Long-term potentiation in the goldfish optic tectum

Field potential recordings were made in the primary retinal synaptic area, the stratum fibrosum et griseum superficiale (SFGS), of an in vitro goldfish optic tectum preparation. Stimulation of the optic nerve at frequencies of 1 and 5 Hz produced a long-term potentiation (LTP) of the synaptic response which developed gradually. No potentiation was seen with lower or higher frequencies. These results demonstrate a significant LTP with a slow time course and restricted low-frequency dependence.     

Putative cholinergic interneurons in the optic tectum of goldfish

We have demonstrated that a significant fraction of cells in layer 1 of goldfish optic tectum are immunoreactive for choline acetyltransferase. These cells constitute 4–7% of type XIV cells, which are thought to be intrinsic neurons and represent 95% of all tectal cells. The number of immunoreactive type XIV cells is unchanged following short-term enucleation. The possible integrative role of these putative cholinergic neurons is discussed in relation to sensory functions of the optic tectum.     

Electrophysiology of neural units in goldfish optic tectum

Axons of retinal ganglion cells showed responses not previously emphasized: (a) many tonic units discharged oscillations, 2--12 spikes per burst, interburst intervals 20--300 msec; (b) phasic units showed concentric or flanking ON and OFF fields, response frequency depended on balance of retinal excitation and inhibition; (c) directional sensitivity was maximal for retinal stimuli moving in naso-temporal direction; (d) in anterior tectum deep afferent layer (DAL) provides for deep electrical sink, fibers of DAL have small fields, mostly in front of fish; (e) color-opponent units are prevalent in the superficial terminal layers, color is spatially and temporally represented. Tectal cell responses were distinguished by large visual fields, spontaneity, multiple spikes and long latencies to optic nerve stimulation, failure to follow above 60 per sec, plasticity of response. Tectal neurons of three classes included (a) cells of one type in upper layers were inhibited in ongoing activity by visual input, receptive fields exceeded 100 degrees, were often oblong, responses did not habituate; (b) cells of second type were excited by visual stimuli, became unresponsive (habituated) or responsive only to stimuli in different position or direction (newness cells); lability precluded field mapping and dishabituation was produced by change in background, extraneous stimulation, and spontaneous firing; (c) pyriform cells in periventricular layer were abundant, difficult to isolate electrically, discharged spontaneously in bursts at intervals of several seconds and responded to visual input by interruption of firing. Some tectal cells responded to non-visual stimuli as well.     

Effects of lesions of the optic nerve, optic tectum and nervus terminalis on rod precursor proliferation in the goldfish retina.

Teleost retinas grow throughout life by proliferation of neuroblasts at the retinal margin and dedicated rod precursors in the outer nuclear layer. Mechanisms regulating this proliferation are largely unknown. Previous investigators observed that rod precursor replication, as detected by incorporation of radioactive thymidine into cells of the outer nuclear layer, is enhanced after optic nerve crush. We attempted to determine whether this was due to severing of the retinopetal (nervus terminalis, n.t.) or retinofugal (retinal ganglion cell) axons in the optic nerve of the goldfish, Carassius auratus. In the first series of experiments, we ablated unilaterally the optic nerve, olfactory bulb (containing n.t. ganglia), or optic tectum (containing retinal ganglion cell axons and n.t. collaterals). Rod precursor proliferation increased dramatically in both retinas as soon as 5 days after surgery; in addition, the numbers of dividing cells were greater in the ipsilateral retina 10-15 days after optic nerve crush or tectal ablation and in the contralateral retina 20-25 days after olfactory bulb ablation. These observations are not accounted for by the known projections of retinal ganglion cells, but are consistent with the projections of the n.t. In the second series of experiments, n.t. projections to the brain and retina were severed bilaterally 7-8 weeks before the unilateral optic nerve crush or hemitectal ablation. Rod precursor proliferation increased as before, but the quantities of dividing cells were always equal in both retinas. We conclude that the n.t. may modulate rod proliferation locally and that injury to (some) brain regions may cause release of mitogens that affect rod precursors in both retinas.     

Retinal terminals in the goldfish optic tectum: Identification and characterization

Retinal terminal profiles in the goldfish optic tectum were identified electron microscopically after (1) labeling with horseradish peroxidase and (2) in the early stages of degeneration in short-term eye enucleates. All labeled terminals shared certain common morphological characteristics which were identical to those of a population of terminals in normal tecta. Terminals of this type disappeared 30 days after enucleation of the contralateral eye. Retinal terminal presynaptic profiles were characterized by (1) round and oval synaptic vesicles; (2) mitochondria with irregular, randomly oriented cristae, large intracristal spaces, dilated membrane spaces, and primarily light matrices; (3) a wide range in profile area, 0.06–6.82 μm²; (4) large numbers of synaptic vesicles per profile area 168± 33 synaptic vesicles per μm²; (5) asymmetric synapses; and (6) multiple synaptic contacts (1.46 ± 0.73 per terminal profile). The postsynaptic elements included both dendritic and, less commonly, pleomorphic vesicle-containing profiles. The majority of postsynaptic dendritic profiles were small (0.01–0.40 μm²). Serial synaptic contacts were occasionally seen. The combination of vesicular and mitochondrial morphology (1 and 2 above) was necessary and sufficient to establish the retinal origin of a terminal, but use of such criteria would underestimate the number of retinotectal terminals by omitting those which did not have a mitochondrion in the plane of section. The number of such terminals was calculated from independent measurements, and the total number of retinal terminal profiles per area of neuropil was estimated.     

Glutamic acid binding in goldfish brain and denervated optic tectum

The binding of L-glutamic acid to goldfish brain membranes and changes in tectal binding following optic nerve denervation and regeneration were investigated. Saturable, reversible, and specific binding occurred to sodium-free washed membranes from goldfish brain at a single population of sites having an apparent Kd of 3.4 microM and a capacity of 10 pM/mg original tissue. Binding was enriched in crude synaptosomal (P2) subcellular fractions. There was a 10-fold regional variation in the concentration of binding sites. In pharmacological studies protection constants (Kp) (the concentration which resulted in a 50% inhibition of binding) ranged from 4 microM for glutamate to greater than 10 mM for GABA. Following eye removal, the total number of tectal glutamic acid binding sites was stable for 4 days, followed by a rapid loss in binding, reaching 40% of control at 24 days. After optic nerve crush and optic nerve regeneration, the number and concentration of binding sites was not different from control. The relationship between glutamate, nicotinic, and muscarinic receptor sites in the retinotectal pathway is discussed.     

The optic tectum regulates the transport of specific proteins in regenerating optic fibers of goldfish

The pattern of rapidly-transported proteins in regenerating optic fibers of the adult goldfish is regulated by interactions between these fibers and their main target, the optic tectum. When the optic fibers are allowed to interact with the tectum, the transport of proteins with molecular weights in the range of 110-145 kilodaltons (kDa) increases, whereas the transport of proteins in the 24-27 kDa range declines from the previously high level which has been induced by axotomy. If the optic fibers are prevented from interacting with the tectum, the transport of the 24-27 kDa proteins remains elevated for months. Amounts of other rapidly-transported retinal proteins (e.g. the acidic 43-49 kDa proteins that increase in regenerating optic fibers after axotomy) are relatively unaffected by tectal ablation.     

Fluorescence recordings of electrical activity in goldfish optic tectum in vitro

Optical methods for recording electrical activity in the goldfish optic tectum were evaluated. Tectal slices, with a short section of the optic nerve attached, were stained with a fluorescent styryl dye. Potential-dependent fluorescence changes following optic nerve stimulation were monitored with a photodiode. We found that large optical signals could be obtained. Experimental manipulations of the slice bathing solution permitted us to identify several events that contributed to the optical response, including activity in afferent fibers, excitatory and inhibitory postsynaptic potentials, and presumptive glial depolarizations. These results suggest that voltage-sensitive dyes can provide a useful alternative method for monitoring synaptic responses in the goldfish tectum, and may prove valuable in studying changes in the functional synaptic organization of the tectum following manipulations of the retinotectal pathway.     

Laminar histochemical and cytochemical localization of cytochrome oxidase in the goldfish retina and optic tectum in response to deafferentation and during regeneration

Cytochrome oxidase (C.O.) histochemistry and cytochemistry were used to examine the effects of optic denervation and subsequent optic fiber regeneration on oxidative metabolism in the retina and optic tectum of the goldfish. In the tectum, there was a dramatic and rapid decrease in C.O. activity within the optic layers 3-4 days after contralateral eye removal or optic nerve crush. At the E.M. level this was correlated with an initial decrease in mitochondrial reactivity within optic terminals followed by the subsequent degradation of mitochondria and phagocytosis of optic terminals. By 1 month after optic nerve crush, the entire tectum was reinnervated. However, the normal dark reactivity of the stratum fibrosum et griseum superficialis (SFGS), the main optic innervation layer, was not restored until after 3-4 months postcrush. The normal intense reactivity of the large-diameter optic axons and terminals at the bottom of the SFGS required an even longer period, about 7-8 months, for full recovery. The delayed restoration of C.O. reactivity was not due to a delay in synaptogenesis or in mitochondrial accumulation within optic terminals but to a delay in the maturation of mitochondrial reactivity. Following regeneration, the normal sublaminar stratification of C.O. bands was reestablished, suggesting that metabolically distinct classes of optic fibers may reinnervate at their original sublaminae. By using a distinct and persistent C.O. reactive sublamina, a of stratum griseum centrale (SGCa), just subjacent to the SFGS, it was possible to measure the thickness of the SFGS following optic denervation and subsequent reinnervation. At 1 week after optic nerve crush, the SFGS shrank by 35%. During regeneration, the thickness of the SFGS gradually increased to about 23% above normal at 2 months postcrush and this was maintained indefinitely. In the retina, ganglion cells were hypertrophic by 1 month postcrush and exhibited elevated levels of C.O. during the same period of time when optic terminals were unreactive. This indicates that oxidative metabolic activity within perikarya and axon terminals of the same neuron may be locally and independently regulated. It also suggests that in spite of the well-known elevation of axonal transport during the initial period of axon elongation and synaptogenesis, that oxidative metabolic energy production within the optic fibers is less than that of the mature projection.(ABSTRACT TRUNCATED AT 400 WORDS)     

Target selection by surgically misdirected optic fibers in the tectum of goldfish

This study tested the capacity of regenerating optic fibers to read tectal markers and thereby grow to their appropriate tectal loci when initial position, optic pathway, and interfiber interactions are eliminated as useful cues. The stability of these markers with long-term optic denervation of the tectum was also examined. In adult goldfish optic fibers innervating lateroposterior optic tectum were dissected free of tectum and inserted into the medial anterior region of the opposite "host" tectum. Normally, fibers at this position either innervate medial anterior tectum or follow the medial division of the optic pathway into medioposterior tectum. Host tectum was denervated of all other optic fibers by enucleating its contralateral eye either at the time of the deflection or at various times up to 18 months prior to deflection. The regeneration of these deflected fibers into host tectum was examined by autoradiography and electrophysiology at 1 to 11 months later. At the insertion site deflected fibers split into two groups of roughly equal size. One group directly entered the optic layers of medial tectum and grew posterolaterally across the medial half of tectum into the lateral half. The second group followed an almost direct path to the lateral tectum, sometimes traversing through the deep cell layers of tectum in which optic fibers are not usually found. These fibers subsequently entered the optic layers at the lateral edge of tectum and grew posteriorly. This second path was not seen in controls in which optic fibers from medioposterior tectum were similarly deflected. Instead growth was almost entirely posteriorly directed. On the average by 1.5 months deflected lateroposterior fibers were preferentially distributed in the lateral half of the tectum. Densitometric measurements indicated nearly a 4-fold difference in lateroposterior compared with medial posterior labeling. By contrast, controls in which medial posterior fibers were deflected had 4 times more grains medially than laterally. There was also a posterior over anterior preference, but this was weak. There was no suggestion that long periods of optic denervation prior to deflection or long postoperative periods after deflection of lateroposterior fibers diminished the lateral over medial preference. These findings support the idea that stable tectal markers exist which are differentially read by medial and lateral optic fibers. However, in no case was the innervation by deflected fibers as selective as in the normal projection.(ABSTRACT TRUNCATED AT 400 WORDS)     

Enhancement by ethanol of visually evoked responses in the goldfish optic tectum

Flash-evoked potentials and multiple unit discharge were recorded from the superficial layers of the goldfish optic tectum, and post-flash histograms were prepared. When goldfish were intoxicated by a level of ethyl alcohol just sufficient to depress maze-learning ability, the first of three components of these evoked responses was consistently enhanced. These results indicate that ethanol can unbalance neural processes near the periphery as well as central brain mechanisms. Furthermore, the data provide a basis for suggestion that the retina might be useful as a model system for further physiological and pharmacological analysis of the mechanisms by which ethanol disrupts normal brain functions.     

Histochemical localization of cytochrome oxidase in the retina and optic tectum of normal goldfish: a combined cytochrome oxidase-horseradish peroxidase study

Cytochrome oxidase (C.O.) was histochemically localized in the normal retina and optic tectum of goldfish in order to examine the laminar and cellular oxidative metabolic organization of these structures. In the optic tectum, C.O. exhibited a distinct laminar, regional, and cellular distribution. The laminae with highest C.O. levels were those that receive optic input, suggesting a dominant role for visual activity in tectal function. This was demonstrated by colocalizing C.O. and HRP-filled optic fibers in the same section. However, the distribution of C.O. within the optic laminae was not uniform. Within the main optic layers, the SFGS, four metabolically distinct sublaminae were distinguished and designated from superficial to deep as sublaminae a, b, c, and d. The most intense reactivity was localized within SFGSa and SFGSd, followed by SFGSb, then SFGSc. In SFGSd, intense reactivity was found to occur specifically within a class of large diameter axons and terminals that were apparently optic since these were also labeled with HRP and cobaltous lysine applied to the optic nerve. Regional C.O. differences across the tectum were also noted. Low levels were found in neurons and optic terminals along the growing immature medial, lateral, and posterior edges of tectum, but were higher at the more mature anterior pole and central regions of tectum. This suggests that the oxidative metabolic activity is initially low in newly formed tectal neurons and optic axons, but gradually increases with neuronal growth and functional axon terminal maturation. Most C.O. staining was localized within neuropil, whereas the perikarya of most tectal neurons were only lightly reactive. Only a few neuron classes, mostly the relatively larger projection neurons, had darkly reactive perikarya. In the retina, intense C.O. reactivity was localized within the inner segments of photoreceptors, the inner and outer plexiform layers, and within certain classes of bipolar and ganglion cells. The large ganglion cells in particular were intensely reactive. Like the large diameter optic terminals in SFGSd, the large ganglion cells were preferentially filled with HRP, suggesting that they may project to tectum and are the source of the darkly reactive large diameter axons and terminals in sublamina SFGSd. We propose a new scheme to describe tectal lamination that integrates laminar differences in C.O. reactivity with classical histological work.(ABSTRACT TRUNCATED AT 400 WORDS)     

Locally correlated activity in the goldfish tectum in the absence of optic innervation

Despite a substantial literature on the role of correlated presynaptic activity in the patterning of nerve connections during synaptogenesis in the central nervous system, few studies have focuses on postsynaptic activity during this process. To address the possibility that the target exhibits correlated activity independently of presynaptic activity, extracellular activity was recorded from the main optic innervation layer stratum fibrosum et griseum superficiale (SFGS) in goldfish in which the optic nerve was crushed or the eye removed. At about 2 weeks after denervation, multiunit recordings revealed phasic temporally correlated discharge between different tectal units. Auto-correlation analysis of these trains showed a broad peak 75-100 ms wide confirming temporal correlation. Using cross-correlation analysis of two simultaneous recordings at different distances across tectum, this correlation was shown to be local. Strong positive correlations were detected over about 200 micron and decrease with greater distances disappearing beyond about 400 micron. These correlograms showed a broad symmetrical peak about 75-100 ms wide. This pattern of activity persisted from the day following nerve crush into the period of activity dependent reinnervation at 1 month. When the eye was removed, the pattern could be demonstrated for up to 3 months of denervation indicating the circuitry responsible for the correlated activity was quite stable in the absence of optic innervation. We conclude that tectal elements are capable of locally correlated discharge independently of optic innervation. We propose that locally correlated discharge represents cooperative groups of tectal cells and that these groups, rather than single cells, are the target of the activity dependent synaptic rearrangement such as ocular dominance columns which occurs during synaptogenesis.     

A quantitative study of the reinnervation of the goldfish optic tectum following optic nerve crush

Neurons of the raccoon main (MCN) and external (ECN) cuneate nuclei having terminal fields in the anterior lobe and paramedian lobule of the cerebellum were studied by means of the retrograde transport of horseradish peroxidase (HRP). In the MCN, neurons of the polymorphic regions, but not of the round cell clusters, were labeled following HRP injections of the anterior lobe. HRP-labeled fusiform, triangular, and stellate cells were observed from 3–4 mm caudal to the obex to 2–3 mm rostral to it. The main body of labeled MCN neurons was located in the 1.25 mm immediately rostral to the obex, there forming a recognizable subnucleus in the ventral MCN. The ECN displayed uniform labeling from the anterior lobe with no particular cell type or nuclear region being labeled preferentially. A minor projection to the anterior lobe appeared to originate in the contralateral MCN and ECN. Injections of the paramedian lobule produced MCN labeling similar to that observed after anterior lobe injections, but the majority of labeled ECN neurons were found in the dorsomedial part of the nucleus. In lobule V of the anterior lobe, sagittal zone c contained the greatest density of cuneocerebellar terminals as judged by the degree of retrograde labeling seen in neurons of both the MCN and ECN. This observation was consistent with evoked potentials recorded in the anterior lobe of the raccoon cerebellum. The position of cuneocerebellar neurons in the ventrolateral MCN suggests that the bulk of the information being transferred to the cerebellumby the raccoon MCN is related to deep, rather than cutaneous, modalities.     

Telencephalic terminals in the major retinal synaptic lamina of the goldfish optic tectum

Light and electron microscopic degeneration studies were used to examine the telencephalotectal pathway in goldfish. Both techniques showed that each telencephalic lobe sent bilateral projections to several tectal laminae. Degenerating synaptic terminals and fibers were observed in the major retinal projection lamina as well as in other tectal laminae. The terminals contained round to oval synaptic vesicles, asymmetric synapses and contacted relatively small postsynaptic profiles.     

Multiple nicotinic acetylcholine receptor genes are expressed in goldfish retina and tectum

cDNAs encoding a novel nAChR structural subunit (GFn alpha-3) and a ligand-binding subunit (GF alpha-3) have been isolated from a goldfish retina cDNA library. The protein encoded by GFn alpha-3 shares 88% amino acid similarity with that encoded by GFn alpha-2, a structural subunit gene previously identified to be expressed in this system (Cauley et al., 1989). The ligand-binding subunit (GF alpha-3) is likely the goldfish homolog of the rat alpha-3 gene (Boulter et al., 1986). Northern blots and S1 protection experiments show that GFn alpha-3 and GF alpha-3 genes are expressed in retina and brain. GFn alpha-3 identifies multiple RNAs differing in their 3' untranslated regions. In situ hybridization analysis demonstrates GFn alpha-3, GFn alpha-2, and GF alpha-3 expression by cells of the retinal ganglion cell layer. Unlike GFn alpha-2 and GF alpha-3, GFn alpha-3 is expressed at highest levels by cells of the retina's inner nuclear layer. In the optic tectum, both GF alpha-3 and GFn alpha-3 genes are expressed by cells of the periventricular zone, as well as more superficial layers. These results suggest the presence of multiple nAChR systems in retina and tectum. In addition, they indicate that tectal nAChRs may arise from remote (ganglion cell) as well as local (tectal cell) synthesis.     

Monocular visual discrimination in the goldfish after unilateral ablation of the optic tectum

Goldfish with unilateral ablation of the optic tectum and trained to the ablated side failed to learn or to transfer interocularly a differential classically conditioned color discrimination. The direct ipsilateral retinal projection to the intact/naive side is therefore shown to contribute little to engram formation. Evidence is also presented demonstrating that the absence of one tectum does not affect the learning of a differential bar discrimination task when trained via the intact brain side.     

Reorganization of retinotectal projection following surgical operations on the optic tectum in goldfish

The pattern of neural reconnection between the retina and surgically operated tectum was studied in juvenile goldfish (8–11 cm long) with electrophysiological methods. The results confirm that the remaining rostral half-tectum reacquires a complete visual projection from the whole retina about 90 days after excision of the caudal half. The same reorganization of visual projection from the whole retina onto the rostral half-tectum was found to occur in the presence of the caudal half of the tectum, if the two halves were separated by a transverse surgical incision down to the level of the optic ventricle regardless of whether the contralateral optic nerve was left intact or crushed to regenerate. The reorganization of retinotectal projection was also found to occur biaxially along the mediolateral as well as the rostrocaudal axis of the tectum following excision of a caudomedial sector of the tectum. It is suggested that the reorganization of retinotectal projection is due to synaptic respecifications of individual tectal neurons in correct retinotopic order, and that the cellular discontinuity between the rostral and the caudal parts of the tectum is sufficient to induce the orderly synaptic respecifications in juvenile goldfish.     

Anomalous retinal projection after removal of contralateral optic tectum in adult goldfish

Retinotectal projections were mapped in a series of adult goldfish at various intervals after complete ablation of one tectum, or after enucleation of one eye and removal of its ipsilateral tectum. In the first case, regenerating optic axons entered the ipsilateral tectum and innervated it retinotopically; the normal and ipsilateral projections overlapped each other. In the second case, the remaining eye projected over the remaining ipsilateral tectum in normal retinotopic order. Thus, the presence of already innervated optic tectum does not deter regenerating axons from innervating it. It is suggested that the visual projection of goldfish tends to retain its completeness in the absence of normal terminal spaces.