Horizontal Cell Dendrites Research Articles

Heterogeneous distribution of AMPA glutamate receptor subunits at the photoreceptor synapses of rodent retina

In the retina the segregation of different aspects of visual information starts at the first synapse in signal transfer from the photoreceptors to the second-order neurons, via the neurotransmitter glutamate. We examined the distribution of the four AMPA glutamate receptor subunits GluR1-GluR4 at the photoreceptor synapses in mouse and rat retinae by light and immunoelectron microscopy and serial section reconstructions. On the dendrites of OFF-cone bipolar cells, which make flat, noninvaginating contacts postsynaptic at cone synaptic terminals, the subunits GluR1 and GluR2 were predominantly found. Horizontal cell processes postsynaptic at both rod and cone synaptic terminals preferentially expressed the subunits GluR2, GluR2/3 and GluR4. An intriguing finding was the presence of GluR2/3 and GluR4 subunits on dendrites of putative rod bipolar cells, which are thought to signal through the sign-inverting metabotropic glutamate receptor 6, mGluR6. Furthermore, at the rod terminals, horizontal cell processes and rod bipolar cell dendrites showed labelling for the AMPA receptor subunits at the ribbon synaptic site or perisynaptically at their site of invagination into the rod terminal. The wide distribution of AMPA receptor subunits at the photoreceptor synapses suggests that AMPA receptors play an important role in visual signal transfer from the photoreceptors to their postsynaptic partners.

The cone pedicle, a complex synapse in the retina.

Cone pedicles, the synaptic terminals of cone photoreceptors, are connected in the macaque monkey retina to several hundred postsynaptic dendrites. Using light and electron microscopy, we found underneath each cone pedicle a laminated distribution of dendritic processes of bipolar and horizontal cells. Superimposed were three strata of glutamate receptor (GluR) aggregates, including a novel layer of glutamate receptors clustered at desmosome-like junctions. They are, most likely, postsynaptic densities on horizontal cell dendrites. GABA(A) and GABA(C) receptors are aggregated on bipolar cell dendrites in a narrow band underneath the cone pedicle. Glutamate released from cone pedicles and GABA released from horizontal cell dendrites act not only through direct synaptic contacts but also (more so) through diffusion to the appropriate receptors.

Photoreceptor classes and transmission at the photoreceptor synapse in the retina of the clawed frog,Xenopus laevis

The photoreceptor population in Xenopus consists of a green-sensitive rod (lambda(max) = 523 nm), a blue-sensitive rod (lambda(max) = 445 nm) and three classes of cone. The largest cone is red-sensitive (lambda(max) = 611 nm). The intermediate cone is presumed to be blue-sensitive based on physiological criteria, whereas the miniature cone may be UV-sensitive. Horizontal cells (HC) are of two sorts: axon-bearing and axonless. The axon-bearing HC is of the luminosity type and probably contacts all types of photoreceptor. The axonless HC is of the chromaticity type and contacts only intermediate (blue) cones and at least one type of rod. During development dendrites of HCs and bipolar neurons penetrate photoreceptor bases. A progressive maturation of HC and bipolar synapses with rods and cones occurs between tadpoles stages 37/8 and 46. Neighboring rods and cones are joined by gap junctions. During this same period, the outer segments are laid down and photopigments synthesized. A linear relation was found between the quantum capturing ability of the rod and its absolute threshold. Mature rods of the Xenopus retina release glutamate in a calcium-dependent manner. Glutamate release was found to be a linear function of calcium influx through L-type calcium channels. Both types of HC possess ionotropic glutamate receptors of the AMPA subtype.

Expression, distribution and ultrastructural localization of the synapse-organizing molecule agrin in the mature avian retina.

At the vertebrate neuromuscular junction the extracellular matrix molecule agrin is responsible for the formation, maintenance and regeneration of most if not all postsynaptic specializations. Several agrin isoforms are generated by alternative splicing which differ in their function and which are all expressed in the CNS. To analyse the role of agrin in the CNS, we investigated the expression and ultrastructural localization of agrin in the posthatched chick retina. In situ hybridization revealed the presence of agrin mRNA in all cellular layers of the mature retina, indicating that most if not all major retinal cell types synthesize agrin. Pan-specific as well as isoform-specific antiagrin antisera stained the optic fibre layer and the outer plexiform layer. However, only the pan-specific antiserum additionally stained the inner limiting membrane. Immunoelectron microscopy showed that in the optic fibre layer agrin was associated with ganglion cell axons and that at least part of this agrin corresponds to a neuronal isoform of agrin. In the outer plexiform layer, agrin was localized in the cleft between the photoreceptor terminals and the invaginating horizontal and bipolar cell dendrites. In the synapse-containing inner plexiform layer both antisera revealed punctate immunoreactivity. This staining corresponded to agrin concentrated in the synaptic cleft of conventional synapses as determined by preembedding immunoelectron microscopy. Agrin is thus concentrated at mature interneuronal synapses as it is at the neuromuscular junction, consistent with a role of agrin during formation and/or maintenance of synapses in the CNS.

AMPA-preferring receptors with high Ca2+ permeability mediate dendritic plasticity of retinal horizontal cells.

The synaptic complex formed by the cone photoreceptor pedicles and the dendrites of horizontal cells in the teleost retina undergoes structural changes during light adaptation. Numerous spinules are formed by the terminal dendrites, and they are subsequently retracted during dark adaptation. In a retina kept under continuous illumination, the retraction process can be initiated by analogues of the neurotransmitter glutamate acting at AMPA/kainate receptors. On the other hand, the retraction process depends on calcium influx and the subsequent activation of CaMkII. We show here that the retraction of spinules induced by AMPA or kainate is not impaired in the presence of cobalt, making an involvement of voltage-gated calcium channels unlikely. Using calcium imaging techniques with isolated horizontal cells, we demonstrate that AMPA and kainate, but not NMDA, increase [Ca2+]i in the presence of nicardipine, caffeine and thapsigargin. The increase of [Ca2+]i under these conditions depends on [Ca2+]o and on the agonist in a dose-dependent manner, suggesting that the increase of [Ca2+]i is largely due to calcium influx through the agonist-gated channel. Pharmacological studies were performed to determine whether AMPA- and/or kainate-preferring receptors mediate the calcium influx. The AMPA-preferring receptor antagonist LY303070 blocked glutamate- and kainate-evoked increases of [Ca2+]i in a concentration-dependent manner, indicating that kainate-preferring receptors contributed little or nothing to the observed [Ca2+]i increase. This was supported by experiments where cyclothiazide (which blocks the desensitization of AMPA receptors) and concanavalin A (which potentiates responses mediated by kainate receptors) were applied. In all cases, LY303070 blocked the agonist-evoked increase of [Ca2+]i. The presence of AMPA-preferring receptors with high Ca2+ permeability on horizontal cells was also supported by measuring agonist-induced currents using whole-cell recording techniques. Furthermore, LY303070 was able to impair the retraction of spinules during dark adaption in the in vivo situation.

Layer VI in cat primary auditory cortex: Golgi study and sublaminar origins of projection neurons.

The organization of layer VI in cat primary auditory cortex (AI) was studied in mature specimens. Golgi-impregnated neurons were classified on the basis of their dendritic and somatic form. Ipsilateral and contralateral projection neurons and the corticogeniculate cells of origin were labeled with retrograde tracers and their profiles were compared with the results from Golgi studies. Layer VI was divided into a superficial half (layer VIa) with many pyramidal neurons and a deeper part (layer VIb) that is dominated by horizontal cells. Nine types of neuron were identified; four classes had subvarieties. Classical pyramidal cells and star, fusiform, tangential, and inverted pyramidal cells occur. Nonpyramidal neurons were Martinotti, multipolar stellate, bipolar, and horizontal cells. This variety of neurons distinguished layer VI from other AI layers. Pyramidal neuron dendrites contributed to the vertical, modular organization in AI, although their apical processes did not project beyond layer IV. Their axons had vertical, intrinsic processes as well as corticofugal branches. Horizontal cell dendrites extended laterally up to 700 microm and could integrate thalamic input across wide expanses of the tonotopic domain. Connectional experiments confirmed the sublaminar arrangement seen in Nissl material. Commissural cells were concentrated in layer VIa, whereas corticocortical neurons were more numerous in layer VIb. Corticothalamic cells were distributed more equally. The cytological complexity and diverse connections of layer VI may relate to a possible role in cortical development. Layer VI contained most of the neuronal types found in other layers in AI, and these cells form many of the same intrinsic and corticofugal connections that neurons in other layers will assume in adulthood. Layer VI, thus, may play a fundamental ontogenetic role in the construction and early function of the cortex.

Retinoic acid has light-adaptive effects on horizontal cells in the retina.

Ambient light conditions affect the morphology of synaptic elements within the cone pedicle and modulate the spatial properties of the horizontal cell receptive field. We describe here that the effects of retinoic acid on these properties are similar to those of light adaptation. Intraorbital injection of retinoic acid into eyes of dark-adapted carp that subsequently were kept in complete darkness results in the formation of numerous spinules at the terminal dendrites of horizontal cells, a typical feature of light-adapted retinae. The formation of these spinules during light adaptation is impaired in the presence of citral, a competitive inhibitor of the dehydrogenase responsible for the generation of retinoic acid in vivo. Intracellularly recorded responses of horizontal cells from dark-adapted eyecup preparations superfused with retinoic acid reveal typical light-adapted spatial properties. Retinoic acid thus appears to act as a light-signaling modulator. Its activity appears not to be at the transcriptional level because its action was not blocked by actinomycin.

Doparninergic control of light-adaptive synaptic plasticity and role in goldfish visual behavior

Dopamine has been implicated in processes of retinal light and dark adaptation. In goldfish retina, horizontal cell dendrites elaborate neurite processes (spinules) into cone terminals, in a light- and dopamine-dependent manner. However, the functions of retinal dopamine and the horizontal cell spinules in visual behavior are unknown. These issues were addressed in behavioral, electroretinographic, and anatomical studies of normal fish and those with unilateral depletion of retinal dopamine induced by intraocular (i.o.) injections with 6-hydroxydopamine (6-OHDA). Dopamine interplexiform cells (DA-IPC) disappear within 2 weeks after 6-OHDA injection; cell bodies appear at the marginal zone within 6 weeks at which time neurites slowly reinnervate the retina with a sparse plexus over the next 12 months. We found that dopamine depletion increased light sensitivity at photopic but not scotopic backgrounds by 2.5 log units, an effect mimicked by i.o. injections of dopamine DI and D2 antagonists. The ERG b-wave increment thresholds were the same for control and dopamine depleted eyes, indicating a normal transition from rod to cone systems in the ON pathway. Light-dependent spinule formation was reduced by about 60% in dopamine-depleted retinas, but returned to normal by 3 months and 9 months after injection in the entire retina, even areas not directly innervated with DA-PC processes. Spinule formation in vivo was inhibited 50% with i.o. injection of SCH 23390 in control retinas as well as throughout 3 month 6-OHDA injected retinas, including DA-IPC free areas. This latter result indicates a volume effect of dopamine, diffusing laterally through the retina over several millimeters, in regulating spinules. We conclude that DA-IPCs regulate sensitivity to background at photopic levels not via the ON pathway, but perhaps the OFF pathway. Goldfish display both increased sensitivity to light and a normal Purkinje shift in the ERG b-wave whether or not horizontal cell spinules are present, indicating that dopamine control of photopic vision in fish is not mediated through light-induced spinule formation of horizontal cell dendrites.

Ca 2+-dependency of spinule plasticity at dendrites of retinal horizontal cells and its possible implication for the functional role of spinules

Calcium is involved in many aspects of synaptic plasticity and we have analyzed its involvement in spinule dynamics at retinal horizontal cell dendrites. We show here that in particular the retraction of spinules is a Ca 2+-dependent process. Inhibiting calmodulin or CaMKII, blocked the retraction that was also impaired in low calcium Ringer. Changes of the cytosolic Ca 2+-concentration through depletion of internal Ca 2+-stores were without effect. This suggested that Ca 2+-influx during dark adaption and subsequent activation of CaMKII is an important step for spinule retraction. Voltage dependent Ca 2+-channels were not responsible for the Ca 2+-influx, rather Ca 2+ leaking through α-amino-3-hydroi:y-5-methyl-4-isoxazolepropionic acid (AMPA)/kainate-gated channels. This suggested a close :local link between AMPA/kainate receptors and CaMKII indicating a possible postsynaptic function of spinules. The distribution of bound, omega-shaped vesicles within the cone pedicles and its dependence on artificial depolarization further supported the idea of a postsynaptic function of spinules.

Clearance of neurotransmitter from the cone synaptic cleft in goldfish retina

Dendrites of bipolar and horizontal cells protrude deeply into the synaptic terminals of cones in goldfish retina. This arrangement gives the impression that the cone synaptic terminal surrounds a morphologically shielded compartment, the cone synaptic cleft, from which clearance of neurotransmitter by diffusion is limited. In this study the time constant of this clearance has been approached in two ways: (1) the morphological parameters determining the clearance (extracellular synaptic volume and leak area), were estimated using morphometric methods. These data were introduced into a diffusion model of the cone pedicle, yielding a time constant for the clearance of <1 msec; (2) the time constant of the light onset response when the glutamate transporter in the cone was blocked with dl-threo-β-hydroxyaspartate or dihydrokainate, was interpreted as the time constant of the clearance, yielding values of almost 650 msec compared to around 90 msec in control conditions. The decay time-constant of the Ca-dependent tail-currents in cones was used, as an approximation of the dynamics of the intracellular Ca-concentration and thus of the glutamate release by the cones. The decay time constant was about 800 msec. This suggests that the intracellular Ca-concentration in the synaptic terminal and hence the glutamate release by the cones drops, with a similar large time constant. These results indicate that the cone pedicle in goldfish does not limit the clearance of neurotransmitters from the synaptic cleft and that the fast light onset response of horizontal cells under control conditions is due to activation of glutamate transporters by hyperpolarization of the cone membrane potential while the glutamate release drops slowly. The slow horizontal cell light onset response in β-hydroxyaspartate or dihydrokainate may be due to a slow reduction of the glutamate release by the cones at light onset.

Retraction of spinule-type neurites from carp retinal horizontal cell dendrites during dark adaptation involves the activation of Ca2+/calmodulin-dependent protein kinase II.

The formation of spinules at the terminal dendrites of retinal horizontal cells with the onset of light and their subsequent retraction during darkness is a remarkable example of synaptic plasticity where sensory experience modifies reversibly, and on a time scale of minutes the ultrastructure of synaptic connectivity. The signals and the subsequent intracellular cascades underlying the prominent morphological alterations are only partially understood. We show here that lowering the external calcium concentration did prevent dark- and AMPA-induced retraction of spinules in a eyecup preparation. Furthermore, spinule retraction was prevented in vivo by the injection of calmidazolium, an inhibitor of calmodulin, into the eyeball, and also by the injection of KN-62, an inhibitor of Ca2+/calmodulin-dependent protein kinase (CaMkII). We conclude that local Ca2+ influx through AMPA-gated channels followed by activation of CaMkII is an important step for spinule retraction during dark adaptation. The phosphorylation patterns of phosphoproteins derived from purified horizontal cells was affected by the inhibitors of calmodulin and CaMkII respectively. Some of the affected phosphoproteins appeared to be cytoskeleton-associated proteins, including GAP-43. Based on these observations, a putative scenario for the retraction of spinules is proposed.

Volume transmission of dopamine may modulate light-adaptive plasticity of horizontal cell dendrites in the recovery phase following dopamine depletion in goldfish retina.

We investigated the recovery of light-adaptive spinule formation following dopamine depletion with intraocular injection of 6-hydroxydopamine (6-OHDA) and subsequent neogeneration of dopamine interplexiform cells (DA-IPC) at the marginal zone. DA-IPCs were gone by 2 weeks postinjection and appeared at the marginal zone by 6 weeks postinjection, at which time DA-IPC neurites grew toward the central retina, reaching within 0.5 mm of the central retina by 1 year. Retinas from day time, light-adapted fish at 2 weeks, 4 weeks, 3 months, and 1 year postinjection with 6-OHDA were processed for pre-embedding tyrosine hydroxylase immunoreactivity (TOH-IR) and compared to sham-injected and control retinas at the electron-microscopical (EM) level. Only 6-OHDA fish that tilted markedly toward the injected eye were used for these experiments. The tilt mimics the dorsal light reaction, indicating a 2-2.5 log unit increase in the photopic sensitivity of the 6-OHDA eye. Spinule formation was reduced by about 60% in the 2- and 4-week 6-OHDA retinas, but returned to control levels throughout the entire retina of 3-month and 1 year 6-OHDA retinas even though the central region of these retinas contained no detectable TOH-IR. Intraocular injection with 10 microM SCH 23390 (a D1 antagonist) reduced light-adaptive spinule formation by 50% both in control eyes as well as those eyes that were 3 months post 6-OHDA injected. The full return of spinule formation with only partial reinnervation of the retina with DA-IPC processes and their subsequent inhibition by SCH 23390 indicates that dopamine diffused large distances within the retina to regulate this synaptic plasticity (i.e. displayed volume transmission). Also, since all 6-OHDA injected fish displayed an increased photopic sensitivity in the injected eye when sacrificed, we suggest that horizontal cell spinules are not required for photopic luminosity coding in the outer retina.

Horizontal cells function normally in ethambutol-treated goldfish

Ethambutol, a tuberculostatic drug, induces red-green colour vision defects in man and goldfish. The ethambutol-induced red-green colour vision deffect in goldfish was argued to originate in the retina because after ethambutol application: (1) inhibitive interactions in red-green (double) opponent ganglion cells are lost [Van Dijk & Spekreijse, 1982 ( Investigative Ophthalmology and Visual Science, 24, 128–133); Wietsma & Spekreijse, 1992 ( Investigative Ophthalmology and Visual Science Suppl., 33, 1032)] and (2) the depolarizing responses to red light in the biphasic horizontal cells are reduced. To account for these findings Spekreijse, Wietsma and Neumeyer [(1991) Vision Research, 31, 551–562] suggested that ethambutol induced dark adaptation in the retina. In this paper the dark adaptation hypothesis is tested with the following results: (1) ethambutol changes only transiently the receptive field size and spectral sensitivity of horizontal cells; (2) the spectral characteristics of horizontal cells do not change in long-term ethambutol-treated goldfish; (3) formation of spinules on horizontal cell dendrites in cone terminals, a parameter for light adaptation, remains unaffected. Therefore we conclude that ethambutol does not induce functional dark adaptation of horizontal cells and that the ethambutol-induced red-green colour vision deficiency does not originate in the horizontal cell layers

Evidence for calcium/calmodulin dependence of spinule retraction in retinal horizontal cells.

Horizontal cells of the carp retina alter their synaptic connections with cones during dark and light adaptation. At light onset, dendrites of horizontal cells, which are positioned laterally at the ribbon synapse, form "spinules," little processes with membrane densities. Spinules are retracted again during dark adaptation. Spinule retraction is also elicited upon glutamate application to the retina. In the present study, we address the question whether calcium/calmodulin-dependent pathways are involved in dark- and glutamate-evoked spinule retraction. Light-adapted retinas were isolated and subsequently dark adapted during incubation in media of different calcium concentrations. Spinule retraction was clearly blocked in low-calcium solutions (5 microM and 50 nM CaCl2). Incubation in medium containing cobalt chloride (2 mM) had the same effect. Both treatments blocked the glutamate-induced spinule retraction as well. These results indicate that spinule retraction is induced by a calcium influx into horizontal cells. To investigate whether calmodulin, the primary calcium receptor in eukaryotic cells, is present at the site of spinule formation, light- and dark-adapted retinas, embedded in LR White resin, were labelled with an antibody against calmodulin and gold-conjugated secondary antibodies. Horizontal cell dendrites at the ribbon synapse revealed strong calmodulin immunoreactivity, which was more than twice as high in light- as in dark-adapted retinas. The incubation of isolated retinas with the calmodulin antagonists W5 and W13 inhibited spinule retraction. In summary, these results suggest that spinule retraction may be regulated by calcium influx into horizontal cells and subsequent calcium/calmodulin-dependent pathways.

Gonadotropin-releasing hormone, a neuropeptide of efferent projections to the teleost retina induces light-adaptive spinule formation on horizontal cell dendrites in dark-adapted preparations kept in vitro

The teleost retina recieves efferent projections from neurons of the nucleus olfactoretinalis at the base of the olfactory bulbs. These fibres contain gonadotropin-releasing hormone (GnRH) immunoreactive material and are presynaptic to retinal dopaminergic interplexiform cells. We have incubated isolated dark-adapted retinae and eyecup preparations of roach with salmon-GnRH and found an increase in horizontal cell spinule numbers to 70% light-adaptive levels. This effect was blocked by addition of haloperidol to the incubation medium suggesting that GnRH acts via stimulation of the dopaminergic interplexiform cells. We conclude that GnRH containing efferent fibres are capable of inducing light-adaptive changes in the retina and discuss their implication in the control of endogenous rhythms.

Spinule formulation in the fish retina: is there an involvement of actin and tubulin? An electronmicroscopic immunogold study.

During light-adaptation dendrites of teleost horizontal cells form finger-like processes, called 'spinules', which are characterized by synaptic membrane densities. To investigate the involvement of cytoskeletal elements in the formation and retraction of spinules, effects of the microtubule and actin inhibitors colchicine and cytochalasin D were examined by injection into the vitreous. Both substances inhibited the light-induced spinule formation. The ultrastructural immunolocalization of tubulin revealed labelling of dendrites only in their proximal parts. The distal parts of dendrites which invaginate into cone pedicles were free of label. Treatment with anti-actin revealed immunoreactivity along the entire length of dendrites up to the dendritic terminals. The spinules, however, showed no labelling. This finding does not support the hypothesis that spinules are protruded by actin polymerization. After cytochalasin D treatment the density of label in the dendritic terminals was enhanced by a factor of three, which suggests an accumulation of actin. Thus, spinule inhibition by cytochalasin D is probably caused by distortion of a functional actin network in the dendritic terminals.

Spinule-type neurite outgrowth from horizontal cells during light adaptation in the carp retina: an actin-dependentprocess

Dendrites of horizontal cells in the carp retina which invaginate the cone pedicles form numerous spinules during light adaptation. We have analyzed the contribution of cytoskeletal elements to this process. Isolated horizontal cells and frozen sections were screened with phalloidin for the existence of F-actin. F-actin was present in all types of horizontal cells and particularly enriched in the distal parts of the dendrites. Electron microscopical analysis demonstrated that interruption of the F-actin polymerization with cytochalasin B inhibited the formation of spinules during light adaptation. The persistence of spinules was also affected. Cytochalasin B also prevented the light-independent, phorbol ester-induced formation of spinules. Cytochalasin B only affected the morphology of the lateral, spinule-forming dendrites of cone horizontal cells within the cone pedicles, leaving the central, non spinule-forming dendrites of cone horizontal cells and the processes of rod horizontal cells within rod spherules unaffected. Whereas cytochalasin B prevented the protrusion of spinules, the spinule-associated membrane densities were only slightly affected. The two main characteristics of spinules, protrusion and membrane densities are therefore independently regulated processes.

Ultrastructural and functional connectivity of intracellularly stained neurones in the vertebrate retina: Correlative analyses

A variety of intracellular recording and staining techniques has been used to establish structure-function and, in some cases, structure-function-neurochemical correlations in fish, turtle, and cat retinae. Cone photoreceptor-horizontal cell connectivity has been studied extensively in the cyprinid fish retina by intracellular staining with horseradish peroxidase (HRP) and subsequent electron microscopy. The available data suggest that horizontal cell dendrites around the ridge of the synaptic ribbon are postsynaptic, whilst finger-like extensions ("spinules") of lateral dendrites function as inhibitory feedback terminals. An interesting feature of this interaction is its plasticity: the feedback pathway is suppressed in the dark and becomes potentiated by light adaptation of the retina. Intracellular recordings and stainings of ganglion cells in both turtle and cat retinae have been possible. Prelabelling of ganglion cells by retrograde transport of rhodamine from the tectum allows ganglion cells to be stained under visual control, and their synaptic inputs determined by electron microscopy. Such studies have been extended to double labelling by using autoradiography or postembedding immunohistochemistry to identify the neurotransmitter content of the labelled cell and/or the neurotransmitter(s) converging upon it. It is envisaged that further applications of intracellular staining followed by double- or even triple-labelling will continue to enhance greatly our understanding of the functional architecture of the vertebrate retina.

CAMP-mediated second messenger mechanisms are involved in spinule formation in teleost cone horizontal cells

A number of light adaptive changes of teleost horizontal cells are mediated by dopamine D 1 receptors coupled positively with the cAMP second messenger system. Spinules, finger-like extensions from horizontal cell dendrites directed towards the cone pedicle cytoplasm, are formed in response to a stimulation of D 1 receptors. We studied the second messenger mechanism associated with this process using isolated dark-adapted cyprinid retinae. Increasing intracellular cAMP concentrations by adding a membrane permeable analogue, or by stimulating the adenylate cyclase and simultaneously blocking the degradation of cAMP, resulted in a significant increase of spinule numbers in spite of the absence of light. In contradistinction to using isolated retinae for pharmacological experiments, injection of drugs into the vitreous had inconsistent or negative results.

The circadian component of spinule dynamics in teleost retinal horizontal cells is dependent on the dopaminergic system.

During the light phase of a light/dark cycle, dendrites of teleost cone horizontal cells display numerous finger-like projections, called spinules, which are formed at dawn and degraded at dusk, and are thought to be involved in chromatic feedback processes. We have studied the oscillations of these spinules during a normal light/dark cycle and during 48 h of constant darkness in two groups of strongly rhythmic, diurnal fish, Aequidens pulcher. In one group the retinal dopaminergic system had been destroyed by the application of 6-OHDA, while in the other (control) group, the dopaminergic system was intact. In control fish, oscillations of spinule numbers were observed under both normal and constant dark conditions, indicating the presence of a robust circadian rhythm. However, spinule dynamics were severely affected by the absence of retinal dopamine. During the normal light phase, the number of spinules in 6-OHDA injected retinae was strongly reduced, and throughout continual darkness, spinule formation was almost completely suppressed. These results indicate that dopamine is essential for both light-evoked and circadian spinule formation; furthermore, we conclude that there is no circadian oscillator within horizontal cells controlling the formation of spinules.