Alternative splicing and genomic structure of the AML1 gene involved in acute myeloid leukemia.

We previously isolated the AML1 gene, which is rearranged by the t(8;21) translocation in acute myeloid leukemia. The AML1 gene is highly homologous to the Drosophila segmentation gene runt and the mouse transcription factor PEBP2 alpha subunit gene. This region of homology, called the Runt domain, is responsible for DNA-binding and protein--protein interaction. In this study, we isolated and characterized various forms of AML1 cDNAs which reflect a complex pattern of mRNA species. Analysis of these cDNAs has led to the identification of two distinct AML1 proteins, designated AML1b (453 amino acids) and AML1c (480 amino acids), which differ markedly from the previously reported AML1a (250 amino acids) with regard to their C-terminal regions, although all three contain the Runt domain. The large C-terminal region common to AML1b and AML1c is suggested to be a transcriptional activation domain. AML1c differs from AML1b by only 32 amino acids in the N-terminal. Characterization of the genomic structure revealed that the AML1 gene consists of nine exons and spans > 150 kb of genomic DNA. Northern blot analysis demonstrated the presence of six major transcripts, encoding AML1b or AML1c, which can all be explained by the existence of two promoters, alternative splicing and differential usage of three polyadenylation sites. A minor transcript encoding AML1a which results from alternative splicing of a separate exon can be detected only by reverse transcription-polymerase chain reaction amplification. The distinct proteins encoded by the AML1 gene may have different functions, which could contribute to regulating cell growth and/or differentiation through transcriptional regulation of a specific subset of target genes.