Whole Exome Sequencing Identifies Novel Recurrently Mutated Genes in Patients with Splenic Marginal Zone Lymphoma

Abstract

The pathogenesis of splenic marginal zone lymphoma (SMZL) remains largely unknown. Recent high-throughput sequencing studies have identified recurrent mutations in key pathways, most notably NOTCH2 mutations in >25% of patients. These studies are based on small, heterogeneous discovery cohorts, and therefore only captured a fraction of the lesions present in the SMZL genome. To identify further novel pathogenic mutations within related biochemical pathways, we applied whole exome sequencing (WES) and copy number (CN) analysis to a biologically and clinically homogeneous cohort of seven SMZL patients with 7q abnormalities and IGHV1-2*04 gene usage. We identified 173 somatic non-silent variants, affecting 160 distinct genes. In additional to providing independent validation of the presence of mutation in several previously reported genes (NOTCH2, TNFAIP3, MAP3K14, MLL2 and SPEN), our study defined eight additional recurrently mutated genes in SMZL; these genes are CREBBP, CBFA2T3, AMOTL1, FAT4, FBXO11, PLA2G4D, TRRAP and USH2A. By integrating our WES and CN data we identified three mutated putative candidate genes targeted by 7q deletions (CUL1, EZH2 and FLNC), with FLNC positioned within the well-characterized 7q minimally deleted region. Taken together, this work expands the reported directory of recurrently mutated cancer genes in this disease, thereby expanding our understanding of SMZL pathogenesis. Ultimately, this work will help to establish a stratified approach to care including the possibility of targeted therapy.

About the Authors:

Marina Parry

Contributed equally to this work with: Marina Parry, Matthew J. J. Rose-Zerilli, Jane Gibson

Affiliation: Cancer Sciences, Faculty of Medicine, University of Southampton, Southampton, United Kingdom

Matthew J. J. Rose-Zerilli

Contributed equally to this work with: Marina Parry, Matthew J. J. Rose-Zerilli, Jane Gibson

Affiliation: Cancer Sciences, Faculty of Medicine, University of Southampton, Southampton, United Kingdom

Jane Gibson

Contributed equally to this work with: Marina Parry, Matthew J. J. Rose-Zerilli, Jane Gibson

Affiliation: Human Development and Health, Faculty of Medicine, University of Southampton, Southampton, United Kingdom

Sarah Ennis

Affiliation: Human Development and Health, Faculty of Medicine, University of Southampton, Southampton, United Kingdom

Renata Walewska

Affiliation: Department of Pathology, Royal Bournemouth Hospital, Bournemouth, United Kingdom

Jade Forster

Affiliation: Cancer Sciences, Faculty of Medicine, University of Southampton, Southampton, United Kingdom

Helen Parker

Affiliation: Cancer Sciences, Faculty of Medicine, University of Southampton, Southampton, United Kingdom

Zadie Davis

Affiliation: Department of Pathology, Royal Bournemouth Hospital, Bournemouth, United Kingdom

Anne Gardiner

Affiliation: Department of Pathology, Royal Bournemouth Hospital, Bournemouth, United Kingdom

Andrew Collins

Affiliation: Human Development and Health, Faculty of Medicine, University of Southampton, Southampton, United Kingdom

David G. Oscier

Affiliations Cancer Sciences, Faculty of Medicine, University of Southampton, Southampton, United Kingdom, Department of Pathology, Royal Bournemouth Hospital, Bournemouth, United Kingdom

Jonathan C. Strefford

* E-mail: [email protected]

Affiliation: Cancer Sciences, Faculty of Medicine, University of Southampton, Southampton, United Kingdom

Introduction

Splenic Marginal Zone Lymphoma (SMZL) is a low grade chronic B cell lymphoproliferative disorder that predominantly affects elderly patients and involves the spleen, bone marrow, and peripheral blood [1]. Although the median survival is around 10 years, approximately 70% of SMZL patients require treatment, of whom 25% experience progressive disease, leading to early death [1].

Our understanding of the molecular pathogenesis of SMZL remains limited. Early cytogenetic studies identified recurrent deletions of 7q31-q32 and duplications of 3q in approx. 30% and 20% of cases, respectively [2], but subsequent molecular investigations have failed to identify causative genes within these regions [3]. Candidate gene studies are limited to mutations in TP53, which is disrupted in 10-15% of cases [2], and to genes within the NF-ƘB pathway, which are mutated in a third of all cases [4,5]. The presence of a highly restricted immunoglobulin gene repertoire, in particular the selective usage of the immunoglobulin heavy chain variable (IGHV) 1-2*04 allele in 20-30% of patients, suggests that antigenic stimulation may be important in the pathogenesis of this disease [6].

The recent application of whole exome sequencing to frozen splenic tissue from 14 patients with SMZL followed by targeted resequencing of recurrent variants in larger cohorts has identified further biologically relevant genes [7,8]. Mutations in NOTCH2, which eliminate the C-terminal PEST domain and result in compromised protein degradation, were identified in 20 - 25% of cases although there was no consensus as to the clinical significance of these mutations between studies [7,8]. Gene mutations in modulators or other members of the Notch signalling pathway and in other pathways, such as chromatin remodelling and transcriptional regulation were also implicated [8].

In view of the relatively small number of patients investigated so far and the biological heterogeneity of SMZL, it is vital to perform additional gene discovery experiments to fully catalogue the molecular lesions that contribute to disease pathogenesis. To this aim, we performed whole exome sequencing and copy number analysis of tumour and germ-line DNA extracted from a clinically homogeneous cohort of SMZL patients. In doing so, we expand the reported directory of recurrently mutated cancer genes in this disease, thereby expanding our understanding of SMZL pathogenesis that will ultimately facilitate improvements in disease management and the promise of novel therapies.

Materials and Methods

Patients and biomarker analysis

Seven patients were included in this current study, all met established diagnostic criteria [1], and 5/7 underwent a splenectomy with histology typical of SMZL in each case and no evidence of transformation to a high-grade lymphoma. Each patients harboured chromosomal aberrations targeting 7q and IGHV1-2*04 usage (Table S1), ensuring the exclusion of other types of splenic lymphoma from our analysis and maximizing the likelihood of identifying pathogenic mutations within related biochemical pathways. Informed patient consent was obtained according to the declaration of Helsinki, and the study was ethically approved by the local REC.

Chromosomal analysis was performed and described according to the International System for Human Cytogenetic Nomenclature [9]. Immunoglobulin variable region genes were sequenced from either cDNA or gDNA as previously described [6]. cDNA was synthesised by reverse transcription according to the manufacturers protocol (Promega). gDNA was extracted using the Qiagen Blood Mini Kit and amplified using the BIOMED 2 protocol [10]. PCR products were sequenced directly using an ABI 310 genetic analyser and sequences were aligned to the IMGT-V-Quest database.

High-throughput sequencing, variant calling and Sanger validation

Using targeted exome capture (SureSelect Human All Exon 51Mb V4, 50Mb V3, Agilent) we prepared sequencing libraries from high-molecular weight genomic DNA from CD19 positive-purified tumour cells (five cases extracted from the spleen and two from peripheral blood) and matched saliva cells (Oragene DNA kit, DNA Genotek) prior to high-throughput sequencing with the Illumina HiSeq system. The paired-end sequencing data were aligned against the human genome reference sequence (hg19/GRCh37) using the Novoalign software (novoalignMPI V2.08.02, Novocraft Technologies, Selangor, Malaysia). Duplicate reads, resulting from PCR clonality or optical duplicates, and reads mapping to multiple locations were excluded from downstream analysis. Depth and breadth of sequence coverage was calculated with custom scripts and the BedTools package (v2.13.2) [11] and is included in table S2.

Germ-line-Tumour paired datasets were analysed to identify single nucleotide variations (SNVs) and small insertion and deletions using Varscan 2.3.3 [12] (http://varscan.sourceforge.net). The minimum variant allele frequency threshold was set to 10% with a minimum read depth of 4. Variants were filtered using the ‘somaticFilter’ command to remove clusters of false positives and SNV calls near indels with the same frequency and depth thresholds.

Variants were annotated with respect to genes and transcripts and filtered using the Annovar software tool (v2012Jun21) [13]. Variants were cross referenced with databases of known variation were downloaded from the Annovar website (June 2012); data from the 1000 Genomes Project (2012 April release)[14], dbSNP135 (and a version with SNPs flagged as rare <1% frequency or clinically associated by NCBI) and data from 4300 European American samples from The National Heart Lung and Blood Institute Exome Sequencing Project Exome Variant Server (http://evs.gs.washington.edu/EVS/), (ESP6500 release). Using conventional Sanger sequencing, we confirmed the presence of 38/45 somatic variants (84.4%) and those non-concordant cases were due to low exome read-depth in the tumour sample.

SNP6.0 array hybridization, data extraction and analysis

Tumour and germ-line DNA was purified, amplified, labelled and hybridized to the Affymetrix SNP6.0 platform (Affymetrix, Santa Clara, CA) as previously described [15]. For copy number analysis, two independent researchers visually inspected parallel copy number profiles (aligned to hg19/GRCh37) from tumour and germ-line samples using Partek Genomics Suite (Partek Inc, Missouri, USA), and lesions were classified as somatic if they were present and absent in the tumour and germ-line material, respectively. Copy number alterations (CNAs) were defined as a deviation of 50 consecutive array features (probes) from a normal value of 2 (±0.3), within a consecutive genomic window of 50 Kilobases. The allele ratio was calculated for each sample using the HapMap Allele Reference baseline (Affymetrix) and copy number neutral loss of heterozygosity (CNNLOH) event were defined as somatic if they were present and absent in the tumour and germ-line material, respectively.

Results and Discussion

Exome-capture and high-throughput sequencing allowed us to align approx. 41.9 million reads per sample at a mean depth of 69x (range, 43-109x). In total, an average of 82.2% (range, 70-95%) of target sequences captured at 20x. Our analytical pipeline identified 176 somatic non-silent variants, affecting 165 distinct genes (Table S2). These variants were base-pair transitions (34%), transversions (28%), insertions (6%) and deletions (31%). Copy number analysis identified 28 somatically-acquired copy number deletions (66%) and duplications (33%), (Table S2). Considering the mutation and copy number data together, our patients exhibited an average of 25 somatic mutations (range, 9-40) and four copy number alterations (range, 2-9) per tumour sample.

We initially investigated our exome sequencing data for the presence of somatic variants in genes known to be recurrently mutated in SMZL. In doing so, we identified mutations in NOTCH2 [exon 34, n=2], TNFAIP3 [n=3], MAP3K14 [n=2], MLL2 [n=1] and SPEN [n=1] (Table 1). As the exome capture efficiency of NOTCH2 can compromise variant identification, we also performed Sanger sequencing of exon 34 as previously reported [8]. In doing so, we found no additional mutations. Furthermore, we identified mutations in six genes that have previously been shown to harbour mutations in single SMZL cases [8] (Table 1). This observation implicates these genes as recurrent mutational targets in SMZL. Mutations in several of these genes have been identified in other tumour types, for example FBXO11, which is recurrently mutated in diffuse large B-cell lymphoma (DLBCL) and promote leukaemogenesis by stabilization of BCL6 [16] (Table 1).

Genes Accession Variant nomenclature SIFT score Polyphen-2 score Case no6. Published

number prediction prediction 1 2 3 4 5 6 7 Study

Nucleotide change4 Amino acid change consequences consequences 7

ESTABLISHED: NOTCH2 NM_024408 c.C7081T° p.Q2361X Damaging Truncating ✓ [7,8]

Genes c.6836delA§ p.H2279fs Truncating Truncating ✓

recurrently MAP3K14 NM_003954 c.C200G* p. A67G Damaging Probably damaging ✓ [8]

mutated TNFAIP3 NM_006290 c.T1343A* p.M448K Tolerated Truncating ✓ [4,5,8]

in SMZL1 c.C1681T* p. P561S Tolerated Truncating ✓

c.A328T* p. T110S Tolerated Truncating ✓

MLL2 NM_003482 c.2507_2508insC p.Q836fs Truncating Truncating ✓ [8]

SPEN NM_015001 c.C5179T° p.Q1727X Tolerated Truncating ✓ [8]

c.10286_10289del§ p.3429_3430del Truncating Truncating ✓

NOVEL: AMOTL1 NM_130847 c.G1270A* p.A424T Tolerated Benign ✓ [8]

Recurrent FAT4 NM_024582 c.G6628A* p.A2210T Tolerated Benign ✓ [8]

genes across FBXO11 NM_001190274 c.G1587C* p.W529C Damaging Truncating ✓ [8]

studies2 PLA2G4D NM_178034 c.23delG§ p.G8fs Truncating Truncating ✓ [8]

TRRAP NM_003496 c.367-10T>-^ Splicing5 Truncating Truncating ✓ [8]

USH2A NM_206933 c.G7553C* p.S2518T Tolerated Benign ✓ [8]

NOVEL: CBFA2T3 NM_175931 c.C464T* p.P155L Damaging Probably damaging ✓ Novel

recurrent c.G1445A p. S482N Damaging Probably damaging ✓

genes in our study3 CREBBP NM_001079846 c.A4349G* p.Y1450C Damaging Truncating ✓ ✓ Novel

Table 1. Summary of recurrently mutated genes in our cases and a comparison with previously published studies.

1 Identifies those genes that have previously been shown to be targeted by recurrent mutations in SMZL. 2 Shows those genes that were mutated in single SMZL cases in both our current study and in previously published work. 3 Shows the novel genes targeted by recurrent mutations in our study.

4 Identified non-synonymous (*), splice-site (^), frameshift (§) and stopgain (°) mutations

5 The TRRAP mutation in case 5 occurred within a splice-site and is predicted to resulted in aberrant splicing

6 Showed the presence (✓) and absence (white box) of each mutation in the patients in our series

7 Highlights the published studies that identified the mutations in each of the genes listed

CSV

Download CSV

Next we investigated our SMZL cases for recurrent mutations in genes that have not been previously identified in SMZL (Table 1). This analysis identified two genes, CREBBP and CBFA2T3, both mutated in two patients, which in the context of the published literature provides a potential prevalence of approx. 10% in SMZL. Both of the CREBBP mutations were the Y1412C variant previously identified in DLBCL [17]. CREBBP is involved in chromatin remodelling and transcription factor recognition, and this mutation has been shown to compromise the protein’s ability to acetylate BCL6 and p53 [17]. The CBFA2T3 gene, a core binding factor from the myeloid translocation gene family, is targeted by recurrent chromosomal rearrangements in both lymphoid and myeloid malignancies. Whilst non-synonymous in nature, our mutations were not located within the key ETO, MTG16 or TAFH functional domains of the protein. In pediatric B-cell lymphoma, CBFA2T3 has been implicated as a cellular proto-oncogene as in rare cases the gene is juxtaposed to the immunoglobulin locus [18]. In AML chromosomal inversions involving CBFA2T3 can directly increase the self-renewal capacity of hematopoietic progenitors [19]. Mutations in both these genes were present in approx. 50% of reads, suggesting they are heterozygous mutations present in the dominant tumour clone.

To further assess the potential biological impact of the mutations observed in our cases, pathway analysis was performed using the Database for Annotation, Visualisation and Integrated Discovery (DAVID) (Table 2). In addition to identifying pathways already implicated in SMZL pathogenesis, such as notch signalling (NOTCH2, NOTCH4), we also show that genes within MAPK signalling pathway are targeted by somatic non-synonymous mutations in the majority of our cases (5/7, 71%). Whilst a biological role of these genes in SMZL required functional confirmation, our data does suggest that the MAPK signalling pathway is a major target for somatic mutations in this sub-group of SMZL.

DAVID Pathway Genes Accession numbers Variant nomenclature SIFT score prediction Polyphen-2 score prediction Case no2.

Nucleotide change Amino acid change consequences consequences 1 2 3 4 5 6 7

MAP kinase CACNA1E NM_001205293 c.G1069C p.E357Q Damaging Damaging ✓

CACNA1H NM_021098 c.391delG p.E131fs Truncating Truncating ✓

CACNA2D2 NM_001174051 c.2837delC p.P946fs Truncating Truncating ✓

FLNC NM_001458 c.C3179T p.P1060L Damaging Probably damaging ✓

MAP3K14 NM_003954 c.C200G p. A67G Truncating Truncating ✓ ✓

MAPK8IP3 NM_001040439 c.743delA p.Q248fs Truncating Truncating ✓

RASA1 NM_002890 c.C142A p. P48T Damaging Truncating ✓

TAOK3 NM_016281 c.438-7-T) Splicing1 Truncating Truncating ✓

Notch NOTCH2 NM_024408 c.C7081T p.Q2361X Truncating Truncating ✓

c.6836delA p.H2279fs Truncating Truncating ✓

PIWIL3 NM_001008496 c.2242delA p.T748fs Truncating Truncating ✓

NOTCH4 NM_004557 c.C5877G p.C1959W Truncating Damaging ✓

MAML3 NM_018717 c.1513_1514del p. 505_505del Truncating Truncating ✓

Cell cycle CUL1 NM_003592 c.T469G p.Y157D Damaging Probably damaging ✓

CREBBP NM_001079846 c.A4349G p.Y1450C Damaging Truncating ✓ ✓

CDC27 NM_001114091 c.A701C p. Y234S Tolerated Benign ✓

Cytokine-cytokine FLT1 NM_002019 c.2594_splice splicing Truncating Truncating ✓

receptor interaction CRLF2 NM_022148 c.G340C p.V114L Tolerated Probably damaging ✓

Table 2. Summary of the pathways in which mutated genes in our SMZL cohort can be found and their predicted functional consequences.

1 The TRRAP mutation in case 5 occurred within a splice-site and is predicted to resulted in aberrant splicing

2 Showed the presence (✓) and absence (white box) of each mutation in the patients in our series

CSV

Download CSV

Finally, we identified somatically acquired mutations in genes also targeted by 7q deletions in our patients. In doing so, we found CUL1, FLNC and EZH2 mutations in individual cases (Table 2). Of these gene mutations, only FLNC was located within the published 7q MDR [3,20]. FLNC mutations have not been previously identified in a series eight del(7q) cases [20], suggesting that the prevalence of FLNC mutation is low in this sub-type of SMZL. However, further research will be required to establish if rare mutations represent only one mechanism of gene deregulation, as repression of FLNC transcription by promoter methylation in prevalent in several other human cancer types [21-24]. The somatic variant we identified in EZH2 (p.K199N), which is located outside the SET protein domain, is not the activating mutation prevalent in follicular lymphoma and DLBCL [25] nor has it been previously reported in AML or MDS [26].

Herein, for the first time, we report the analysis of a homogeneous cohort of SMZL cases using whole exome sequencing and copy number analysis. In doing so, we validate the presence of recurrent mutations in several genes with established importance in SMZL. Furthermore, we expand the reported directory of recurrently mutation cancer genes in this disease, with the most significant observation being the identification of recurrent mutations in CREBBP and CBFA2T3. The importance of CREBBP is further strengthen by the presence of a single SMZL case in the literature with a small deletion that juxtaposes 16 exons of CREBBP with the ZNF434 gene, resulting in loss of the acetyltransferase domain of the CREBBP protein [8]. Furthermore, we show the majority of cases in our series carried mutations within MAPK signalling genes, suggesting that mutations in these genes are strongly associated with 7q-rearranged SMZL with IGHV1-2*04 usage. Whilst our analysis identifies a series of novel genes mutated in SMZL, a larger study is required to determine the frequency of these events and any utility in the risk-adapted stratification of SZML patients. To this aim, we are currently coordinating a pan-European study into the presence of somatic mutations in approx. 750 genes with a known or postulated role in SMZL pathophysiology in a cohort of more than 300 SMZL cases. This will ultimately establish the frequency and clinical importance of gene mutations in SMZL and help to establish a stratified approach to care including the possibility of targeted therapy.

Supporting Information

[Figure omitted. See PDF.]

Table S1.

Clinical characteristics of each patient included in the study.

https://doi.org/10.1371/journal.pone.0083244.s001

(DOCX)

Table S2.

Shows the 176 non-silent somatic mutations identified in our cases.

https://doi.org/10.1371/journal.pone.0083244.s002

(XLSX)

Author Contributions

Conceived and designed the experiments: JS DO. Performed the experiments: MP MRZ JF HP ZD AG. Analyzed the data: JG SE AC. Contributed reagents/materials/analysis tools: JG SE AC DO RW. Wrote the manuscript: MP DO JS.

Citation: Parry M, Rose-Zerilli MJJ, Gibson J, Ennis S, Walewska R, Forster J, et al. (2013) Whole Exome Sequencing Identifies Novel Recurrently Mutated Genes in Patients with Splenic Marginal Zone Lymphoma. PLoS ONE 8(12): e83244. https://doi.org/10.1371/journal.pone.0083244

References

1. Matutes E, Oscier D, Montalban C, Berger F, Callet-Bauchu E et al. (2008) Splenic marginal zone lymphoma proposals for a revision of diagnostic, staging and therapeutic criteria. Leukemia 22: 487-495. doi:10.1038/sj.leu.2405068. PubMed: 18094718.

2. Salido M, Baró C, Oscier D, Stamatopoulos K, Dierlamm J et al. (2010) Cytogenetic aberrations and their prognostic value in a series of 330 splenic marginal zone B-cell lymphomas: a multicenter study of the Splenic B-Cell Lymphoma Group. Blood 116: 1479-1488. doi:10.1182/blood-2010-02-267476. PubMed: 20479288.

3. Watkins AJ, Hamoudi RA, Zeng N, Yan Q, Huang Y et al. (2012) An integrated genomic and expression analysis of 7q deletion in splenic marginal zone lymphoma. PLOS ONE 7: e44997. doi:10.1371/journal.pone.0044997. PubMed: 23028731.

4. Rossi D, Deaglio S, Dominguez-Sola D, Rasi S, Vaisitti T et al. (2011) Alteration of BIRC3 and multiple other NF-κB pathway genes in splenic marginal zone lymphoma. Blood 118: 4930-4934. doi:10.1182/blood-2011-06-359166. PubMed: 21881048.

5. Novak U, Rinaldi A, Kwee I, Nandula SV, Rancoita PM et al. (2009) The NF-{kappa}B negative regulator TNFAIP3 (A20) is inactivated by somatic mutations and genomic deletions in marginal zone lymphomas. Blood 113: 4918-4921. doi:10.1182/blood-2008-08-174110. PubMed: 19258598.

6. Bikos V, Darzentas N, Hadzidimitriou A, Davis Z, Hockley S et al. (2012) Over 30% of patients with splenic marginal zone lymphoma express the same immunoglobulin heavy variable gene: ontogenetic implications. Leukemia 26: 1638-1646. doi:10.1038/leu.2012.3. PubMed: 22222599.

7. Kiel MJ, Velusamy T, Betz BL, Zhao L, Weigelin HG et al. (2012) Whole-genome sequencing identifies recurrent somatic NOTCH2 mutations in splenic marginal zone lymphoma. J Exp Med 209: 1553-1565. doi:10.1084/jem.20120910. PubMed: 22891276.

8. Rossi D, Trifonov V, Fangazio M, Bruscaggin A, Rasi S et al. (2012) The coding genome of splenic marginal zone lymphoma: activation of NOTCH2 and other pathways regulating marginal zone development. J Exp Med 209: 1537-1551. doi:10.1084/jem.20120904. PubMed: 22891273.

9. ISCN (2009) An International System for Human Cytogenetic Nomenclature: Recommendations of the International Standing Committee on Human Cytogenetic Nomenclature Shaffer LG, Slovak ML, Campbell LJ, editors. Basel: S.Karger.

10. Matthews C, Catherwood M, Morris TC, Alexander HD (2004) Routine analysis of IgVH mutational status in CLL patients using BIOMED-2 standardized primers and protocols. Leuk Lymphoma 45: 1899-1904. doi:10.1080/10428190410001710812. PubMed: 15223652.

11. Quinlan AR, Hall IM (2010) BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics 26: 841-842. doi:10.1093/bioinformatics/btq033. PubMed: 20110278.

12. Koboldt DC, Zhang Q, Larson DE, Shen D, McLellan MD et al. (2012) VarScan 2: somatic mutation and copy number alteration discovery in cancer by exome sequencing. Genome Res 22: 568-576. doi:10.1101/gr.129684.111. PubMed: 22300766.

13. Wang K, Li M, Hakonarson H (2010) ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res 38: e164. doi:10.1093/nar/gkq603. PubMed: 20601685.

14. Consortium. GP. , Durbin RM, Abecasis GR, Altshuler DL, Auton A et al. (2010) A map of human genome variation from population-scale sequencing. Nature 467: 1061-1073. doi:10.1038/nature09534. PubMed: 20981092. Available online at: Available online at: PubMed: 20981092.

15. Parker H, Rose-Zerilli MJ, Parker A, Chaplin T, Chen X et al. (2011) 13q deletion anatomy and disease progression in patients with chronic lymphocytic leukemia. Leukemia 25: 489-497. doi:10.1038/leu.2010.288. PubMed: 21151023.

16. Duan S, Cermak L, Pagan JK, Rossi M, Martinengo C et al. (2012) FBXO11 targets BCL6 for degradation and is inactivated in diffuse large B-cell lymphomas. Nature 481: 90-93. PubMed: 22113614.

17. Pasqualucci L, Dominguez-Sola D, Chiarenza A, Fabbri G, Grunn A et al. (2011) Inactivating mutations of acetyltransferase genes in B-cell lymphoma. Nature 471: 189-195. doi:10.1038/nature09730. PubMed: 21390126.

18. Salaverria I, Akasaka T, Gesk S, Szczepanowski M, Burkhardt B et al. (2012) The CBFA2T3/ACSF3 locus is recurrently involved in IGH chromosomal translocation t(14;16)(q32;q24) in pediatric B-cell lymphoma with germinal center phenotype. Genes Chromosomes Cancer 51: 338-343. doi:10.1002/gcc.21919. PubMed: 22420028.

19. Gruber TA, Larson Gedman A, Zhang J, Koss CS, Marada S, et al. (2012) An Inv(16)(p13.3q24.3)-encoded CBFA2T3-GLIS2 fusion protein defines an aggressive subtype of pediatric acute megakaryoblastic leukemia. Cancer Cell 22: 683-697.

20. Fresquet V, Robles EF, Parker A, Martinez-Useros J, Mena M et al. (2012) High-throughput sequencing analysis of the chromosome 7q32 deletion reveals IRF5 as a potential tumour suppressor in splenic marginal-zone lymphoma. Br J Haematol 158: 712-726. doi:10.1111/j.1365-2141.2012.09226.x. PubMed: 22816737.

21. Yi JM, Dhir M, Van Neste L, Downing SR, Jeschke J et al. (2011) Genomic and epigenomic integration identifies a prognostic signature in colon cancer. Clin Cancer Res 17: 1535-1545. doi:10.1158/1078-0432.CCR-10-2509. PubMed: 21278247.

22. Vanaja DK, Ehrich M, Van den Boom D, Cheville JC, Karnes RJ et al. (2009) Hypermethylation of genes for diagnosis and risk stratification of prostate cancer. Cancer Invest 27: 549-560. doi:10.1080/07357900802620794. PubMed: 19229700.

23. Kim JH, Jung EJ, Lee HS, Kim MA, Kim WH (2009) Comparative analysis of DNA methylation between primary and metastatic gastric carcinoma. Oncol Rep 21: 1251-1259. PubMed: 19360301.

24. Imura M, Yamashita S, Cai LY, Furuta J, Wakabayashi M et al. (2006) Methylation and expression analysis of 15 genes and three normally-methylated genes in 13 Ovarian cancer cell lines. Cancer Lett 241: 213-220. doi:10.1016/j.canlet.2005.10.010. PubMed: 16303245.

25. Morin RD, Johnson NA, Severson TM, Mungall AJ, An J et al. (2010) Somatic mutations altering EZH2 (Tyr641) in follicular and diffuse large B-cell lymphomas of germinal-center origin. Nat Genet 42: 181-185. doi:10.1038/ng.518. PubMed: 20081860.

26. Ernst T, Chase AJ, Score J, Hidalgo-Curtis CE, Bryant C et al. (2010) Inactivating mutations of the histone methyltransferase gene EZH2 in myeloid disorders. Nat Genet 42: 722-726. doi:10.1038/ng.621. PubMed: 20601953.