An Lndoleamine System in Photoreceptor Cell Terminals of the Long-Evans Rat Retina

The Journal of Neuroscience, February 1987, 7(2): 319-329

An lndoleamine System in Photoreceptor Cell Terminals of the Long-Evans Rat Retina

Dianna A. Redburn and Lynn ChurchiW Department of Neurobiology and Anatomy, University of Texas Medical School at Houston, Houston, Texas 77030

Uptake of 3H-serotonin is localized to the outer plexiform serotonin systemcoincident with degenerationof photoreceptor layer in Long-Evans rat retinas. Autoradiographic accumu- cellsin the retinal dystrophic rat (the Noel1line of Royal College lation is seen only after in vitro incubation in the light, with of Surgeonsrats) and (2) EM autoradiographic localization of retinas isolated from the underlying sclera. Potassium stim- 3H-serotonin accumulation in rod and cone terminals, we sug- ulates the release of 3H-serotonin. In this species, amacrine gest that the serotonin system is of photoreceptor origin. In cells do not accumulate these compounds; thus the outer addition, we demonstrate the regulation of serotonin by light, plexiform layer appears to be the only site of uptake and dark, and depolarizing levels of potassium.This raisesthe strong release of this indoleamine. possibility that serotonin and/or related indoles play a physi- The age-related loss of 3H-serotonin accumulation in the ologically important role in photoreceptors. outer plexiform layer of retinal dystrophic rats coincides A preliminary report ofthis work waspresented at the meeting temporally with the spontaneous degeneration of photore- of the American Society for Pharmacology and Experimental ceptor cells that occurs in this species. Electron-micro- Therapeutics, April 11, 1983. scopic autoradiography of 3H-serotonin accumulation further confirms that uptake is localized to rod and cone Materials and Methods terminals in the outer plexiform layer. The specific accu- Animals. Long-Evansrats, obtainedfrom CharlesRiver, Portage,MI, mulation of indoleamines into rod and cone terminals that werebred and raisedon a 14 hr light : 10 hr dark cycle.Retinal dys- is observed in the light but is absent in darkness suggests trophic (rdy) ratsfrom the Noel1line of the RoyalCollege of Surgeons ratswere bred and raisedunder a red safelight.The 20-d-oldrats-were that indoles have an important physiological role in photo- housedin a litter with the mother.The loo-d-oldrats were housed in receptors. a litter separatefrom the motherafter 20-22 d of age.Since indole turnover (i.e., releaseand/or metabolism)appears to be activatedby Serotonin (5hydroxytryptamine, 5-HT) is present in the brain darkness,rats were kept in the dark 2 hr prior to decapitationin an as a neurotransmitter (Bloom et al., 1972) and a precursor for attemptto depleteendogenous stores of serotonin. Uptake of 3H-serotonin. Animalswere decapitated and rapidlyenu- melatonin in the pineal gland (Sitaram and Lees, 1978). Sero- cleatedin the light.Isolated retinas were incubated in oxygenatedKrebs tonin is alsopresent in retina, although its levels are low (Thom- bicarbonatebuffer containing 10 mM glucose.Portions of eachretina as and Redburn, 1979; Osborne et al., 1982; Tornqvist et al., were incubatedin a 37°Cwater bath -underambient light or in total 1983).Synthesis, metabolism, and uptake systemsfor serotonin darknessin buffercontainina 100 &/ml ‘H-SHT (1.2-3H-(n)-hvdroxv- have been demonstratedin retinas of several species(Osborne, tryptamine creatinesulfate~26.7’Wmmol speci&‘activiiy, New En- glandNuclear, Boston, MA), 100 MM ascorbic acid, and 10 PM pargyline. 1980). The retinal serotonin system has been previously local- In selected samples, CaCl, was omitted and 10 mM MgCl, was added ized to a specificset of amacrine cells in many species,although to the buffer. After 20 min, the tissue was prepared for analysis by tbin- some important exceptions have been noted. For example, layer chromatography (TLC) or autoradiography. Tomqvist et al. (1983) report that guinea pig and albino rat Thin-layer chromatography. Retinas were homogenized in 50% ethanol and the particulate fraction removed by centrifugation. The soluble retinaslack serotonin-accumulatingamacrine cells. In this study extracts were chromatographed on silica gel plates in butanol, acetic we find that retinas from the pigmented Long-Evans rat are acid, and water (254: 10). More than 80% of the recovered radioactivity likewisedevoid of serotonin-accumulatingelements in the inner corn&rated with the ‘H-5HT standard. retina. In the outer retina, however, we find evidence for a Autorudiogruphy. Isolated retinas or eyecups were removed and placed in 3% glutaraldehyde buffered with 0.05 M sodium cacodylate @H 7.2) serotonin system that has not been previously described. Our for 30 min at room temperature and subsequently refrigerated overnight. study includesan analysisof autoradiography and histofluores- The tissue was rinsed with buffer and postfixed in 1% 0~0, buffered cence of indole compounds. On the basisof (1) the loss of the with 0.05 M sodium cacodylate for 1% hr at 0°C. After dehydration with a graded ethanol series, the tissue was embedded in Epon 8 12 epoxy resin. Received Oct. 22, 1985; revised June 30, 1986; accepted Aug. 21, 1986. Blocks were sectioned for both light and electron microscopy and This work was supported by NE1 Grant 1655-10 to D.A.R. We wish to thank autoradiography. Sections were coated with Kodak Nuclear Track Dr. Werner Noel1 for providing specimens from the Noel1 line of Royal College Emulsion, type NTB-2, for light-microscopic autoradiography and Il- of Surgeon rats; Ms. Cheryl Mitchell and Ms. Yvonne Blocker for technical as- ford L-4 emulsion for electron-microscopic autoradiography. Sections sistance; Dr. Richard Wilkerson for printing the micrographs; and Ms. Diana were incubated for 2-4 weeks for light-microscopic autoradiographs Parker and Ms. Mona Wilhich for typing the manuscript. Correspondence should be addressed to Dianna A. Redburn, Ph.D., Department and for 2-4 months for electron-microscopic autoradiographs. Some of Neurobiology and Anatomy, University of Texas Medical School at Houston, sections were stained with toluidine blue for comparison with unstained P.O. Box 20708, Houston, TX 77030. sections. a Present address: Department of Anatomy, University of Kansas Medical School, Treatment with 5,7-dihydroqt~ptamin (DHT). Long-Evans rats were MRRC-37-115, Kansas City, KS 66103. anesthetized with an Acepromazine-ketamine-Rompun cocktail (0.7, Copyright 0 1987 Society for Neuroscience 0270-6474/87/020319-l 1$02.00/O 2 1.4, and 4.3 mg/kg, respectively) via an intramuscular injection in the 320 Redburn and Churchill - lndoleamine System in Photoreceptor Ceils of Rat Retina hindlimb. Each animal also received an injection of 50 mg/kg, i.p., ed during the 20 min incubation period, a specificaccumulation pargyline. Thirty minutes after the injections, the eyes were bathed with pattern is observed. A thin band of label is specificallyassociated 1% lidocaine hydrochloride and a 27-gauge x 0.5 in. needle was used with the outer portion of the OPL (Fig. 1, upper right). Cell for intraocular injections. The injection volume was 50 pl and contained 50 pg DHT in 0.9% NaCl, with ascorbic acid (1 mg/ml) as an antiox- bodies of the photoreceptor cells also showlight labeling in this idant. The first group of animals was killed 4 hr after the intraocular section. There is a light and uniform distribution of grains injection and retinas were prepared for fluorescence microscopy. Single throughout the inner half of the retina, with no evidence of injections of DHT were previously shown by Ehinger and Floren (1978) specificaccumulation of the ligand within any cellsin the GCL to cause indoleamine-accumulating neurons to appear fluorescent in rabbit retina. Thus, the first group of animals was used to determine or INL. The only exception is accumulation by elements as- whether there were any indoleamine-accumulating cells in normal rat sociated with a blood vesselseen in the GCL. These observa- retina. A second intraocular injection was given to the remaining ani- tions confirm previous reports that amacrine cells of the rat mals on the following day. One week later, a third intraocular injection retina do not accumulate serotonin (Tornqvist et al., 1983). of DHT was given 4 hr before killing the remaining animals. Cells that The specificlabeling pattern in the OPL is seenonly if the in accumulated the toxin would be allowed to degenerate during the week- long interval, and their loss would be obvious through the specific loss vitro incubation occursin the absenceof underlying sclera.Au- of cells with DHT-induced fluorescence. toradiogramsof attached retina in an isolatedeyecup after a 20 Histojluorescence.Localization of catecholamines and indoleamines min in vitro incubation showno specificaccumulation of seroto- was visualized by histofluorescence using the method of de la Torre and nin in either inner or outer portions of the retina, with light Surgeon (1976). Animals were decapitated and quickly enucleated. The globes w&e hemisected and the vitreous gently removed. Eyecups were uniform labelingthroughout, except for labelingof vascularele- immediatelv nlaced in OCT. crvostat embedding medium (Lab-Tek. ments (Fig. 1, lower left). This observation suggeststhat removal Miles Laboiaiories, Naperville, iL) and frozen on dry ice. Frozen 2d of the sclerais necessaryto allow adequatedifision of molecules pm sections were cut in a cryostat set at -20°C. The sections were among all cell types of both inner and outer retina. placed on clean glass slides and immediately dipped into fresh glyoxylic acid stain (pH 7.4). The slides were blotted, dried under a stream of Regulation of ‘H-serotonin accumulation by light, dark, and cold air for 3 min, then heated in an 80°C oven for 5 min. Coverslips were mounted with mineral oil and the sections observed under a Zeiss magnesium Fluorescence Photomicroscope III. The specific uptake of 3H-serotonin into the outer plexiform In an attempt to enhance serotonin-induced histofluorescence, ani- layer wasfound to be particularly sensitiveto lighting conditions mals were pretreated with colchicine to inhibit axoplasmic transport, with tryptophan to provide precursor, and with pargyline to inhibit during the incubation period. When isolated retinas are incu- metabolism, according to the method described by Rassmussen and bated in 3H-serotonin for 20 min in the dark and prepared for Bunney (1982). Rats were anesthetized with chloral hydrate (400 mg/ autoradiography, we do not observe the specificaccumulation kg, i.p.). Ten microliters of saline containing 3 pg colchicine was injected of label in the OPL that is normally seenafter incubation in the into the right eye of each animal. Tryptophan methyl ester (100 mg/kg) and pargyline (100 mg/kg) were injected intraperitoneally, 22 hr later. light (Fig. 2, top). Under dark incubation conditions,only sparse, After an additional 2 hr, retinas were removed and prepared for his- uniform labeling is seenthroughout the retina, except for la- tofluorescence. beling of vascularelements in the inner retina. Incubated retinas Unless otherwise noted, all chemicals were obtained from Sigma, St. that are subsequentlyhomogenized and prepared for scintilla- Louis, MO. tion counting show a 50% inhibition of 3H-serotonin accumu- 3H-serotoninrelease. Isolated retinas were incubated in 0.5 ml oxy- lation in the dark (72 13 cpm/mg protein/min + 527 SEM; n = genated Krebs bicarbonate buffer containing 10 mM glucose, 100 PM ascorbate, 10 PM pargyline, and 50 pCi 3H-5-HT for 15 min at 37°C 3) as compared to in the light (13,906 cpm/mg protein/ under a continuous stream of 95% 0, and 5% CO,. Retinas were then min f 822 SEM; n = 3). rinsed with buffer and placed in a 2 ml superfusion chamber connected The lack of specific labeling could be due to either dark- to a continuous flow pump. All perfusion buffers contained 6 PM imipra- induced inhibition of uptake or dark-induced loss of 3H-sero- mine to inhibit reuptake. One milliliter of the perfusate was collected per minute. After a 10 min perfusion period in standard buffer, the tonin through metabolism or calcium-dependent release. Ad- potassium concentration was increased to 56.2 mM for 5 min. After a dition of magnesium and removal of calcium to inhibit release return to standard buffer for 12 min, the potassium concentration was during a dark incubation period results in only modest auto- increased a second time. radiographic accumulation of 3H-serotonin in the OPL (Fig. 2, center). Scintillation counting showsno difference after dark Results incubation in high-magnesium,low-calcium buffer (6475 cpm/ Light-microscopic autoradiography of ‘H-serotonin mg protein/min f 447 SEM; n = 3), as compared to dark accumulation controls (seeabove), presumablybecause the increaseis so small In vitro incubation of isolatedLong-Evans rat retinas in control compared to background accumulation. High-magnesium,low- buffer for 20 min doesnot significantly alter normal morphology calcium buffer had no effect on incubations in the light, as as- (Fig. 1, upper left). Cell bodies are characteristically arranged sessedby autoradiography (Fig. 2, bottom) or scintillation in 3 major layers. The outer nuclear layer (ONL) contains cell counting (11,575 cpm/mg protein/min f 769 SEM; n = 3). bodies of the photoreceptor cell. The inner nuclear layer (INL) Control values are listed above. Magnesiumis a potent blocker contains cell bodiesof horizontal cells,bipolar cells, Muller cells, of synaptic releaseand has little known effect on uptake (Doug- and amacrine cells. The ganglion cell layer (GCL) contains cell las, 1968; Miledi, 1973; Levy et al., 1974); thus the modest bodiesof displacedamacrine cells and ganglioncells. Two plexi- magnesiumenhancement of accumulation in the dark seenau- form layers are clearly discernible,with the outer plexiform layer toradiographically indicates that releaseof 3H-serotoninduring (OPL) representedby a narrow band containing synaptic endings the dark incubation period can only account in small part for of photoreceptor cells, horizontal cell neurites, and bipolar cell the lack of net accumulation under theseconditions. dendrites. The inner plexiform layer (IPL) is seenas a much broader band and contains synaptic contacts between ama- Histofluorescence crines, bipolars, and ganglion cells. Treatment of neuronal tissueswith glyoxylic acid resultsin the When micromolar concentrations of 3H-serotonin are includ- formation of blue fluorescentderivatives of catecholamines,such The Journal of Neuroscience, February 1997, 7(2) 321

Figure 1. Light microscopicautoradiographs of ‘H-serotoninaccumulation in retinasof Long-Evansrat. Top Iej?, Isolatedretina incubatedin controlbuffer for 20 min showsnormal morphology. ONL, Outer nuclearlayer; OPL, outerplexiform layer; ZNL, innernuclear layer; ZPL, inner plexiformlayer; CCL, ganglioncell layer. x 375. Top right, Isolatedretina incubatedin ‘H-serotoninshows specific accumulation of labelin the OPL, and perhapsin the ONL. Arrows denotelabeled vascular elements. Unstained. x375. Bottom leji, Attached retina (eyecuppreparation) incubatedin ‘H-serotoninshows no specificneuronal accumulation. Dark pigment granules areobserved in the pigmentepithelium. Note labeling of bloodvessels (arrows). Unstained.x 275. Bottom right, Stainedautoradiograph of attachedretina for comparisonwith photoat left. x 275. as dopamine, and yellow fluorophores of indoleamines,such as of endogenoushistofluorescence is also observed in the outer serotonin (de la Tort-e and Surgeon, 1976). As has previously retina, in a location correspondingto the OPL. In comparison been reported (Ehinger, 1966), treatment of rat retinas with to the histofluorescencein the inner retina, the fluorescencein glyoxylic acid producesblue fluorescence,characteristic of do- the outer retina is very faint and difficult to visualize under these pamine, in a singlediffuse band in the IPL (Fig. 3). Fluorescence conditions. It is yellow in color, which is characteristic of an appearsin a punctate fashion and is presumedto be associated indoleamine derivative, and appearsto be more uniform rather with cell processesor varicosities from amacrinecells. Fluores- than punctate in its distribution. Prominent fluorescentpatterns cent cell bodiesare sparselydistributed within the INL. A band are also observed in associationwith the sclera. The histoflu- 322 Redburn and Churchill l lndoleamine System in Photoreceptor Cells of Rat Retina

Figure 3. Endogenous histofluorescence of rat retina after glyoxylic acid treatment. Dopamine histofluorescence appears in the ZPL, indoleamine histofluorescence appears in the OPL. Prominent fluorescent patterns are also present in the sclera. x 275.

orescenceof the OPL remains faint even after the application of a variety of in vivo enhancementtechniques that have been shown to enhance serotonin histofluorescencein brain (Rass- mussenand Bunney, 1982). Theseinclude pretreatment of the animal in vivo with colchicine, pargyline, and tryptophan. DHT is a specificneurotoxin for serotonin-accumulatingneu- rons and has thus been widely used as a specific marker for serotonergicneurons in various areasof the CNS (Lachenmayer and Nobin, 1973; Baumgartenand Bjorklund, 1976),including retina (Ehinger and Floren, 1978). We have previously shown that intraocular injections of DHT causednecrosis of serotonin- accumulating amacrine cells in the rabbit retina (Mitchell and Redbum, 1985). Under these same conditions, however, we observe no specific neurotoxic effects of this compound in reti- nasfrom Long-Evans rat (Fig. 4, top and center). Photoreceptor cell nuclei appear normal. There is no evidence of swellingin either the IPL or OPL, no significant necrosisis observed in either the INL or the GCL. DHT is an intracellular toxin and therefore must be accumulated intracellularly in sufficient quantities to expressneu- rotoxic effects (Lachenmayer and Nobin, 1973). DHT is accumulated via the serotonin-uptake systemand thus expressesits effects primarily in serotonin-accumulating neurons. The lack of neurotoxic effects in the Long-Evans rat retina may suggest that the DHT is not accumulatedat sufficient concentrations to produce a neurotoxic effect. In order to test this hypothesis, we

c Figure 2. Effect on light, dark, and magnesiumon 3H-serotoninaccumulation. Top, Isolated retina incubated in ‘H-serotonin for 20 min maanesium chloride added. There is a very modest accumulation of in the dark shows no specific neuronal accumulation. Arrow denotes 3H-ierotonin in the OPL under these conditions. Unstained. x275. labeled blood vessel similar to that seen under light conditions. Un- Bottom, Isolated retina incubated in 3H-serotonin.a. 1 for. 20_. min in the stained. x 275. Center, Isolated retina incubated in )H-serotonin for 20 light in low-calcium, high-magnesium nuner. Accumurauon pattern IS min in the dark in buffer with calcium chloride omitted and 10 mM similar to that seen in control buffer in the light. x 275. The Journal of Neuroscience, February 1987, 7(2) 323 400 h

MIN Figure 5. Releaseof 3H-serotoninfrom isolated,intact rat retina.After preincubationin ‘H-5-HT, retinaswere perfused with buffercontaining 6.2 mM potassium(open bars) for 10 min. Addition of depolarizing levelsof potassium(56.2 mM final concentration)for 5 min causedan increasein releaserates (striped bars). After return to originalbuffer, a secondapplication of 56.2 mM potassiumdid not stimulaterelease. Valuesshown are from a representativeexperiment; n = 3.

examined histofluorescent patterns of rat retina 4 hr after intraocular injection of DHT. Since DHT forms a fluorescent derivative when treated with glyoxylic acid, uptake of DHT can be visualized by histofluorescence(Ehinger and Floren, 1978). Under these conditions, we observed no significant changein the fluorescenceassociated either with inner or outer plexiform layers (Fig. 4, bottom). These data indicate that DHT is not accumulated by cells of the rat retina in amounts sufficient to elicit neurotoxic effects or to reach the limits of detection by histofluorescencetechniques. 3H-serotonin release Isolated retinas were incubated in 3H-serotonin for 20 min, rinsed, and placed in a superfusionchamber in the light. The perfusion buffer contained imipramine to inhibit reuptake. Per- fusateswere collected at the rate of 1 ml per min and counted. A stable, basalefflux rate can be establishedwithin 10 min (Fig. 5). Addition of depolarizing levels of potassium(56.2 mM final concentration) approximately doubles the releaserate during a 5 min exposure period. A return to control buffer (6.2 mM potassium) lowers the releaserate to prestimulation levels. A second exposure to high levels of potassium fails to elicit a significant increasein releaserates. 3H-serotonin accumulation in dystrophic rats To determine whether or not photoreceptor cell terminals are responsiblefor serotonin accumulation in the OPL, we analyzed serotonin-uptake patterns in retinas from the rat rdy, which exhibit spontaneousretinal degeneration.In theseanimals, photoreceptor cells slowly degenerate,sparing the other cell types associatedwith the OPL (Noell, 1965; Cotter and Noell, 1985). Figure 4. Morphologyand histofluorescenceof DHT-treated retinas. Retinas appearrelatively normal at 20 d of age, with only subtle Top, Retinal morphologyis not significantlyaltered after intraocular injectionof DHT. (Seecontrol tissue in Fig. i , top left, for comparison.) Center, In control tissuea brightband of histofluorescence,correspond- t ing to dopaminefluorescence, is seen in the ZPL. (Thehistofluorescence in the OPL shownin Fig. 3 cannotbe appreciatedunder these photo- the DHT, no changein fluorescencepatterns is observed.All magnifi- graphicconditions.) Bottom, Four hoursafter intraocularinjection of cationsx 450. 324 Redburn and Churchill l lndoleamine System in Photoreceptor Cells of Rat Retina

Figure 6. ‘H-serotonin accumulation in retinas of rdy rat. Top left, Morphology of 20-d-old rdy rat is similar to that of Long-Evans control rats. x450. Top right, Accumulation of ‘H-serotonin in the OPL of the 20-d-old rdy rat is similar to that in the Long-Evans control rat. Unstained. x 450. Centerleft, Retina from loo-d-old rdy rat shows loss of inner and outer segments as well as of most cell bodies in the ZNL.The inner retina appears relatively normal. x 500.Center right, Retina from 100-d rdy rat does not specifically accumulate label. x 500.Bottom, Retina from age- matched (loo-d-old) Long-Evans control rat displays uptake similar to that of young adult controls. x 225. The Journal of Neuroscience, Februaty 1987, 7(2) 325

Figure 7. EM autoradiography of ‘H-serotonin accumulation. Top, Individual grains appear in close proximity to virtually every photoreceptor terminal. Other elements of the OPL are not labeled. Grains are also present in photoreceptor cell bodies. x 4000. Bottom, Cone terminals (arrow), although few in number, are also labeled. x 8000. signs of degeneration in rod photoreceptor cells. By 100 d of cumulation appears to result specifically from the loss of pho- age, the majority of rod photoreceptor cells have undergone toreceptor cells and their terminals, which suggests that these degeneration and only a few cone cell remnants remain in the elements are indeed responsible for serotonin accumulation. ONL. Horizontal and bipolar cell processes in the OPL appear to be intact at 100 d of age. Electron-microscopic autoradiography of 3H-serotonin Accumulation of 3H-serotonin by retinas of 20-d-old rdy rat accumulation in the outer plexiform layer retinas (Fig. 6, upper left and right) was virtually indistinguish- EM autoradiography further confirms the uptake of ‘H-sero- able from that seen in control retinas of Long-Evans rats (Fig. tonin specifically by terminals of photoreceptor cells. In retinas 1, upper right). Specific accumulation of label is observed in a incubated for 20 min in 3H-serotonin and processed for EM band corresponding to the location of the OPL. In contrast, autoradiography, grains are observed in close association with retinas from IOO-d-old rdy rats do not show accumulation of photoreceptor cell terminals of both rods and cones (Fig. 7 in 3H-serotonin in the OPL (Fig. 6, center, left and right). The loss Table 1). There is relatively little labeling associated with the of serotonin uptake in the OPL in the loo-d-old rdy rat retina remaining elements of the OPL. Grains are also associated with is not a simple function of age. Age-matched control retinas photoreceptor cell bodies, although their density is significantly from 1OO-d-old Long-Evans control rats show accumulation of lower (Table 1). Grain density in the sparse population of cone 3H-serotonin in the OPL similar to that seen in the 20-d-old terminals observed appeared to be equivalent to that seen in control rat (Fig. 6, bottom). Thus, the loss of 3H-serotonin ac- rod terminals. 326 Redburn and Churchill l lndoleamine System in Photoreceptor Cells of Rat Retina

Table 1. Distribution of 3H-serotonin autoradiographic grains in initial target of this degenerative disease(LaVail et al., 1974; photoreceptor terminals and cell bodies Kaitz, 1976; Ciceroneet al., 1979; Noel1et al., 1981). 3H-serotonin uptake is normal at this stage,which suggeststhat the loss Averageno. of serotonin uptake is not a major feature of the onset of the of grainsin disease. each Unla- Cellular compartment Density beled Number At an advanced stageof dystrophy (100 d), rod cells are vir- compartment (mean+ SEM) (grainslj3) (W counted tually absent and there is no detectable accumulation of 3H- serotonin in the outer plexiform layer. A small population of Photoreceptor partially degenerative cones is present. The loss of serotonin terminals 3.5 + 0.5 0.75 t 0.1 23.4 47 accumulation in the outer plexiform layer thus coincides with Photoreceptor the disappearanceof the degeneratedrod cells and the partial cellbodies 3.6 + 0.6 0.16 +- 0.2 22.6 53 degenerationof cone cells. These data, taken with EM autora- Inner portion diographic observations,support the notion that photoreceptor of the OPL 0.05 + 0.01 cells are indeed the site of serotonin accumulation in the outer plexiform layer. Animals at this stageof degenerationdo retain minimal visual capabilities, suggestingthat some degreeof visual information Discussion processingis occurring via the remainingcone terminals (LaVail Autoradiographic localization of an indoleaminesystem in rat et al., 1974; Noel1et al., 1981). The lack of specificaccumulation retina of serotonin by cone terminals in late-stagerdy retinas suggests In this study, we provide evidence for the presenceof an in- that either cones do not normally take up serotonin or their doleamine system within rod and cone terminals in rat retina. accumulation of serotonin at this degenerative stageis below The localization of the indoleamine system is basedprimarily the level of autoradiographic detection. Our EM autoradio- on the ability of rod and cone terminals to accumulate and graphic resultsin control Long-Evans rats, which show specific release3H-serotonin. At the light-microscopic level, autoradio- accumulation by cone terminals, suggestthat the latter possi- graphic localization of 3H-serotonin is associatedspecifically bility is more likely. with the outer portion of the outer plexiform layer. With par- Many vertebrate retinas demonstrate specific accumulation gyline and ascorbateincluded in the incubation buffer to inhibit of serotonin by a subsetof amacrine cell bodies and processes metabolism and oxidation, the tritium label remains as au- within specific sublaminaof the inner plexiform layer (Ehinger thentic 3H-serotonin when the retina is maintained in the light. and Floren, 1980; Osborne, 1982; Mitchell and Redburn, 1985). This is not the casewhen the retina is exposed to darkness,as However, 3H-serotonin is not accumulated by any element in will be discussedbelow. the inner retina of the rat. In particular, there are no amacrine Electron-microscopic autoradiography demonstrates that cells that are specifically labeled by 3H-serotonin. The autora- specific accumulation of 3H-serotonin in the outer plexiform diographic grains of 3H-serotonin observed in the inner retina layer is associatedwith the terminal region of rods and cones. are uniformly distributed and are significantly reduced com- There is no specific accumulation in inner or outer segments. pared to labeling in the outer plexiform layer. These data con- Accumulation by photoreceptor cell bodies is only slightly en- firm previous reports (Ehinger and Floren, 1979, 1980)that the hancedcompared to that seenin the remaining retina. Other rat retina doesnot possess3H-serotonin-accumulating amacrine elementsof the outer plexiform layer, including horizontal and cells. Thus serotonergicproperties in these retinas can be at- bipolar cells neurites, do not appear to accumulate serotonin. tributed solely to photoreceptor cell terminals. Although this representsthe most likely interpretation of our The retina of the squirrel monkey provides yet another vari- autoradiographs,other interpretationsmight be considered.Since ation in the distribution of serotonin-accumulatingcells. It has both bipolar and horizontal cell neurites form deep invagina- recently been reported that horizontal cells accumulate sero- tions within rod and coneterminals, it is possiblethat the labeled tonin in this species(Floren and Hendrickson, 1984).Our results compound is associatedwith either of thesepostsynaptic ele- in Long-Evans rat retinas did not demonstrate any specificac- ments and, as a result, producesan autoradiographic labeling cumulation of serotonin by horizontal cells. Both the EM au- pattern coincident with photoreceptor terminals. Although no toradiography and the experiments with the rdy rat essentially statistical analyseswere made of the intraterminal distribution rule out the possibility that the labelingwe observedin the outer of grains,our subjective interpretation is that the grains do not plexiform layer is associatedwith any element other than the appearto be associatedwith postsynaptic elementsin the triad. photoreceptor terminal. In addition, grains were rarely seenin the inner portion of the outer plexiform layer occupied by neurites of horizontal and Characteristicsof indoleamineaccumulation bipolar cells, which are postsynaptic to photoreceptors. These Accumulation of serotonin within photoreceptor terminals has observations, although not conclusive, do suggestthat the la- not previously been reported for retinas of any species.This beling observedis associatedwith the presynaptic elementrath- apparent discrepancybetween our findingsand previous reports er than the postsynaptic element of the photoreceptor cell syn- by others may in fact be related to a number of characteristics apse. of the serotonin-uptake systemin the outer plexiform layer that Further support for this interpretation is provided by studies differ significantly from the serotonin-uptake system observed that demonstrate a loss of serotonin accumulation in the outer in amacrine cells. One of the more important technical consid- plexiform layer coincident with the degeneration of photore- erations is the fact that specific accumulation is observed only ceptor cells in the rdy rat retina. At 20 d of age, retinas begin whenthe isolatedretina is incubated in vitro in buffer containing to show signsof degenerationin rods, which appear to be the the label. Intraocular injections or in vitro incubations of eyecup The Journal of Neuroscience, February 1987, 7(2) 327 preparations with retina attached do not allow free diffusion of DHT-induced fluorescence. Thus, the absence of neurotoxic the label through both retinal surfaces. We assume that the effects may be due to lack of sufficient accumulation of DHT. attached retinal preparations used by other investigators pre- cluded the distribution of serotonin to the outer portions of the The fate of accumulated serotonin in photoreceptor terminals retina, which would be required for uptake by photoreceptor There is no specific accumulation of serotonin in retinas exposed terminals. On the other hand, attached retinal preparations are to darkness. Since the photoreceptor cell depolarizes and re- sufficient for labeling of serotonin-accumulating amacrine cells. leases its neurotransmitter in response to darkness (for a review, One important characteristic of the serotonin-uptake system see Rodieck, 1973), the lack of observed accumulation may in photoreceptor terminals is the high degree of sensitivity to result from a dark-stimulated release of serotonin. In support lighting conditions during the incubation period. Exposure of of this notion, our data demonstrate that in the light, depolar- the rat retina to darkness during in vitro incubation with the izing levels of potassium stimulate the release of 3H-serotonin label blocks the specific accumulation of serotonin in photo- as measured by scintillation counting. On the basis of these receptor cell terminals. In contrast, lighting conditions have no observations, it is tempting to speculate that photoreceptors may effect on the low, uniform distribution of autoradiographic grains use an indoleamine compound as a transmitter. However, recent observed over the remaining portions of the retina. Even under evidence describing a melatonin system in retina may provide the most favorable lighting conditions used, however, we were alternative interpretations. unable to enhance serotonin accumulation in photoreceptor ter- Several reports over the past 2 decades have documented the minals to the level previously observed by autoradiography or presence of melatonin in retinas from various species (Quay, histofluorescence in amacrine cells and their processes in other 1965; Cardinali and Rosner, 197 la; Bubenik et al., 1974). As species, such as rabbit (Mitchell and Redburn, 1985). The up- in the pineal gland, melatonin synthesis in retina is known to take by photoreceptor terminals simply appears to be less robust increase in the dark and decrease in the light (Pang et al., 1977; than that observed in amacrine cells. Whether this property is Bubenik et al., 1978). The increase in both retinal and pineal a result of low transport affinity, low number of transport sites, melatonin synthesis results from dark activation of N-acetyl smaller pool size, or higher turnover rate cannot be determined transferase (NAT), the rate-limiting enzyme for melatonin syn- on the basis of these experiments. thesis (Binkley et al., 1979, 1980; Iuvone and Besharse, 1983). Melatonin synthesis in the pineal gland has been localized to Localization of indoleamine histojluorescence pinealocytes (Bubenik et al., 1974). Since pinealocytes in the A very pale yellow band of endogenous fluorescence, indicative pineal gland and the photoreceptors in the retina are analogous of indoleamines, is observed in the outer plexiform layer in in many ways, photoreceptors are implicated as the source of untreated retinas. We were unsuccessful in our attempts to en- retinal melatonin. Hydroxy indole-O-methyl transferase hance this fluorescence by treatment of the animal and/or iso- (HIOMT), one of the non-rate-limiting enzymes for melatonin lated retina with tryptophan, serotonin, pargyline, ascorbate, synthesis, has been localized to photoreceptor terminals (Car- colchicine, or varying lighting conditions. Serotonin accumu- dinali and Rosner, 197 lb; Wiechmann et al., 1985). In a pre- lation, clearly demonstrated by autoradiography, is not ob- liminary report, we showed that the 3H-serotonin accumulated served by histofluorescence. These findings are in agreement in rat retina in the light is rapidly converted to melatonin when with the widely recognized observation that histofluorescence the retina is subsequently exposed to darkness (Redbum and is a less sensitive method for detecting indoleamines than is Mitchell, 1986). Therefore the decrease in 3H-serotonin accu- autoradiography. For example, uptake of serotonin into ama- mulation seen in darkness may be due, at least in part, to dark- crine cells of cow and pig retina is seen only with autoradiog- stimulated activation of the NAT enzyme, which converts sero- raphy, not with histofluorescence (Ehinger et al., 1982; Osborne tonin to melatonin. et al., 1982). Nevertheless, the endogenous fluorescence in the Newly synthesized melatonin may be released by diffusion rat retina outer plexiform layer does not provide convincing from photoreceptors (Yu and Chow, 1982), as is the case for evidence for an endogenous indoleamine pool in photoreceptor pinealocyes (Cardinali, 198 1). Melatonin is highly lipophilic terminals. Until there is a clear demonstration of the localization (Lemer et al., 1959) and its release is not considered to be of an endogenous indoleamine, we cannot be certain as to the depolarization- or calcium-dependent, in contrast to most clas- exact nature of the indole compound that functions as the en- sical neurotransmitter release systems. Rather, its release (dif- dogenous ligand for the indoleamine-accumulating system in fusion) rate is directly linked to its synthesis rate, as regulated photoreceptor terminals. by NAT (Cardinali, 198 1). Serotonin is presumably released Uptake of the serotonin analog DHT is not detectable in any from pinealocytes via stimulation-secretion coupling, and it has cell of the rat retina. DHT is a neurotoxin specifically accu- neurotransmitter-neurohormone effects on other pineal cells and mulated by serotonergic neurons in the CNS, including sero- other parts of the brain (Quay, 1963, 1965; Koella, 1969; Ow- tonin-accumulating amacrine cells in retinas of other species man, 1971). The indole system in the photoreceptor cell may (Lachenmayer and Nobin, 1973; Baumgarten and Bjorklund, mimic that of the pinealocyte in releasing both melatonin and 1976; Ehinger and Floren, 1978). This analog also produces a serotonin under varying conditions and for different physiolog- fluorescent derivative when exposed to glyoxylic acid. As pre- ical purposes. viously reported (Mitchell and Redbum, 1985), the uptake of It is not possible to determine at this point whether any indole DHT by serotonergic amacrine cells in the rabbit retina after compound functions as the primary neurotransmitter released intraocular injection can be monitored by histofluorescence after from photoreceptor cells, which subsequently stimulates and short exposure (hours) and by cell necrosis and cell death after inhibits horizontal and bipolar cells. Primarily on the basis of longer exposure (days). Intraocular injections of DHT equiva- electrophysiological and pharmacological experiments, the ma- lent to those used in rabbit had no toxic effect on cells in the jor photoreceptor neurotransmitter is currently thought to be outer retina of the rat. Likewise, we were unable to detect any an amino acid, such as glutamate or aspartate. (For a review, 328 Redburn and Churchill l lndoleamine System in Photoreceptor Cells of Rat Retina

see Miller and Slaughter, 1985.) However, it should be noted Ehinger, B., and I. Floren (1980) Retinal indoleamine accumulating that electrophysiological analyses of serotonin receptors in ret- neurons. Neurochemistry 1: 209-229. Ehinger, B., I. Floren, and K. Tomqvist (1982) Autoradiography of ina have previously focused on the effects of indoles on elements (3H)-5-hydroxytryptamine uptake in the retina of some mammals. of the inner plexiform layer (amactine and ganglion cells) rather Albrecht Von Graefes Arch. Klin. Exp. Ophthalmol. 218: 1-18. than on the outer plexiform layer (Straschill, 1968; Ames and Floren, I., and A. Hendrickson (1984) Indoleamine-accumulating hor- Pollen, 1969; Straschill and Perwein, 1969; Thier and Wassle, izontal cells in the squirrel monkey retina. Invest. Ophthalmol. Vis. 1984). Another complicating factor is that all of the reported Sci. 25: 997-1006. Iuvone, P. M., and J. C. Besharse (1983) Regulation of indoleamine studies have used species that contain indoleamine-accumulat- N-acetvltransferase activitv in the retina: Effects of light and dark, ing amacrine cells. A more thorough assessment of the possible protein synthesis inhibitors and cyclic nucleotide anzogues. Brain role of the indole system in photoreceptors would be greatly Res. 273: 11 l-l 19. aided by an in-depth electrophysiological analysis of indoles in Kaitz, M. (1976) The effect of light on brightness perception in rats with retinal dystrophy. Vision Res. 16: 141-145. the outer plexiform layer of a species such as rat, which does Koella, W. P. (1969) Serotonin and sleep. Exp. Med. Surg. 27: 157- not possess indoleamine-accumulating amacrine cells. None- 168. theless, the available data do provide strong biochemical and Lachenmayer, L., and A. Nobin (1973) Evaluation of the effects of morphological evidence for an indoleamine system in photo- 5,7-dihydroxytryptamine on serotonin and catecholamine neurons in receptors. Indoleamine uptake, synthesis, and release processes cat CNS. Acta Physiol. Stand. (Suppl.) 391: 1-19. LaVail, M. M., M. Sidman, R. Rausin, and R. L. Sidman (1974) Dis- are localized in photoreceptor cells and appear to be regulated crimination of light intensity by rats with inherited retinal degener- by light and dark. This regulation argues in favor of a physio- ation. Vision Res. 14: 693-670. logically important role for this indoleamine system. Lemer, A. B., J. D. Case, and R. V. Heinzelman (1959) Structure of melatonin..J. Am. Chem. Sot. 81: 6084-6085. Iew. W. B.. J. W. Havcock. and C. W. Cotman (1974) Effects of References polyvalentcations on stimulus coupled secretion of14C-gamma-ami- Ames, A., and D. A. Pollen (1969) Neurotransmission in central ner- nobutyric acid from isolated brain synaptosomes. Mol. Pharmacol. vous tissue: A study of isolated rabbit retina. J. Neurophysiol. 32: 10: 438-449. 424-442. Miledi, R. (1973) Transmitter release induced by injection of calcium Baumgarten, H. G., and A. Bjorklund (1976) Neurotoxic indoleamines ions into nerve terminals. Proc. R. Sot. Lond. [Biol.] 183: 421-425. and monoamine neurons. Annu. Rev. Pharmacol. Toxicol. 16: lOl- Miller, R. F., and M. M. Slaughter (1985) Excitatory amino acid 111. receptors in the vertebrate retina. In Retinal Transmitters and Mod- Binkley, S., M. Hryshchyshyn, and M. Reilly (1979) NAT responds ulators: Models for the Brain, vol. II, W. W. Morgan, ed., pp. 123- to environmental lighting in the eye as well as in the pineal gland. 160, CRC Press, FL. Nature 281: 479-48 1. Mitchell, C. K., and D. A. Redbum (1985) Analysis of pre- and post- Binkley, S., K. B. Reilly, and M. Hryshchyshyn (1980) N-Acetyltrans- synaptic factors of the serotonin system in rabbit retina. J. Cell Biol. ferase in the chick retina. I. Circadian rhythms controlled by envi- 100: 64-73. ronmental lighting are similar to those in the pineal gland. J. Comp. Noell, W. K. (1965) Aspects of experimental and hereditary retinal Physiol. Biochem. Syst. Environ. Physiol. 139: 103-108. degeneration. In Biochemistry of the Retina, C. N. Graymore, ed., Bloom, F. E., B. J. Hoffer, G. R. Siggens, J. L. Barker, and R. A. Nicoll ppy 5 l-72, Academic, New York. (1972) Effects of serotonin on central neurons, microiontophoretic Noell. W. K.. M. C. Salinskv. R. A. Stockton. S. B. Schnitzer. and V. administration. Fed. Proc. 31: 97-106. Kan (198 1) Electrophy&logical studies of the visual capacities at Bubenik, G. A., G. M. Brown, L. Uhlir, and L. J. Grota (1974) Im- advanced stages of photoreceptor degeneration in the rat. Dot. munohistochemical localization of N-acetylindolealkylamines in pi- Ophthalmol. 27: 175. neal gland, retina and cerebellum. Brain Res. 81: 233-242. Osborne, N. N. (1980) In vitro experiments on the metabolism, uptake Bubenik, G. A., R. A. Purtill, G. M. Brown, and L. J. Grota (1978) and release of 5-hydroxytryptamine in bovine retina. Brain Res. 184: Melatonin in the retina and the Harderian gland Ontogeny, diurnal 283-297. variations and melatonin treatment. Exp. Eye Res. 27: 323-333. Osborne, N. N. (1982) Evidence for serotonin being a neurotransmitter Cardinali, D. P. (198 1) Melatonin: A mammalian pineal hormone. in the retina. In Biology of Serotonin Transmission, N. N. Osborne, Endocrine Rev. 2: 327-346, ed., pp. 401-430, Wiley, New York. Cardinali, D. P., and J. M. Rosner (1971a) Metabolism of serotonin Osborne, N. N., T. Nesselhut, D. A. Nicholas, S. Patel, and A. C. Cue110 by the rat retina in vitro. J. Neurochem. 18: 1769-l 770. (1982) Serotonin-containing neurons in vertebrate retinas. J. Neu- Cardinali, D. P., and J. M. Rosner (197 1b) Retinal localization of the rochem. 39: 1519-1528. hvdroxvindole-O-methvl transferase (HIOMT) in the rat. Endocri- Owman, C. (197 1) General discussion: Role of serotonin in the pineal. nblogy89: 301-303. - In The Pineal Gland, G. E. W. Wolstenholme and J. Knight,-. eds., Cicerone, C. M., D. G. Green, and L. J. Fisher (1979) Cone inputs to pp. 361-363, Churchill Livingstone, London. ganglion cells in hereditary retinal degeneration. Science 203: 1113. Pann. S.. G. M. Brown. L. J. Grota. J. W. Chambers. and R. L. Rodman Cotter, J. R., and W. K. Noel1 (1985) Ultrastructure of remnant pho- (1577) Determination ofN-acetyl-serotonin and melatonin activities toreceptors in advanced hereditary retinal degeneration. Invest. in the pineal gland, retina, Harderian gland, brain and serum of rats Ophthal. Vis. Sci. 25: 1366-1375. and chickens. Neuroendocrinology 23: 1-13. de la Terre, J. C., and J. W. Surgeon (1976) A methodological approach Quay, W. B. (1963) Circadian rhythm in rat pineal serotonin and its to rapid and sensitive monoamine’histolhrorescence using a modified modifications by estrous cycles and photoperiod. Gen. Comp. En- alvoxvlic acid techniaue: The SPG method. Histochem. J. 49: 8 l-93. docrinol. 3: 473-479. D&&as, W. W. (1968) Stimulus-secretion coupling: The-concept and Quay, W. B. (1965) Retinal and pineal hydroxyindole-O-methyl trans- the clues from chromaffin and other cells. Br. J. Pharmacol. 34: 45 l- ferase activity in vertebrates. Life Sci. 4: 983-99 1. 474. Rassmussen, S. A., and B. S. Bunney (1982) A modification of the Ehinger, B. (1966) Adrenergic neurons in the retina. Life Sci. 5: 129- rapid glyoxylic acid technique permits visualization of serotonergic 133. and hypothalamic dopaminergic neurons. J. Neurosci. Methods 6: Ehinger, B., and I. Floren (1978) Chemical removal of indoleamine 139-144. accumulating terminals in rabbit and goldfish retina. Exp. Eye Res. Redbum, D. A., and C. K. Mitchell (1986) Endogenously synthesized 26: 321-328. serotonin and melatonin in rat retina. Invest. Ophthalmol. Vis. Sci. Ehinger, B., and I. Floren (1979) Absence of indoleamine-accumu- (Suppl.) 1: 183. lating neurons in the retina of humans and cynomolgus monkeys. Rodieck, R. W. (1973) The Vertebrate Retina, W. H. Freeman, San Albrecht Von Graefes Arch. Klin. Exp. Ophthalmol. 209: 145-153. Francisco. The Journal of Neuroscience, February 1987, 7(2) 329

Sitaram, B. R., and G. J. Lees (1978) Diurnal rhythms and turnover Thomas, T. N., and D. A. Redburn (1979) 5-HT: A neurotransmitter of tryptophan hydroxylase in the pineal gland of the rat. J. Neuro- of bovine retina. Exp. Eye Res. 28: 55-61. them. 31: 1021-1026. Tornqvist, K., C. H. Hansson, and B. Ehinger (1983) Immunohisto- Straschill, M. (1968) Actions on single neurons in the cat’s retina. chemical and quantitative analysis of 5-hydroxytryptamine in the Vision Res. 8: 35-47. retina of some vertebrates. Neurochem. Int. 5: 299-308. Straschill, M., and J. Perwein (1969) The inhibition ofretinal ganglion Wiechmann, A. F., D. Bok, and J. Horowitz (1985) Localization of cells by catecholamines and y-aminobutyric acid. Pfluegers Arch. 312: hydroxyindole-O-methyltransferase in the mammalian pineal gland 42-54. and retina. Invest. Ophthalmol. Vis. Sci. (Suppl.) 26: 253. Thier, P., and H. Wassle (1984) Indoleamine-mediated reciprocal Yu, H. S., and D. H. Chow (1982) Effect of light and darkness on the modulation of on-centre and off-centre ganglion cell activity in the in vitro release of N-acetylserotonin and melatonin by the retina of retina of the cat. J. Physiol. (Lond.) 351: 613-630. guinea pigs. Neuroendocrinology 34: 265-268.