Letters to the Editor 1881 13 Zhang M, Zamore PD, Carmo-Fonseca M, Lamond AI, Green MR. Cloning and 15 Emanuel PD. Juvenile myelomonocytic leukemia and chronic myelomonocytic intracellular localization of the U2 small nuclear ribonucleoprotein auxiliary factor leukemia. Leukemia 2008; 22: 1335 -- 1342. small subunit. Proc Natl Acad Sci USA 1992; 89: 8769 -- 8773. 14 Edmond V, Brambilla C, Brambilla E, Gazzeri S, Eymin B. SRSF2 is required for This work is licensed under the Creative Commons Attribution- sodium butyrate-mediated p21(WAF1) induction and premature senescence in NonCommercial-No Derivative Works 3.0 Unported License. To view a human lung carcinoma cell lines. Cell Cycle 2011; 10: 1968 -- 1977. copy of this license, visit http://creativecommons.org/licenses/by-nc-nd/3.0/

Supplementary Information accompanies the paper on the Leukemia website (http://www.nature.com/leu)

Sequencing -modifying identifies UTX mutations in acute lymphoblastic leukemia

Leukemia (2012) 26, 1881--1883; doi:10.1038/leu.2012.56 AML patients, and, where available, used bone marrow samples obtained in complete remission to validate the somatic nature of the mutations. Samples had been collected with patient/parental Mutations affecting epigenetic regulators have long been known informed consent from patients enrolled on Dana--Farber Cancer to have a crucial role in cancer and, in particular, hematological Institute protocols for childhood ALL (DFCI 00-001 (NCT00165178), malignancies.1,2 One of the earliest epigenetic factors described DFCI 05-001 (NCT00400946)) or AML treatment protocols of the altered in leukemia was the mixed lineage leukemia (MLL) German-Austrian AML Study Group (AMLSG) for younger adults which is found translocated in 10% of adult acute myeloid (AMLSG-HD98A (NCT00146120), AMLSG 07-04 (NCT00151242)), leukemia (AML), 30% of secondary AML and 475% of infants with and the study was approved by the IRB of the participating both AML and acute lymphocytic leukemia (ALL). MLL is a SET centers. domain-containing protein, which is recruited to many promoters Using conventional Sanger sequencing of primary leukemia and mediates histone 3 lysine 4 (H3K4) methyltransferase activity, sample-derived genomic DNA, we first screened all coding exons 4,5 thought to promote expression.3 in which mutations have been reported previously. Initially, we In addition to MLL fusions, recently, somatic mutations of UTX analyzed a total of 36 of 174 exons (KDM3B (2/24), KDM5C (9/26), (also known as KDM6A), encoding an H3K27 , were UTX (7/29), MLL2 (8/54), EZH2 (1/20) and SETD2 (9/21)) and found described in multiple hematological malignancies, including 7 non-synonymous tumor-specific aberrations. In AML, we found multiple myeloma and many types of leukemia cell lines.4,5 one EZH2 mutation (p.G648E) in a t(8;21)-positive, and two MLL2 H3K27 methylation is generally thought to cause gene repression. missense mutations (p.R5153Q and p.Y5216S; Table 1) and one Complimentary to UTX, mutations of EZH2, a H3K27 methyltrans- ferase, have been reported in both lymphoid and myeloid tumors (Figure 1).6,7 These mutations lead to altered EZH2 activity and influence H3K27 in tumor cells. Mutations in EZH2, EED and SUZ12, which all cooperate in Polycomb repressive complex 2 have been recently described in early T-cell precursor ALL.8 Similarly, point mutations affecting the functional jumonji C (jmjC) domain of UTX inactivates its H3K27 demethylase activity. In addition, UTX associates with MLL2 in a multiprotein complex, which promotes H3K4 methylation, and recently MLL2 has also been found mutated in cancer, further pointing to a common and complex epigenetic deregulation in cancer.9 In line with the growing evidence for epigenetic regulators as important in tumorigenesis, additional mutations affecting epigenetic regula- tors such as SETD2, a H3K36 methyltransferase, KDM3B, a H3K9 demethylase, and KDM5C, a H3K4 demethylase, have been reported and are associated with distinct patterns (Figure 1).4 Though the clinical significance of these findings remains to be explored, it is evident that epigenetic deregulation is having an important role in both lymphoid and myeloid leukemo- genesis. Furthermore, with novel drugs at hand, such as histone deacetylase inhibitors or demethylating agents that can target and reverse epigenetic alterations, understanding the under- lying molecular aberrations is of growing interest.10 We therefore undertook an effort to examine the prevalence of somatic Figure 1. Histone 3 methylation and selected histone mutations in encoding histone-modifying , in and methyltransferases. Cancers are shown in italics next to the particular, KDM3B, KDM5C, UTX, MLL2, EZH2 and SETD2, which 4,5 mutated protein they are associated with. MM, multiple myeloma; previously were reported mutated in cancer. FL, follicular lymphoma; DLBCL, diffuse large B-cell lymphoma; RCC, For an initial screen, we analyzed banked diagnostic primary renal cell carcinoma; CCC clear cell carcinoma; MPN, myeloproli- leukemia samples from 44 childhood B-cell ALL and 50 adult ferative neoplasm; MB, medulloblastoma.

Accepted article preview online 1 March 2012; advance online publication, 3 April 2012

& 2012 Macmillan Publishers Limited Leukemia (2012) 1879 -- 1898 Letters to the Editor 1882 Table 1. Overview of variants found in AML and ALL

Patient Gender Disease Cytogenetics Gene DNA Protein Somatic

560-D Male AML t(8;21) EZH2 2136G4Aa p.G648Eb Confirmedc 232-D Female AML CN MLL2 15458G4Ad p.R5153Qe Confirmedc 692-D Female AML CN MLL2 15647A4Cd p.Y5216Se Confirmedc 00-171 Female ALL CN MLL2 4498_4499insGGd p.G1500fsX6e Confirmedc 00-A04 Female ALL Not available UTX 4076G4Af p.C1234Yg NAh 05-091 Male ALL t(3;8),+8,dic(9;12) UTX 3711_3721 insCCTTCCGGGGf p.V1113fsX40g Confirmedc 00-D10 Female ALL Not available UTX 3964C4Tf p.L1197Fg NAh 05-357 Male ALL der(19),t(1;19)(q23;q13) UTX 4331G4Af p.W1319Xg Confirmedc 05-046 Female ALL CN UTX 2981A4Tf p.D869Vg Confirmedc Abbreviations: CN, cytogenetically normal; AML, acute myeloid leukemia; ALL, acute lymphocytic leukemia. aEZH2: cDNA reference: NM_004456.4. bEZH2: protein reference: NP_004447.2. cGermline material (remission bone marrow) was available and variant was confirmed to only be present in tumor material. dMLL2: cDNA reference: NM_003482.3. eMLL2: protein reference: NP_003473.3. fUTX: cDNA reference: NM_021140.2. gUTX: protein reference: NP_066963.2. hGermline material was not available for this patient, thus this variant cannot be confirmed to be a somatic mutation.

(c.3711-3721 insCCTTCCGGGG) at codon 1113, leading to 40 missense amino acids before a stop codon (p.V1113fsX40) and one heterozygous nonsense mutation (p.W1319X). On the basis of these findings, we sequenced the remaining 22 UTX exons in our first cohort of 44 ALL cases and screened the entire coding region of UTX in an additional 94 B-cell ALL diagnosis samples. This analysis identified 1 additional missense (p.D869V), making 5 variants in 138 samples (4%; Table 1 and Figure 2). None of these variants were present in dbSNP build 131, which includes 1000 Genomes data. These mutations were validated as somatic in those with germline DNA available (three of five patients). RNA from the patient with the D869V mutation was extracted and RT-PCR performed with UTX transcript-specific primers. Sanger sequencing of the PCR product demonstrated expression of both the wild-type and mutant alleles, in approximately equal amounts (data not shown). In addition, according to profiles banked at the NIH Gene Expression Omnibus, UTX is expressed at high levels in both primary ALL and AML patient samples.11,12 The frameshift observed at codon 1113 of UTX truncates the protein in the jmjC domain (Figure 2b). Similarly located frameshift mutations leading to truncation were described in bladder transitional cell carcinoma, colorectal adenocarcinoma, multiple myeloma and renal cell carcinoma samples.5 The mutation p.W1319X leads to a stop codon located in the C-terminal end, truncating 81 amino acids with a preserved jmjC domain (Figure 2c). Similarly located mutations were described in colorectal adenocarcinoma and renal cell carcinoma samples.5 Of the missense mutations observed in the jmjC domain, p.C1234Y was predicted by PolyPhen13 to be damaging, whereas p.L1197F was predicted to be benign based on the possible impact of the amino-acid substitution on the structure and function of the protein (Figure 2d). UTX is found on the X and, interestingly, 2 of the 5 variants we found were in males and thus hemizygous (p.V1113fsX40 and p.W1319X). In the patient with W1319X, the single-nucleotide variant traces show the remaining wild-type allele, which is likely to be a small amount of normal hematopoietic cells, Figure 2. UTX variants in ALL samples. (a) schematic representation or a tumor cell subclone lacking the mutation. A homolog of UTX, of the protein structure of UTX. Mutations identified in ALL samples called UTY, exists on the Y chromosome; however, in vivo studies are indicated by the asterisks. (b) (top) germline sequence, (bottom) show that purified UTY does not completely recapitulate the forward sequence. (c, f) (top) Germline sequence, (middle) forward 14 (bottom) reverse. (d, e) (top) hg19 reference chromatograph gen- activity of UTX. The other mutations were found heterozygous in erated by mutation surveyor (no germline material available), female patients; however, this gene has been shown to escape X 15 (middle) forward (bottom) reverse. inactivation, consistent with the finding that both alleles are expressed in the patient with the D869V mutation. As most mutations in UTX are thought to cause loss of function, it is possible MLL2 insertion leading to a frameshift in cytogenetically normal that UTX gene dosage may be critical. Alternatively, these mutants AML cases. UTX mutations (Figure 2 and Table 1) were found at a have the potential to act as gain of function dominant negatives, as higher incidence (n ¼ 4), which included two heterozygous they preserve the protein-interacting tetratricopeptide repeats at missense (p.L1197F and p.C1234Y), one hemizygous frameshift the N terminus of UTX.

Leukemia (2012) 1879 -- 1898 & 2012 Macmillan Publishers Limited Letters to the Editor 1883 In conclusion, our analysis in acute leukemia revealed mutations REFERENCES in the histone-modifying enzymes recently identified to be altered 1 Claus R, Plass C, Armstrong SA, Bullinger L. DNA methylation profiling in acute in other cancers, particularly UTX in ALL and MLL2 in AML, myeloid leukemia: from recent technological advances to biological and clinical although not at a high incidence. Important to note is that UTX insights. Future Oncol 2010; 6: 1415 -- 1431. and MLL2 are part of the same protein complex and both MLL2 2 Neff T, Armstrong SA. Chromatin maps, histone modifications and leukemia. and UTX mutations may lead to similar phenotypic consequences Leukemia 2009; 23: 1243 -- 1251. in cancer cells. Also, MLL2 was not fully sequenced in our analysis, 3 Slany RK. The molecular biology of mixed lineage leukemia. Haematologica 2009; leaving open the possibility of additional mutations in AML. 94: 984 -- 993. 4 Dalgliesh GL, Furge K, Greenman C, Chen L, Bignell G, Butler A et al. Systematic Nevertheless, our findings warrant further analyses within larger sequencing of renal carcinoma reveals inactivation of histone modifying genes. studies and most likely more comprehensive studies based on Nature 2010; 463: 360 -- 363. targeted or whole genome/exome next-generation sequencing 5 van Haaften G, Dalgliesh GL, Davies H, Chen L, Bignell G, Greenman C et al. approaches. Of interest, most of the UTX mutations observed in Somatic mutations of the histone H3K27 demethylase gene UTX in human cancer. our study were found in clinically defined high-risk patients and at Nat Genet 2009; 41: 521 -- 523. least 1 of those patients (05-046) eventually relapsed. Therefore, 6 Ernst T, Chase AJ, Score J, Hidalgo-Curtis CE, Bryant C, Jones AV et al. Inactivating this observation might be of particular interest with regard mutations of the histone methyltransferase gene EZH2 in myeloid disorders. to potential epigenetic treatment approaches, although the Nat Genet 2010; 42: 722 -- 726. exact mechanisms of transformation in UTX-mutated ALL remain 7 Morin RD, Johnson NA, Severson TM, Mungall AJ, An J, Goya R et al. Somatic mutations altering EZH2 (Tyr641) in follicular and diffuse large B-cell lymphomas to be elucidated. of germinal-center origin. Nat Genet 2010; 42: 181 -- 185. 8 Zhang J, Ding L, Holmfeldt L, Wu G, Heatley SL, Payne-Turner D et al. The genetic basis of early T-cell precursor acute lymphoblastic leukaemia. Nature 2012; 481: CONFLICT OF INTEREST 157--163. The authors declare no conflict of interest. 9 Parsons DW, Li M, Zhang X, Jones S, Leary RJ, Lin JC et al. The genetic landscape of the childhood cancer medulloblastoma. Science 2011; 331: 435 -- 439. ACKNOWLEDGEMENTS 10 Chin L, Gray JW. Translating insights from the cancer genome into clinical This work was supported by NCI P01CA684841 and the Charles Hood Foundation. practice. Nature 2008; 452: 553 -- 563. LB was supported in part by the Deutsche Forschungsgemeinschaft (Heisenberg- 11 Marston E, Weston V, Jesson J, Maina E, McConville C, Agathanggelou A et al. Stipendium BU 1339/3-1). Stratification of pediatric ALL by in vitro cellular responses to DNA double-strand breaks provides insight into the molecular mechanisms underlying clinical 1,2,5 3,5 4 1,2 1,2 response. Blood 2009; 113: 117 -- 126. BG Mar , L Bullinger , E Basu , K Schlis , LB Silverman , 12 Metzeler KH, Hummel M, Bloomfield CD, Spiekermann K, Braess J, 3 1,2 KDo¨hner and SA Armstrong Sauerland MC et al. An 86-probe-set gene-expression signature predicts 1 Department of Pediatric Oncology, Dana Farber Cancer Institute, survival in cytogenetically normal acute myeloid leukemia. Blood 2008; 112: Boston, MA, USA; 4193 -- 4201. 2Department of Hematology/Oncology, Children’s Hospital Boston, 13 Adzhubei IA, Schmidt S, Peshkin L, Ramensky VE, Gerasimova A, Bork P et al. Boston, MA, USA; A method and server for predicting damaging missense mutations. Nat Methods 3Department of Internal Medicine III, University of Ulm, 2010; 7: 248 -- 249. Ulm, Germany and 14 Lan F, Bayliss PE, Rinn JL, Whetstine JR, Wang JK, Chen S et al. A histone H3 4 lysine 27 demethylase regulates animal posterior development. Nature 2007; 449: Department of Pediatrics, Memorial Sloan Kettering, 689 -- 694. New York, NY, USA 15 Greenfield A, Carrel L, Pennisi D, Philippe C, Quaderi N, Siggers P et al. The E-mail: [email protected] UTX gene escapes X inactivation in mice and humans. Hum Mol Genet 1998; 7: 5 These authors contributed equally to this work. 737 -- 742.

Evaluating surface erythropoietin in multiple myeloma

Leukemia (2012) 26, 1883--1886; doi:10.1038/leu.2012.58 treatment and clinical outcomes would provide a means for directly testing the clinical significance of tumor EpoR expression. Although EpoR mRNA levels can be reliably measured in formalin- Anemia is common in cancer patients and is especially frequent fixed, paraffin-embedded tumor tissue,7 the detection of EpoR in myeloma patients because of the high prevalence of renal protein in these specimens using immunohistochemistry has been insufficiency. Although erythropoietin (Epo) can reduce the need for confounded by the low level of EpoR protein and by inadequate blood transfusions, its use has been tempered by adverse effects on antibody specificity.8 Using a sensitive and specific monoclonal venous thromboembolism, tumor progression, and cancer survival antibody (A82), and semi-quantitative western blotting with in recent clinical trials.1 Although some myeloma patients were whole-cell lysates, scientists at Amgen recently demonstrated included in these trials, no prospective study has evaluated the absence or low expression of EpoR protein (o100 EpoR dimers effect of Epo on survival in myeloma. Epo use was associated with per cell) in 54 human tumor cell lines, whereas 7 other non- significantly reduced myeloma survival in one retrospective study,2 erythroid cell lines expressed between 400 and 3200 EpoR dimers whereas other studies showed that Epo caused myeloma regression per cell.9 However, this study did not examine EpoR levels in in mice,3 that Epo did not impair time to progression or overall myeloma cell lines or primary tumor cells. survival in patients treated with melphalan and prednisone with Unlike most solid tumors, malignant cells from the bone or without bortezomib,4 or that Epo improved survival.5 marrow of patients with multiple myeloma can be examined as Whether tumor epithelial and/or endothelial cells express single cells with minimal processing. This affords the opportunity functional Epo receptor (EpoR) is the subject of controversy.6 to analyze EpoR surface protein levels by flow cytometry. The ability to unequivocally link tumor EpoR expression with Epo A previous study failed to detect surface EpoR in primary

Accepted article preview online 1 March 2012; advance online publication, 16 March 2012

& 2012 Macmillan Publishers Limited Leukemia (2012) 1879 -- 1898