Beyond the neckless phenotype: influence of reduced retinoic acid signaling on motor neuron development in the zebrafish hindbrain.

Retinoic acid (RA) has been identified as a key signal involved in the posteriorization of vertebrate neural ectoderm. The main biosynthetic enzyme responsible for RA signaling in the hindbrain and spinal cord is Raldh2. However, neckless/raldh2-mutant (nls) zebrafish exhibit only mild degrees of anteriorization in the neural ectoderm, compared to full vitamin A deficiency in amniotes and the Raldh2-/- mouse. Here we investigated the role of RA during neuronal development in the zebrafish hindbrain and anterior spinal cord using DEAB, an inhibitor of retinaldehyde dehydrogenases. We show that the nls hindbrain and spinal cord are not fully devoid of RA, since blocking Raldh-mediated RA signaling leads to a more severe hindbrain phenotype than in nls. The anteroposterior distribution of branchiomotor neurons in the facial and more posterior nuclei depends on full RA signaling throughout early and late gastrula stages. In contrast, inhibition of RA synthesis after gastrulation reduces the number of branchiomotor neurons in the vagal nucleus, but has no effect on anteroposterior cell fates. In addition, blockage of RA-mediated signaling not only interferes with the differentiation of branchiomotor neurons and their axons in the hindbrain, but also affects the development of the posterior lateral line nerve.

Dual function of polysialic acid during zebrafish central nervous system development.

Polysialic acid (PSA), a carbohydrate epitope attached to the neural cell adhesion molecule, serves as a modulator of axonal interactions during vertebrate nervous system development. We have used PSA-specific antibodies and whole-mount immunocytochemistry to describe the spatiotemporal expression pattern of PSA during zebrafish central nervous system development. PSA is transiently expressed on all cell bodies and, except for the posterior commissure, it is not found on axons. Floorplate cells in the spinal cord and hindbrain strongly express PSA throughout development. Enzymatic removal of PSA leads to a defasciculated growth pattern of the posterior commissure and also affects distinct subsets of commissural axons in the hindbrain, which fail to cross the midline. Whereas the disordered growth pattern of hindbrain commissures produced by PSA-removal could be mimicked by injections of soluble PSA, the growth of axons in the posterior commissure was unaffected by such treatment. These results suggest that there are distinct mechanisms for PSA action during axon growth and pathfinding in the developing zebrafish CNS.

Ectopic expression of the neural cell adhesion molecule L1 in astrocytes leads to changes in the development of the corticospinal tract.

The cell recognition molecule L1, of the immunoglobulin superfamily, participates in the formation of the nervous system and has been shown to enhance cell migration and neurite outgrowth in vitro. To test whether ectopic expression of L1 would influence axonal outgrowth in vivo, we studied the development of the corticospinal tract in transgenic mice expressing L1 in astrocytes under the control of the GFAP-promoter. Corticospinal axons innervate their targets by extending collateral branches interstitially along the axon shaft following a precise spatio-temporal pattern. Using DiI as an anterograde tracer, we found that in the transgenic animals, corticospinal axons appear to be defasciculated, reach their targets sooner and form collateral branches innervating the basilar pons at earlier developmental stages and more diffusely than in wild type littermates. Collateral branches in the transgenic mice did not start out as distinct rostral and caudal sets, but they branched from the axon segments in a continuous rostrocaudal direction across the entire region of the corticospinal tract overlying the basilar pons. The ectopic branches are transient and no longer present at postnatal day 22. The earlier outgrowth and altered branching pattern of corticospinal axons in the transgenics is accompanied by an earlier differentiation of astrocytes. Taken together, our observations provide evidence that the ectopic expression of L1 on astrocytes causes an earlier differentiation of these cells, results in faster progression of corticospinal axons and influences the branching pattern of corticospinal axons innervating the basilar pons.

Function of Neurolin (DM-GRASP/SC-1) in guidance of motor axons during zebrafish development.

Neurolin (zf DM-GRASP), a transmembrane protein with five extracellular immunoglobulin domains, is expressed by secondary but not primary motoneurons during zebrafish development. The spatiotemporally restricted expression pattern suggests that Neurolin plays a role in motor axon growth and guidance. To test this hypothesis, we injected zebrafish embryos with function-blocking Neurolin antibodies. In injected embryos, secondary motor axons form a broadened bundle along the common path and ectopic branches leave the common path at right angles. Moreover, the formation of the ventral and the rostral projection of secondary motor axons is inhibited during the second day of development. Pathfinding errors, resulting in secondary motor axons growing through ectopic regions of the somites, occur along the common path and in the dorsal and rostral projection. Our data are compatible with the view that Neurolin is involved in the recognition of guidance cues and acts as a receptor on secondary motor axons. Consistent with this idea is the binding pattern of a soluble Neurolin-Fc construct showing that putative ligands are distributed along the common path, the ventral projection, and in the area where the rostral projection develops.

The retinal axon's pathfinding to the optic disk.

Retinal ganglion cell (RGC) axons travel in radial routes unerringly toward the optic disk, their first intermediate target in the center of the eye. The path of the RGC growth cone is restricted to a narrow zone subjacent to the endfeet of Müller glial cells and the vitreal basal lamina. The present survey indicates that RGC growth cones are guided by many molecular cues along their pathway which are recognized by receptors on their surface. Growth-promoting molecules on Müller glial endfeet and in the basal lamina assist growth cones in maintaining contact with these elements. The repellant character of deeper retinal laminae discourages them from escaping the RGC axon layer. Cell adhesion/recognition proteins enable growth cones to fasciculate with preformed axons in their vicinity. It is still unclear whether the optic disk emits long range guidance components which enable the growth cones to steer toward it. Recent evidence in fish indicates the existence of an axonal receptor (neurolin) for a guidance component of unknown identity. Receptor blockade causes RGC axons to course in aberrant routes before they reach the disk. At the disk, axons receive signals to exit the retina. Contact with netrin-1 at the optic disk/nerve head encourages growth cones to turn into the nerve. This response requires the axonal netrin receptor DCC, laminin-1, beta-integrin and most likely the UNC5H netrin receptors which convert the growth encouraging signal into a repulsive one which drives growth cones into the nerve.

Optical measurements of invasive forces exerted by appressoria of a plant pathogenic fungus

Many plant pathogenic fungi, such as the cereal pathogen Colletotrichum graminicola, differentiate highly specialized infection structures called appressoria, which send a penetration peg into the underlying plant cell. Appressoria have been shown to generate enormous turgor pressure, but direct evidence for mechanical infection of plants by fungi is lacking. A microscopic method was developed that uses elastic optical waveguides to visualize and measure forces locally exerted by single appressoria. By this method, the force exerted by appressoria of C. graminicola was found to be about 17 micronewtons.

Imaging of cell/substrate contacts of living cells with surface plasmon resonance microscopy.

We have developed a new method for observing cell/substrate contacts of living cells in culture based on the optical excitation of surface plasmons. Surface plasmons are quanta of an electromagnetic wave that travel along the interface between a metal and a dielectric layer. The evanescent field associated with this excitation decays exponentially perpendicular to the interface, on the order of some hundreds of nanometers. Cells were cultured on an aluminum-coated glass prism and illuminated from below with a laser beam. Because the cells interfere with the evanescent field, the intensity of the reflected light, which is projected onto a camera chip, correlates with the cell/substrate distance. Contacts between the cell membrane and the substrate can thus be visualized at high contrast with a vertical resolution in the nanometer range. The lateral resolution along the propagation direction of surface plasmons is given by their lateral momentum, whereas perpendicular to it, the resolution is determined by the optical diffraction limit. For quantitative analysis of cell/substrate distances, cells were imaged at various angles of incidence to obtain locally resolved resonance curves. By comparing our experimental data with theoretical surface plasmon curves we obtained a cell/substrate distance of 160 +/- 10 nm for most parts of the cells. Peripheral lamellipodia, in contrast, formed contacts with a cell substrate/distance of 25 +/- 10 nm.

Neurolin Ig domain 2 participates in retinal axon guidance and Ig domains 1 and 3 in fasciculation.

The optic disk-directed growth of retinal ganglion cell axons is markedly disturbed in the presence of polyclonal antineurolin antibodies, which mildly affect fasciculation (Ott, H., M. Bastmeyer, and C.A.O. Stuermer, 1998. J. Neurosci. 18:3363-3372). New monoclonal antibodies (mAbs) against goldfish neurolin, an immunoglobulin (Ig) superfamily cell adhesion/recognition molecule with five Ig domains, were generated to assign function (guidance versus fasciculation) to specific Ig domains. By their ability or failure to recognize Chinese hamster ovary cells expressing recombinant neurolin with deletions of defined Ig domains, mAbs were identified as being directed against Ig domains 1, 2, or 3, respectively. Repeated intraocular injections of a mAb against Ig domain 2 disturb the disk-directed growth: axons grow in aberrant routes and fail to reach the optic disk, but remain fasciculated. mAbs against Ig domains 1 and 3 disturb the formation of tight fascicles. mAb against Ig domain 2 significantly increases the incidence of growth cone departure from the disk-oriented fascicle track, while mAbs against Ig domains 1 and 3 do not. This was demonstrated by time-lapse videorecording of labeled growth cones. Thus, Ig domain 2 of neurolin is apparently essential for growth cone guidance towards the disk, presumably by being part of a receptor (or complex) for an axon guidance component.

E587 antigen is upregulated by goldfish oligodendrocytes after optic nerve lesion and supports retinal axon regeneration.

The properties of glial cells in lesioned nerves contribute quite substantially to success or failure of axon regeneration in the CNS. Goldfish retinal axons regenerate after optic nerve lesion (ONS) and express the L1-like cell adhesion protein E587 antigen on their surfaces. Goldfish oligodendrocytes in vitro also produce E587 antigen and promote growth of both fish and rat retinal axons. To determine whether glial cells in vivo synthesize E587 antigen, in situ hybridizations with E587 antisense cRNA probes and light- and electron microscopic E587 immunostainings were carried out. After lesion, the goldfish optic nerve/tract contained glial cells expressing E587 mRNA, which were few in number at 6 days after ONS, increased over the following week and declined in number thereafter. Also, E587-immunopositive elongated cells with ultrastructural characteristics of oligodendrocytes were found. Thus, glial cells synthesize E587 antigen in spatiotemporal correlation with retinal axon regeneration. To determine the functional contribution of E587 antigen, axon-oligodendrocyte interactions were monitored in co-culture assays in the presence of Fab fragments of a polyclonal E587 antiserum. E587 Fabs in axon-glia co-cultures prevented the normal tight adhesion of goldfish retinal growth cones to oligodendrocytes and blocked the preferential growth of fish and rat retinal axons on the oligodendrocyte surfaces. The ability of glia in the goldfish visual pathway to upregulate the expression of E587 antigen and the growth supportive effect of oligodendrocyte-associated E587 antigen in vitro suggests that this L1-like adhesion protein promotes retinal axon regeneration in the goldfish CNS.

Neurolin, the goldfish homolog of DM-GRASP, is involved in retinal axon pathfinding to the optic disk.

Young axons of new retinal ganglion cells (RGCs) in the continuously growing goldfish retina fasciculate with one another and their immediate forerunners on their path toward the optic disk and along the optic nerve. They express the immunoglobulin superfamily cell adhesion molecules (CAMs) neurolin (DM-GRASP) and the L1-like E587 antigen. Repeated injections of Fab fragments from polyclonal antisera against neurolin (neurolin Fabs) into the eye of 3. 4-cm-long and rapidly growing goldfish caused highly aberrant pathways of young RGC axon subfascicles in the dorsal retina. Many axons grew in circles and failed to reach the optic disk. In contrast, E587 Fabs, used in parallel experiments, disrupted the fascicles but did not interfere with the disk-directed growth. Neurolin Fabs also disturbed axonal fasciculation in vivo as well as in vitro but less severely than E587 Fabs. Coinjections of both Fabs increased defasciculation of the dorsal axons in both aberrant and disk-directed routes. They also disrupted the order of young RGC axons in the optic nerve more severely than E587 Fabs alone. This demonstrates that the development of tight and orderly fascicles in the dorsal retina and in the optic nerve requires both E587 antigen and neurolin. More importantly, our results suggest an involvement of neurolin in RGC axonal guidance from the retinal periphery to the optic disk. Because disrupted fascicles and errant axon routes were found only in the dorsal retinal half, a cooperation with so-called positional markers may be conceived.

Collateral branch formation related to cellular structures in the axon tract during corticopontine target recognition.

The corticopontine projection develops exclusively by collateral branches that form along the length of corticospinal axons days after they have passed their hindbrain target, the basilar pons. In vitro evidence suggests that the basilar pons releases a diffusible activity that initiates and directs the growth of collateral branches. This study investigates whether contact-dependent mechanisms may also influence the formation of collateral branches. By using immunocytochemistry, electron microscopy, and neuronal tracing techniques, we examined the region of the axon tract, the cerebral peduncle, overlying the basilar pons for cellular structures that correlate spatially and temporally with collateral branch formation. We found that radial glia are excluded from the tract. Oligodendrocyte precursors are found only at low density. Although mature astrocytes are absent, immature astrocytes are present throughout the tract. However, our evidence does not suggest a direct role for glial cell types in collateral branch formation. In contrast, dendrites of basilar pontine neurons are transiently present in the tract during the time of collateral branch formation. Although collateral branches are observed in regions of the tract devoid of dendrites, the orientation and location of most collateral branches correlates at the light microscopic level with dendrites. Electron microscopy reveals sites of increased collateral branch formation near neuronal cell bodies or dendrites. However, cell processes, whether dendritic or otherwise, are rarely found in direct contact with collateral branch points. A common and unexpected feature is the bundles of corticopontine collateral branches, oriented transversely to their parent corticospinal axons and directed across the tract to the basilar pons. Dendrites were often apposed to or embedded within the transverse bundles. These findings suggest that dendrites are not essential for collateral branch formation but that they may enhance this process and define discrete preferred locations for collateral branch initiation and elongation within the cerebral peduncle.

Expression of an L1-related cell adhesion molecule on developing CNS fiber tracts in zebrafish and its functional contribution to axon fasciculation.

E587 antigen, an L1-related cell adhesion molecule, is expressed by growing axons and has previously been shown to enhance axon growth and to mediate fasciculation of axons from newborn retinal ganglion cells in goldfish. In zebrafish, the monoclonal antibody E17 against E587 antigen stains all axons in the primary tracts and commissures from 17 h postfertilization (pf) onward and axons which are added subsequently to this scaffold. Moreover, Fab fragments of an E587 antiserum (E587 Fabs) injected into the ventricle of 30-h pf zebrafish embryos caused a marked defasciculation of distinct axon bundles in the posterior commissure, in hindbrain commissures, and in longitudinal tracts of the hindbrain, where they also caused increased crossings between fascicles. The regulated expression of E587 antigen by all developing axons and the effects caused by E587 Fabs show that E587 antigen contributes to the formation of tight and orderly fascicles in the developing CNS.

Spatially restricted increase in polysialic acid enhances corticospinal axon branching related to target recognition and innervation.

The polysialic acid (PSA) modification of the neural cell adhesion molecule (NCAM) has been shown to alter the responses of developing axons to their environment. We have studied the potential role of PSA in regulating the innervation of the spinal cord by corticospinal axons, which occurs by a delayed formation of collateral branches from the parent axons. Developmental changes in the distribution of PSA were examined immuno-histochemically using light and electron microscopy. Whereas NCAM is distributed along the entire pathway of rat corticospinal axons as they grow from the cortex to the spinal cord, PSA-modified NCAM does not become evident until later. When PSA becomes evident, it is restricted to the distal segment of these axons from the caudal hindbrain through the spinal cord. The increase in PSA on corticospinal axons coincides with the time that they begin to form collateral branches in the spinal cord. This unique spatiotemporal distribution of PSA suggests its involvement in corticospinal axon branching. To test this hypothesis, PSA was selectively removed by an in vivo injection of endoneuraminidase N. This treatment did not seem to interfere with the pathfinding of corticospinal axons; however, PSA removal delayed the onset of collateral branching by corticospinal axons within the spinal cord and later diminished the magnitude of branching. These findings indicate a role for PSA in the regulation of interstitial axon branching, a crucial step in the process of target recognition and innervation by corticospinal axons.

Optic nerve glia secrete a low-molecular-weight factor that stimulates retinal ganglion cells to regenerate axons in goldfish.

The ability of lower vertebrates to regenerate an injured optic nerve has been widely studied as a model for understanding neural development and plasticity. We have recently shown that, in goldfish, the optic nerve contains two molecules that stimulate retinal ganglion cells to regenerate their axons in culture: a low-molecular-weight factor that is active even at low concentrations (axogenesis factor-1) and a somewhat less active polypeptide of molecular weight 10,000-15,000 (axogenesis factor-2). Both are distinct from other molecules described previously in this system. The present study pursues the biological source and functional significance of axogenesis factor-1. Earlier studies have shown that cultured goldfish glia provide a highly favorable environment for fish or rat retinal ganglion cells to extend axons. We report that the glia in these cultures secrete high levels of a factor that is identical to axogenesis factor-1 in its chromatographic properties and biological activity, along with a larger molecule that may coincide with axogenesis factor-2. Axogenesis factor-1 derived from either goldfish glial cultures or optic nerve fragments is a hydrophilic molecule with an estimated molecular weight of 700-800. Prior studies have reported that goldfish retinal fragments, when explanted in organ culture, only extend axons if the ganglion cells had been "primed" to begin regenerating in vivo for one to two weeks. However, axogenesis factor-1 caused the same degree of outgrowth irrespective of whether ganglion cells had been induced to regenerate new axons in vivo. Moreover, ganglion cells primed to begin regenerating in vivo continued to extend axons in culture only when axogenesis factor-1 was present. In summary, this study shows that glial cells of the goldfish optic nerve secrete a low-molecular-weight factor that initiates axonal regeneration from retinal ganglion cells.

Dynamics of target recognition by interstitial axon branching along developing cortical axons.

Corticospinal axons innervate their midbrain, hindbrain, and spinal targets by extending collateral branches interstitially along their length. To establish that the axon shaft rather than the axonal growth cone is responsible for target recognition in this system, and to characterize the dynamics of interstitial branch formation, we have studied this process in an in vivo-like setting using slice cultures from neonatal mice containing the entire pathway of corticospinal axons. Corticospinal axons labeled with the dye 1,1'-dioctodecyl-3,3,3',3'-tetramethylindocarbocyanine perchlorate (or Dil) were imaged using time-lapse video microscopy of their pathway overlying the basilar pons, their major hindbrain target. The axon shaft millimeters behind the growth cone exhibits several dynamic behaviors, including the de novo formation of varicosities and filopodia-like extensions, and a behavior that we term "pulsation," which is characterized by a variable thickening and thining of short segments of the axon. An individual axon can have multiple sites of branching activity, with many of the branches being transient. These dynamic behaviors occur along the portion of the axon shaft overlying the basilar pons, but not just caudal to it. Once the collaterals extend into the pontine neuropil, they branch further in the neuropil, while the parent axon becomes quiescent. Thus, the branching activity is spatially restricted to specific portions of the axon, as well as temporally restricted to a relatively brief time window. These findings provide definitive evidence that collateral branches form de novo along corticospinal axons and establish that the process of target recognition in this system is a property of the axon shaft rather than the leading growth cone.

Reevaluation of the growth-permissive substrate properties of goldfish optic nerve myelin and myelin proteins.

To determine whether optic nerve myelin of goldfish carries mammalian-like neurite growth inhibitory proteins which can be neutralized by the antibody IN-1, myelin fractions of fish optic nerves were used as substrates for fish retinal ganglion cell axons and rat dorsal root ganglia (DRG). Axonal growth was monitored and compared with that of IN-1 treated preparations. Growth of fish retinal axons and rat DRG neurites was substantial on goldfish optic nerve myelin and no improvement was observed with IN-1. In contrast, rat CNS myelin allowed only poor growth, and number of axons and length of DRG neurites increased significantly with IN-1. In addition, proteins of fish optic nerve myelin and bovine CNS myelin were extracted, reconstituted in liposomes and applied to growth cones. When goldfish myelin proteins in liposomes were seeded onto growth cones, 77% of fish and 89% of rat DRG growth cones continued to elongate, and the proportion of elongating fish growth cones (80%) did not significantly change when liposomes were pretreated with IN-1. But 73% of fish and 93% of rat growth cones collapsed with liposomes containing proteins from bovine CNS myelin. Upon IN-1 treatment, only 24% of fish growth cones collapsed. Thus, axon growth in vitro indicates that goldfish optic nerves, which permit successful axon regeneration in vivo, lack mammalian-like neurite growth inhibitors which are neutralized by IN-1.

Fish E587 glycoprotein, a member of the L1 family of cell adhesion molecules, participates in axonal fasciculation and the age-related order of ganglion cell axons in the goldfish retina.

Axons derived from young ganglion cells in the periphery of the retinae of larval and adult goldfish are known to fasciculate with one another and their immediate forerunners, creating the typical age-related order in the retinotectal pathway. Young axons express the E587 antigen, a member of the L1 family of cell adhesion molecules. Repeated injections of Fab fragments from a polyclonal E587 antiserum (E587 Fabs) into the eye of 3.4 cm goldfish disrupted the orderly fascicle pattern of RGC axons in the retina which was preserved in controls. Instead of bundling tightly, RGC axons crossed one another, grew between fascicles and arrived at the optic disk in a broadened front. When added to RGC axons growing in vitro, E587 Fabs neutralized the preference of growth cones to elongate on lanes of E587 protein, caused defasciculation of axons which normally prefer to grow along each other when explanted on polylysine, and prevented clustering of E587 antigen at axon-axon contact sites. Monoclonal E587 antibody disturbed axonal fasciculation moderately but led to a 30% reduction in growth velocities when axons tracked other axons. Therefore we conclude that E587 antigen mediates axonal recognition, selective fasciculation and the creation of the age-related order in the fish retina.

Astrocytes from adult rat optic nerves are nonpermissive for regenerating retinal ganglion cell axons.

We have directly compared the abilities of astrocytes from newborn and adult rats to support or inhibit the growth of regenerating axons in vitro. Astrocytes prepared from newborn rats were able to promote retinal ganglion cell (RGC) axon growth from embryonic and adult rat and from adult fish retinal explants. Retinal axons from E16 rat retinae grew significantly faster on astrocytes from neonatal rats than those from E18 or adult rat retinae with growth rates comparable to RGC axons from adult fish retinae. RGC regeneration from adult rat retinae was almost completely inhibited on adult rat optic nerve astrocytes. Only axons from adult fish retinae were able to extend onto monolayers from these reactive astrocytes, although their growth rates were significantly reduced. We conclude that the failure of mammalian RGC axons to regrow within the lesioned optic nerve environment is, at least in part, due to nonpermissive aspects of adult "reactive" optic nerve astrocytes. However, the cell intrinsic growth potential of RGCs also appears to influence their ability to extend axons on cellular substrates.

Similarities and differences between fish oligodendrocytes and Schwann cells in vitro.

In light of the striking differences between oligodendrocytes of the optic nerve/tract of adult goldfish and their mammalian counterparts, a further characterization of goldfish oligodendrocytes was performed. A comparison with Schwann cells was included because fish optic nerve/tract-derived oligodendrocytes bear remarkable similarities to this type of glial cell. Fish optic nerve/tract-derived oligodendrocytes that had differentiated into 04 and 6D2-positive cells and thus expressed early myelin marker molecules were found to incorporate BrdU and to divide in vitro over a period of weeks. For the induction of more advanced markers of myelinogenesis such as the CNS-specific myelin protein 36K, oligodendrocytes required extensive contact with axons. Other agents, such as fetal calf or carp serum, substrate components, or forscolin failed, however, to induce 36K expression. 04/6D2-positive oligodendrocytes could be distinguished from fish 6D2-positive Schwann cells derived from cranial nerves by their antigenic phenotype: Schwann cells but not oligodendrocytes exhibited the low affinity NGF receptor. While both cell types carry the cell adhesion molecules NCAM, E 587 antigen, and the L2/HNK-1 epitope, only Schwann cells possess a further adhesion molecule, Neurolin.

Fish optic nerve oligodendrocytes support axonal regeneration of fish and mammalian retinal ganglion cells.

Segments from adult fish and rat retinae were explanted on myelin-marker expressing oligodendrocytes derived from the regenerating goldfish optic nerve. Fish axons grew in high density and even rat retinal axons regenerated to considerable length on the surface of the fish oligodendrocytes, suggesting that this type of fish glia has axon-growth promoting surface components that exert their influence across species boundaries. One interesting surface component of the fish oligodendrocytes as demonstrated here is the E 587 antigen, which is related to the L1 family of cell adhesion molecules. In long term cocultures of oligodendrocytes and retinal axons, the fish glial cells were found to enwrap rat axons. This suggests that the oligodendrocytes of the regenerating goldfish optic nerve/tract may, despite striking differences, represent the equivalent to mammalian optic nerve oligodendrocytes.