Eph Receptors and Ephrins Are Key Regulators of Morphogenesis 127
The identification of two ephrins showing a graded distribution from
posterior to anterior in the developing chick and mouse tectum suggested a role for Eph signalling in establishing appropriate connections
in the retinotectal system. Drescher et al. purified an Eph ligand, ephrinAS, from the chick tectum, based on its expression in the tectum and the
fact that it is a GPI-linked protein (Drescher et al. 1995). Bioassays,
devised to characterise the growth of chick retinal axons over the tectum, showed that membranes isolated from tectal cells possessed a
collapsing activity for growth cones from temporal but not nasal retinal
ganglion cells. Ephrin-AS mimicked this collapsing activity. The existence of a second ligand in the tectum, ephrin-A2, was shown in the
embryo by binding of a chimaeric protein in which the extracellular
domain of a receptor was linked to alkaline phosphatase. The ephrin-A2
cDNA was then cloned using an expression library (Cheng and Flanagan 1994).
In the chick the ephrin-AS and ephrin-A2 ligands are assumed to
interact with two Eph receptors: EphA4, which is expressed uniformly
across the retina, and EphA3, whose expression is graded across the
retina, with a high point in the temporal region (Cheng et al. 1995). In
the tectum, the expression domains of the ligands differ from each other,
with ephrin-A2 extending more anteriorly than ephrin-AS. With variations in ligand-binding specificities and the graded distributions of
both the ligands and receptors (Monschau et al. 1997), it is feasible that
sufficient information can be provided by Eph signalling to resolve the
retinotopic map. In addition, Eph receptors and ligands are spatially
regulated with respect to the dorsal-ventral axis of the eye. For instance,
EphB2 is expressed more strongly in the ventral than dorsal retina
(Hoi ash and Pasquale 1995; Kenny et al. 1995). Similarly, ephrins of the
A class exist in a high-nasal-to-Iow-temporal gradient and ephrins of the
B class are expressed higher dorsally than ventrally in the retina (Marcus et al. 1996). Such localisation ofligands in the eye also occurs in the
zebrafish where three ephrins are differentially expressed in retinal
ganglion cells prior to and during the projection ofaxons to the midbrain
(Brennan et al. 1997).
Nakomoto et al. have provided direct evidence that Eph signalling is
involved in retinotectal map formation by misexpressing ephrin-A2 in
the developing chick tectum and showing that this leads to abnormal
retinotectal axon growth (Nakamoto et al. 1996). Furthermore, a mouse
The identification of two ephrins showing a graded distribution from
posterior to anterior in the developing chick and mouse tectum suggested a role for Eph signalling in establishing appropriate connections
in the retinotectal system. Drescher et al. purified an Eph ligand, ephrinAS, from the chick tectum, based on its expression in the tectum and the
fact that it is a GPI-linked protein (Drescher et al. 1995). Bioassays,
devised to characterise the growth of chick retinal axons over the tectum, showed that membranes isolated from tectal cells possessed a
collapsing activity for growth cones from temporal but not nasal retinal
ganglion cells. Ephrin-AS mimicked this collapsing activity. The existence of a second ligand in the tectum, ephrin-A2, was shown in the
embryo by binding of a chimaeric protein in which the extracellular
domain of a receptor was linked to alkaline phosphatase. The ephrin-A2
cDNA was then cloned using an expression library (Cheng and Flanagan 1994).
In the chick the ephrin-AS and ephrin-A2 ligands are assumed to
interact with two Eph receptors: EphA4, which is expressed uniformly
across the retina, and EphA3, whose expression is graded across the
retina, with a high point in the temporal region (Cheng et al. 1995). In
the tectum, the expression domains of the ligands differ from each other,
with ephrin-A2 extending more anteriorly than ephrin-AS. With variations in ligand-binding specificities and the graded distributions of
both the ligands and receptors (Monschau et al. 1997), it is feasible that
sufficient information can be provided by Eph signalling to resolve the
retinotopic map. In addition, Eph receptors and ligands are spatially
regulated with respect to the dorsal-ventral axis of the eye. For instance,
EphB2 is expressed more strongly in the ventral than dorsal retina
(Hoi ash and Pasquale 1995; Kenny et al. 1995). Similarly, ephrins of the
A class exist in a high-nasal-to-Iow-temporal gradient and ephrins of the
B class are expressed higher dorsally than ventrally in the retina (Marcus et al. 1996). Such localisation ofligands in the eye also occurs in the
zebrafish where three ephrins are differentially expressed in retinal
ganglion cells prior to and during the projection ofaxons to the midbrain
(Brennan et al. 1997).
Nakomoto et al. have provided direct evidence that Eph signalling is
involved in retinotectal map formation by misexpressing ephrin-A2 in
the developing chick tectum and showing that this leads to abnormal
retinotectal axon growth (Nakamoto et al. 1996). Furthermore, a mouse