Expression of Cell–Cell Interacting Genes Distinguishes HLXB9/TEL from MLL-Positive Childhood Acute Myeloid Leukemia
Total Page:16
File Type:pdf, Size:1020Kb
Letters to the Editor 1657 an 8 base-pair guanine mononucleotide repeat sequence Conflict of interest makes this reported variant suspicious for an artifact of PCR amplification rather than a true somatic mutation. This is a The authors declare no conflict of interest. well-known phenomenon that is commonly seen as an artifact 1,2 1 1 after PCR amplification of a region of DNA with homopolymer O Abdel-Wahab , O Kilpivaara , J Patel , runs.6,7 Moreover, slipped-strand mispairing of the PCR L Busque3 and RL Levine1,2 1 polymerase resulting in this in vitro frameshift does not Leukemia Service, Memorial Sloan Kettering Cancer Center, 6,7 New York, NY, USA; necessarily occur in every PCR amplification product, which 2 explains the variable presence of this allele in different patients Human Oncology and Pathogenesis Program, Memorial Sloan Kettering Cancer Center, New York, NY, USA and as assessed by Sanger resequencing, which has limited 3Research Centre, Maisonneuve-Rosemont Hospital, sensitivity. To improve our ability to detect this variant, Montreal, Quebec, Canada we perfomed amplification of this region of ASXL1 followed E-mail: [email protected] by sensitive mass spectrometry (Sequenom, San Diego, CA, USA) to distinguish between PCR products with 8 versus 9 guanine nucleotides in paired tumor and normal DNA (Figure 1b). When we performed PCR amplification followed References by (Seqeunom, San Diego, CA, USA) mass spectrometry for this variant in 10 paired samples from samples with myelodysplastic 1 Gelsi-Boyer V, Trouplin V, Adelaide J, Bonansea J, Cervera N, syndrome, myeloproliferative neoplasm and chronic myelomo- Carbuccia N et al. Mutations of polycomb-associated gene ASXL1 in nocytic leukemia, this variant was detected in tumor and normal myelodysplastic syndromes and chronic myelomonocytic leukaemia. Br J Haematol 2009; 145: 788–800. DNA in every instance. This again strongly suggests that this 2 Boultwood J, Perry J, Pellagatti A, Fernandez-Mercado M, alteration is not a somatic mutation. Finally, we performed Fernandez-Santamaria C, Calasanz MJ et al. Frequent mutation of Sanger sequencing of ASXL1 in granulocyte DNA extracted from the polycomb-associated gene ASXL1 in the myelodysplastic 96 individuals with no evidence of any hematologic disorder. syndromes and in acute myeloid leukemia. Leukemia 2010; 24: The c.1934dupG p.Gly646TrpsfsX12 variant was readily appar- 1062–1065. ent in 425% of samples from patients without hematologic 3 Boultwood J, Perry J, Zaman R, Fernandez-Santamaria C, Littlewood T, Kusec R et al. High-density single nucleotide polymorphism array disease (Figure 1c). Although mutations in ASXL1 have been analysis and ASXL1 gene mutation screening in chronic myeloid reported in individuals without clinical evidence of a hemato- leukemia during disease progression. Leukemia 2010; 24: 1139–1145. logic disorder at the time of DNA acquisition, the fact that this 4 Carbuccia N, Murati A, Trouplin V, Brecqueville M, Adelaide J, sample is found repeatedly in paired tumor and normal DNA Rey J et al. Mutations of ASXL1 gene in myeloproliferative makes this unlikely to be a somatic mutation.8 neoplasms. Leukemia 2009; 23: 2183–2186. The findings reported above indicate that the most commonly 5 Carbuccia N, Trouplin V, Gelsi-Boyer V, Murati A, Rocquain J, Adelaide J et al. Mutual exclusion of ASXL1 and NPM1 mutations in reported mutation in ASXL1, to date, is not a somatic mutation. a series of acute myeloid leukemias. Leukemia 2010; 24: 469–473. Much of the literature on the mutational frequency and clinical 6 Fazekas A, Steeves R, Newmaster SG. Improving sequencing quality correlates of ASXL1 mutations should be reanalyzed with this in from PCR products containing long mononucleotide repeats. mind. The 8 mononucleotide guanine repeat sequence in the BioTechniques 2010; 48: 351–355. reference sequence for ASXL1 in this region may confound 7 Clarke LA, Rebelo CS, Goncalves J, Boavida MG, Jordan P. PCR delimitation of the true repeat number in this region. These amplification introduces errors into mononucleotide and dinucleo- tide repeat sequences. Mol Pathol 2001; 54: 351–353. findings also further highlight the importance of the use of paired 8 Abdel-Wahab O, Manshouri T, Patel J, Harris K, Yao J, Hedvat C et al. normal tissue for accurate detection of true somatic mutations Genetic analysis of transforming events that convert chronic myelo- rather than polymorphisms of PCR/sequencing artifacts. proliferative neoplasms to leukemias. Cancer Res 2010; 70: 447–452. Expression of cell–cell interacting genes distinguishes HLXB9/TEL from MLL-positive childhood acute myeloid leukemia Leukemia (2010) 24, 1657–1660; doi:10.1038/leu.2010.146; a separate cohort of MLL-negative infant AML characterized by published online 1 July 2010 an early disease onset (o2 years) as well as t(7;12) HLXB9/TEL ( ¼ MNX1/ETV6) rearrangement and with concomitant high Molecular characterization of leukemic cells is a continuously expression of HLXB9 (MNX1). Surprisingly, all patients relapsed emerging field and has become fundamental for therapy having a 3-year EFS of 0%.2,3 The role of HLXB9, a transcription stratification and prediction of event-free survival (EFS). Infant factor of the family of homeobox proteins is rarely studied in acute myeloid leukemia (AML) is in 460% cases characterized hematopoiesis and the data regarding its ability to cause by a genomic rearrangement involving the mixed lineage malignant transformation of hematopoietic stem cells (HSCs) is leukemia (MLL) locus (11q23) and the expression of a fusion not yet available. Interestingly, germline mutations of HLXB9 protein (450 fusion partners are described). Patients lead to annorectal malformations and Currarino syndrome in with primary diagnosis of MLL-positive leukemia are young children, but hematopoietic abnormalities are not described.4 (o2 years) and have generally an inferior outcome compared The poor clinical outcome in this HLXB9/TEL-positive leukemia with MLL-negative patients.1 We and others recently described subset prompted us to comprehensively characterize the two Leukemia Letters to the Editor 1658 t(7;12) CCGACTTCAACTGCT TGCAGC CA Actin HLXB9 Exon1 TEL Exon3 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 breakpoint t(7;12) vs 11q23 down-regulated up-regulated PBX3 EDIL3 HOXA6 CNTNAP5 HOXA5 ANGPT1 HOXA10 DSG2 HOXA3 ITGA9 HOXA4 ITGAV HOXA7 KDR HOXA9 SIGLEC6 HOXA2 HOX genes cell-cell interacting genes 123456123456 11q23 t(7;12) HLXB9HOXA9 MEIS1 C-MYB HLXB9/ 0.20 0.10 0.50 HLXB9/ neg. 1.40 MLL/ENL HLXB9 TEL+ 0.40 TEL contr. 1.20 0.15 0.08 HLXB9 1.00 0.06 0.30 0.80 0.10 0.60 0.04 0.20 to actin 0.05 0.40 0.02 0.10 0.20 relative expression expression relative 0.00 0.00 0.00 0.00 CFU after 3rd replating t(7;12) 11q23*t(7;12) 11q23* t(7;12) 11q23* t(7;12) 11q23* Figure 1 Gene expression profiling and transformation process. Bone marrow or peripheral blood mononuclear cells were obtained from the leukemia laboratory, Giessen, Germany and informed consent was obtained from all families. (a) RT–PCR amplifying the HLXB9/TEL-fusion transcript. Indicated is the in-frame fragment variant at 194 bp as well as the out-of-frame variant at 330 bp (nested PCR primer: HLXB9-F1 50-CTTCCAGCTGGACCAGTGGCTG-30; TEL-R1 50-CTGAAGGAGTTCATAGAGCACATC-30; HLXB9-F2 50-CACCGCGGGCATGATCCTGC-30; TEL-R2 50-ATCGATAGCGAAAGTCCTCTT-30). Shown are 18 representatives out of 42 screened samples. (b) Sequence of the HLXB9/TEL breakpoint fusing HLXB9 exon 1 and TEL exon 3. (c) Microarray analysis was carried out on RNA samples (Trizol; Invitrogen, Darmstadt, Germany) of six HLXB9/TEL- and six MLL/AF9-positive patients using Human Gene 1.0 ST Array (Affymetrix). In all, 1554 significant differentially regulated genes are illustrated. Red indicates relative upregulation and green downregulation. (d) Listed are selected significantly (Po0,05) altered genes in t(7;12) patients compared with t(11q23) patients representing a group of downregulated HOX genes and upregulated cell–cell interacting genes. (e) Relative expression of HLXB9, HOXA9, MEIS1 and C-MYB in t(7;12) and 11q23 patients. Quantitative real-time PCR was carried out in triplicates using TaqMan primer probe assays (Applied Biosystems, Darmstadt, Germany). Samples were normalized to b-ACTIN and DCt values were calculated. A significant difference was found for HLXB9 expression (Po0,05, Mann–Whitney U-test). t(7;12)-positive AML: n ¼ 5; t(11q23) AML/ALL: n ¼ 8, nMLL-positive ALL and AML patients. (f) Bone marrow of 5-FU-treated wild-type mice was collected, transduced with a g-retroviral pMSCV vector with the transgenes MLL/ENL, HLXB9, HLXB9/TEL, HLXB9/TEL þ HLXB9 or an empty vector and plated on methylcellulose. After three rounds of replating, cells were analyzed for their ability to form colonies. entities of MLL- and HLXB9/TEL-positive AML regarding their polymerase chain reaction, French-American-British (FAB) cellular morphology, transcriptional profile and transformation subtype M5), who met the following criteria: diagnosis of process. AML, ageo2 years, blast content 460%, no trisomy 21. The MLL-positive cohort comprised six patients (t(9;11); Immunologic analysis revealed AML with coexpression of MLL-AF9, verified by fluorescence in situ hybridization and T-cell-associated antigens CD4 (mean: 81.25%), CD7 (mean: 26%) Leukemia Letters to the Editor 1659 Table 1 Patient characteristics Age (years) Karyotype Morphology Immunology V(D)J CD34 CD117 HLXB9/TEL 0.4 47, XX, t(7;16) (q36;q12), +mar M2 CD4/CD7 Neg 72 68 0.7 48, XY, t(7;12)(q36;p13),