Search

MEF2D fusion genes are newly discovered recurrent gene abnormalities that are detected in approximately 5% of acute lymphoblastic leukemia cases. We previously demonstrated that the vector-driven expression of MEF2D fusion proteins was markedly stronger than that of wild-type MEF2D; however, the underlying mechanisms and significance of this expression have yet to be clarified. We herein showed that the strong expression of MEF2D fusion proteins was caused by the loss of the target site of miRNA due to gene translocation. We identified the target region of miRNA located in the coding region and selected miR-122 as a candidate of the responsible miRNA. Mutations at a putative binding site of miR-122 increased MEF2D expression, while the transfection of its miRNA mimic reduced the expression of wild-type MEF2D, but not MEF2D fusion proteins. We also found that MEF2D fusion proteins inhibited the transcriptional activity of PAX5, a B-cell differentiation regulator in a manner that depended on fusion-specific strong expression and an association with histone deacetylase 4, which may lead to the differentiation disorders of B cells. Our results provide novel insights into the mechanisms underlying leukemia development by MEF2D fusion genes and the involvement of the deregulation of miRNA-mediated repression in cancer development.

Deletions of the 3p arm have been detected in various solid tumors, but no study to date has investigated this deletion in diffuse large B-cell lymphoma (DLBCL). Recently, we demonstrated that 3p14.2 was deleted in approximately 30% of DLBCL cases by use of a genome-wide array-comparative genomic hybridization (CGH). For a more detailed examination of the genomic losses at 3p14.2, here we made use of contig BAC array for 3p14.2, and found that 12 DLBCL samples displayed losses. All of the deleted regions were located within the fragile histidine triad (FHIT) gene, and the most frequent region of loss was mapped to 0.4 Mbp of the region encompassing the introns 4 and 5 and exon 5 of the FHIT gene. Concomitant analysis of transcripts showed that the FHIT gene was aberrantly transcribed in 31% of the DLBCL samples examined and that the lost exons of the aberrant transcripts were correlated with genomic deletions. These findings indicate that (1) loss of genomic material at 3q14.2 is responsible for exon losses of the FHIT gene, and (2) genomic loss of the FHIT gene is one of the causes of the generation of aberrant transcripts.

t(11; 18) (q21; q21) is a chromosomal aberration specific to low-grade mucosa-associated lymphoid tissue (MALT) lymphoma, and generates the chimeric product apoptosis inhibitor 2 (API2)-MALT1, which has been suggested to play an important role in MALT lymphomagenesis. However, little is known about the characteristics of API2, MALT1, and API2-MALT1 proteins. We therefore investigated the subcellular localization and stability of these products. API2 was localized in the nucleus and the cytoplasm, and MALT1 and API2-MALT1 in the cytoplasm only. Western blot analysis showed that the products of API2 and MALT1 were unstable, while the API2-MALT1 product was stable. The API2 deletion mutants at the end of the C-terminal and the MALT1 deletion mutants at the end of the N-terminal were stable compared with the full-length products. These results indicate that the domains responsible for protein instability are located in the end of the C-terminal of API2 and in that of the N-terminal of MALT1, and also that API2-MALT1 became stable because it lacks these domains. It has been suggested that NF-kappaB activation plays an important role in the tumorigenesis of MALT lymphoma. Our findings further suggest that the stabilized expression of API2-MALT1 products may continuously stimulate the NF-kappaB activating pathway, thus leading to MALT lymphomagenesis.

PAX5 is a transcription factor required for B-cell development and maintenance. PML is a tumor suppressor and a pro-apoptotic factor. A fusion gene, PAX5-PML, was found in acute lymphoblastic leukemia (ALL) with chromosomal translocation t(9;15)(p13;q24), but no functional analysis has been reported. Here, we demonstrate that PAX5-PML had a dominant-negative effect on both PAX5 and PML. PAX5-PML dominant negatively inhibited PAX5 transcriptional activity in the luciferase reporter assay and suppressed the expression of the PAX5 transcriptional targets in B-lymphoid cell line. Surprisingly, PAX5-PML hardly showed DNA-binding activity in vitro although it retained the DNA-binding domain of PAX5. Additional experiments, including chromatin immunoprecipitation (ChIP) assay, suggested that PAX5-PML bound to the promoter through the association with PAX5 on the promoter. On the other hand, coexpression of PAX5-PML inhibited PML sumoylation, disrupted PML nuclear bodies (NBs), and conferred apoptosis resistance on HeLa cells. Furthermore, treatment with arsenic trioxide (ATO) induced PML sumoylation and reconstitution of PML NBs, and overcame the anti-apoptotic effect of PAX5-PML in HeLa cells. These data suggest the possible involvement of this fusion protein in the leukemogenesis of B-ALL in a dual dominant-negative manner and the possibility that ALL with PAX5-PML can be treated with ATO.