B. Neelam et al., STRUCTURE-FUNCTION STUDIES OF LIGAND-INDUCED EPIDERMAL GROWTH-FACTOR RECEPTOR DIMERIZATION, Biochemistry, 37(14), 1998, pp. 4884-4891
We present a novel 96-well assay which we have applied to a structure-
function study of epidermal growth factor receptor dimerization. The b
asis of the assay lies in the increased probability of EGFRs being cap
tured as dimers by a bivalent antibody when they are immobilized in th
e presence of a cognate ligand. Once immobilized, the antibody acts as
a tether, retaining the receptor in its dimeric state with a resultan
t 5-7-fold increase in binding of a radiolabeled ligand probe. When th
e assay was applied to members of the EGF ligand family, murine EGF, t
ransforming growth factor alpha, and heparin-binding EGF-like growth f
actor were comparable with human EGF (EC50 = 2nM); betacellulin, which
has a broader receptor specificity, was slightly less effective. In c
ontrast, amphiregulin (AR(1-84)), which has a truncated C-tail and lac
ks a conserved leucine residue, was ineffective unless used at >1 mu M
. We further probed the involvement of the C-tail and the conserved le
ucine residue in receptor dimerization by comparing the activities of
two genetically modified EGFs (the chimera mEGF/TGF alpha(44-50) and t
he EGF point mutant L47A) and a C-terminally extended form of AR (AR(1
-90)) with those of two other unrelated EGF mutants (I23T and L15A). T
he potency of these ligands was in the order EGF > I23T > mEGF/TGF alp
ha(44-50) > L47A = L15A much greater than AR(1-90) > AR(1-84). Althoug
h AR was much worse than predicted from its affinity, this defect coul
d be partially rectified by co-localization of the immobilizing antibo
dy with heparin. Thus, it seems likely that AR cannot dimerize the EGF
R unless other accessory molecules are present to stabilize its functi
onal association with the EGFR.