Somatic mutations in c-myc intron I cluster in discrete domains that define protein binding sequences

The activated c-myc allele in Burkitt's lymphoma tumor cells is associated with a clustering of somatic mutations within intron I near the exon I boundary. We have identified several discrete protein binding sites within this region of c-myc intron I designated as myc intron factor-1 (MIF-1), MIF-2, and MIF-3. In addition to our previous characterization of a 20-nucleotide binding site for MIF-1, we now have identified adjacent 20-nucleotide and 34-nucleotide binding sites for MIF-2 and MIF-3, respectively. All three elements are protected from exonuclease digestion by nuclear protein extracts, and each gives rise to a distinct migration pattern on mobility shift assays. In addition, MIF-1, 2, and 3 share a 5-nucleotide (TTATG) internal sequence, which may account for cross-competition of these binding sites in the exonuclease protection experiment. Deletion mutant analyses showed that selective removal of the MIF-3 binding site alone was sufficient to enhance chloramphenicol acetyltransferase reporter activity similar to that observed with larger deletions of myc intron I. We have demonstrated that somatic mutations in activated c-myc alleles are frequently clustered in discrete domains that define protein recognition sequences.