Evi-l was originally identified as a common site of viral integation in murine myeloid tumors. Evi-l encodes a 120-kDa polypeptide containing 10 zinc finger motifs located in two domains 380 amino acids apart and an acidic domain located carboxy terminal to the second set of zinc fingers. These features suggest that Evi-l is a site-specific DNA-binding protein involved in the regulation of RNA transcription. We have purified Evi-l protein from E. coli and have employed a gel shift-polymerase chain reaction method using random oligonucleotides to identify a high-affinity binding site for Evi-l. The consensus sequence for this binding site is TGACAAGATAA. Evi-l protein specifically protects this motif from DNase I digestion. By searching the nucleotide sequence data bases, we have found this binding site both in sequences 5' to genes in putative or known regulatory regions and within intron sequences.Many eukaryotic transcriptional regulatory proteins contain one of several DNA-binding motifs, such as a helix-turnhelix or zinc fingers, which confer site-specific, high-affinity binding to transcriptional control sequences in genomic DNA. This binding, under proper conditions, is thought to result in changes in rates of transcriptional initiation by RNA polymerases by mechanisms that are not yet entirely clear (25,37). The first two eukaryotic transcriptional regulatory proteins to be characterized, TFIIIA (17) and Spl (6), were identified by a combination of genetic and biochemical means, by which the binding sites were identified initially as cis-acting regulatory elements. Now that common DNAbinding motifs within proteins have been identified, putative transcriptional factors are often identified on the basis of protein sequence homology.Evi-J was first identified as a common site of ecotropic retrovirus integration in virally induced myeloid tumors of AKXD mice (40). Subsequent cloning and sequencing of the cDNA corresponding to this locus (39) showed that this locus encodes a 120-kDa polypeptide that contains 10 regions with extensive homology to the zinc-binding domains found in several transcriptional regulatory factors and first identified in the Xenopus transcription factor IIIA (36). This amino acid sequence consists of 27 to 30 residues with the consensus sequence (Y/F)XCX2 CX3FX5LX2_3HX34HX5.Current structural models propose that the cysteine and histidine residues chelate a single zinc atom, constraining the intervening amino acids into a looped domain, which is further stabilized by hydrophobic interactions among (Y/F), F, and L (1, 30, 44). DNA binding then occurs between charged amino acids on the exposed face of an alpha helix (which extends from the center of the motif through the histidines) and an estimated 3 to 5 bases along the major groove (1, 7, 21). DNA-binding studies with purified zinc finger regions containing one or several fingers suggest that multiple fingers are necessary to obtain high-affinity, sitespecific DNA binding, perhaps through cooperativity (18,41,44).These sequence compari...