Localization of hydroxyindole-O-methyltransferase-like immunoreactivity in photoreceptors and cone bipolar cells in the human retina: a light and electron microscope study.

The localization of the melatonin-synthesizing enzyme hydroxyindole-O-methyltransferase (HIOMT) was examined by light and electron microscopic immunocytochemistry in the human retina. HIOMT-like immunoreactivity was observed in the photoreceptor layers and the inner nuclear layer (INL). The immunoreactive cells in the INL were more numerous in the central retina than in the peripheral retina and sent processes to both the outer plexiform and inner plexiform layers. The HIOMT immunoreactivity in the inner plexiform layer (IPL) appeared as punctate terminals in the proximal and distal one-thirds of that layer. At the ultrastructural level, HIOMT-like immunoreactivity was localized to the cytoplasm of rod and cone photoreceptors and to a population of cone bipolar cells. HIOMT-immunoreactive bipolar cell dendrites were observed to make both invaginating and flat synaptic contacts with cone pedicles. No immunoreactive invaginating contacts in rod spherules were observed. HIOMT immunoreactivity was observed in the bipolar cell cytoplasm in the INL, and in the bipolar synaptic terminals in the IPL. These terminals contained synaptic ribbons, which formed synaptic contacts with unlabeled cells in the IPL. HIOMT radioenzymatic assays confirmed the presence of HIOMT in the human retina. Average HIOMT activity of eight donors was determined to be 15.0 pmol/mg protein/hour +/- 7.2 S.D. The ultrastructural localization of HIOMT observed in this study, combined with reports from other laboratories, suggests that the cytoplasm of the photoreceptors and a population of cone bipolar cells may be the sites of melatonin synthesis in the human retina.

Diffuse bipolar cells provide input to OFF parasol ganglion cells in the macaque retina.

Parasol retinal ganglion cells are more sensitive to luminance contrast and respond more transiently at all levels of adaptation than midget ganglion cells. This may be due, in part, to differences between bipolar cells that provide their input, and the goal of these experiments was to study these differences. Midget bipolar cells are known to be presynaptic to midget ganglion cells. To identify the bipolar cells presynaptic to parasol cells, these ganglion cells were intracellularly injected with Neurobiotin, cone bipolar cells were immunolabeled, and the double-labeled material was analyzed. In the electron microscope, we found that DB3 diffuse bipolar cells labeled by using antiserum to calbindin D-28k were presynaptic to OFF parasol cells. In the confocal microscope, DB3 bipolars costratified with OFF parasol cell dendrites and made significantly more appositions with them than expected due to chance. Flat midget bipolar cells were labeled with antiserum to recoverin. Although they made a few appositions with parasol cells, the number was no greater than would be expected when two sets of processes have overlapping distributions in the inner plexiform layer. DB2 diffuse bipolar cells were labeled with antibodies to excitatory amino acid transporter 2, and they also made appositions with OFF parasol cells. These results suggest that DB2 bipolar cells are also presynaptic to OFF parasol ganglion cells, but midget bipolar cells are not. We estimate that midperipheral OFF parasol cells receive approximately 500 synapses from 50 DB3 bipolar cells that, in turn, receive input from 250 cones.

Distribution of melatonin receptors in the brain of the frog Rana pipiens as revealed by in vitro autoradiography.

Melatonin binding sites were identified in the brain of the frog Rana pipiens using in vitro autoradiography. Coronal sections were incubated for 1 h in 100 pM 2-125I-iodomelatonin. Specific binding was displaced with 1 microM nonradioactive melatonin. Autoradiographic labeling of 3H-Hyperfilm was observed in areas that receive primary, secondary, and tertiary visual input: the superficial layers of the optic tectum, anterior and posterior thalamic nuclei, striatum, medial pallium, and interpeduncular nucleus. Other areas that demonstrated binding included the medial and lateral septal nuclei, medial preoptic area, suprachiasmatic region, and anterodorsal tegmental nucleus. Binding was also apparent in the distribution of the lateral olfactory tract (lateral pallium), and in tracts associated with visual pathways: optic nerve, chiasm and tract, and lateral and medial forebrain bundles. A high degree of melatonin binding was observed in the left habenular nucleus, but not in the habenulum of the right side of the brain. Radioreceptor binding assays of frog whole-brain homogenate demonstrated specific saturable melatonin binding (Kd = 70 pM, Bmax = 0.80 fmol/mg protein). Melatonin and 6-chloromelatonin were potent displacers of 2-125I-iodomelatonin, while 5-hydroxytryptamine, 5-methoxytryptamine, and N-acetylserotonin were much less potent. Melatonin inhibited the forskolin-stimulated increase in cAMP synthesis in optic tectum explants. These results suggest that high-affinity melatonin receptor binding sites are widely distributed in the telencephalon, diencephalon, and mesencephalon and are very prominent in areas of the frog brain that are associated with visual processing.

Localization of hydroxyindole-O-methyltransferase in the mammalian pineal gland and retina.

The pineal hormone, melatonin, has been reported to be synthesized in the retina by the enzyme, hydroxyindole-O-methyltransferase (HIOMT). Several laboratories have suggested that melatonin may be involved in photoreceptor outer segment disc shedding, photomechanical movements, and neuromodulation, but the cellular location of the retinal synthesizing enzymes has not been determined yet. Antiserum to HIOMT was obtained from rabbits immunized with bovine pineal extract. The monospecific immunoglobulins to HIOMT were isolated by positive-negative selection using pineal extract-sepharose and brain extract-sepharose affinity chromatography. The purity and specificity of the antibody to HIOMT was confirmed by immunodiffusion, electroblot immunolabeling, SDS-PAGE, and immunoprecipitin titration. Using the peroxidase-antiperoxidase (PAP) technique, HIOMT was localized in the pinealocytes of bovine and human pineal glands obtained during the light period. Rat pineal glands obtained during the dark period exhibited HIOMT immunoreactivity, whereas rat pineal glands obtained during the light period did not. Some pinealocytes of the bovine pineal did not exhibit HIOMT immunoreactivity, suggesting that not all pinealocytes are actively involved in melatonin synthesis. HIOMT was localized in the photoreceptors of bovine, rat, and human retinas, and some labeling also was observed in the inner retina, although the latter showed some species variation. This observation supports the hypothesis that photoreceptors are capable of melatonin synthesis.

Melatonin increases photoreceptor susceptibility to light-induced damage.

Melatonin is an indolamine hormone synthesized in the retina and pineal gland. It is thought to act as a paracrine neurohormone in the mammalian retina. Pinealectomy has been shown to protect photoreceptors from light-induced damage, and melatonin treatment has been reported to increase the degree of photoreceptor damage in albino rats. To determine how melatonin influences photoreceptor survival, the effect of melatonin administration on light-induced retinal damage was studied. Melatonin was administered to albino rats by intraperitoneal injections at various times before or after light exposure. The rats were exposed to high-intensity illumination (1600 lux) for 24 hr to induce photodamage, then returned to cyclic lighting for 12 days. After this, they were killed, and their eyes were removed and examined histologically. Measurements of the outer nuclear layer (ONL) thickness were taken at 12 different loci around the circumference of the retinal sections. The animals that received daily melatonin injections (100 micrograms) in the late afternoon (3 hr before lights off) for 1-3 days before photodamage showed an approximate 30% greater reduction compared with sham control animals in ONL thickness in the superior quadrant, the area most susceptible to light damage. Melatonin injections given after the photodamage did not affect ONL thickness. Although retinal susceptibility to light damage varied with time of day, the degree to which melatonin increased the degree of damage appeared unaffected by the time of day. These results suggest that melatonin may be involved in some aspects of photoreceptor sensitivity to light damage.

Melatonin-binding in the frog retina: autoradiographic and biochemical analysis.

Binding of melatonin was examined in the retina of Rana pipiens. When intact frog retinas were incubated with 3H-melatonin and processed for autoradiography, most of the radioactivity was localized to the melanosomes of the retinal pigment epithelium-choroid (RPE-choroid) and to the outer plexiform layer of the retina. Melanosome-enriched fractions of the RPE-choroid and membrane-enriched fractions of the neural retina demonstrated saturable melatonin binding when incubated with increasing melatonin concentration. Thin-layer chromatography showed that greater than 98% of the bound radioactivity was authentic melatonin. Scatchard analysis revealed a single population of binding sites with apparent Kd values of 6 X 10(-7) M for both the RPE-choroid and neural retina. When various indole analogs were tested for their ability to inhibit 3H-melatonin binding to the neural retina, both 5-methoxytryptophol and 6-chloromelatonin demonstrated complete displacement of melatonin binding. Endogenous retinal melatonin levels were measured by radioimmunoassay. A twofold increase in melatonin levels was observed during the dark period with peak levels at 384.5 +/- 28.8 pgms melatonin/pair retinas. Melatonin levels persisted in constant darkness, but were suppressed in constant light. Our data suggest that in the frog, the sites of action of retinal melatonin are the melanosomes of the RPE-choroid and the outer plexiform layer of the neural retina.

Genetic and physical mapping of human recoverin: a gene expressed in retinal photoreceptors.

PURPOSE: Recoverin is a calcium-binding protein that may be involved in phototransduction in mammalian retinal photoreceptors, and is considered to be a candidate gene for retinitis pigmentosa. This study was undertaken to develop the recoverin locus into a polymorphic marker for future linkage studies on retinitis pigmentosa families. METHODS: A human genomic cosmid clone was isolated and used to map the recoverin gene to a human chromosome through hybridization to a panel of somatic hybrid cell line DNAs, and to human metaphase chromosomes by fluorescence in situ hybridization. A dinucleotide repeat polymorphism located within the coding region of the recoverin gene was identified, and used to genetically map the recoverin gene relative to index markers. In addition, three restriction fragment length polymorphisms revealed by the cosmid clone were identified and characterized. RESULTS: Hybridization to the somatic hybrid cell line DNAs localized the recoverin gene to chromosome 17. Recoverin was further localized to 17p12-p13 by fluorescence in situ hybridization. The dinucleotide repeat polymorphism and restriction fragment length polymorphisms at the recoverin locus have a cumulative polymorphic information content = 0.71. CONCLUSIONS: These polymorphic markers and additional closely linked markers will be useful for linkage analysis of families with retinitis pigmentosa.

Diurnal expression of recoverin in the rat retina.

The levels of expression of recoverin mRNA and protein was examined during a 24-h period in the rat retina. Northern blot analysis revealed that rat recoverin mRNA expression was consistently high during the light period, then decreased after onset of darkness, and gradually increased later during the dark period. The cyclic rhythm in recoverin protein expression was consistent with the cyclic rhythm in recoverin mRNA expression, insofar as recoverin mRNA and protein levels were lowest soon after lights off, and there was an increase in expression late in the dark period. These observations suggest that the rate of recoverin transcription may occur maximally during the period of greatest exposure to light, presumably when it is most needed by the photoreceptor to fulfill its role in visual transduction, then decreases at night, when high levels of expression are not required.

Melatonin receptor RNA is expressed in photoreceptors and displays a diurnal rhythm in Xenopus retina.

Melatonin is an output signal of an endogenous circadian clock of retinal photoreceptors, with highest levels occurring at night. Melatonin synthesized in the retina appears to act as a paracrine signal by binding to specific receptors in the eye. We have previously demonstrated that RNA encoding the Mel(1b) and Mel(1c) melatonin receptor subtypes is expressed in the Xenopus laevis retina. The goal of this study was to determine the distribution of the Mel(1b) and Mel(1c) receptor subtype RNA expression in the retina, and to determine if the level of expression of these receptors exhibits a diurnal rhythm. Sections of frog neural retina were analyzed by in situ hybridization with 35S-labeled Xenopus Mel(1c) and Mel(1b) riboprobes. Hybridization was present in cells of the inner nuclear layer and the ganglion cell layer. Moreover, there was hybridization in the photoreceptors, which has not been previously reported. To test the hypothesis that retinal melatonin receptor mRNA undergoes a diurnal rhythm of expression, total RNA was isolated from frog neural retinas obtained at 3-h intervals during a 24-h period. The total RNA was used in real-time PCR assays to quantify the differences in Mel(1b) and Mel(1c) receptor mRNA expression at various circadian times. Both the Mel(1b) and Mel(1c) receptor RNA demonstrated a diurnal rhythm of expression, with peak levels occurring late in the light period, and lowest levels late in the dark period. These results support the hypothesis that RNA encoding melatonin receptors undergo a diurnal rhythm of expression. To further investigate the possible expression of the Mel(1a) receptor subtype in Xenopus retina, we generated Mel(1a) PCR products in genomic DNA, and in reverse-transcribed neural retina and retinal pigment epithelium (RPE) RNA. The identity of the PCR product was confirmed by sequencing. Therefore, all three known Xenopus melatonin receptor subtypes appear to be expressed in the neural retina and RPE.

Melatonin receptor RNA expression in Xenopus retina.

Melatonin is an indolamine hormone presumably synthesized by retinal photoreceptors, and may act as a paracrine signal of darkness within the retina. Previous studies have suggested that melatonin, acting through specific receptors, may be involved in cyclic retinal functions such as photoreceptor outer segment disc shedding and phagocytosis, and modulation of neurotransmitter release in the inner retina. The goal of this study was to determine if melatonin receptor mRNA is expressed in the neural retina and retinal pigment epithelium (RPE) of Xenopus laevis. Sheets of RPE, devoid of contaminating cells, were obtained from Xenopus eyes, and epithelial cultures were subsequently established on microporous membrane filters in a defined medium. Total RNA was isolated from whole brain, neural retina, fresh RPE sheets, and cultured RPE cells. RNA expression of the three known Xenopus melatonin receptor subtypes (MEL1A, 1B, and 1C) was determined by reverse-transcription/polymerase chain reaction (RT/PCR) amplification, followed by Southern hybridization with RNA probes. PCR-amplified cDNA encoding melatonin receptor subtypes 1B and 1C, but not 1A, were detected in reverse-transcribed RNA obtained from brain, neural retina and RPE. RPE cells grown in culture for two weeks also demonstrated 1B and 1C receptor RNA expression. This study suggests that RNA encoding the 1B and 1C melatonin receptor subtypes is expressed in the neural retina and RPE of Xenopus retina, and the expression persists in RPE cells when grown in culture. The expression of melatonin receptor RNA in the RPE may reflect a regulatory role for melatonin in some diurnal events that occur in this tissue, such as phagocytosis of photoreceptor outer segment membranes, and intracellular migration of pigment granules.

Hydroxyindole-O-methyltransferase in rat retinal bipolar cells: persistence following photoreceptor destruction.

The presence of hydroxyindole-O-methyltransferase (HIOMT) activity and localization of HIOMT immunoreactivity was examined in albino rat retinas following photoreceptor destruction. Male Sprague-Dawley rats were exposed to high intensity fluorescent light for 4 consecutive days, then placed on a 14:10 h light:dark cycle for two weeks to allow for phagocytic removal of damaged cells from the retina. Histologic examination revealed almost complete destruction and removal of all photoreceptors. The damaged retinas exhibited an increase in HIOMT activity relative to controls, when expressed as activity per mg of protein. HIOMT activity in the pineal glands was not affected. When control and light damaged retinas were examined for HIOMT localization by immunocytochemistry, the control retinas displayed intense HIOMT immunoreactivity in all photoreceptors, and a somewhat lighter labeling in a population of bipolar cells, whereas the light damaged retinas (lacking photoreceptors) showed intense HIOMT immunoreactivity in bipolar cells. These results suggest that the increase in HIOMT activity following photoreceptor destruction is due to increased synthesis of this enzyme in a population of bipolar cells. These HIOMT-immunoreactive bipolar cells may perhaps respond in a compensatory manner to changing levels of melatonin in the retina.

Asymmetric distribution of melatonin receptors in the brain of the lizard Anolis carolinensis.

The pineal hormone melatonin may regulate seasonal reproduction and entrainment of circadian rhythms by binding to specific brain receptors. An analysis of melatonin receptor distribution in the lizard brain revealed an asymmetry of melatonin binding in the diencephalon. A high degree of melatonin binding was present in the left habenular nucleus, but no binding was observed in the habenulum of the right brain hemisphere. It is intriguing that the left habenular nucleus, in contrast to the right habenulum, both possesses a high density of melatonin receptors and receives primary photic input from the parietal eye. Similarly, the optic tectum, which receives primary visual input from the retina, is also rich in melatonin receptors. These observations suggest that the left habenulum is under dual control (neuronal and hormonal) of the parietal eye/pineal complex, and that melatonin may play a significant role in neural processing of visual information.

Melatonin enhances horizontal cell sensitivity in salamander retina.

Intracellular electrophysiological recording techniques were utilized to investigate the possible function of retinal melatonin in the larval tiger salamander. Endogenous retinal melatonin was present and appeared to bind a membrane-enriched fraction of the salamander retina, as determined by radioimmunoassay and receptor binding studies. Melatonin added through the perfusion bath to flat-mounted retinas resulted in a horizontal cell (HC) hyperpolarization of 10-20 mV. Additionally, the amplitude of HC responses to short test flashes increased in the presence of melatonin. Voltage-intensity plots revealed that application of 500 microM of melatonin caused an increase of the HC light sensitivity and this effect was reversible. These results suggest that melatonin synthesized and released during the dark period of the diurnal cycle may alter the sensitivity of second-order neurons at a time of day when photopic input is at its lowest level.

Localization of mRNA encoding the indolamine synthesizing enzyme, hydroxyindole-O-methyltransferase, in chicken pineal gland and retina by in situ hybridization.

Hydroxyindole-O-methyltransferase (HIOMT) is the enzyme that catalyzes the final step in the synthesis of the hormone melatonin. We have examined the localization of expression of the mRNA encoding HIOMT by in situ hybridization in the 3-day-old and adult chicken retina and pineal gland. The riboprobe utilized for this study was transcribed from the complete coding region of HIOMT cDNA synthesized from retina RNA and amplified by polymerase chain reaction (PCR). High levels of HIOMT mRNA were present in the pinealocytes of the pineal gland. In the retina, the hybridization signal was localized to the photoreceptors. The retinal photoreceptors of the 3-day-old chick displayed a much lower level of hybridization than did the photoreceptors of the adult chicken. This study strongly suggests that the photoreceptors are the sites of melatonin synthesis in the retina.

Effect of inhalation anaesthetics on the phase behaviour, permeability and order of phosphatidylcholine bilayers.

We have used differential scanning calorimetry and fluorescence anisotropy measurements to investigate the effect of five inhalation anaesthetics of diverse chemical structure (halothane, enflurane, n-pentane, chloroform and diethylether) on the phase behaviour of liposomes prepared from dimyristoylphosphatidylcholine (DMPC) and dipalmitoylphosphatidylcholine (DPPC), respectively. The incorporation of these anaesthetics induced a decrease of the phase transition temperature and/or a broadening of the phase transition peak depending on the transverse localisation of the investigated anaesthetic. At high anaesthetic concentrations we observed the disappearance of the pretransition peak and the appearance of a shoulder on the main phase transition peak due to the domain formation of the anaesthetics. An anaesthetic induced carboxyfluorescein efflux from the vesicle lumen was completed within a few minutes after the addition of the anaesthetics, probably resulting from a transient formation of membrane holes. All results are discussed with regard to the physicochemical properties of the anaesthetics applied.

Early expression of recoverin in a unique population of neurons in the human retina.

The calcium-binding protein recoverin has been reported as present in photoreceptors, cone bipolar cells and sparse cells in the ganglion cell layer in the adult retinae of various vertebrate species. The present study was undertaken to clarify the developmental pattern of recoverin-immunoreactive cells in the human retina with particular attention to the cells in the inner retinal layers. In the adult human retina, small populations of recoverin-containing cells are present in the ganglion cell and nerve fiber layers. However, the precursors of these cells are quite numerous on the inner and outer borders of the nerve fiber layer in the fetal retina. By 13 weeks of gestation these cells express recoverin very intensely. By 24 weeks they are mature-looking with relatively large soma sizes (mean = 118 microns 2) and appear round, oval or multipolar in shape, with varying numbers of short processes. There follows a noticeable reduction of the mean soma size, but little change in morphology and process number during the remaining gestational stages up to and after birth. The mean numerical density of the recoverin-positive cells in the fetal inner retinal layers is gradually reduced from the high level at 13 weeks until birth, when there is a great drop to the adult level. The recoverin-immunoreactive cells in the ganglion cell layer demonstrate distinctively different developmental and morphological features from the principle neurons and glial cells in the retina. They are probably the neurons derived from the marginal zone of the retinal primordium that reside in the inner and outer borders of the nerve fiber layer due to the invasion of ganglion cell axons. The expression of recoverin in the neurons may be significant in maintaining an inside-out and centroperipheral gradient of calcium concentration in the premature retina, thereby playing a role in determining the polarity of the differentiating ganglion cells and the growth of their axons in a centrifugal spatiotemporal order.

On the evaluation of data from flow-dialysis experiments.

Flow dialysis can be used to measure (i) ligand binding to macromolecules and (ii) the size of transmembrane ion gradients. Generally an approximate method is used to calculate the binding or gradient parameters from the raw data. Here we present a simple but exact method and evaluate the errors that may arise when the approximate method is used to calculate the magnitude of ion gradients. In addition, equations are presented that allow for a correction for sampling from or additions to the upper compartment of a flow-dialysis vessel during the measurements. Setty and Hendler [(1982) J. Biochem. Biophys. Methods 7, 35-46] have reported artifacts in the measurement of ion-gradients caused by the addition of electron donors to the upper compartment of a flow-dialysis cell. Here we extend their observations and suggest additional methods to prevent such artifacts.

Recoverin in cultured human retinoblastoma cells: enhanced expression during morphological differentiation.

Recoverin is a calcium-binding protein expressed in retinal photoreceptors. It appears to delay the termination of the phototransduction cascade by blocking the phosphorylation of photoexcited rhodopsin. The goal of this study was to determine if recoverin mRNA and protein are expressed in cultured human Y79 retinoblastoma cells, so that this cell line could be used as a model to study the mechanism of recoverin gene expression in the retina. A cDNA encoding human recoverin was PCR cloned and used for prokaryotic expression of recoverin protein. Polyclonal antibodies raised against pure recombinant recoverin were used for western blotting and immunocytochemistry of Y79 cells grown as attachment cultures in the presence of the differentiating agents dibutyryl cyclic AMP (dbcAMP) or butyrate. Northern blot analysis was performed on mRNA extracted from Y79 cells that were also treated with the differentiating agents. In Y79 cell monolayer cultures, recoverin was immunolocalized to the cell cytoplasm, and immunoreactivity was increased dramatically by the addition of 2 mM butyrate to the culture medium. Butyrate treatment also caused an increase in the development of neurite-like cellular processes. Addition of 4 mM dbcAMP resulted in a moderate increase in both recoverin immunoreactivity and number of cellular processes. Western and northern blots of butyrate and dbcAMP-treated Y79 cell cultures demonstrated an increase in recoverin protein and RNA expression, respectively, comparable with that observed with immunocytochemistry. These data suggest that, under the influence of the differentiating agent butyrate, Y79 cells exhibit an increase in expression of the photoreceptor protein recoverin and a concomitant morphological differentiation toward a neuronal phenotype.