Search

We identified a novel serine/arginine (SR)-rich-related protein as a binding partner of Clk4 mutant lacking kinase activity (Clk4 K189R) in the two-hybrid screen and designated it Clasp (Clk4-associating SR-related protein). Northern blot analysis revealed that Clasp mRNA was highly expressed in brain, and in situ hybridization of a mouse brain sagittal section hybridized with antisense probes revealed that both Clasp and Clk4 mRNAs were expressed in the hippocampus, the cerebellum, and the olfactory bulb. Two forms of Clasp were produced by a frameshift following alternative splicing. The staining of an HA-tagged short form of Clasp (ClaspS) showed a nucleoplasmic pattern, while the long form of Clasp (ClaspL) was localized as nuclear dots. In vitro protein interaction assay demonstrated that Clk4 K189R was bound to Clasp while wild Clk4 was not. Overexpression of ClaspL promoted accumulation of Clk4 K189R in the nuclear dots and the exon EB inclusion from CR-1 and CR-2 pre-mRNA of Clk1. These data suggest that Clasp is a binding partner of Clk4 and may be involved in the regulation of the activity of Clk kinase family.

The AML1-MTG8 fusion transcription factor generated by t(8;21) translocation is thought to dysregulate genes that are crucial for normal differentiation and proliferation of hematopoietic progenitors to cause acute myelogenous leukemia (AML). Although AML1-MTG8 has been shown to repress the transcription of AML1 targets, none of the known targets of AML1 are probably responsible for AML1-MTG8-mediated leukemogenesis. In this study, 24 genes under the downstream control of AML1-MTG8 were isolated by using a differential display technique. Analysis with deletion mutants of AML1-MTG8 demonstrated that the regulation of the majority of these genes requires the region of 51 residues (488-538) containing the Nervy homology region 2 (NHR2), through which AML1-MTG8 interacts with MTGR1. Among the 24 genes identified, 10 were considered to be genes under the control of AML1, because their expression was altered by AML1b or AML1a or both. However, the other 14 genes were not affected by either AML1b or AML1a, suggesting the possibility that AML1-MTG8 regulates a number of specific target genes that are not normally regulated by AML1. Furthermore, an up-regulated gene, TIS11b (ERF-1,cMG1), was highly expressed in t(8;21) leukemic cells, and the overexpression of TIS11b induced myeloid cell proliferation in response to granulocyte colony-stimulating factor. These results suggest that the high-level expression of TIS11b contributes to AML1-MTG8-mediated leukemogenesis.

The FUS (TLS)-ERG chimeric protein associated with t(16;21)(p11;q22) acute myeloid leukemia is structurally similar to the Ewing's sarcoma chimeric transcription factor EWS-ERG. We found that both FUS-ERG and EWS-ERG could induce anchorage-independent proliferation of the mouse fibroblast cell line NIH 3T3. However, only FUS-ERG was able to inhibit the differentiation into neutrophils of a mouse myeloid precursor cell line L-G and induce its granulocyte colony-stimulating factor-dependent growth. We constructed several deletion mutants of FUS-ERG lacking a part of the N-terminal FUS region. A deletion mutant lacking the region between amino acids 1 and 173 (exons 1 to 5) lost the NIH 3T3-transforming activity but retained the L-G-transforming activity. On the other hand, a mutant lacking the region between amino acids 174 and 265 (exons 6 and 7) lost the L-G-transforming activity but retained the NIH 3T3-transforming activity. These results indicate that the N-terminal region of FUS contains two independent functional domains required for the NIH 3T3 and L-G transformation, which we named TR1 and TR2, respectively. Although EWS intrinsically possessed the TR2 domain, the EWS-ERG construct employed lacked the EWS sequence containing this domain. Since the TR2 domain is always found in chimeric proteins identified from t(16;21) leukemia patients but not in chimeric proteins from Ewing's sarcoma patients, it seems that the TR2 function is required only for the leukemogenic potential. In addition, we identified three cellular genes whose expression was altered by ectopic expression of FUS-ERG and found that these are regulated in either a TR1-dependent or a TR2-dependent manner. These results suggest that FUS-ERG may activate two independent oncogenic pathways during the leukemogenic process by modulating the expression of two different groups of genes simultaneously.