Reorganization of retinotopic cortical maps in adult mammals after lesions of the retina.

The organization of the visual cortex has been considered to be highly stable in adult mammals. However, 5 degrees to 10 degrees lesions of the retina in the contralateral eye markedly altered the systematic representations of the retina in primary and secondary visual cortex when matched inputs from the ipsilateral eye were also removed. Cortical neurons that normally have receptive fields in the lesioned region of the retina acquired new receptive fields in portions of the retina surrounding the lesions. The capacity for such changes may be important for normal adjustments of sensory systems to environmental contingencies and for recoveries from brain damage.

Brain stem topography of vagus nerve to the greater curvature of the stomach.

If preganglionic vagus nerve fibers enter the stomach via all of its neurovascular bundles, then proximal gastric vagotomy that divides only the bundles along the lesser curvature of the stomach neglects a potential source of innervation to the parietal cells. To determine whether or not these bundles contained preganglionic efferent vagal nerve fibers, horseradish peroxidase was applied to the central cut end of selected neurovascular bundles along the greater curvature of the stomach in rats and ferrets. Cells in the dorsal motor nucleus of the vagus (dmnX) of the rat were labeled after horseradish peroxidase applications to the right gastroepiploic, the splenic, and the short gastric bundles. The ferrets had horseradish peroxidase applied to the right gastroepiploic bundle and they also had cellular labeling of the dmnX. The labeling in cells of the dorsal motor nucleus of the vagus had a distinct topographic, rostrocaudal distribution in both species, and was maximal in the vicinity of the obex. Cells of the bilateral dmnX were labeled after horseradish peroxidase applications at all bundles. This study showed (1) that the bundles along the greater curvature of the stomach contained preganglionic efferent vagus nerve fibers, (2) that the cells of origin of these fibers were represented in the localized rostrocaudal position of the dmnX, and (3) that these fibers had their origins in the bilateral dmnX. Such nerve fibers may account for incomplete vagal denervation of the parietal cells after proximal gastric vagotomy.

Pyloroplasty divides vagus nerve fibers to the greater curvature of the stomach. An axonal tracing study.

Recent studies have demonstrated the location in the dorsal motor nucleus of the vagus nerve (dmnX) of nerve cells that project preganglionic efferent vagus nerve fibers to the greater curvature of the stomach. Although it is clear that these fibers are contained within the vagus nerve trunks, the intra-abdominal pathways of these fibers are unknown. When a neurotracer was applied to the right gastroepiploic pedicle, nerve cells in the bilateral dmnX were labeled. If a preliminary anterior or posterior pyloroplasty was performed before the application of the neurotracer, cellular labeling was seen on the right or left side of the dmnX, respectively. Furthermore, division of the anterior Latarjet nerve eliminated labeling in cells of the left dmnX. This study demonstrates that the preganglionic vagus nerve fibers within the right gastroepiploic pedicle traverse an intramural course across the pylorus and are contained in the Latarjet nerve.

Cell birthdays and rate of differentiation of ganglion and horizontal cells of the developing cat's retina.

Tritiated thymidine autoradiography experiments demonstrated that three cell classes are produced by ventricular cells during the first phase of neurogenesis: retinal ganglion cells, A-type horizontal cells, and cone photoreceptors. Light microscopy and scanning electron microscopy were used to study the migration and morphological differentiation of these three cell classes. The patterns of postmitotic migration are of interest because these three classes of cells are found in three different layers of the adult retina. Cones retain their position at the outer limiting membrane (OLM) throughout life and do not migrate. Ganglion cells migrate immediately to the proximal (vitread) layer of the retina and begin to differentiate. In contrast, A-type horizontal cells migrated away from the OLM within 10-14 days after their final mitosis but were morphologically relatively undifferentiated at that time. Subsequent differentiation of the A-type horizontal cell is also protracted; dendrites are not observed until approximately 3 weeks after the final mitosis. These observations suggest that there are several interacting mechanisms involved in neurogenesis: a sequence that produces a specific cohort of committed cells at a specific time, the subsequent migration of postmitotic neuroblasts to an appropriate position in the retina, and a spatial gradient of differentiation increasing from distal to proximal layers. While this distribution of differentiated cells early in fetal development is striking, the existence of underlying time-dependent processes that might cause this apparent spatial phenomenon cannot be eliminated.

Labeling of nerve cells in the dorsal motor nucleus of the vagus of rats by retrograde transport of Fluoro-Gold.

Nerve cells in the dorsal motor nucleus of the vagus (dmnX) were identified by retrograde axonal transport after injections of a fluorescent tracer, Fluoro-Gold, into the anterior gastric wall. The intramural injection resulted in labeling of cells in the medial half of the left dmnX. These observations were contrasted with the diffuse (mediolateral and rostrocaudal) and bilateral distribution of labeled cells after Fluoro-Gold solution was dripped onto the stomach. In comparison with other neurotracers, the advantages of Fluoro-Gold are that (1) it can be visualized without the chemical reaction with chromogen, thereby allowing better reproducibility, and (2) it does not fade up to one year.

Preganglionic vagus nerve fibers also enter the greater curvature of the stomach in rats and ferrets.

The efferent gastric vagus nerve fibers appear to enter the stomach by several routes. For example, the rate of gastric acid secretion is directly affected by the nerves of the greater curvature of the stomach. Specifically, acid secretion decreases abruptly after division of the gastroepiploic nerve(s). To determine whether efferent vagus nerve fibers are contained in the gastroepiploic nerve(s), horseradish peroxidase, a protein that undergoes retrograde axonal transport, was applied to these nerves; the brainstem locus of the nuclei of the vagus nerves was examined 2 days later. Typical peroxidase labeling was observed in the dorsal motor nucleus of the vagus nerve in 5 of 6 rats and 3 of 3 ferrets; the hypothesis that efferent vagus nerves enter the greater curvature of the stomach was thus supported in two vertebrate species. These previously unrecognized nerves should be considered in the interpretation of experimental and clinical phenomena.

The topography of ganglion cell production in the cat's retina.

The ganglion cells of the cat's retina form several classes distinguishable in terms of soma size, axon diameter, dendritic morphology, physiological properties, and central connections. Labeling with [3H]thymidine shows that the ganglion cells which survive in the adult are produced as several temporally shifted, overlapping waves: medium-sized cells are produced before large cells, whereas the smallest ganglion cells are produced throughout the period of ganglion cell generation (Walsh, C., E. H. Polley, T. L. Hickey, and R. W. Guillery (1983) Nature 302: 611-614). Large cells and medium-sized cells show the same distinctive pattern of production, forming rough spirals around the area centralis. The oldest cells tend to lie superior and nasal to the area centralis, whereas cells in the inferior nasal retina and inferior temporal retina are, in general, progressively younger. Within each retinal quadrant, cells nearer the area centralis tend to be older than cells in the periphery, but there is substantial overlap. The retinal raphe divides the superior temporal quadrant into two zones with different patterns of cell addition. Superior temporal retina near the vertical meridian adds cells only slightly later than superior nasal retina, whereas superior temporal retina near the horizontal meridian adds cells very late, contemporaneously with inferior temporal retina. The broader wave of production of smaller ganglion cells seems to follow this same spiral pattern at its beginning and end. The presence of the area centralis as a nodal point about which ganglion cell production in the retinal quadrants pivots suggests that the area centralis is already an important retinal landmark even at the earliest stages of retinal development. This sequence of ganglion cell production differs markedly from that seen in the retinae of nonmammalian vertebrates, where new ganglion cells are added as concentric rings to the retinal periphery, and also bears no simple relationship to the cat's retinal decussation line. However, it can be related in a straightforward manner to the organization of axons in the cat's optic tract, suggesting that the fiber order in the tract represents a grouping of fibers by age.

A technique for flat embedding and en face sectioning of the mammalian retina for autoradiography.

A method of tissue preparation is described which allows analysis of large areas of the mammalian retina in relatively few 2-5 micrometers sections, and which permits [3H]autoradiography with good resolution and low background. In contrast to standard methods of paraffin embedding and radial sectioning, we have embedded flattened retinas in glycol methacrylate resin and cut sections parallel to the plane of the retinal layers, i.e. en face. As previously shown (Sidman, 1970; Carter-Dawson and LaVail, 1979a, b), methacrylate embedding results in excellent cellular preservation and allows relatively thin sectioning appropriate for [3H]autoradiography and demonstration of cytological detail. Sectioning en face reduces the number of tissue sections required to survey large areas of the retina, and allows reconstruction of the topography of each retinal layer from serial sections. Thus, studies of retinal topography using [3H]autoradiography can now be more easily compared to studies of retinal topography using the whole-mount technique (see Stone, 1981).

Generation of cat retinal ganglion cells in relation to central pathways.

The ganglion cells of the cat retina form classes distinguishable in terms of perikaryal size, dendritic morphology and functional properties. Further, the axons differ in their diameters, patterns of chiasmatic crossing and in their central connections. Here we define, by 3H-thymidine autoradiography, the order of production of cells of each class and relate the order of the 'birthdates' to the known axonal pathways. The ganglion cell classes are produced in broad waves, which overlap as cells are produced first for central then for peripheral retina. Medium-sized cells are produced before the largest cells, and small ganglion cells are produced throughout the period of cell generation. This sequence of cell production relates to the orderly arrangement of axons in the optic tract, and can also be related to the rules of chiasmatic crossing observed for each ganglion cell class.

Studies of retinal representations within the cat's optic tract.

The manner in which each retina can be mapped onto a single cross section of the optic tract of the cat has been defined by neuroanatomical methods. It has been found that the contralateral nasal hemi-retina and both temporal hemi-retinae are represented in each tract by multiple, rough maps which partially overlap one another. All maps show the same general orientation, with area centralis represented dorsomedially, lower retina represented dorsolaterally, and upper retina represented ventromedially. The peripheral part of the horizontal meridian is represented ventrolaterally. Labeling all of the fibers from one eye by axonal degeneration or autoradiographic methods shows that the crossed map is displaced dorsally and medially relative to the uncrossed map, leaving a dorsomedial crescent of pure crossed fibers. Localized retinal lesions or injections of 3H-amino acid show the general orientation of the maps. Lesions within the dorsomedial pure crossed crescent show that fibers in this crescent arise from retinal areas close to the optic disc, near the site of the early fetal fissure. Localized injections of horseradish peroxidase into the optic tract show the relationships of the several maps in terms of the retinal distribution of retrogradely labeled retinal ganglion cells. They show that axons of large and small cells map ventrolaterally in the tract while intermediate sizes map dorsomedially. They confirm that the crossed map is displaced relative to the uncrossed maps. It is suggested that the optic tract develops by fibers taking a position in the tract in accordance with their time of arrival at the chiasm. The several maps are displaced because they develop sequentially and the optic tract can be read as a developmental record, the most dorsomedial axons being the oldest.

The arrangement of axons according to fiber diameter in the optic tract of the cat.

The topographic distribution of axons according to diameter classes in the optic tract of the cat has been studied by light and electron microscopic methods. The subdivision of the tract into a ventral and a dorsal sector, with the former containing the thickest fibers of the tract, has been confirmed. In addition, there is a ventrolateral marginal zone of the tract which contains many of the finest axons and there are gradients of fiber diameters definable for thick and intermediate axons. The gradients show the thinner fibers lying rostral to the thicker fibers. The major points demonstrable by fiber diameter measurements have been confirmed by counts of myelin lamellae. Comparison of fiber distributions in the optic tract with the distributions in the optic nerve on the one hand and with the geniculate terminations on the other shows that, in the region of the optic chiasm, retinofugal axons are sorted not only according to their crossed or uncrossed pathway but also according to their fiber diameter class and thus are presorted in terms of their final destinations.

A demonstration of several independent, partially overlapping, retinotopic maps in the optic tract of the cat.

We show here that the retina in the cat's optic tract by several separate, retino-topically organized maps. Fibers in different diameter classes form independent maps, and both hemi-retinae of the opposite side, as well as the temporal retina of the same side, are mapped in the tract. These several maps form rough topographic representations significantly out of register with each other.

An analysis of the retinal afferents to the cat's medial interlaminar nucleus and to its rostral thalamic extension, the "geniculate wing".

The retinal afferents to the medial interlaminar nucleus and to its rostro-dorsal extension at the edge of the pulvinar have been studied in cats by fiber degeneration and autoradiographic methods. Fiber degeneration following section of one optic nerve shows two distinct retinal inputs: One is coarse-fibered and goes to the medial interlaminar nucleus itself; the second, which is fine-fibered, goes to the rostral and medial borders of the medial interlaminar nucleus and continues into the pulvinar. The regions in receipt of these fine fibers have been called the "geniculate wing". The topography of retinal representations and the degree of binocular overlap within the medial interlaminar nucleus and the wing have been studied by combining intraocular injections of 3H proline with local lesions of the injected eye, or with removal of the non-injected eye. In the medial interlaminar nucleus three distinct laminae are recognizable and are particularly clearly shown in horizontal sections. Rostrally and medially, lamina 1 maps the contralateral nasal retina as a mirror reversal of lamina A. Posterior and lateral to this, lamina 2 maps the ipsilateral temporal retina as a mirror reversal of lamina A1. Lamina 3 lies closest to the optic tract and receives crossed afferents from the temporal retina. In this lamina, which is the smallest, vertical retinal dimensions map as in the other layers, but we were unable to determine the mapping of the horizontal dimensions. In the geniculate wing, as in the other geniculate layers, vertical lines of the visual field are mapped as corresponding vertical diencephalic dimensions; horizontal retinal dimensions are mapped as horizontal lines in the wing, with central retinal areas represented furthest from the optic tract. In the geniculate wing the contralateral nasal and ipsilateral temporal retina are mapped with considerable binocular overlap. The crossed temporal retina has no demonstrable representation in the wing.

Absence of sprouting by retinogeniculate axons after chronic focal lesions in the adult cat retina.

Focal lesions were placed in the retina of adult cats in order to denervate partially the laminae of the lateral geniculate nucleus (LGN). Retinogeniculate projections were assessed after survival times of from 5 days to 2 years by means of either reduced silver staining for degeneration or autoradiographic labelling. Filling of the lesion-denervated zones by 'sprouts' from the intact retinofugal fibers was not observed, even in the brains of animals with long-term lesions. It was concluded that the retinogeniculate projection in adult cat does not display any significant ability to sprout into denervated regions.

Experimental argon laser photocoagulation. II. Effects on the optic disc.

Eight monkeys and five humans were subjected to photocoagulation with single argon laser burns of varying intensity on the optic nerve head. Energies greater than 400 milliwatts, 0.2 second and 100 mu spot size consistently caused neural parenchymal damage, even in the absence of heat-generating pigment epithelium. Papillary burns in monkeys and humans created with powers less than 400 mW showed only small, focal submicroscopic areas of degeneration, and were therefore considered relatively safe except for the occasional involvement of critical foveal fibers in which scotomas of disproportionately greater severity than one would expect from the size of the lesion could result. The threshold for toxicity following peripapillary photocoagulation is lower, because the burned pigment epithelium can radiate heat into the adjacent optic nerve fibers. Normal capillaries within the optic nerve were seldom destroyed or occluded, even after high energy densities. In normal vessels, we have found it impossible to create a highly selective lesion without concurrently causing neural damage about the vessel.

Experimental argon laser photocoagulation. I. Effects on retinal nerve fiber layer.

Eight rhesus monkeys were photocoagulated with varying intensities of argon laser energy to determine the effects on the retinal nerve fiber layer. Photocoagulation of arterioles and venules always resulted in largely irreversible and permanent perivascular axonal destruction. Photocoagulation of areas free of major vessels showed variable responses to identical dosages, ranging from minimal destruction of the outer part of the retina to full-thickness destruction involving the nerve fiber layer. Variations in retinal thickness, vascularity, pigment epithelial pigmentation, and focus of the laser beam are responsible for differences in severity of each burn. Even when major vessels are avoided, it is impossible to predict accurately the degree of damage. However, our studies suggest that much of the initial damage is extracellular, sparing nerve fiber layer axons. Therefore, some acute damage might be reversible following resorption of the edema.