relatively highly selective ligand [3H]-[D-Pen2.5]-enkephalin(DPDPE, strongly immunoreactive fibers as previously deNew England Nuclear, Boston, MA; 43 Ci/mm).¢ The incubation medium consisted of radioligand at a concen- scribed) 9 Apparently normal patterns of enkephalintration of 2 nM and 50 mM Tris buffer at pH 7.4 containing like immunoreactivity also persisted in the pretectal 0.3% bovine serum albumin. Sections were incubated for region and in the ventral lateral geniculate nucleus 90 rain at 25°C and then rinsed in four 20 s changes of ice despite the unilateral retinal deafferentation. cold 10 mM Tris buffer. Controls for non-specific binding The survival of the immunoreactivity in this case were prepared and incubated as above except that 10#M ruled out crossed retinal afferents as the exclusive unlabeled levallorphan was added to the incubation media. To determine the specificityof binding, selected sections and source of the enkephalin-like immunoreactivity in the matching controls were scraped off slides and radioactivity SGS. However, such survival left open the possibility was assessed by liquid scintillation spectrophotometry. that retinal afferents contribute a part of the immunoLevels of specific binding ranged from 60 to 71% for etorphine, 67-77% for DAGO, and 56-63% for DPDPE. reactivity, because in that case the enucleation would Slides were apposed to tritium-sensitive film (LKB have produced only a modest reduction in immunoUltrofilm) for 3.5-15 weeks before being developed (D-19, staining. Such subtle reductions would not have been Kodak). Selected sections were postlLxed following auto- detectable by comparisons with control brains, due to radiography with paraformaldehyde vapor under reduced pressure (2 h, 80°C) and counterstained with Cresyl Violet. the variable staining from brain to brain, but would Photographs were printed directly from the films. Quantitat- presumably have been observable through bilateral ive densitometry was carried out on selected autoradiograms comparisons. Retinal input to the upper SGS is using a Microcomputer Imaging Device (MCID, Imaging overwhelmingly crossed, 3,4.~5.ss so eye removal would
~'ig. I. ~ght-fi~d phor,~gapl~ 'illustrating c t ~ l i n - l i k e immunor0activity in a tramverse section through the superior c O l ~ l ~ of an adult :at (O',411) which had ~ o ~ e a ~eft m ~ e r a ] on~cleation 14 da~s l:~or to being killed. COUiculm i ~ l to h em~le~io~ is l~:)wn n A and C; corn~x)nJng ~ o g r a ; , ~ (made under identical conditions) for the coUiculm contralmeral to the e~ucJeatk,n are shown in B todD. hmnunostaiaing1~.t teyns.api~r g r o s ~ normal on both sides. Note extremely ~ i m m t m o r ~ t y "m~lls and n~uropfl o f the . t ~ r ~GS. Patches o f ~ g m the i.n!er~__late ggay layer, though prormnent more cau0talty, are.uarcuy vlm~e, at,uns u ~ a l u ~ , ~vel. ~.t mgnc t a~agnific,ation, staining of the upper SGS appears slightly denser contralateral to the enuc|eation (D) than tpsnaterauy tu)- t.Jense stamtng a pial surface of the superior colliculus is a non-specific edge artifact. Scale bar in D represents 1 mm for A and B and lOOgm for C and D.
7
©
Opioid peptidcs and binding sites in superior colliculus have been expected to reduce immunostaining more in the contralateral than in the ipeilateral upper SGS. Paradoxically, however, the immunoreactivity in the IIi tier of the SGS appeared to be slightly more intense on the contralateral side (Fig. I B, D) than on the ipsilateral side (Fig. l A, C). This difference was subtle enough to be barely detectable in the high-contrast photomicrographs of Fig. 1. Nonetheless, the discrepancy in staining intensity on the two sides was discernible in both the neurons and neuropil throughout the caudal two-thirde of the colliculus and was especially clear medially. It was observable both in the staining of individual cells and in the density of neuropil labeling in the upper SGS. Although these results make it highly unlikely that any of the collicular enkephalin is contained in retinal afferents, long-term interruption of these afferents can apparently change the pattern of distribution of
,.' '~'
517
collicular enkephalin-like immunoreactivity. Figure 2 illustrates enkephalin immunmtaining in the SGS of the neonataUy enucleated cat (CT-46) killed more than 10 years after eye removal. Contralateral to the eye enucleation (Fig. 2B), considerable immunoreactivity persisted in both the neuropil and neuronal somata of the SGS. However, in sharp contrast to the normal pattern, which persisted on the~psilateral side (Fig. 2A), immunostalning was markedly reduced in the upper SGS. In addition, there appeared to be a reduction both in the number of enkephalin-immunoreactive neurons and in the intensity of immunostaining in individual neurons of the contralateral SGS. Further work is needed to determine whether the reduction in immunostaining in this animal is attributable to the long duration of the deafl'erentat.ion or to the early age at which the enucleation was performed.
~•
Fig. 2. Photographs from a single transverse section illustrating enkephalin-like immtmoreactivity in the upper superficial gray layer of the superior collicttlus of an adult cat (CP-46) which had undergone a left unilateral enucleation within the first few weeks of life. (A) CoUiculus ipsilateral to enucleation (left). (B) Matching part of contralateral (right) colliculus. Note reduction in number of stained perikarya and in density of neuropil staining in the contralateral colliculus. The band of staining at the pial surface of the superior colliculus is a non-specific edge artifact. Scale bar ffi 100~m.
518
D.M. BERSONet al.
Effects of excitotoxin lesions. To determine whether the enkephalin-like immunoreactivity was of intrinsic collicular origin, we injected the superior colliculus in five cats with mixtures of the excitotoxins ibotenic acid and N-methyl-o-aspartate, excitatory amino acid analogs that locally destroy neurons but spare axons of passage, s°'s7 Figure 3 illustrates results from CENK-4, one of three cases in which we made multiple large deposits of excitotoxins in order to kill neurons in as much of the SGS as possible. Eight days after the deposit, the excitotoxin lesion was clearly visible in Nissl-stained sections (Fig. 3A; arrowheads). The lesion was largely confined to the superficial layers, but encroached slightly upon the intermediate gray layer (not illustrated) and spread across the midline into the medial edge of the contralateral (left) superior colliculus. Within the lesion, there was a nearly total loss of neuronal somata, disruption of normal cytoarchitecture, and an increased number of small presumably glial cells, Enkephalin-like immunoreaetivity was dramatically reduced in the superficial gray layer: the sheet of dense neuropil immunostaining in the upper tier of the SGS was virtually abolished, and there was a complete loss of enkephalin-positive cell bodies (Fig. 3B-D). Two sorts of faint residual immunostaining could be identified in some sections. First, the upper half of the SGS contained diffuse staining slightly above background intensity. In addition, a few well-stained immunoreactive puncta and fibers were scattered throughout the superficial gray layer (inset in Fig. 3D). Neither type of immunostaining exhibited the precise laminar distribution characteristic of the contralateral superior colliculus. No such immunostaining was visible in control sections in which the primary antiserum was omitted. Immunoreactivity in the deeper collicular layers, largely spared from the lesion, appeared normal. Discrete patches of fiber labeling persisted in the intermediate gray layer despite virtually complete neuronal cell loss in the overlying superficial layers (Fig. 3B). Comparable widespread loss of upper SGS immunoreactivity occurred in both of the other cases of large superficial excitotoxin injection (CENK-2 and CENK-5). Cell loss in all three of these cases of superficial injection extended to the deeper collieular layers and to varying degrees to neighboring nuclei, but the critical site of the lesion for producing the effect appeared to be the SGS itself. In a fourth cat (CENK-I), a small excitotoxin lesion apparently confined to the SGS produced a circumscribed zone of reduced immunostaining there, whereas in a fifth cat (CENK-3) a much larger deposit centered in the deeper tectal layers and sparing the SGS had no detectable effect on upper SGS immunostaining (not illustrated). The enkephalin-immunoreaetive patches in the deeper collicular layers survived excitotoxin lesions of the deeper layers (Fig. 4).
The excitotoxin-induced loss of enkephahn-like immunoreactivity in the SGS appeared attributable to destruction of collicular neurons rather than to unintended damage to afferent fibers, in addition to the surviving immunopositive fibers within the lesion, retinotectal axons were still present and capable of anterograde transport throughout the zone of induced neuronal death as shown by anterograde labeling after intraocular HRP injection (not illustrated). Finally, in sections stained for cytochrome oxidase, staining was remarkably normal within the lesions despite the loss of stained neurons, making non-specific tissue damage an unlikely explanation for the effect.
Ligand binding Our observations in untreated colliculi confirmed the pattern reported previously for opiate receptor binding in the cat's superior colliculus?° Very dense labeling marked the SGS. Ligand binding was particularly dense in the upper third of this layer, the tier of densest enkephalin-like immunostaining. Labeling was moderately dense in the intermediate gray layer. but did not exhibit the pronounced patchiness of the enkephalin4ike immunoreaetivity. Binding was weak in the optic layer and in the intermediate white and deeper layers. Effects of eye removal. If the opiate binding sites in the SGS were mainly on retinal terminals, eye enucleation should reduce collicular opiate binding, and because retinocollicular input is predominantly crossed, the effect should be especially dramatic in the contralateral colliculus. Instead, enucleation had no apparent effect on opiate receptor binding patterns. Figure 5 illustrates the distribution of labeling by thc non-selective opiate receptor ligand [3H]etorphine in case COR-1 14 days after right unilateral enucleation. The pattern was indistinguishable from that in normal animals. Labeling density in the superficial layers contralateral to the enucleation (left coUiculus) was within I% of that in the ipsilateral (right) colliculus. Ligand binding patterns also exhibited normal bilateral symmetry with the mu-selective ligand [3I-I]DAGO (and with the delta-selective ligand [3H]DPDPE, though the binding was extremely weak, as previously reportedt°). We examined other primary retinal targets tbr possible effects of the enucleation but found none with any of the ligands. Figure 5C and D shows that there was no obvious reduction in receptor binding in the layers of the dorsal lateral geniculate nucleus innervated by the enucleated eye (e.g. layer A of the left and layer AI of the right lateral geniculate nucleus). Labeling in the ventral division of the lateral geniculate nucleus and in the pretectal region also appeared unaffected by the eye removal. The labeling patterns in the accessory optic nuclei were of particular interest to us, because, in the rat. retinal afferents in this system possess presynaptic opiate receptors. ~ However, we found little if any binding in
520
D . M . BERSON et al.
..~ ~ ~ ~ ~ . ~
~. .~ ~ ,..~
~
