BIOLOGICAL OVERVIEW

The Eph receptors constitute the largest subfamily of receptor tyrosine kinases; they have been implicated in pattern formation, cell migration, and axon pathfinding (for reviews see Bruckner, 1998; Hsueh, 1998; Lemke, 1998; Ernst, 1998; Flanagan, 1998; O'Leary, 1999; McLaughlin, 1999, and Holder, 1999). To date, over 15 distinct Eph receptors and 8 of their ligands, the ephrins (see Ephrin), have been identified. Members of the Eph subfamily are characterized by their extracellular structure, which consists of a globular domain, a cysteine-rich region unrelated to those of other receptor protein tyrosine kinases (RTKs), and two fibronectin (FN) type III repeats. The cytoplasmic structure of Eph receptors contains conserved catalytic domains and a unique sterile alpha motif (SAM) domain, which is thought to be involved in binding SH2 proteins and is not found in other RTKs. The Eph subfamily is divided into two classes, EphA and EphB, based on the amino acid sequence homology of their extracellular domains. Ligands of the Ephrin-A subclass are attached to the membrane via a glycosylphosphatidylinositol linkage, while the Ephrin-B ligands span the membrane via a transmembrane domain. A Drosophila Eph receptor has now been characterized. Drosophila Eph possesses all the domains characteristic of the Eph subfamily of RTKs and is equally similar in sequence to both the EphA and the EphB subclasses. Drosophila Eph is expressed on the growth cones and axons of embryonic interneurons and larval photoreceptor cells (Scully, 1999).

Eph was isolated in a screen designed to isolate novel Drosophila RTKs. An adult cDNA library was screened by polymerase chain reaction (PCR) using degenerate primers corresponding to amino acid sequences HRDLAARN and DVWSYGV within the conserved catalytic domain of receptor tyrosine kinases. An amino acid comparison of the open reading frame (ORF) of one of the isolated clones revealed homology to the vertebrate Eph subfamily (Scully, 1999).

In order to begin to address its potential role in axon pathfinding, Eph was misexpressed using the UAS-GAL4 system. To reliably follow the complete translation of Eph, five c-myc epitopes were added in frame to the C-terminal end of the Eph coding sequence. Transformants carrying the full-length myc-tagged Eph construct, UAS-Eph-myc, were initially screened for high expression levels of the c-myc epitope by crossing them to a panneural GAL4 line. Individuals carrying four copies of the highest-expressing UAS-Eph-myc insertions were crossed to a number of GAL4 lines that drive expression in various tissues at different developmental times. The progeny were assayed for phenotypes at embryonic, larval, and adult stages. Immunostaining with anti-Eph and anti-myc antibodies reveals that Eph-myc is expressed at high levels when crossed to these GAL4 lines; the protein is targeted to axons and growth cones in a manner indistinguishable from endogenous Eph protein. The axonal targeting of Eph-myc allows the examination of subsets of neurons and their processes without an additional tau-based axon-targeted reporter. Despite the observation that Eph-myc protein is being expressed at robust levels, misexpression in motor neurons or overexpression in interneurons of Eph-myc protein does not appear to generate a phenotype in any of the GAL4 assays (Scully, 1999).

The SAM domain, as well as the terminal PDZ binding motif, may have biological functions in Eph (Hock, 1998; Stein, 1996; Stein, 1998b) and may be disrupted by the addition of the c-myc epitopes. Therefore, a full-length version of Eph without an epitope tag was tested. As with Eph-myc, UAS-Eph individuals over- and mis-expressing Eph with the panel of GAL4 drivers are viable and appear normal. The non-myc-tagged Eph does not grossly affect axon pathfinding in the VNC, although axon fascicles appear somewhat defasciculated when UAS-Eph is driven by scabrous (sca)-GAL4 together with elav-GAL4. This sca-GAL4; elav-GAL4 combination initiates GAL4 expression at the neuroblast stage and continues its expression in postmitotic neurons throughout the CNS. Overexpression of Eph, using pGMR-GAL4 (which drives expression in the photoreceptor cells of the larval eye disc) does not affect axon pathfinding in the developing visual system. These results indicate that altering the levels or patterns of Eph in the CNS by ectopic expression has little effect on axon pathfinding (Scully, 1999).

An attempt was made to alter the normal function of endogenous Eph by expressing a kinase-inactive form of the protein containing the critical lysine (K759) changed to methionine. To assay expression levels, the mutant derivative was myc-tagged at the C-terminus. Flies were transformed with UAS-Eph-K759M-myc and screened as described above for high expression levels. Individuals carrying four copies of the highest-expressing UAS-Eph-K759M- myc insertions were crossed to a panel of GAL4 lines and assayed for phenotypes at embryonic, larval, and adult stages. Although Eph-K759M-myc protein, like Eph-myc, is expressed at robust levels, its expression does not appear to generate a phenotype in any of the assays (Scully, 1999).

In spite of the negative results obtained from misexpression of Eph, the highly localized expression of Eph on axons in the developing nervous system suggests that the kinase will have a role in axon pathfinding, especially in light of the evidence from vertebrate Eph studies that has demonstrated a role for Eph RTKs in axon guidance. It has been suggested that the tagged proteins may not have given an overexpression phenotype due to the interference of the epitopes with the proper functioning of the Eph protein. There could be several reasons for the lack of phenotypes with full-length Eph. (1) The level of Eph expression may not be sufficiently high to affect axon pathfinding or other putative biological functions in which Eph may be involved; (2) for misexpression in motor neurons, the Eph ligand may not be present on those cells that interact with motor neurons; therefore, the signaling cascade of Eph might not be initiated upon misexpression of Eph in those cells;(3) for overexpression in interneurons, the levels of Eph may not be critical, in contrast to systems in which disturbing gradients of Eph receptors or ephrins gives rise to pathfinding phenotypes, and (4) though less likely, considering the evidence from vertebrate Eph receptor studies, the endogenous Eph receptor may not play a role in axon guidance; in this case, misexpression of Eph would not be expected to perturb axon pathfinding. In vertebrates, redundancy has been an issue in uncovering the function of the Eph RTK subfamily and its signaling pathway(s). Analysis of Eph receptors and their ligands may be more amenable in a simpler genetic system such as C. elegans or Drosophila. Currently 14 distinct vertebrate Eph receptors have been identified, but Eph represents the only Drosophila Eph receptor isolated to date. A single Eph receptor, VAB-1, has been identified in the recently completly sequenced C. elegans genome (George, 1998). Therefore, although additional Drosophila Eph receptors may exist, it is unlikely that this organism will have the large number of Eph receptors present in vertebrates. Further studies of Eph function in Drosophila employing loss-of-function mutants may produce insight into the roles of Eph receptors in the developing nervous system (Scully, 1999).

GENE STRUCTURE

Genomic length - 10 kb

Exons - 10

PROTEIN STRUCTURE

Amino Acids- 1080

Structural Domains

Like other members of the family, the predicted Eph protein possesses a single globular domain, a cysteine-rich region, and two FN type III repeats in the extracellular domain, as well as a putative transmembrane domain and conserved catalytic region. Eph also contains a conserved stretch of amino acids in the juxtamembrane region, which includes two tyrosine residues known to be major sites of autophosphorylation in vertebrate Eph receptors; these residues appear to interact with SH2-containing proteins. The cytoplasmic region of Eph contains the conserved SAM domain found in all vertebrate Eph receptors. A conserved tyrosine present in the SAM domain of EphB1, also conserved in Eph, has been shown to be required for the interaction of EphB1 with the adapter protein Grb10 (Ellis, 1996; Holland, 1997; Stein,1998a; Zisch, 1998) and the low-molecular-weight protein tyrosine phosphatase (LMW-PTP: Stein, 1996; Stein, 1998b). The last three residues of Eph, T-I-I, follow immediately after the SAM domain. This sequence fits the consensus motif of S/T-X-V/I, that binds one group of PDZ domains. Many PDZ-domain-containing proteins have been shown to be involved with clustering or localization of membrane proteins, while some allow for the assembly of signaling complexes. The Eph subfamily is classified into two groups, EphA and EphB, based on amino acid similarity within the extracellular domain. The extracellular portion of Eph from the globular domain through the FN type III repeats shows 32% and 35% identity to EphA3 and EphB2, respectively, and the cytoplasmic region spanning the catalytic domain shows an equal 71% identity to both EphA3 and EphB2. Thus, Eph cannot readily be placed in either the A or the B Eph subclass based on amino acid similarity. (Scully, 1999 and references therein).

date revised: 20 November 99

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.