1. ¹Department of Developmental Medicine (Pediatrics), Graduate School of Medicine, D-5, Osaka University, Osaka, Suita City, Japan
2. ²Osaka Medical Center and Research Institute for Maternal and Child Health, Izumi City, Osaka, Japan
3. ³Toyonaka Municipal Hospital, Osaka, Toyonaka City, Japan; and 4-14-1, Shibaharacho, Osaka, Japan
4. ⁴Department of Stem Cell Biology, Research Institute for Radiation Biology and Medicine, Hiroshima, Japan

Correspondence: Dr Y Matsuda-Hashii, Department of Developmental Medicine, Graduate School of Medicine, D-5, Osaka University, 2-2, Yamadaoka, Suita, Osaka. Fax 81-6-6879-3939; E-mail: matsuda@ped.med.osaka-u.ac.jp

Received 8 March 2004; Accepted 11 June 2004; Published online 29 July 2004.

TO THE EDITOR

Most MLL partner proteins are novel and contain various functional domains, but little is known about the function of these fusion proteins. The critical role of partner proteins in converting MLL to an oncogenic protein via a dominant gain of MLL function has been proposed.¹ Two different mechanisms for regulating target genes by fusion proteins have been proposed. Fusions with cytoplasmic partners promote dimerization of MLL proteins and are suggested to increase the affinity to the target genes and cause the constitutive activation of target genes. On the other hand, transcription factor partners introduce novel DNA-binding domains implicated in transcriptional regulation.^{2, 3} The essential target genes of these chimeric transcription factors for leukemogenesis are unknown. Ayton and Cleary³ previously demonstrated that MLL oncoproteins function as upstream constitutive regulators, aberrantly maintaining expression of HoxA genes. HoxA7 and HoxA9 were required for leukemogenesis in MLL/ENL translocation. However, HoxA9 was reported to be important in defining the phenotype but not the induction of leukemia.⁴ Furthermore, neither HoxA7 nor HoxA9 was reported to be necessary for MLL/GAS7-mediated leukemogenesis.⁵ This indicates that the transformation mechanism of MLL oncoproteins is determined by fusion partners. Thus, identifying the partner genes is important to elucidate the mechanism by which the MLL translocations cause leukemia.

We report identification of a novel MLL fusion partner, CIP29 (a 29-kDa cytokine-inducible protein), in an infant with AML-M4, karyotype t(11;12)(q23; q13). CIP29 was originally isolated from lysates of erythropoietin-stimulated UT-7/Epo cells.⁶ CIP29 has two nuclear localization signals and is localized to the nucleus. Furthermore, CIP29 has a SAP motif, which is involved in chromosomal organization. SAP proteins contribute to the regulation of transcription, DNA repair, RNA processing, and apoptotic chromatin degradation. In fact, CIP29 is involved in DNA transcription accompanying cell-cycle progression or DNA synthesis.⁷ These features of CIP29 suggest MLL/CIP29 fusion protein introduces novel target genes and contribute to activation of MLL as an oncoprotein. Identification of this novel partner protein suggests the need for further study of the molecular mechanisms of leukemic transformation mediated by MLL fusion proteins.

A 7-month-old boy was admitted because of fever and poor complexion. Bone marrow (BM) aspirate was indicative of acute myelomonocytic leukemia (M4-FAB). The karyotype was 46XY, t(11;12)(q23;q13) in 20 BM cells. IRB approval regarding the use of the sample of this patient was obtained. Southern blot analysis of BamHI-digested DNA hybridized with the MLL cDNA probe (0.9 kbp exon 59) demonstrated a rearranged band, suggesting that MLL was involved in the chromosome translocation (data not shown). Furthermore, EcoRI-digested DNA hybridized with a probe covering exons 8–9 demonstrated no rearranged band (data not shown). However, probe encoding exon 10–11 revealed a rearrangement band and this indicated that the breakpoint was located between exons 9 and 10 of MLL (Figure 1a). As all previous reports showed the N-terminal amino acids of the MLL protein were retained in oncogenic fusion proteins, exons 9 of MLL was supposed to fuse to the C-terminal of the partner gene.¹ To identify the C-terminal region of the partner gene, 3' prime rapid amplification of cDNA ends (3' RACE) was performed. This product was subjected to cloning and sequencing, showing that these novel sequences were fused in frame to the N-terminal portion of exon 9 of MLL (Figure 2a). A BLAST SEARCH showed that these sequences were highly homologous to CIP29 (NM033082). To confirm the presence of the fusion gene of MLL/CIP29 in the patient's cells, we carried out PCR with specific primers for MLL and CIP29, MLL-4136 derived from exon 9 and CIP29-256, respectively. PCR amplification of cDNA from the patient's cells yielded products of the predicted sizes (Figure 2b).

Figure 1.

(a) High molecular weight DNA was extracted from BM cells obtained at diagnosis. After digestion with BamHI and EcoRI, DNA was hybridized with DNA probes named probe 8–9 and 10–11 (Figure 1b). Southern blot analysis of genomic DNA extracted from the normal donor (C) and the patient (pt). Genomic DNA was digested with EcoRI and hybridized with a radiolabelled probe derived from exon 10–11. No rearranged band hybridized with probe 8–9 was found. (b) Partial restriction map of the MLL gene spanning from exons 8 to 14 with location of the probes used in this study. B, BamHI, E, EcoRI. (c) Total RNA was extracted from CD34⁺, CD3⁺, or CD19⁺ cells that were isolated from PBMNCs or BM cells from healthy donors, using anti-CD3⁺, CD34⁺, or CD19⁺ MicroBeads (Miltenyi Biotec GmbH, Bergisch Gladbach, Germany). Half of the isolated CD3⁺ cells were stimulated with ionomycin (1 g/ml) (Sigma, St Louis, USA) and phorbol 12-myristate 13-acetate (25 ng/ml) (Sigma, St Louis, MO, USA) for 24 h. Semiquantitative RT-PCR was performed with CIP29-specific primers, CIP29-256 and CIP29-2. -actin was used as a control of semiquantitative PCR.

Full figure and legend (46K)

Figure 2.

(a) Nucleotide sequences and deduced amino-acid sequences of fused MLL/CIP29 cDNAs. After detection of the genomic fusion point between exon 9 and 10, we performed 3' RACE to characterize the 3' sequence of the fusion gene. After obtaining mRNA, SalI+Hex primer (5'-GGG TCG TCG ACA GTC CNN NNN NNN N-3') was used for the first strand of cDNA synthesis. The first PCR was performed using an MLL-3937 primer derived from exon 8 and a SalI primer. For nested PCR, an MLL-4136 primer derived from exon 9 and a SalI primer were used. A 493-bp band obtained was cloned and sequenced. Arrows show the fusion point. An asterisk indicates the termination codon. MLL exon 9 is underlined. A SAP motif is double underlined. Two nuclear localization signals are boxed. (b) To confirm the presence the MLL/CIP29 fusion mRNA, RT-PCR was performed with MLL- and CIP29-specific primers, MLL-4136 and CIP29-256. C: normal control cDNA, Pt: patient cDNA. (c) Schematic representation of the MLL and CIP 29 fusion protein. AT hooks, AT hook DNA-binding motif; MT, DNA methyltransferase homology region; PHD, PHD zinc-finger; TA, transactivation domain; SET, SET domain; NLS nuclear localization signal.

Full figure and legend (142K)

Among peripheral blood mononuclear cells (PBMNCs) as determined by RT-PCR, an increased expression of CIP29 was observed in CD34⁺ cells. In other PBMNC subsets, there were no differences between CD3⁺, CD19⁺, and whole PBMNCs, and comparing stimulated vs resting CD3⁺ lymphocytes (Figure 1c).

We have shown CIP29 is a fusion partner of the MLL gene in an AML-M4 infant with t(11;12)(q23;q13) and the SAP domain of CIP29 is included in the MLL partner protein. CIP29 was originally identified in lysates of erythropoietin-stimulated human UT-7/Epo cells.⁶ CIP29 possesses a single SAP domain and two nuclear localization signals, and this structure suggests that CIP29 is localized primarily to the nucleus (Figure 2c). SAP domain is a putative DNA-binding motif predicted to be involved in chromosomal organization and regulate transcription, DNA repair, RNA processing, and apoptotic chromatin degradation. Fukuda et al⁶ reported that CIP29 was involved in DNA transcription accompanying cell-cycle progression or DNA synthesis. The significance of the SAP domain in leukemogenesis is that it has DNA-binding capacity.

The translocation in which SAP participates was reported in acute megakaryocytic leukemia with t(1;22).⁸ This chromosomal rearrangement results in RBM15-MKL1 fusion. RBM15 encodes RNA-recognition motif (RRM), and spen paralog and ortholog C-terminal (SPOC) domain. MKL1 has a single SAP motif. In RMB15/MKL1, the MKL1 SAP domain would be expected to aberrantly relocalize the RRM and SPOC motifs of RBM15 to transcriptionally active sites on chromatin, deregulating RNA processing and/or Hox and Ras/MAP kinase signalling and altering the normal proliferation and differentiation of megakaryoblasts.

Translocation of CIP29 to MLL introduces a novel DNA-binding domain, the SAP domain. In an analogous way to RBM15/MKL1, fusion of the SAP domain of CIP29 to MLL might relocalize MLL to aberrant genes leading to leukemogenesis. Future studies of delineating molecular functions of the predicted chimeric protein will provide important insight into understanding the mechanisms of MLL-mediated transformation.