RUNX1 rearrangements are common genetic abnormalities in acute leukemia. The t(7;21)(p22;q22) translocation, recently described in three cases of myeloid neoplasias, fuses the ubiquitin specific peptidase 42 gene, USP42, a member of the deubiquitinating enzyme family, to RUNX1. In this study, we characterized the semicryptic t(7;21)(p22;q22) translocation, identified by fluorescent in situ hybridization and spectral karyotyping, in a novel case of acute myeloid leukemia. Sequence analysis of the reverse transcription-polymerase chain reaction products confirmed the presence of two in-frame RUNX1-USP42 and one reciprocal in-frame USP42-RUNX1 fusion transcripts. Bioinformatic analysis of the genomic translocation breakpoints revealed microhomologies and insertion of shared nucleotides at the junctions. A topoisomerase II sequence was also detected near the break site. Additionally, we demonstrated a significant overexpression of the rearranged USP42 gene in t(7;21) positive cells using quantitative real-time PCR. Our results provide the first evidence of the possible involvement of the nonhomologous end-joining mechanism in the origin of the recurrent t(7;21) translocation. Moreover, presence of the complete catalytic USP site in the putative chimeric proteins and the upregulated expression of USP42 suggest a role of the deubiquitinating enzyme in the pathogenesis of this leukemia.

Chromosomal translocations in lymphoid tumors frequently result from recombination between a normally rearranging antigen receptor gene and a normally non-rearranging second locus. The possibility that the lymphocyte recombinase apparatus plays a role in determining the position of breakage at the second locus has been a matter of controversy because of the inconsistent presence of heptamer-like recognition sequences adjoining breakpoints at this site. To further investigate this issue, sites of DNA recombination were analyzed in both the der(9) and der(7) products of t(7;9)(q34;q32), a recurrent translocation of human acute lymphoblastic leukemias (T-ALL). In each of three separate cases, the translocation has divided the TCR-beta locus, juxtaposing chromosome 9 DNA 5\' to a J-region in the der(9) product and 3\' to a D-region in the der(7) product, with variably sized N-insertions and small deletions detectable at the junctions. All three cases contain breakpoints in chromosome 9 DNA tightly clustered between two closely spaced, and oppositely oriented heptamer sequences, CAC(A/T)GTG, which perfectly match the consensus heptamer sequence recognized by the lymphocyte recombinase apparatus in normal antigen receptor gene rearrangement. In no case was there evidence of directly duplicated sequences in the two reciprocal products, as is often associated with recombination involving random staggered breakage of DNA. Taken together, these results support a mechanism for this particular translocation proceeding by recombinase-mediated breakage of both participating chromosomes.

The identification of genes in genomic DNA presents challenging technical difficulties. We show here the feasibility of using short oligonucleotides based on the consensus sequences surrounding intron-exon junctions to detect random phage and cosmid clones containing genes both through the analysis of DNA blots and by direct screening. Three degenerate oligonucleotides, a 10-mer corresponding to the 5\' splice junction and a 9-mer and a 15-mer corresponding to the 3\' splice junction, were tested on the known intron-exon boundaries of the cloned human proteolipid protein (PLP) gene at hybridization and washing temperatures appropriate to their length and composition. All predicted hybridizations were observed. The oligonucleotides were also used to identify random genomic plasmid and cosmid clones containing putative intron-exon junctions; the presence of genes in these clones was supported by RNA blot analysis and by cross-hybridization to DNA from other species. This technique should facilitate the identification of genes for inherited diseases by positional cloning studies and will assist in the identification of genes in random clones for the human genome project.