The nuclear hormone receptors regulate target gene expression in response to hormones of extracellular origin. The DNA binding specificity of these receptors therefore plays the critical role of defining the precise repertoire of target genes that respond to a given hormone. We report here an analysis of the DNA binding specificity of the thyroid hormone receptor (c-ErbA protein) and that of an oncogenic derivative, the v-ErbA protein. These otherwise closely similar proteins exhibit quite divergent DNA sequence specificities at multiple positions within the DNA binding site. The thyroid hormone receptor (c-ErbA protein) exhibits a particularly broad DNA specificity, whereas the v-ErbA protein is comparatively quite specific. Intriguingly, these differences in DNA recognition largely map to an N-terminal receptor domain not traditionally implicated in DNA binding, and are further influenced by heterodimer formation with retinoid X receptors. We propose that the N terminus of nuclear hormone receptors plays an critical role in DNA recognition by altering the conformation of the receptor domains that make the actual base-specific contacts.The nuclear hormone receptors are a family of ligand-regulated transcription factors that mediate cellular responses to a broad range of small hydrophobic hormones (1-5). Members of the family include the steroid receptors, the retinoid acid receptors (RARs), 1 retinoid X receptors (RXRs), and the thyroid hormone receptors (T 3 Rs). Additional diversity is generated within individual receptor classes by the expression of multiple receptor isoforms; for example, T 3 Rs are encoded by two separate loci, ␣ and ␤ (3, 4). Despite this diversity, the different nuclear hormone receptors share a common mode of action, functioning by binding to specific DNA sequences and regulating expression of nearby target genes (1-4, 6, 7).The DNA recognition properties of each receptor play the critical role of defining the specific repertoire of target genes that respond to a given hormone. Most receptors bind to DNA as protein dimers, with each receptor molecule binding to a "half-site," a conserved 6 -8-nucleotide DNA sequence (7-11). Recognition of the sequence of each half-site has generally been believed to be mediated exclusively by a zinc-finger motif within the center of each receptor (Fig. 1) (12-14). Amino acids in the P-box helix within this zinc-finger motif make direct contact with bases in the major grove of the DNA half-site, and altering amino acids in the P-box can alter the half-site specificity of the receptor (14 -22). An additional ␣-helix, the A-box located at the C-terminal extreme of the zinc-finger motif, makes minor groove contacts with bases at the 5Ј end of the half-site (14, 23).Aberrant forms of nuclear hormone receptors are involved in several forms of cancer. The v-erbA oncogene, for example, is a mutated copy of the host cell gene (c-erbA) for T 3 R␣-1 (24, 25). V-ErbA has sustained a number of mutations relative to its T 3 R␣ progenitor (Fig. 1); as a result, v-ErbA...

The avian erythroblastosis virus v-erbA locus potentiates the oncogenic transformation of erythroid and fibroblast cells and is derived from a host cell gene encoding a thyroid hormone receptor. We report here the use of site-directed mutagenesis to identify and characterize functional domains within the v-erbA protein. Genetic lesions introduced into a putative hinge region or at the extreme C-terminus of the v-erbA coding domain had no significant effect on the biological activity of this polypeptide. In contrast, mutations introduced within the cysteine-lysine-arginine-rich center of the v-erbA coding region, a DNA-binding domain in the thyroid and steroid hormone receptors, abolished or severely compromised the ability of the viral protein to function. Our results suggest that the mechanism of action of the v-erbA protein in establishing the neoplastic phenotype is closely related to its ability to interact with DNA, presumably thereby altering expression of host target genes by either mimicking or interfering with the action of the normal c-erbA gene product.

The avian erythroblastosis virus v-erbA locus potentiates the oncogenic transformation of erythroid and fibroblast cells and is derived from a host cell gene encoding a thyroid hormone receptor. We report here the use of site-directed mutagenesis to identify and characterize functional domains within the v-erbA protein. Genetic lesions introduced into a putative hinge region or at the extreme C-terminus of the v-erbA coding domain had no significant effect on the biological activity of this polypeptide. In contrast, mutations introduced within the cysteine-lysine-arginine-rich center of the v-erbA coding region, a DNA-binding domain in the thyroid and steroid hormone receptors, abolished or severely compromised the ability of the viral protein to function. Our results suggest that the mechanism of action of the v-erbA protein in establishing the neoplastic phenotype is closely related to its ability to interact with DNA, presumably thereby altering expression of host target genes by either mimicking or interfering with the action of the normal c-erbA gene product.