Optic Tract Lesions Research Articles

Human Umbilical Cord Blood–Derived Mesenchymal Stem Cells Do Not Differentiate Into Neural Cell Types or Integrate Into the Retina After Intravitreal Grafting in Neonatal Rats

This study investigated the ability of mesenchymal stem cells (MSCs) derived from full-term human umbilical cord blood to survive, integrate and differentiate after intravitreal grafting to the degenerating neonatal rat retina following intracranial optic tract lesion. MSCs survived for 1 week in the absence of immunosuppression. When host animals were treated with cyclosporin A and dexamethasone to suppress inflammatory and immune responses, donor cells survived for at least 3 weeks, and were able to spread and cover the entire vitreal surface of the host retina. However, MSCs did not significantly integrate into or migrate through the retina. They also maintained their human antigenicity, and no indication of neural differentiation was observed in retinas where retinal ganglion cells either underwent severe degeneration or were lost. These results have provided the first in vivo evidence that MSCs derived from human umbilical cord blood can survive for a significant period of time when the host rat response is suppressed even for a short period. These results, together with the observation of a lack of neuronal differentiation and integration of MSCs after intravitreal grafting, has raised an important question as to the potential use of MSCs for neural repair through the replacement of lost neurons in the mammalian retina and central nervous system.

Retinal nerve fiber layer loss documented by optical coherence tomography in patients with optic tract lesions

PURPOSE: To report abnormalities of retinal nerve fiber layer (RNFL) thickness using Stratus - optical coherence tomography (OCT) in two patients with optic tract lesions. METHODS: Two patients with long standing homonymous hemianopia from optic tract lesions were submitted to a complete neuro-ophthalmic evaluation and to Stratus -optical coherence tomography examination. RESULTS: Both patients revealed diffuse loss of the RNFL at Stratus - OCT in both eyes. In the eyes with the temporal hemianopia, RFNL loss was diffuse but predominantly in the nasal and temporal areas of the optic disc, the classic pattern of band atrophy of the optic nerve. In the eyes with nasal hemianopia RNFL loss could be documented in the superior and inferior quadrants of the optic disc. RNFL loss correlated well with visual field loss and the expected pattern of RNFL loss in optic tract lesions. CONCLUSION: Stratus-Optical coherence tomography can provide useful information in the diagnosis of optic tract lesions by identifying the characteristic pattern of RNFL loss that occurs in both eyes in this condition.

Contour-integration deficits on the intact side of the visual field in hemianopia patients

Objective Visual impairments in hemianopia are thought to be exclusively caused by the reduced visual field size. However, the primary lesion may affect the contralateral hemisphere through damage of interhemispheric projections. The question therefore arises if the presumed “intact” hemifield is perceptually impaired. Methods Three hemianopia patients and three matched controls carried out a Yes/No figure detection task with their intact side of the visual field. The figure (square) contours were composed of non-contiguous Gabor patches embedded in a random patch array of different background densities (low, Δ = 2; high, Δ = 1). Response accuracy and reaction times were recorded. Results A temporal–parietal patient revealed figure detection impairments, with accuracy rate, 77% ( Δ = 2) and 53% ( Δ = 1), below compared control values. An occipital patient was comparable to his match: 99% ( Δ = 2); 84% ( Δ = 1). Both patients exhibited frequent false alarms to random patterns and required longer presentation times to perform the task. In the third patient, with optic tract lesion, figure detection was nearly normal at low density (92%, Δ = 2) but impaired at noisy background (62%, Δ = 1). Conclusion The “intact” visual field in hemianopes is impaired in detection of incomplete figures embedded in a noisy background. This deficit may be caused by damage to higher visual centers and/or loss of interhemispheric interactions.

Combined Therapies in the Treatment of Neurotrauma: Polymers, Bridges and Gene Therapy in Visual System Repair

Background: The mature central nervous system (CNS) has limited capacity for self-renewal and repair after injury or neurodegeneration, and therapeutic strategies are needed to promote the viability of damaged neurons and the regrowth of their axons. The retina and optic nerve (ON) are part of the CNS, and the visual system is widely used in experimental studies on injury and repair. Objective:To test various cellular and molecular approaches in attempts to replace retinal ganglion cells (RGCs) in depleted retinas or, more usually, promote the survival of endogenous injured RGCs and stimulate axonal regeneration after ON or intracranial optic tract (OT) injury. Methods and Results: Intraocular injections of brain-derived neurotrophic factor and ciliary neurotrophic factor (CNTF) temporarily increase RGC survival after ON injury. More sustained neuroprotection is obtained using adeno-associated viral vectors to transfect RGCs with brain-derived neurotrophic factor or CNTF genes. After ON crush, intravitreal adeno-associated viral CNTF injections also increase RGC axonal regrowth. Additional protective and growth effects are obtained after intraocular elevation of cAMP and by manipulation of protein kinase signalling pathways in RGCs. Regeneration is increased by transplanting a segment of peripheral nerve onto the cut ON. Schwann cells in peripheral nerve grafts can be genetically modified using lentiviral vectors to over-express CNTF, resulting in increased regrowth of RGC axons. After OT lesions, hydrogels have been used to bridge the injury, sometimes with the incorporation of signalling peptides or cells genetically modified to express neurotrophic factors. Conclusions: There is now a general consensus that combinatorial approaches are needed to elicit sustained and effective regenerative responses in injured adult CNS neurons.

Congruency in Homonymous Hemianopia

To evaluate the value of congruency in the localization of brain lesions in patients with homonymous hemianopia (HH). Retrospective observational study. Charts of all patients with HH seen over 15 years were reviewed. Only patients with incomplete HH documented on formal visual field testing and neuroimaging were included. HH was said to be congruent when the fields of both eyes were identical in shape, depth, and size. Patients were divided into two groups based on congruency of HH; demographic, clinical, and radiological characteristics were compared. Five hundred and thirty patients with 548 incomplete HH were included (373 congruent HH and 175 incongruent HH). Demographic variables were similar in both groups. Stroke caused 75% of congruent HH and 55.8% of incongruent HH; trauma and tumors caused 20.5% of congruent HH and 34.5% of incongruent HH (P < .001). The lesion locations in congruent HH vs incongruent HH included occipital lobe in 47.9% vs 21.3%, occipital lobe and optic radiations in 8.3% vs 5.6%, optic radiations in 32.4% vs 50.6%, optic tract in 7.2% vs 16.3%, and other locations in 4.2% vs 6.3% (P < .0001). Although there was a trend toward more congruent HH for lesions of the posterior visual pathways (P < .001), 50% of optic tract lesions and 59% of optic radiation lesions produced congruent HH. Although lesions involving the occipital lobe characteristically produce congruent HH, at least 50% of lesions in other locations also produce congruent HH, especially if these lesions are stroke-related. The rule of congruency should be used cautiously and may not apply to optic tract lesions.

Origin of the Relative Afferent Pupillary Defect in Optic Tract Lesions

To determine the percent decussation of pupil input fibers in humans and to explain the size and range of the log unit relative afferent pupillary defect (RAPD) in patients with optic tract lesions. Experimental study. Five patients with a unilateral optic tract lesion. The pupil response from light stimulation of the nasal hemifield, temporal hemifield, and full field of each eye of 5 patients with a unilateral optic tract lesion was recorded using computerized binocular infrared pupillography. Six stimulus light intensities, separated by 0.5-log unit steps, were used; 12 stimulus repetitions were given for each stimulus condition. For each stimulus condition, the pupil response of each eye was characterized by plotting the mean pupil contraction amplitude as a function of stimulus light intensity. The percentage of decussating afferent pupillomotor input fibers was calculated from the ratio of the maximal pupil contractions elicited from each eye. The RAPD was determined pupillographically from full-field stimulation to each eye. In all patients, the pupil response from the functioning temporal hemifield ipsilateral to the tract lesion was greater than that from the functioning contralateral nasal hemifield. This temporal-nasal asymmetry increased with increasing stimulus intensity and was similar in hemifield and full-field stimuli, eventually saturating at maximal light intensity. The log unit RAPD did not correlate with the estimated percentage of decussating pupil fibers, which ranged from 54% to 67%. In patients with a unilateral optic tract lesion, the pupillary responses from full-field stimulation to each eye are the same as comparing the functioning temporal field with the functioning nasal field. The percentage of decussating fibers is reflected in the ratio of the maximal pupil contraction amplitudes resulting from stimulus input between the two eyes. The RAPD that occurs in this setting reflects the difference in light sensitivity between the intact temporal and nasal hemifields. Its magnitude does not correlate with the difference in the number of crossed and uncrossed axons, but its sidedness contralateral to the side of the optic tract lesion is consistent with the greater percentage of decussating pupillomotor input.

Irene E. Loewenfeld, PhD Physiologist of the Pupil

Irene E. Loewenfeld, PhD has devoted a long and vigorous professional life to understanding the workings of the pupil of the human eye. Her interest in the pupil began in 1940 when she went to work as a technician in the pupillography laboratory of Professor Otto Lowenstein at New York University. It culminated in her widely admired textbook The Pupil, published in 1993. Among her many contributions, Loewenfeld provided rigorous observations about Adie tonic pupil, anisocoria in optic tract lesions, Argyll Robertson pupil, oculomotor paresis with cyclic spasms, and innovations in electronic recordings of pupil movement.

The 57th Annual Meeting of the American Academy of Neurology, Miami, Florida, April 9-16, 2005

The 57th Annual Meeting of the American Academy of Neurology, Miami, Florida, April 9-16, 2005

Retinal Nerve Fiber Layer Thickness in Optic Tract Syndrome

Optic tract syndrome (OTS) is characterized by incongruous homonymous hemianopia and a perpendicular pattern of bilateral optic atrophy due to the optic tract lesion. However, loss of retinal nerve fiber layer thickness (RNFLT) associated with OTS has not been quantitatively assessed. A 20-year-old woman with blunt head trauma showed normal visual acuity, color vision, ocular motility, and intraocular pressure. Because of a relative afferent pupillary defect in her left eye and left-sided homonymous hemianopia, we suspected right-sided optic tract damage, although magnetic resonance imaging detected no intracranial lesion. Using optical coherence tomography (OCT), the RNFLT of this case was measured at 31 months after the trauma and compared with age-matched normal controls (n = 41). Nasal, temporal, superior, and inferior quadrant RNFLT was reduced by 22%, 21%, 5%, and 46% in the right eye and 76%, 64%, 25%, and 27% in the left eye, respectively. The reduction was > 3 x the standard deviation of the normal mean values in the nasal and temporal quadrants of the left eye and in the inferior quadrant of the right eye. OCT can determine the RNFLT reduction corresponding to the characteristic patterns of optic atrophy of OTS.

0919 Electrodiagnostic correlation of Uhthoff's phenomenon in a patient with an optic tract lesion due to multiple sclerosis

0919 Electrodiagnostic correlation of Uhthoff's phenomenon in a patient with an optic tract lesion due to multiple sclerosis

A new type of biocompatible bridging structure supports axon regrowth after implantation into the lesioned rat optic tract.

We have developed a new type of polymer/cell/matrix implant and tested whether it can promote the regrowth of retinal ganglion cell (RGC) and other axons across surgically induced tissue defects in the CNS. The constructs, which consisted of 2-2.5-mm-long polycarbonate tubes filled with lens capsule-derived extracellular matrix coated with cultured neonatal Schwann cells, were implanted into lesion cavities made in the left optic tract (OT) of 18-21-day-old rats. In one group, to promote Schwann cell proliferation and perhaps also to stimulate axon regrowth, basic fibroblast growth factor (bFGF) was added to the lens capsule matrix prior to implantation. In another group, to determine whether application of growth factors to the somata of cells enhances the regrowth of distally injured axons, the neurotrophin NT-4/5 was injected into the eye contralateral to the OT lesion. NT-4/5 and bFGF treatments were combined in some rats. After medium-term (4-10 weeks) or long-term (15-20 weeks) survivals, axon growth into implants was assessed immunohistochemically using a neurofilament (RT97) antibody. RGC axons were visualized after injection of WGA/HRP into the right eye. Viable Schwann cells were present in implants at all times after transplantation. Large numbers of RT97+ axons were consistently found within the bridging implants, often associated with the peripheral glia. Axons were traced up to 1.7 mm from the nearest CNS neuropil and there was immunohistochemical evidence of myelination by Schwann cells and by host oligodendrocytes. There were fewer RGC axons in the implants, fibers growing up to 1.6 mm from the thalamus. Neither NT-4/5 nor bFGF, alone or in combination, significantly increased the extent of RGC axon growth within the implants. A group of OT-lesioned rats was implanted with polymer tubes filled with 2-2.5-mm-long pieces of predegenerate peripheral nerve. Surprisingly, polymer/cell/matrix constructs contained comparatively greater numbers of RGC and other axons and supported more extensive axon elongation. Thus, implants of this type may potentially be useful in bridging large tissue defects in the CNS.

Opiate receptor avidity is increased in rhesus monkeys following unilateral optic tract lesion combined with transections of corpus callosum and hippocampal and anterior commissures

Opiate receptor avidity ( B max′/ K D) was measured in four rhesus monkeys following unilateral lesioning of the optic tract combined with transection of the corpus callosum and the hippocampal and anterior commissures depriving one hemisphere of visual input (Tract and Split), two animals with transection of commissures only (Split), and nine healthy monkeys with positron emission tomography (PET) and 6-deoxy-6-β-[ 18F]fluoronaltrexone (cyclofoxy, CF), a μ- and κ-opiate receptor antagonist. Opiate receptor avidity was found to be significantly higher in the Tract and Split animals, only, bilaterally, throughout the lateral cortex and in the cingulate and posterior putamen (41–117%). Ipsilateral changes were consistently greater than those contralateral, but this asymmetry was of statistical significance only in the parietal and occipital cortices. Cyclofoxy avidity was decreased in the medial cortex of both the Tract and Split and Split animals (∼25%). The results suggest that opiate pathways undergo extensive alteration in response to changes in brain functional activities brought about through hemispheric visual deprivation.

Differential expression of beta-amyloid precursor and Bcl-2 proto-oncogene proteins in the developing dog retina.

Previous studies of rat retinas have not only provided evidence that beta-amyloid precursor (APP) and B-cell lymphoma proto-oncogene (Bcl-2) proteins are colocalized in retinal Müller glial cells, but have also indicated that common mechanisms regulate their expression in these cells (Chen, S.T., Garey, L.J., Jen, L.S., 1994. Bcl-2 proto-oncogene protein immunoreactivity in normally developing and axotomised rat retinas. Neurosci. Lett. 172, 11 14; Chen, S.T., Gentleman, S.M., Garey, L.J., Jen, L.S., 1996. Distribution of beta-amyloid precursor and B-cell lymphoma proto-oncogene proteins in the rat retina after optic nerve transection or vascular lesion. J. Neuropathol. Exp. Neurol. 55, 1073-1082; Chen, S.T., Garey, L.J., Jen, L.S., 1997. Expression of beta-amyloid precursor protein immunoreactivity in the retina of the rat during normal development and after neonatal optic tract lesion. NeuroReport 8, 713-717). This investigation attempts to resolve whether or not the pattern observed in rats also applies to other higher mammalian species by examining the expression of immunoreactivity to APP and Bcl-2 in developing as well as mature dog retinas using immunocytochemical methods. Experimental results indicate that the immunoreactivity of both APP and Bcl-2 is located primarily in the inner retina, particularly in the ganglion cells and their axons in late fetal and neonatal stages. From the second postnatal week (the time of eye opening) onwards, immunoreactivity to APP, but not Bcl-2, is localized primarily in the endfeet and proximal part of the radial process of retinal Müller glial cells. Although the findings show both APP and Bcl-2 are expressed in ganglion cells and their processes suggest that the molecules have a role in the differentiation of neurons in the central nervous tissue, the lack of Bcl-2 in the Müller glial cells in dog retinas further suggests that the two molecules may have different biological roles with respect to glial function.

Retinal axon regeneration in peripheral nerve, tectal, and muscle grafts in adult rats.

This study examined whether prior regenerative growth through peripheral nerve (PN) bridging grafts influenced the specificity with which lesioned adult rat retinal ganglion cell (RGC) axons grew into co-grafts of developing target tissue (fetal superior colliculus). Growth into nontarget (muscle) tissue was also examined. Autologous PN was grafted onto the transected optic nerve. After 14 days, the distal ends of the PNs were placed next to, or inserted into, embryonic tectal tissue or into autologous muscle grafts placed in frontal cortex cavities. Host retinal projections were examined 3-8 months later using anterograde and retrograde tracing techniques. In rats in which there was good apposition between PN and tectal tissue, small numbers of RGC axons were observed growing into the tectal grafts (maximum distance of 180 microm). No evidence of specific innervation of appropriate target regions within tectal grafts was detected, even though such regions (identified by acetylcholinesterase histochemistry) were often located close to the PN grafts. In rats with PN/muscle co-grafts, the extent of retinal axon outgrowth was greater (up to 465 microm from the PN tip) and labelled profiles that resembled motor endplates were seen contacting muscle fibres. Previous studies have shown that spontaneously regenerating RGC axons consistently and selectively innervate appropriate target areas in fetal tectal tissue grafted directly into optic tract lesion cavities. Together, the data suggest that exposure to a PN environment may have reduced the extent of adult retinal axon growth into fetal tectal transplants and affected the way regenerating axons responded to specific developmental cues expressed by target cells in the co-grafted tissue.

Expression of beta-amyloid precursor protein immunoreactivity in the retina of the rat during normal development and after neonatal optic tract lesion.

Immunoreactivity to beta-amyloid precursor protein (APP) was present in the inner plexiform, ganglion cell and optic fibre layers, as well as in blood vessels, at birth in normally developing rat retinas. In the inner plexiform layer immunoreactivity disappeared by postnatal day (P) 14. A small population of ganglion cells was immunoreactive at birth, but none were visible at P7. From P14 onwards, however, there was weak immunoreactivity in ganglion cells again, and strong staining in Müller glia. Retinas affected by neonatal optic tract lesions contained more immunoreactive ganglion cells at P4 than did controls, but by P14 there was a severe loss of ganglion cells. These observations are consistent with APP being involved in retinal differentiation, including maturation of glia and neurones, synaptogenesis and possibly neuronal survival.

Development of abnormal lamination and binocular segregation in the retinotectal pathways of the rat

The uncrossed retinotectal pathway of pigmented rats originates from a small fraction of the retinal ganglion cell population. This projection terminates deeply in discrete patches within the upper grey layers where crossed and uncrossed inputs overlap. However, after the experimental enlargement of the uncrossed pathway, the ipsilateral fibers are also found in a superficial tier of the upper grey layers where binocular inputs segregate [36]. We studied the development of the retinotectal projections in rats after the enlargement of the uncrossed pathway as a result of a contralateral (left) optic tract lesion (OTL) made at birth. Horseradish peroxidase (HRP) was used as an anterograde tracer. An abnormal uncrossed projection from the right eye to the collicular surface appeared at postnatal day 3 (P3). Between P5 and P10, this projection developed the bilaminar pattern seen in similarly operated adults. The laminar arrangement of the aberrant terminal fields did not change significantly after an ipsilateral visual cortex ablation on the day of birth. Despite the early development of the aberrant uncrossed pathway, binocular segregation was incipient at P10. At P14, 46% of the operated rats presented gaps in the terminal labeling at the tectal surface. This figure increased to 55.5% at 6 weeks, a proportion still smaller than in adult animals of the same group (69%). Eyelid suture had no effect on segregation. This projection remains plastic for at least 3 weeks, since the removal of the ipsilateral input at either P14 or P21 resulted in the absence of gaps in the contralateral projection. We conclude that the laminar selection of retinotectal projections depends on binocular interactions and that the abnormal segregation of retinal inputs to the superior colliculus has an unusually protracted development which can be reversed long after the previously defined critical period in this system.

Bcl-2 proto-oncogene protein immunoreactivity in normally developing and axotomised rat retinas

This study demonstrates immunocytochemically that bcl-2 (B cell lymphoma) proto-oncogene protein is expressed by two populations of retinal cells, one in the inner retina representing developing neurons or glia at perinatal stages, and the other identified as Müller cells from postnatal day 10 onward. An increase in bcl-2 immunoreactivity was detected in the Müller cell processes after neonatal optic tract lesion or optic nerve transection in adult rats. These results suggest that bcl-2 protein may function on cellular differentiation, maturation and homeostasis in the inner retina, and its expression by Müller cells can be up-regulated by the degeneration of ganglion cells.

Protein gene product 9.5-immunoreactive retinal neurons in normal developing rats and rats with optic nerve or tract lesion

The present immunocytochemical study indicates that protein gene product 9.5 (PGP 9.5) in the rat retina first appears in a population of neurons in the inner and central part of the neuroblast layer at embryonic day (E) 14. Presumptive horizontal cells which are PGP 9.5 positive were observed at E17. At birth, cells in the inner nuclear layer and ganglion cell layer (GCL) as well as the inner plexiform layer (IPL) were positive. Further differentiation, particularly the appearance and the formation of immunoreactive sublaminae in the IPL, was observed in the first 2 postnatal weeks. This pattern reached adult levels by postnatal day 14. In rats with unilateral neonatal optic tract lesion or optic nerve transection as young adults, 43–45% of the immunoreactive cells were lost in the GCL. However, only minor changes were detected in the IPL, suggesting that amacrine cells contribute mainly to the PGP 9.5 immunoreactivity in this laminar zone of the retina.

Regrowth of axons within Schwann cell-filled polycarbonate tubes implanted into the damaged optic tract and cerebral cortex of rats

The efficacy of Schwann cell-filled polycarbonate tubes as a bridging substrate for regrowing axons following lesions of the rat optic tract or cerebral cortex has been assessed after short (11-31 days) or long (82-119 days) survival times. Tubes were impregnated with Iaminin and poly-L-lysine, soaked in basic FGF and filled with Schwann cells. They were implanted into optic tract lesions in 34 rats aged 15-21 days and into cortical lesion cavities in 3 adult rats. Gelfoam soaked in basic FGF and Schwann cell conditioned medium was placed over the tubes. In one group of rats, axon regrowth into implants was assessed using neurofilament antibody RT97; antibodies to proteolipid protein, Po, Iaminin, the low-affinity nerve growth factor receptor (NGFr), S-100 and EDI were also used to study myelination and the cellular content of the tubes. In a second group of rats, anterograde tracing techniques were used to specifically identify host retinal axons within the implanted polymers. After long survival times, the relationships between regrown axons and cells inside the tubes were also examined ultrastructurally. In all implants examined immunohistochemically at short survival times, large numbers of RT97+ axons were found throughout the tubes, usually in association with Iaminin+, NGFr+ Schwann cells. At longer survival times, viable Schwann cells were still present, but tubes contained fewer axons and less cellular material. This material often formed a cord (200-250 μm thick) which extended the length of the implant. In the second group of rats, labelled retinal axons were found in 11 of the 16 implants that were attached to the dLGN. Axons regrew up to 1 mm but did not reach the distal (tectal) end of the implants. Interestingly, there was no evidence of myelinogenesis by either implanted Schwann cells or by host-derived oligodendroglia which had migrated into the tubes. Oligodendroglia were usually encircled by processes, many of which originated from Schwann cells, suggesting that the grafted cells may have been involved in isolating the central glia. The data show that Schwann cell-filled polycarbonate tubes provided a favourable milieu for axonal regeneration in the short term; however over time there was a decrease in the cellular and fibre content of the tubes. After intracranial implantation, an additional supporting matrix inside the polycarbonate tubes may aid in providing an environment conducive to the long term maintenance of regenerated retinal and other axons.

Dendritic competition in the developing retina: ganglion cell density gradients and laterally displaced dendrites.

Dendrites of retinal ganglion cells (RGCs) tend to be distributed preferentially toward areas of reduced RGC density. This, however, does not occur in the retina of normal pigmented rats, in which it has been suggested that the centro-peripheral gradient of RGC density is too shallow to provide directional guidance to growing dendrites. In this study, laterally displaced dendrites of RGCs retrogradely labeled with horseradish peroxidase were related to cell density gradients induced experimentally in the rat retina. Neonatal unilateral lesions of the optic tract produced retrograde degeneration of contralaterally projecting RGCs, but spared ipsilaterally projecting neurons in the same retina. These lesions created an anomalous temporal to nasal gradient of cell density across the decussation line, opposite to the nasal to temporal gradient found along the same axis in either normal rats or rats that had the contralateral eye removed at birth. RGCs in rats that received optic tract lesions had their dendrites displaced laterally toward the depleted nasal retina, while in either normal or enucleated rats there was no naso-temporal asymmetry. The lateral displacement affected both primary dendrites and higher-order branches. However, the gradient of cell density after optic tract lesions was less steep than the gradient in either normal or enucleated rats. To test for the presence of steeper gradients at early stages of development, RGC density gradients were also examined at postnatal day 5 (P5). In normal rats, the RGCs were homogeneously distributed throughout the retina, while rats given optic tract lesions at birth already showed a temporo-nasal density gradient at P5. Still, this anomalous gradient was less steep than that found in normal adults. It is concluded that the time course, rather than the steepness of the RGC density gradient, is the major determinant of the lateral displacement of dendritic arbors with respect to the soma in developing RGCs. The data are consistent with the idea that the overall shape of dendritic arbors depends in part on dendritic competition during retinal development.