The role of tangential dispersion in retinal mosaic formation

https://doi.org/10.1016/S1350-9462(01)00024-6Get rights and content

Abstract

Individual types of retinal nerve cell are spaced across the retina in an orderly manner, ensuring a uniform sampling of the visual field. This regularity in cellular spacing has been commonly attributed to fate determination mechanisms operating around the time of cell birth, an hypothesis presuming that the position of a nerve cell is fixed within the plane of the retina from the time of its determination. At odds with this view, recent results from X-inactivation mosaic mice indicate that certain classes of retinal nerve cell, those known to form orderly mosaics in the adult retina, disperse tangentially during development. Furthermore, studies defining the spatial characteristics of developing and mature retinal mosaics suggest that cell–cell interactions around the time of morphological differentiation lead to mutual repulsion. Modelling studies in turn show that nothing more than a simple minimal spacing rule between neighboring cells of the same type is sufficient for the creation of the global patterning observed in biological retinal mosaics. For some cell types, the size of this “exclusion zone” surrounding individual cells is shown to be an intrinsic characteristic of each cell type, invariant across the retina, and accounting for the variation in mosaic regularity across changes in cell density. These results show how short-distance movements driven by intercellular interactions at the local level may mediate the emergence of the global patterning characteristic of retinal mosaics during development.

Introduction

The architectural complexity of the retina is brought about by various developmental events including proliferation, fate determination, migration, process outgrowth, target cell recognition, synaptogenesis and cell death. These events gradually establish a precisely layered structure in which distinct types of retinal neuron become situated in three cellular layers, interconnected by two intervening synaptic layers. Within this layered architecture, certain cell types are distributed in orderly arrays across a given layer so that they ensure a uniform sampling of the retinal surface. These orderly arrays, or “mosaics”, of retinal nerve cells are a prime example of pattern formation during development, but the developmental mechanism responsible for the establishment of this patterning remains to be defined. Indeed, multiple developmental processes may contribute to the assembly of such orderly distributions of nerve cells (Wässle and Riemann, 1978; Cook and Chalupa, 2000).

This regularity in the cellular spacing of retinal mosaics has often been interpreted as being due to the action of lateral inductive and inhibitory events that determine cellular fate. The precisely organized and repeating nature of cellular positioning within the eye of Drosophila, produced by cellular interactions governing fate determination (Lawrence, 1992; Wolff et al., 1997), has been regarded as a model for understanding vertebrate retinal mosaic formation (Wikler and Rakic (1991), Wikler and Rakic (1994)), and recent studies have begun to identify homologous molecular determinants of cell fate in the vertebrate retina (Dorsky et al (1995), Dorsky et al (1997)). Indeed, a myriad of recent studies have revealed multiple ligand–receptor interactions during vertebrate retinal development that influence cellular fate (Lillien, 1995; Ezzeddine et al., 1997; Lillien and Wancio, 1998; McFarlane et al., 1998; Dyer and Cepko, 2001). These studies are consistent with a model in which the competence of multipotent progenitors varies in the presence of an increasingly heterogeneous microenvironment to generate retinal neurons and glia in reproducible ratios across the retinal surface (Cepko et al., 1996; Jasoni and Reh, 1996; Alexiades and Cepko, 1997), with later generated cell types spacing out those produced earlier in development (Harman and Beazley, 1989; Wikler and Rakic, 1994). Subtle alterations in these interactions have been proposed to account for species-typical variations upon a vertebrate theme, producing, for example, the changing ratio of ganglion cells to other retinal neurons across eccentricity in primates, or the characteristic differences in the retinas of nocturnal versus diurnal rodents (Harman and Beazley, 1989; Reichenbach et al., 1998). If such fate-determining events occur periodically across the developing retina, as has been shown at the leading edge of retinal proliferation (e.g: McCabe et al., 1999), this may be sufficient to establish the periodicity in retinal mosaics observed in maturity (Fig. 1).

Consistent with this view is the widespread assumption that retinal neurons become postmitotic at a particular locus on the retinal surface and subsequently move only radially to occupy an appropriate depth in the retina for a given cell type (Luskin, 1994), preserving their relative positioning at the time of fate determination. By contrast with the phenomenal distances migrated by cortical neuroblasts, employing radial glial guides in their migration out of a germinal zone to form the cortical plate (Rakic, 1972; Gadisseux et al., 1990; Fishell and Hatten, 1991; Zheng et al., 1996), newborn retinal neurons move only modest distances, or not at all, and the germinal zone is progressively transformed into the mature retinal structure (Sidman, 1961; Zimmerman et al., 1988; Polley et al., 1989). The use of lineage tracing techniques to study clonal relationships provided the first clear evidence for this presumption, when individual retinal progenitors were either labelled with heritable cytoplasmic markers or were transfected with replication-deficient retroviruses encoding reporter genes (Turner and Cepko, 1987; Holt et al., 1988; Wetts and Fraser, 1988; Turner et al., 1990), and similar results were subsequently reported for chimeric mice in which highly imbalanced chimeras revealed the distribution of presumed clones (Williams and Goldowitz, 1992). These studies all reported that, not only were progenitor cells multipotent, but that their progeny were distributed as precise radial columns within the retina. The implication of the latter observation was that these cells, upon their birth, must move exclusively in the radial axis. This too was in contrast with experimental observations in the developing neocortex, in which siblings in a clone become dispersed by the time they arrive within the cortical plate (Walsh and Cepko, 1993; O’Rourke et al., 1995; Reid et al., 1997; Tan et al., 1998).

Those same retinal lineage studies occasionally detected individual clonally related cells that were separated from the clonal column by a few cell diameters. These individual cells were often interpreted as being either artifactual or ectopic; the fact that they were extremely rare led them to be ignored or dismissed (Turner and Cepko, 1987; Holt et al., 1988; Wetts and Fraser, 1988; Turner et al., 1990; Williams and Goldowitz, 1992). Subsequent interpretations of their organizational insignificance were that they became displaced from their clonal columns by passive forces generated during retinal expansion (Goldowitz et al., 1996), or that they were produced as a consequence of their mitotic sibling moving laterally within the germinal zone before continuing proliferation of the remaining clonal column (Fekete et al., 1994), analogous to the movements of dividing cells detected within the telencephalic germinal zone (Fishell et al., 1993). In fact, we now know that tangential dispersion is a universal feature of certain cell types within the retina, occurring amongst all members of a given type. The significance of this lateral displacement is that it may reflect an active movement of the postmitotic cell in the presence of cell–cell interactions which ultimately contributes to the regularity within a retinal mosaic, an hypothesis originally proposed by Wässle and Riemann (1978) but which has received relatively little support over the years. The present review considers the evidence for this perspective.

The key experimental observation for this hypothesis came from studies using X-inactivation mosaic mice (Tan et al., 1993), in which the lacZ reporter transgene, under the control of a promoter for a ubiquitous housekeeping gene (Gautier et al., 1989), had been serendipitously integrated on the X-chromosome (Tan et al., 1995). Because of the natural and random phenomenon of X-inactivation, the transgene is switched off in ≈50% of all retinal progenitors in female hemizygous mice, and the clonal descendents of these cells inherit the X-active status of their ancestor at the time of X-inactivation (Fig. 2). Since X-inactivation occurs prior to the onset of retinal neurogenesis, the benefit of this approach is that it enables the marking of roughly half of all retinal clones. In the adult retina, the organization of most of the cells in these clones is precisely columnar (Fig. 3a), as might have been expected from the above lineage-tracing studies. Surprisingly however, certain cell types are often observed to be displaced laterally with respect to their clonal columns of origin. These cell types include the cone, horizontal, amacrine and ganglion cells (Reese et al., 1995). As these cell types are composed of subclasses that are known to be assembled into orderly mosaics, their tangential dispersion may contribute to the assembly of this mosaic order. The following discussion in this section will address a number of aspects associated with these tangentially dispersing neurons: How common is this phenomenon of tangential dispersion? How far do retinal neurons disperse tangentially? When does tangential dispersion occur? And what is its relationship to morphological differentiation? The subsequent section will then consider its consequences for the establishment of the mature retinal architecture: Ultimately, might it play a role in the formation of orderly retinal mosaics?

Section snippets

Tangential dispersion of retinal neuroblasts

The use of these X-inactivation transgenic mice has been key to recognizing the frequency of tangential dispersion within individual cell types, because 50% of the progenitor population is labelled, and because the labelling is established before any retinal neurons had become post-mitotic (i.e: X-inactivation occurs around E−8.5, Tan et al., 1993). These two features overcome problems associated with other single-cell lineage-tracing approaches. First, those approaches have been generally less

The formation of retinal mosaics

Do these experimental results documenting tangential dispersion and its developmental time-course have anything to do with retinal mosaic formation? Understanding the formation of retinal mosaics has been hindered by the fact that most markers of retinal cells that display orderly arrays in adulthood are not expressed during the period of mosaic formation (but see Wikler and Rakic, 1994; Raymond et al., 1995, for exceptions amongst the photoreceptor populations). It has recently been

Future directions

The nature of the homotypic cell–cell interactions in relation to the differentiation of tangentially oriented processes remains to be determined. The adult morphology of a cell is certainly no indication of whether that cell type will disperse tangentially, and nor is density. The cholinergic amacrine cells and the horizontal cells in the mouse retina are similarly dense and comparably regular, yet their mature morphologies and dendritic overlap are conspicuously different. Imaging the

Acknowledgements

This research program was supported by the NIH (EY-11087), the CNR and the Telethon Foundation.

References (75)

  • T.-L. Ling et al.

    The development of astrocytes in the cat retinaevidence of migration from the optic nerve

    Dev. Brain Res.

    (1988)
  • J. Provis et al.

    Ontogeny of the primate foveaa central issue in retinal development

    Prog. Neurobiol.

    (1998)
  • B.E. Reese et al.

    Neurogenesis in the retinal ganglion cell layer of the rat

    Neuroscience

    (1992)
  • B.E. Reese et al.

    Clonal boundary analysis in the developing retina using X-inactivation transgenic mosaic mice

    Sem. Cell Dev. Biol.

    (1998)
  • S.-S. Tan et al.

    Separate progenitors for radial and tangential cell dispersion during development of the cerebral neocortex

    Neuron

    (1998)
  • D.L. Turner et al.

    Lineage-independent determination of cell type in the embryonic mouse retina

    Neuron

    (1990)
  • P.K. Ahnelt et al.

    Irregular S-cone mosaics in felid retinas. Spatial interaction with axonless horizontal cells, revealed by cross-correlation

    J. Opt. Soc. Am., A

    (2000)
  • M.R. Alexiades et al.

    Subsets of retinal progenitors display temporally regulated and distinct biases in the fates of their progeny

    Development

    (1997)
  • L.D. Carter-Dawson et al.

    Rods and cones in the mouse retina. II. Autoradiographic analysis of cell generation using tritiated thymidine

    J. Comp. Neurol.

    (1979)
  • A. Cellerino et al.

    Retinal ganglion cells with NADPH-diaphorase activity in the chick form a regular mosaic with a strong dorsoventral asymmetry that can be modelled by a minimal spacing rule

    Eur. J. Neurosci.

    (2000)
  • C.L. Cepko et al.

    Cell fate determination in the vertebrate retina

    Proc. Natl. Acad. Sci.

    (1996)
  • J.E. Cook

    Getting to grips with neuronal diversitywhat is a neuronal type?

  • R.I. Dorsky et al.

    Regulation of neuronal diversity in the Xenopus retina by Delta signalling

    Nature

    (1997)
  • U. Dräger

    Birth dates of retinal ganglion cells giving rise to the crossed and uncrossed optic projections in the mouse

    Proc. R. Soc., London

    (1985)
  • M.A. Dyer et al.

    p27Kip1 and p57Kip2 regulate proliferation in distinct retinal progenitor cell populations

    J. Neurosci.

    (2001)
  • S.J. Eglen et al.

    Lateral cell movement driven by dendritic interactions is sufficient to form retinal mosaics

    Network: Comput. Neural Syst.

    (2000)
  • Z.D. Ezzeddine et al.

    Postmitotic cells fated to become rod photoreceptors can be respecified by CNTF treatment of the retina

    Development

    (1997)
  • G. Fishell et al.

    Astrotactin provides a receptor system for CNS neuronal migration

    Development

    (1991)
  • G. Fishell et al.

    Dispersion of neural progenitors within the germinal zones of the forebrain

    Nature

    (1993)
  • L. Galli-Resta

    Local, possibly contact-mediated signalling restricted to homotypic neurons controls the regular spacing of cells within the cholinergic arrays in the developing rodent retina

    Development

    (2000)
  • Galli-Resta, L., Novelli, E., 2000. The effects of natural cell loss on the regularity of the retinal cholinergic...
  • L. Galli-Resta et al.

    Mosaics of islet-1 expressing amacrine cells assembled by short range cellular interactions

    J. Neurosci.

    (1997)
  • L. Galli-Resta et al.

    Modelling the mosaic organization of rod and cone photoreceptors with a minimal-spacing rule

    Eur. J. Neurosci.

    (1999)
  • C. Gautier et al.

    A ubiquitous mammalian expression vector, pHMG, based on a housekeeping gene promoter

    Nucleic Acids Res.

    (1989)
  • A.E. Hendrickson

    Primate foveal developmentA microcosm of current questions in neurobiology

    Invest. Ophthalmol. Visual Sci.

    (1994)
  • J.W. Hinds et al.

    Differentiation of photoreceptors and horizontal cells in the embryonic mouse retinaan electron microscopic, serial section analysis

    J. Comp. Neurol.

    (1979)
  • J.W. Hinds et al.

    Development of retinal amacrine cells in the mouse embryoevidence for two modes of formation

    J. Comp. Neurol.

    (1983)
  • Cited by (64)

    • 1.19 - Retinal Mosaics

      2020, The Senses: A Comprehensive Reference: Volume 1-7, Second Edition
    • Metabolic profiling of the mouse retina using amino acid signatures: Insight into developmental cell dispersion patterns

      2013, Experimental Neurology
      Citation Excerpt :

      Previous analysis of cell dispersion patterns indicates that two distinct migration patterns co-operate to form the retinal mosaic during retinogenesis (Rapaport et al., 2004; Reese and Tan, 1998; Reese et al., 1995, 1999). Rod photoreceptors, Müller cells and bipolar cells participate in radial dispersion whereas cone photoreceptors, horizontal cells, amacrine cells and ganglion cells exhibit radial and tangential dispersions (Reese and Galli-Resta, 2002). Macromolecular markers have distinguished the identity of some radially migrating cell populations such as dopaminergic amacrine cells and other tangentially dispersed populations such as cholinergic amacrine cells (Acosta et al., 2008; Eglen et al., 2003; Galli-Resta et al., 1997, 2000; Raven and Reese, 2002; Reese et al., 1999).

    View all citing articles on Scopus
    View full text