Relationship of retinotopic ordering of axons in the optic pathway to the formation of visual maps in central targets.

We examined in rats the relationship between the ordering of retinal axons in the optic pathway and the formation of a retinotopically organized projection to their primary target, the contralateral superior colliculus (SC). We have previously found that axons labeled by focal injections of 1,1'-dioctadecyl 3,3,3',3'-tetramethylindocarbocyanine perchlorate (DiI) made in temporal or nasal retina of perinatal rats commonly mistarget along the medial-lateral and rostral-caudal axes of the SC. By postnatal day (P) 11-12, the retinocollicular projection attains an adult-like topography. Incorrectly targeted axons or axon segments are removed; axons that persist terminate in the topographically appropriate part of the SC (Simon and O'Leary: Dev Biol 137:125, 1990). In the present study, we made similar DiI injections, covering less than 2% of the retinal area, in peripheral temporal, nasal, superior, or inferior retina, in rats of two age groups, embryonic day (E) 21 to P (postnatal day) 2 and P11-P17. Whole mounts of retina, optic nerve and tract, and SC, and cross sections of the optic nerve, were examined. In E21-P2 rats, retinal axons labeled from each retinal site are diffusely distributed in the SC, and poorly ordered in the optic pathway. In retina, labeled axons travel in fascicles directly from the injection site to the optic disc, but neighbor relationships begin to degrade as fascicles split and mix. Retinotopic order is virtually lost in the optic nerve; axons labeled from each injection site disperse throughout its cross-sectional area, but the labeled axons tend to be concentrated toward a specific half of the nerve depending upon their retinal origin. This slight tendency toward retinotopic order increases in the optic tract, but axons are still poorly ordered as they leave the tract and enter the SC. Targeting errors along the medial-lateral axis of the SC, but apparently not along its rostral-caudal axis, are related to the positioning of axons across the width of the optic tract. In P11-P17 rats, axons labeled from each injection site arborize only in a small, topographically correct part of the SC. However, the distributions of labeled retinal axons observed in whole mounts of the retina and optic pathway have a degree of disorder similar to those in E21-P2 rats. Further, the scatter of labeled axons in optic nerve cross sections is comparable in both age groups. Therefore, the emergence of topographic order in the retinocollicular projection is not accompanied by an emergence of a retinotopic ordering of axons in the optic nerve.(ABSTRACT TRUNCATED AT 400 WORDS)     

Development of topographic order in the mammalian retinocollicular projection.

We have used the anterograde axon tracer 1,1'-dioctodecyl-3,3,3',3'-tetramethylindocarbocyanine perchlorate (Dil) to characterize the development of topographic order in the rat retinocollicular projection. Retinal axons were labeled by Dil injections covering 0.15-2% of peripheral temporal, nasal, superior, or inferior retina, or more central retina, in rats ranging in age from embryonic day 20 to postnatal day (P) 19. At P11-P12 and later, such injections label retinal axons that form overlapping arbors restricted to a topographically correct terminal zone covering about 1% of the superior colliculus (SC) area. At perinatal ages, though, axons labeled from each retinal site are distributed in the SC over much of its medial-lateral axis and extend caudally well beyond the rostral-caudal location of their correct terminal zone; some continue caudally into the inferior colliculus. Axons typically form side branches and often arborize at topographically incorrect positions throughout the SC; however, they appear to branch preferentially in a region that includes, but is much larger than, their correct terminal zone. The mature, retinotopically ordered projection emerges during an early postnatal remodeling period through the rapid remodeling of the early, diffuse projection. This process involves the large-scale removal of axons, axon segments, branches, and arbors from topographically inappropriate positions concurrently with a dramatic increase in branching and arborization at topographically correct locations. Quantitative measurements show that elimination of aberrant branches without loss of the primary axons contributes substantially to the development of order. By P6, fewer mistargeted axons persist, but those that do persist tend to branch or arborize more extensively in topographically inappropriate regions. By P8, the labeling patterns begin to approximate those seen at maturity. Further refinement leads to an adultlike topographic ordering of axonal arborizations by P11-P12. At maturity, some axons take very indirect routes to reach their correct terminal zone. However, such trajectory changes typically correct only small positional inaccuracies, indicating that axons and axon segments that make larger targeting errors do not survive the remodeling phase. Previous retrograde labeling studies indicate that some retinal axons make topographic targeting errors (O'Leary et al., 1986; Yhip and Kirby, 1990), but none have suggested the degree of diffuseness revealed by anterograde labeling with Dil. Our findings show that directed axon growth is inadequate as a mechanism to develop the topographic ordering of retinal axons in the rat SC. Rather, mechanisms that control the removal of mistargeted axons and promote the arborization of correctly positioned axons are critical for the development of retinotopic order.(ABSTRACT TRUNCATED AT 400 WORDS)     

Inaccuracies in initial growth and arborization of chick retinotectal axons followed by course corrections and axon remodeling to develop topographic order

The retinotectal projection is organized in a precise retinotopic manner. We find, though, that during development the growth and arborization of temporal retinal axons within the optic tectum of chick embryos is initially imprecise. Axonal targeting errors occur along the rostral-caudal and medial-lateral tectal axes, and arbors are formed at topographically inappropriate positions. Subsequent course corrections along both tectal axes and large-scale axonal remodeling lead to the retinotopic ordering of terminal arborizations characteristic of the mature projection. The trajectories and branching patterns of temporal retinal axons labeled with Dil or DiO were determined in whole mounts of retina and tectum from chicks ranging in age from embryonic day 9 to posthatching. Within the retina, labeled retinofugal axons travel in a compact bundle but do not maintain strict neighbor relations, as they course to the optic fissure. The axons enter the contralateral tectum at its rostral edge and grow caudally. Many extend well past their appropriate terminal zone within rostral tectum; a proportion of these later reverse their direction of growth. Many axons grow onto the tectum at incorrect positions along the medial-lateral tectal axis. Some correct this error in a directed manner by altering their trajectory or extending collateral branches at right angles. About 80% of the positional changes of this type are made in the direction appropriate to correct axon position, and thus are likely a response to tectal positional cues. After maturation of retinotopic order, about half of the axons that project to a mature terminal zone have made abrupt course corrections along one or both tectal axes, indicating that initially mistargeted axons can establish appropriately positioned arbors and survive. The development of temporal axons within the tectum is characterized by 3 phases: elongation, branch and arbor formation, and remodeling. After considerable rostrocaudal elongation, an axon typically develops numerous side branches and arbors, many at inappropriate locations. Most arbors are formed by side branches that develop as interstitial collaterals; few axons grow directly to their appropriate terminal zone and arborize. Aberrant arbors, and axons and axon segments that fail to form arbors in the appropriate terminal zone, are rapidly eliminated over about a 2 d period. Axon degeneration appears to play a role in this remodeling process.     

Development of the hippocamposeptal projection in the rat

We analyzed the development of the hippocamposeptal projection and the morphology of the neurons giving rise to this projection. The fluorescent tracer Dil was injected into the septal region or the hippocampus in fixed brains of embryonic and early postnatal rats. Anterogradely labeled hippocampal axons first reached the septal region at E16. They ran along the midline of the brain, thereby approaching the medial septum. Axons to the lateral septum were first observed around E18/19. The lateral septum is partly innervated by collaterals of axons that travel to the medial septum. The projection to the lateral septal nuclei becomes more massive during early postnatal stages, whereas that to the medial septum becomes smaller. Cells in the medial septum retrogradely labeled by injection into the hippocampus were first observed at E18. Thus, the hippocamposeptal projection is established earlier than the septohippocampal projection. The first hippocampal projection neurons are nonpyramidal neurons that appear to pioneer the pathway to the septum. Pyramidal cell axons follow this first cohort of axons into the medial septum. Pyramidal cells could be retrogadely labeled from the medial septum during the perinatal period but then diminished in number. At P10, only nonpyramidal cells were labeled by medial septal injections. This indicates that the pyramidal component of this projection is transient and is removed shortly after birth. However, as is known from ther studies, hippocampal pyramidal cells give rise to a powerful projection to the lateral septum in adult animals. Our results show that there is a considerable remodeling of the projection from the hippocampus to the septum during ontogenetic development. © 1995 Willy-Liss, Inc.     

Prenatal development of the optic projection in albino and hooded rats

The development of retinofugal projections has been examined in albino and hooded rat embryos from embryonic day 16 to birth (E21.5). Horseradish peroxidase (HRP) was injected intraocularly through the uterine wall and its anterograde transport revealed with TMB and DAB. The retrograde transport of HRP or the fluorescent dyes Nuclear yellow, Fast blue and propidium iodide from optic tract, superior colliculus (SC) or lateral geniculate body (LG) injections was used to demonstrate the origin of the projections. Superficial projections to the contralateral SC were first identified at E16. A light projection to the entire medio-lateral extent of the ipsilateral SC could be detected a day later. The optic axons grow over the surface of the diencephalon at E16 and it was only at later stages that the fibers were observed to invade successively deeper parts of the LG. A superficial projection to the ipsilateral LG could first be detected at E17. Both the ipsilateral and contralateral projections grew through the entire dorso-ventral extent of the lateral geniculate body: some restriction of the axons to their normal adult termination zones could be detected by E21. No difference in the distribution of projections could be detected between the albino and pigmented rats although the projections were lighter, and possibly because of this were detected later, in the albino rats. At all the ages examined in this study labeled retinal ganglion cells were observed in the non-injected eyes after injection of label into the contralateral eye. The use of persistent fluorescent dyes showed that these retinal ganglion cells did not survive for more than 5 days postnatally. The projection to the uninjected eye came preferentially from ganglion cells in the lower nasal retina while the ipsilateral central projections came predominantly but not exclusively from the lower temporal retina of the injected eye. It appears, therefore, that the initial projections of optic axons in the rat are not limited to their normal termination zones and that the choice of pathway at the chiasm appears to be only loosely controlled.     

Refinement of the ipsilateral retinocollicular projection is disrupted in double endothelial and neuronal nitric oxide synthase gene knockout mice

Development of retinal connections to the superior colliculus (SC) requires an activity dependent refinement process in which axons gradually become restricted to appropriate retinotopic locations. Nitric oxide has been implicated in this process. We tested this possibility by studying the refinement of the ipsilateral retinocollicular projections (IRP) in normal C57-BL/6 mice and in double knockout mice in which the genes for the edothelial and neuronal isoforms of nitric oxide synthase (e, nNOS) were disrupted. Mice aged between P19 and adulthood were perfused 44-48 h after anterograde injections of WGA-HRP into one eye in order to measure the distribution of the labeled IRP. In normal mice, segregation of the IRP was complete at P21, with the ipsilateral projection restricted to the rostro-medial SC. By contrast, the ipsilateral projection was spread over much more of the SC in double e, nNOS knockouts at P21 with patches of label distributed across the entire medio-lateral axis of the rostral 700 microm. Although the distribution of the ipsilateral projection became more restricted in knockout animals at later ages, it was still more extensive than that of normal mice of the same age at P28 and P42. In the adult, the distribution of axons was similar in both normal and double knockout animals. These results show that refinement of the IRP is delayed when expression of eNOS and nNOS is disrupted, presumably to axons with uncorrelated activity because nitric oxide serves as a repellant molecule during normal development.     

Axonal arborization in the developing chick retinotectal system

The growth and arborization of chicken retinal ganglion cell axons have been investigated by means of an intraaxonally transported fluorescent marker in the developing retinotectal system. The fluorescent dye D282 or diI from the carbocyanine group of dyes is taken up by ganglion cells and labels the axon as well as the axonal growth cones and the terminal arborizations on the tectum. Branching and arborization start in the chick retinotectal system on embryonic day 9 (E9). At this stage retinal axons leave the stratum opticum (SO) and invade the stratum griseum et fibrosum superficiale (SGFS), where arborization takes place. On day E12 several axons were found to arborize in the SGFS. At this stage arbors appear to have small branches with less than 4 branching points. The extension of terminal arbors in the anterior/posterior (A/P) and in the dorsal/ventral (D/V) direction was determined for 50 axonal trees at days E13-14 and for 24 arbors at days E15-16. Few axonal terminals were investigated at day E18. The mean A/P extent of axonal terminal trees increases from 0.23 +/- 0.12 to 0.36 +/- 0.22 mm from E13-14 to E15-16 and seems to stay at this order of magnitude on E18. The mean D/V extent increases from 0.23 +/- 0.17 to 0.30 +/- 0.18 mm in the same embryonic period of development. The number of branching points calculated from the same number of axonal trees increases from 7.50 +/- 2.98 at E13-14 to 11.70 +/- 4.10 at E15-16. This number seems to increase further after day E16 achieving values of about 20 to 25 at E18. This was, however, not quantifiable by the technique used here and represents an approximate value estimated from 6 completely labeled terminal fields at E18. The data presented here suggest that the modeling of the final branching pattern in the chick retinotectal system takes place within a relatively short period of embryonic development. Prior to the beginning of terminal arborization two important events contribute to the formation of a retinotopic projection. One event is the change of the D/V position by a minority of axons lying ectopic in terms of retinotopy. Some axons turn at right angles and change their D/V position. The other event is the appearance of side branches along the A/P axis.     

Axonal regeneration and synapse formation in the superior colliculus by retinal ganglion cells in the adult rat

In adult rats, one optic nerve was transected and replaced by a 4 cm segment of autologous peripheral nerve (PN) that linked one eye and the superior colliculus (SC) along a predominantly extracranial course. Retrograde and orthograde studies with the tracers HRP or rhodamine-B-isothiocyanate (RITC), as well as immunocytochemical neuronal labels, indicated the following: (1) Regenerating axons from the axotomized retinal ganglion cells extended along the entire PN grafts, covering a distance nearly twice that of the normal retinotectal projection of intact rats. (2) Some of these axons penetrated the SC and formed terminal arborizations up to 500 microns from the end of the graft. (3) By electron microscopy, the arborizations of these regenerated axons in the SC were seen as small HRP-labeled axonal profiles that contacted neuronal processes in the SC; some of these contacts showed pre- and postsynaptic membrane specializations. These findings indicate that injured retinal ganglion cells in the adult rat are not only able to regrow lengthy axons, but may also form synapses in the SC.     

Development of the projection from the nucleus of the brachium of the inferior colliculus to the superior colliculus in the ferret

Neurons in the deeper layers of the superior colliculus (SC) have spatially tuned receptive fields that are arranged to form a map of auditory space. The spatial tuning of these neurons emerges gradually in an experience-dependent manner after the onset of hearing, but the relative contributions of peripheral and central factors in this process of maturation are unknown. We have studied the postnatal development of the projection to the ferret SC from the nucleus of the brachium of the inferior colliculus (nBIC), its main source of auditory input, to determine whether the emergence of auditory map topography can be attributed to anatomical rewiring of this projection. The pattern of retrograde labeling produced by injections of fluorescent microspheres in the SC on postnatal day (P) 0 and just after the age of hearing onset (P29), showed that the nBIC-SC projection is topographically organized in the rostrocaudal axis, along which sound azimuth is represented, from birth. Injections of biotinylated dextran amine-fluorescein into the nBIC at different ages (P30, 60, and 90) labeled axons with numerous terminals and en passant boutons throughout the deeper layers of the SC. This labeling covered the entire mediolateral extent of the SC, but, in keeping with the pattern of retrograde labeling following microsphere injections in the SC, was more restricted rostrocaudally. No systematic changes were observed with age. The stability of the nBIC-SC projection over this period suggests that developmental changes in auditory spatial tuning involve other processes, rather than a gross refinement of the projection from the nBIC.     

Neonatally elevated serotonin levels alter terminal arbors of individual retinal ganglion cells in superior colliculus of hamsters

Previous studies from this laboratory showed that sprouting of serotoninergic (5-HT) axons in the hamster's superior colliculus (SC), induced by a single subcutaneous injection of 5,7-dihydroxytryptamine (5,7-DHT) at birth (postnatal day 0 [P-0]), resulted in an increased terminal distribution of the uncrossed retinocollicular projection that was not associated with any changes in the number or distribution of ipsilaterally projecting retinal ganglion cells. The present study was undertaken to determine what effect this manipulation had on the terminal arbors of such axons. Retinocollicular axons of normal and 5,7-DHT-treated animals were anterogradely labeled with small intraretinal injections of the lipophilic dye 1,1'-dioctadecyl-3,3,3',3'-tetramethylindocarbocyanine perchlorate (DiI) on P-16. After tissue processing on P-19, single retinocollicular axon arbors were reconstructed by using confocal microscopy. Quantitative analysis indicated that arbors from 5,7-DHT-treated hamsters had significantly greater total fiber lengths, areas, and volumes than those from normal animals. There were no differences between axons from the two groups in number of branch points, distribution of relative branch lengths, and numbers of bouton-like swellings. These results support the hypothesis that increased SC concentrations of 5-HT alter development of the uncrossed retinocollicular pathway such that a greater territory is covered by individual terminal arbors but that the number of synaptic contacts per arbor remains constant. This may explain, at least in part, the abnormally widespread distribution of the aggregate ipsilateral projection.     

Prenatal development of retinocollicular projections in the rabbit: an HRP study

The prenatal development of the rabbit's retinal projections to the superior colliculus (SC) was studied by using anterograde transport of horseradish peroxidase injected intraocularly. Fetuses aged embryonic day 21 (E21) to E29 and an adult rabbit were examined. Gestation in the rabbit is 30-31 days. On E21 contralaterally projecting retinal fibers invade across the entire SC. Their distribution is initially diffuse within the superficial laminae, but by E29 they have a distinct stratified appearance. Ipsilaterally projecting retinal fibers invade the rostral half of the SC on E21. By E23 they cover the entire SC and overlap the contralateral fibers both tangentially and radially. The ipsilateral fibers for the most part are sparsely distributed, but they form a dense focal distribution in the rostrolateral quarter of the SC. This focus straddles the stratum griseum superficiale/stratum opticum (SGS/SO) border. On E25 the ipsilateral fibers maintain their widespread distribution and focal rostrolateral concentration. By E27 they are excluded almost entirely from the caudal half of the SC and are reduced in density in the rostromedial quarter of the nucleus. On E29 the ipsilateral terminal field forms distinct patches and bands that are restricted to the rostrolateral quarter of the SC and are confined to the SGS/SO border. Thus, a few days before birth the pattern and location of the ipsilateral retinocollicular projection resemble those seen in the adult. The early widespread distribution of the ipsilaterally projecting retinal fibers to the SC and their eventual restriction in the fetal rabbit are consistent with the development of this projection in other mammalian orders.     

Postnatal innervation of the rat superior colliculus by axons of late-born retinal ganglion cells

Rat retinal ganglion cells (RGCs) are generated between embryonic day (E) 13 and E19. Retinal axons first reach the superior colliculus at E16/16.5 but the time of arrival of axons from late-born RGCs is unknown. This study examined (i) whether there is a correlation between RGC genesis and the timing of retinotectal innervation and (ii) when axons of late-born RGCs reach the superior colliculus. Pregnant Wistar rats were injected intraperitoneally with bromodeoxyuridine (BrdU) on E16, E18 or E19. Pups from these litters received unilateral superior colliculus injections of fluorogold (FG) at ages between postnatal (P) day P0 and P6, and were perfused 1-2 days later. RGCs in 3 rats from each BrdU litter were labelled in adulthood by placing FG onto transected optic nerve. Retinas were cryosectioned and the number of FG, BrdU and double-labelled (FG+/BrdU+) RGCs quantified. In the E16 group, the proportion of FG-labelled RGCs that were BrdU+ did not vary with age, indicating that axons from these cells had reached the superior colliculus by P0/P1. In contrast, for the smaller cohorts of RGCs born on E18 or E19, the proportion of BrdU+ cells that were FG+ increased significantly after birth; axons from most RGCs born on E19 were not retrogradely FG-labelled until P4/P5. Thus there is a correlation between birthdate and innervation in rat retinotectal pathways. Furthermore, compared to the earliest born RGCs, axons from late-born RGCs take about three times longer to reach the superior colliculus. Later-arriving axons presumably encounter comparatively different growth terrains en route and eventually innervate more differentiated target structures.     

EphB regulates L1 phosphorylation during retinocollicular mapping

Interaction of the cell adhesion molecule L1 with the cytoskeletal adaptor ankyrin is essential for topographic mapping of retinal ganglion cell (RGC) axons to synaptic targets in the superior colliculus (SC). Mice mutated in the L1 ankyrin-binding motif (FIGQY(1229)H) display abnormal mapping of RGC axons along the mediolateral axis of the SC, resembling mouse mutant phenotypes in EphB receptor tyrosine kinases. To investigate whether L1 functionally interacts with EphBs, we investigated the role of EphB kinases in phosphorylating L1 using a phospho-specific antibody to the tyrosine phosphorylated FIGQY(1229) motif. EphB2, but not an EphB2 kinase dead mutant, induced tyrosine phosphorylation of L1 at FIGQY(1229) and perturbed ankyrin recruitment to the membrane in L1-transfected HEK293 cells. Src family kinases mediated L1 phosphorylation at FIGQY(1229) by EphB2. Other EphB receptors that regulate medial-lateral retinocollicular mapping, EphB1 and EphB3, also mediated phosphorylation of L1 at FIGQY(1229). Tyrosine(1176) in the cytoplasmic domain of L1, which regulates AP2/clathrin-mediated endocytosis and axonal trafficking, was not phosphorylated by EphB2. Accordingly mutation of Tyr(1176) to Ala in L1-Y(1176)A knock-in mice resulted in normal retinocollicular mapping of ventral RGC axons. Immunostaining of the mouse SC during retinotopic mapping showed that L1 colocalized with phospho-FIGQY in RGC axons in retinorecipient layers. Immunoblotting of SC lysates confirmed that L1 was phosphorylated at FIGQY(1229) in wild type but not L1-FIGQY(1229)H (L1Y(1229)H) mutant SC, and that L1 phosphorylation was decreased in the EphB2/B3 mutant SC. Inhibition of ankyrin binding in L1Y(1229)H mutant RGCs resulted in increased neurite outgrowth compared to WT RGCs in retinal explant cultures, suggesting that L1-ankyrin binding serves to constrain RGC axon growth. These findings are consistent with a model in which EphB kinases phosphorylate L1 at FIGQY(1229) in retinal axons to modulate L1-ankyrin binding important for mediolateral retinocollicular topography.     

Development of primary visual projections occurs entirely postnatally in the fat-tailed dunnart,a marsupial mouse,Sminthopsis crassicaudata

We have examined the development of retinal projections in a diminutive polyprotodont marsupial, the fat-tailed dunnart, Sminthopsis crassicaudata. Here, we document the most immature mammalian visual system at birth described to date. At postnatal day (P) 0, the retinal ganglion cell layer has yet to form, and axons have not entered the optic stalk. By P4, the retinal ganglion cell layer could be distinguished at the posterior pole, and the front of growing axons extended one-third the length of the optic stalk, a distance of approximately 150 μm; a few pioneer growth cones had grown beyond the main axon group but had still to reach the midline. Axons had decussated at the optic chiasm by P10 to penetrate the base of the contralateral optic tract and, by P15, had reached the dorsal lateral geniculate nucleus (dLGN), superior colliculus (SC), and accessory optic system (AOS); ipsilaterally projecting axons matured slightly later. From P20, axons had reached the caudal SC both contralaterally and ipsilaterally and terminated throughout the depth of the retinorecipient layers. After P30, the projections gradually refined. Within the rostral dLGN, segregation into four contralateral and four ipsilateral bands occurred by P50, approximately 5 days after eye opening. The projection to the ipsilateral SC underwent refinement by P50, becoming restricted to its rostral pole, and presented as discrete patches within the stratum opticum. At birth, the dunnart visual system is comparable to early to midembryonic stages [embryonic day (E) 12, E14, E19, E24, and E30, respectively] in the mouse, rat, ferret, cat, and monkey. The extreme immaturity of the neonatal dunnart together with the observation that the entire development of the primary optic pathway occurs postnatally over a protracted period make this marsupial especially valuable for investigating factors that control pathway formation in the early developing mammalian primary visual system. J. Comp. Neurol. 384:26–40, 1997. © 1997 Wiley-Liss, Inc.     

Development of the crossed retinocollicular projection in the mouse

Changes in the distribution of axons of the crossed retinal projection within the superior colliculus of the developing mouse were studied by means of normal fiber and Golgi impregnations and by anterograde horseradish peroxidase labelling. Retinal axons advance along the optic tract from gestational days E12 to E14 and first invade the superior colliculus on E15. Over the subsequent days until birth (E19), the retinal axons extend within rostrocaudally oriented fascicles that distribute through the full thickness of the uppermost collicular layer, the stratum superficiale (SS). A dramatic transformation of this fiber stratification pattern into the mature pattern occurs over the first postnatal week. The fiber bundles are progressively cleared from the upper half of SS, identified as the future stratum griseum superficiale (SGS). Concurrently, the fiber bundles in the deep SS, identified as the stratum opticum (SO), give rise to individual, nonfasciculated fibers, which arborize within SGS. The contralateral retinal origin of the transient population of axons in SGS as well as the majority of axons that persist in SO is evident from the observation that they degenerate following neonatal enucleation. The number of fiber bundles lost is estimated to be 40-50% of the total population present in the superficial layers at birth. The combined set of observations indicates that axon elimination plays a major role in shaping the laminar pattern of retinal innervation of the colliculus. Retinal ganglion cell death, and not axon pruning, is proposed as the most probable mechanism by which axon fascicles are eliminated from SGS.     

Morphology of ganglion cells which project to the dorsal lateral geniculate and superior colliculus in the ground squirrel

We wished to determine whether retinal ganglion cells that have axons terminating in the dorsal lateral geniculate and/or the superior colliculus have specific sizes of somata, comprising only part of the entire size range of ganglion cell somata. If so, then perhaps the specific functional types described by Michael might be associated with morphological types based on soma size. HRP was injected into either the superior colliculus (SC) or dorsal lateral geniculate nucleus (LGd) of thirteen-lined ground squirrels. Soma diameter of labeled ganglion cells was measured and the relation between cell size and frequency determined. After SC injections HRP-filled cells were mostly small and medium-sized. They ranged in diameter from 3 to 14 microns and the mean diameter of labeled neurons was 7.35 microns. Cells labeled after SC injections were often distributed as doublets or triplets in the retina. After LGD injections the majority of labeled cells were medium and large-sized. They ranged from 4 to 18 microns in diameter with a mean of 9.1 microns and were more regularly spaced within the retinal region of labeled cells. Thus, the present results provide reason to believe that functional classes of ganglion cells in ground squirrels may be correlated with particular morphological types.     

Transient retinal ganglion cells in the developing rat are characterized by specific morphological properties

To determine whether dendritic development of mammalian retinal ganglion cells (RGCs) is affected by axonal target specificity, the morphology of three populations of maturing RGCs was examined. These included RGCs that exhibited either a transient, topographically incorrect, projection to the caudal superior colliculus (SC), or a transient projection to the caudal inferior colliculus (IC), in addition to a control group that exhibited a topographically correct projection to the caudal SC. Projection populations were identified by retrograde transport of rhodamine labeled latex microspheres injected into target nuclei. Labeled RGCs were then injected in vitro with Lucifer yellow to reveal the details of their dendritic morphology. Retinal ganglion cells making target errors, most of which ultimately die, were found to undergo a remarkable degree of morphological differentiation and could be categorized according to the adult type I, II, or III criteria. However, the relative proportions of these cell types were different among RGCs making transient connections versus those whose projections were preserved. Approximately half of the RGCs making topographically incorrect projections to the SC belonged to type III, in contrast to 6% that made a topographically correct projection. In addition, the population of cells sending axons to caudal IC did not include type III RGCs, but consisted of small type II neurons. The development of the basic dendritic form of each RGC type was only modestly influenced by its projection pattern; dendritic trees of cells making transient projections were essentially normal with only a slight, but statistically significant, reduction in dimensions. Moreover, dendritic remodeling was evident during maturation of neurons making either transient or normal projections. Together, these findings indicate that target specificity plays a relatively minor role on dendritic development of retinal ganglion cells. © 1996 Wiley-Liss, Inc.     

Developmental changes in the pattern of retinal projections in pigmented and albino rabbits

The distribution of retinal axons and/or terminals in the retino-recipient nuclei of pigmented and albino rabbits varying in age from the 24th postconceptional day (24PCD) to adulthood was examined following unilateral intraocular injections of the enzyme horseradish peroxidase. Both in pigmented and albino rabbits contralateral retinal axons and/or terminals in the dorsal and ventral lateral geniculate nuclei (DLG and VLG), superior colliculi (SC), pretecta (PT) and accessory optic tract nuclei (AON) were already present on 24PCD. In the period 26-30PCD the contralateral projection occupied the entire volume of the DLG, VLG and SC. Although 32PCD (the day of birth) the proportions of the volumes of DLG and VLG occupied by the contralateral projections were slightly reduced, their volume continued to increase in absolute terms up to adulthood. In pigmented rabbits the ipsilateral projections to all retino-recipient nuclei were most dense and extensive on 26PCD. From 26PCD, the relative extent of the ipsilateral projections was gradually reduced, but a reduction in their absolute extent did not become evident until 32PCD. By 32PCD the ipsilateral projection to the AON had disappeared completely. The distribution of ipsilateral axons and/or terminals and the relative proportion of the nuclei occupied by the ipsilateral projection in all other retino-recipient nuclei had become adult-like by 34PCD. In albino rabbits only a sparse ipsilateral projection to the presumptive superficial collicular layers was present on 24PCD. In the remaining retino-recipient nuclei an ipsilateral projection was present on 26PCD. From 26PCD the relative extent and from 30PCD the absolute extent of ipsilateral retinal axons and/or terminals was gradually reduced. The relative extent of the ipsilateral projection had become almost adult-like by 34PCD. Throughout development ipsilateral projections in albinos were consistently less dense and less extensive than those in pigmented rabbits, and unlike in pigmented rabbits, the ipsilateral projections to the VLG and PT were only transient. The differences between the two strains in the pattern of retinofugal projections were further enhanced during the period of segregation of the ipsilateral and contralateral projections. Considering the fact that in both strains there is a partial correspondence between the period in which the spatial extent of the ipsilateral projections is reduced and the period of retinal ganglion cell (RGC) death, it is likely that RGC death plays a role in the process of segregation of the retinal afferents into ocular domains. However, our data suggest that other mechanism(s) also play an important role in the process.     

Alignment of multimodal sensory input in the superior colliculus through a gradient-matching mechanism

The superior colliculus (SC) is a midbrain structure that integrates visual, somatosensory, and auditory inputs to direct head and eye movements. Each of these modalities is topographically mapped and aligned with the others to ensure precise behavioral responses to multimodal stimuli. While it is clear that neural activity is instructive for topographic alignment of inputs from the visual cortex (V1) and auditory system with retinal axons in the SC, there is also evidence that activity-independent mechanisms are used to establish topographic alignment between modalities. Here, we show that the topography of the projection from primary somatosensory cortex (S1) to the SC is established during the first postnatal week. Unlike V1-SC projections, the S1-SC projection does not bifurcate when confronted with a duplicated retinocollicular map, showing that retinal input in the SC does not influence the topography of the S1-SC projection. However, S1-SC topography is disrupted in mice lacking ephrin-As, which we find are expressed in graded patterns along with their binding partners, the EphA4 and EphA7, in both S1 and the somatosensory recipient layer of the SC. Together, these data support a model in which somatosensory inputs into the SC map topographically and establish alignment with visual inputs in the SC using a gradient-matching mechanism.     

The development of corticocollicular projections in anophthalmic mice.

To determine the role of retinal axons in the development of the corticocollicular projection in mice, the lipophilic fluorescent dye, DiI, was used to compare the development of the cortical projections in phenotypically normal (C57BL/6J) mice to that of anophthalmic 129SV/CPorJ mice. Cortical axons in anophthalmic mice found their targets and established a laminar specificity similar to those of cortical axons in normal mice despite the absence of the retinal projection. Cortical axons in normal mice reached the superior colliculus before those in anophthalmic mice and also had a faster rate of growth within the colliculus. Unlike cortical axons in normal mice in early postnatal ages, those in anophthalmic mice formed a disperse bundle in the stratum opticum. Axons labeled by focal applications of DiI into area 17 terminated in a larger and more medial area in anophthalmic mice than in normal mice. Thus, retinal axons are not essential for cortical axons to reach the superior colliculus, but they may have a role in organizing the growth of later-arriving cortical axons. Furthermore, cortical axons can terminate in the superior colliculus with a coarse topography when retinal axons are absent, but they cannot form a topographically refined projection.