Search

Acute myeloid leukemia (AML) is a molecularly complex disease characterized by heterogeneous tumor genetic profiles and involving numerous pathogenic mechanisms and pathways. Integration of molecular data types across multiple patient cohorts may advance current genetic approaches for improved subclassification and understanding of the biology of the disease. Here, we analyzed genome-wide DNA methylation in 649 AML patients using Illumina arrays and identified a configuration of 13 subtypes (termed “epitypes”) using unbiased clustering. Integration of genetic data revealed that most epitypes were associated with a certain recurrent mutation (or combination) in a majority of patients, yet other epitypes were largely independent. Epitypes showed developmental blockage at discrete stages of myeloid differentiation, revealing epitypes that retain arrested hematopoietic stem-cell-like phenotypes. Detailed analyses of DNA methylation patterns identified unique patterns of aberrant hyper- and hypomethylation among epitypes, with variable involvement of transcription factors influencing promoter, enhancer, and repressed regions. Patients in epitypes with stem-cell-like methylation features showed inferior overall survival along with up-regulated stem cell gene expression signatures. We further identified a DNA methylation signature involving STAT motifs associated with FLT3-ITD mutations. Finally, DNA methylation signatures were stable at relapse for the large majority of patients, and rare epitype switching accompanied loss of the dominant epitype mutations and reversion to stem-cell-like methylation patterns. These results show that DNA methylation-based classification integrates important molecular features of AML to reveal the diverse pathogenic and biological aspects of the disease.

Background Acute myeloid leukemia (AML) is a clinically and molecularly heterogeneous disease with poor survival. Recurrent genetic aberrations, such as chromosomal rearrangements and gene mutations, are currently used for patient classification and prognosis, and form the basis of our current understanding of pathogenic mechanisms. However, these markers incompletely predict disease behavior and outcomes. Alterations in DNA methylation patterns are a major hallmark of cancer and recent studies have demonstrated differential global DNA methylation patterns among AML patients. Here we sought to define a novel approach to classify AML patients using genome-wide DNA methylation patterns and to uncover the biological basis and features associated with these epigenetic patterns by integrating mutation, gene expression, and functional data. Methods We analyzed genome-wide DNA methylation data from Illumina arrays on 649 AML cases combined from the Beat AML consortium housed at the Oregon Health & Science University (OHSU) (n=226), The Ohio State University (n=27), TCGA (n=190), and other published studies (n=206). Small molecule inhibitor response data was obtained from the Beat AML consortium. RNA sequencing and survival analyses were performed by Kaplan-Meier analysis in the TCGA and OHSU cohorts. Results K-mediods based clustering on the 500 most-variable CpGs identified 13 distinct subtypes, we termed AML 'epitypes' (Fig. 1A). Comparative analyses incorporating normal hematopoietic cells revealed that AML epitypes retain unique combinations of biological features including the degree of myeloid development, chromatin landscapes, transcription factor binding and gene expression patterns/pathways. Eleven of 13 epitypes were associated with somatic genetic aberrations, i.e. mutations (or combinations thereof) in NPM1, IDH1/2, TET2, DNMT3A, CEBPA, and chromosomal rearrangements t(8;21), t(15;17) and inv(16) accounted for the majority of patients within each respective epitype. The remaining two epitypes were not associated with dominant genetic events, rather they clustered closely with normal CD34+ hematopoietic stem and progenitor cells. These 'stem-like' epigenetic patterns accounted for 27% of patients and showed limited correlation to other leukemic stem cell gene expression signatures. Stem-like epitypes demonstrated broad resistance to a range of small molecule inhibitors. Differential gene expression analysis between normal cells and epitypes revealed enrichment of the JAK-STAT pathway along with other inflammatory pathways, such as interleukin and interferon signaling, selectively in the stem-like epitypes. Patients with stem-like epitypes exhibited decreased overall survival relative to others independent of age (P We next investigated the impact of STAT pathway activation on AML epigenetic patterns. FLT3 mutations are known to selectively activate the STAT pathway in AML and are associated with inferior outcomes. Controlling for the distribution of FLT3 mutations across epitypes, we uncovered a hypomethylation signature enriched in STAT transcription factor binding and FLT3 mutations which we have termed the 'STAT hypomethylation signature' (SHS). SHS-positivity was not restricted to FLT3 mutations, as 34% of SHS+ AMLs lacked FLT3 mutations and, conversely, SHS was absent in 24% of FLT3 mutated patients. SHS-positivity did not impact outcome in the stem-like epitypes; however, within non-stem-like epitypes, SHS+ patients displayed significantly poorer outcomes (P=0.021; Fig 1C). Conclusion Applying our integrative approach of using global DNA methylation profiles to subclassify AML patients, we have uncovered new insights into potential pathogenic mechanisms by which genetic and epigenetic aberrations may contribute to discrete avenues of AML development. Distinct global DNA methylation patterns are commonly driven by genetic aberrations, while others were associated with stem cell-like features and inflammatory pathways. The identification of JAK/STAT pathway upregulation in the stem-like epitype and a subset of non-stem-like samples highlights that nearly half of AML patients exhibit evidence of JAK/STAT activation and experience poor outcomes. Disclosures Druker: Dana-Farber Cancer Institute (antibody royalty): Patents & Royalties: #2524, antibody royalty; Bristol-Myers Squibb: Patents & Royalties, Research Funding; Pfizer: Other: PI or co-investigator on clinical trial(s) funded via contract with OHSU., Research Funding; Bristol-Myers Squibb: Other: PI or co-investigator on clinical trial(s) funded via contract with OHSU., Research Funding; Merck & Co: Patents & Royalties: Dana-Farber Cancer Institute license #2063, Monoclonal antiphosphotyrosine antibody 4G10, exclusive commercial license to Merck & Co; Celgene: Consultancy; Vivid Biosciences: Membership on an entity's Board of Directors or advisory committees, Other: Stock options; The RUNX1 Research Program: Membership on an entity's Board of Directors or advisory committees; Patient True Talk: Consultancy; GRAIL: Equity Ownership, Other: former member of Scientific Advisory Board; Beat AML LLC: Other: Service on joint steering committee; Cepheid: Consultancy, Honoraria; Burroughs Wellcome Fund: Membership on an entity's Board of Directors or advisory committees; Blueprint Medicines: Consultancy, Equity Ownership, Membership on an entity's Board of Directors or advisory committees; Beta Cat: Membership on an entity's Board of Directors or advisory committees, Other: Stock options; Aptose Biosciences: Consultancy, Equity Ownership, Membership on an entity's Board of Directors or advisory committees; Amgen: Equity Ownership, Membership on an entity's Board of Directors or advisory committees; ALLCRON: Membership on an entity's Board of Directors or advisory committees; ICON: Other: Scientific Founder of Molecular MD, which was acquired by ICON in Feb. 2019; Gilead Sciences: Other: former member of Scientific Advisory Board; CureOne: Membership on an entity's Board of Directors or advisory committees; Pfizer: Research Funding; Aileron Therapeutics: #2573, Constructs and cell lines harboring various mutations in TNK2 and PTPN11, licensing fees , Membership on an entity's Board of Directors or advisory committees; Monojul: Other: former consultant; Novartis: Other: PI or co-investigator on clinical trial(s) funded via contract with OHSU., Patents & Royalties: Patent 6958335, Treatment of Gastrointestinal Stromal Tumors, exclusively licensed to Novartis, Research Funding; OHSU (licensing fees): Patents & Royalties: #2573, Constructs and cell lines harboring various mutations in TNK2 and PTPN11, licensing fees . Tyner:Petra: Research Funding; Janssen: Research Funding; Aptose: Research Funding; Constellation: Research Funding; Genentech: Research Funding; Takeda: Research Funding; Syros: Research Funding; Array: Research Funding; AstraZeneca: Research Funding; Array: Research Funding; Incyte: Research Funding; Gilead: Research Funding; Janssen: Research Funding; Takeda: Research Funding; Seattle Genetics: Research Funding; AstraZeneca: Research Funding; Syros: Research Funding; Seattle Genetics: Research Funding; Constellation: Research Funding; Aptose: Research Funding; Incyte: Research Funding; Agios: Research Funding; Petra: Research Funding; Gilead: Research Funding; Genentech: Research Funding; Agios: Research Funding. Byrd:Acerta: Research Funding; Novartis: Other: Travel Expenses, Speakers Bureau; Genentech: Research Funding; Pharmacyclics LLC, an AbbVie Company: Other: Travel Expenses, Research Funding, Speakers Bureau; Janssen: Consultancy, Other: Travel Expenses, Research Funding, Speakers Bureau; Janssen: Consultancy, Other: Travel Expenses, Research Funding, Speakers Bureau; Novartis: Other: Travel Expenses, Speakers Bureau; Gilead: Other: Travel Expenses, Research Funding, Speakers Bureau; Gilead: Other: Travel Expenses, Research Funding, Speakers Bureau; Novartis: Other: Travel Expenses, Speakers Bureau; Ohio State University: Patents & Royalties: OSU-2S; Ohio State University: Patents & Royalties: OSU-2S; BeiGene: Research Funding; Acerta: Research Funding; BeiGene: Research Funding; Pharmacyclics LLC, an AbbVie Company: Other: Travel Expenses, Research Funding, Speakers Bureau; Genentech: Research Funding; Gilead: Other: Travel Expenses, Research Funding, Speakers Bureau; TG Therapeutics: Other: Travel Expenses, Research Funding, Speakers Bureau; TG Therapeutics: Other: Travel Expenses, Research Funding, Speakers Bureau; Pharmacyclics LLC, an AbbVie Company: Other: Travel Expenses, Research Funding, Speakers Bureau; Acerta: Research Funding; Ohio State University: Patents & Royalties: OSU-2S; BeiGene: Research Funding; Genentech: Research Funding; Janssen: Consultancy, Other: Travel Expenses, Research Funding, Speakers Bureau; TG Therapeutics: Other: Travel Expenses, Research Funding, Speakers Bureau.

DNA methylation is associated with malignant transformation, but limitations imposed by genetic variability, tumor heterogeneity, availability of paired normal tissues and methodologies for global assessment of DNA methylation have limited progress in understanding the extent of epigenetic events in the initiation and progression of human cancer and in identifying genes that undergo methylation during cancer. We developed a mouse model of T/natural killer acute lymphoblastic leukemia that is always preceded by polyclonal lymphocyte expansion to determine how aberrant promoter DNA methylation and consequent gene silencing might be contributing to leukemic transformation. We used restriction landmark genomic scanning with this mouse model of preleukemia reproducibly progressing to leukemia to show that specific genomic methylation is associated with only the leukemic phase and is not random. We also identified Idb4 as a putative tumor-suppressor gene that is methylated in most mouse and human leukemias but in only a minority of other human cancers.

Aberrant DNA methylation patterning occurs in many cancers, including chronic lymphocytic leukemia (CLL). Methylation programming changes as B cells differentiate from immature progenitors to mature B cells. CLL clones originate from a continuum of B cell maturation states, which differ in their "methylome." Emerging evidence suggests that aberrant methylation programming occurs at numerous genes that are differentially expressed in IGHV-defined CLL subtypes. Nuclear receptor interacting protein 1 (NRIP1) reproducibly shows significant gene expression differences in IGHV unmutated (U-CLL) and mutated CLL (M-CLL), with lower transcript levels observed in U-CLL cells. Low expression of NRIP1 is associated with poorer overall survival and time-to-treatment. NRIP1 encodes a cofactor that interacts with various nuclear receptors and other proteins to regulate transcription of genes involved in cellular proliferation, survival, metabolism, inflammation, and other processes. Published data reveal several CpG dinucleotides near the NRIP1 promoter that undergo increasing methylation during the maturation and differentiation of normal peripheral blood B cells (NBC), concordant with a stepwise decrease in transcript levels of NRIP1 over the course of maturation. Interestingly, our analysis of publically available expression data revealed significantly lower NRIP1 expression in U-CLL cells (the less mature CLL subtype) than NBC, while M-CLL cells (the more mature CLL subtype) express levels comparable to NBC. Given these observations, our objective was to interrogate the promoter and regulatory regions at the NRIP1 locus for differences in CpG methylation in genomic DNA isolated from U-CLL, M-CLL, and NBC. We hypothesized that the NRIP1 locus is more highly methylated in U-CLL than in M-CLL and NBC, and that this increased methylation represses NRIP1 gene transcription, contributing to its aberrantly low expression in U-CLL compared to M-CLL and NBC. We bisulfite-converted genomic DNA from negatively-selected CD19+ treatment-naïve CLL samples (n=74 U-CLL, n=37 M-CLL) and NBC from healthy donors (n=5). We PCR-amplified regions in or near the CpG island of NRIP1 and used Agena Bioscience's EpiTyper kit to prepare the samples for analysis on a MassARRAY spectrometer. EpiTyper software was used to detect mass differences that indicate either non-methylated or methylated CpG-containing fragments, and to calculate their relative frequency in each sample. For each CpG locus that was assayed, we fit a beta-regression model to detect differential methylation between M-CLL and U-CLL. We also computed Pearson correlation coefficients comparing methylation levels to gene expression levels. We assessed 47 CpGs mapping to the promoter and nearby regulatory elements of NRIP1 for methylation frequency and correlation with NRIP1 expression. From our comparison of methylation frequency in U-CLL and M-CLL, 3 adjacent CpG loci were significant at false discovery rate (FDR) =1% (CpG 58, p=0.00016; CpG 144, p=0.00062; and CpG 155, p=0.00287). The three most significant correlation coefficients arose from the same three CpG loci (CpG 58, R=0.677, p=2.2e-16; CpG 144, R=0.539, p=7.16e-10; CpG 155, R=0.651, p=5.77e-15). Unexpectedly, all 3 loci were methylated at higher levels in M-CLL and at lower levels in U-CLL (on average), and expression was positively correlated with methylation levels. These CpGs map near a documented binding site for the repressive transcription factor RUNX3, prompting us to hypothesize that differential methylation in U-CLL versus M-CLL may affect the binding affinity of RUNX3 and contribute to the robust NRIP1 expression differences in these CLL subtypes. Future work will test this hypothesis using gel-shift and luciferase reporter assays followed by chromatin immunoprecipitation experiments. Few studies have focused on aberrant methylation patterning in CLL and the consequences of deregulated gene expression. We are the first to look at NRIP1 for differences in DNA methylation across IGHV subtypes. Future studies aimed at understanding how methylation at this locus affects NRIP1 transcription, as well as how levels of NRIP1 affect cell survival pathways, will enhance our understanding of the role of this gene in CLL, and the contributions of deregulated methylation to disease pathophysiology. Disclosures No relevant conflicts of interest to declare.

The outcome of older (≥ 60 years) acute myeloid leukemia (AML) patients is poor, and novel treatments are needed. In a phase 2 trial for older AML patients, low-dose (20 mg/m2 per day for 10 days) decitabine, a DNA hypomethylating azanucleoside, produced 47% complete response rate with an excellent toxicity profile. To assess the genome-wide activity of decitabine, we profiled pretreatment and post treatment (day 25/course 1) methylomes of marrow samples from patients (n = 16) participating in the trial using deep-sequencing analysis of methylated DNA captured by methyl-binding protein (MBD2). Decitabine significantly reduced global methylation compared with pretreatment baseline (P = .001). Percent marrow blasts did not correlate with global methylation levels, suggesting that hypomethylation was related to the activity of decitabine rather than to a mere decrease in leukemia burden. Hypomethylation occurred predominantly in CpG islands and CpG island-associated regions (P ranged from .03 to .04) A significant concentration (P < .001) of the hypomehtylated CpG islands was found in chromosome subtelomeric regions, suggesting a differential activity of decitabine in distinct chromosome regions. Hypermethylation occurred much less frequently than hypomethylation and was associated with low CpG content regions. Decitabine-related methylation changes were concordant with those previously reported in distinct genes. In summary, our study supports the feasibility of methylome analyses as a pharmacodynamic endpoint for hypomethylating therapies.

Epigenetic inactivation of genes by DNA hypermethylation plays an important role in carcinogenesis. An in vitro model of human breast epithelial cell transformation was used to study epigenetic changes induced by estradiol during the neoplastic process. Different stages of tumor initiation and progression are represented in this model being MCF-10F the normal stage; trMCF cells, the transformed stage; bsMCF cells, the invasive stage and, caMCF cells, the tumor stage. Global methylation studies by restriction landmark genomic scanning (RLGS) showed an increased DNA methylation during the in the invasive and tumor stages. Expression studies showed that NRG1 (neuregulin 1), CSS3 (chondroitin sulfate synthase 3) and SNIP (SNAP-25-interacting protein) were downregulated in the invasive and tumor cells. The transformed cells showed low expression of STXBP6 (amysin) compared to the parental cells MCF-10F. The treatment of these cells with the demethylating agent 5-aza-dC alone or in combination with the histone deacetylase inhibitor trichostatin increased the expression of NRG1, STXBP6, CSS3 and SNIP confirming that DNA methylation plays an important role in the regulation of the expression of these genes. The NRG1 exon 1 has a region located between −136 and +79 (considering +1, the translational initiation site) rich in CpG sites that was analyzed by methylation specific PCR (MSP). NRG1 exon 1 showed progressive changes in the methylation pattern associated with the progression of the neoplastic process in this model; NRG1 exon 1 was unmethylated in MCF-10F and trMCF cells, becoming hypermethylated in the invasive (bsMCF) and tumor (caMCF) stages. Studies of human breast tissue samples showed that NRG1 exon 1 was partially methylated in 14 out of 17 (82.4%) invasive carcinomas although it was unmethylated in normal tissues (8 out of 10 normal breast tissue samples). Furthermore, NRG1 exon 1 was partially methylated in 9 out of 14 (64.3%) morphologically normal tissue samples adjacent to invasive carcinomas.

Aberrant methylation of CpG islands in gene promoters often represents an early clonal event in carcinogenesis. Accordingly, defining methylation profiles may be useful for developing marker panels for early detection or predicting the risk of cancer precursors. To identify specific genes frequently methylated in cervical cancer, we conducted methylation profiling of 20 primary human cervical cancers using NotI-based restriction landmark genomic scanning (RLGS). Of 2,172 RLGS fragments analyzed (average, 1,753 CpG islands per patient), 186 RLGS fragments were lost in at least one tumor and 40 were lost in three or more. Methylation was identified in 19 (95%) of 20 tumor samples compared with normal DNA. Bisulfite sequencing was conducted to confirm RLGS results. Of the confirmed markers frequently methylated, we developed Methylight assays for two corresponding genes, nucleolar protein 4 (NOL4), and lipoma HMGIC fusion partner–like protein 4 (LHFPL4), which were methylated in 85% and 55% of cancers, respectively. Using these assays, we further confirmed frequent CpG island methylation in the original cancers and in another independent series of 15 cervical cancers. We also showed methylation at a reduced frequency in a set of carefully reviewed cytology specimens demonstrating cells exfoliated from cancer precursor lesions. In summary, we identified, for the first time, NOL4 and LHFPL4 as novel methylation targets specific for cervical cancer. Inclusion of NOL4 and LHFPL4 in evaluating methylation panels for early detection, risk prediction, and etiologic research on cervical cancer is warranted. [Cancer Res 2008;68(7):2489–97]

Background Previous studies of individual genes have shown that in a self-enforcing way, dimethylation at histone 3 lysine 9 (dimethyl-H3K9) and DNA methylation cooperate to maintain a repressive mode of inactive genes. Less clear is whether this cooperation is generalized in mammalian genomes, such as mouse genome. Here we use epigenomic tools to simultaneously interrogate chromatin modifications and DNA methylation in a mouse leukemia cell line, L1210. Results Histone modifications on H3K9 and DNA methylation in L1210 were profiled by both global CpG island array and custom mouse promoter array analysis. We used chromatin immunoprecipitation microarray (ChIP-chip) to examine acetyl-H3K9 and dimethyl-H3K9. We found that the relative level of acetyl-H3K9 at different chromatin positions has a wider range of distribution than that of dimethyl-H3K9. We then used differential methylation hybridization (DMH) and the restriction landmark genome scanning (RLGS) to analyze the DNA methylation status of the same targets investigated by ChIP-chip. The results of epigenomic profiling, which have been independently confirmed for individual loci, show an inverse relationship between DNA methylation and histone acetylation in regulating gene silencing. In contrast to the previous notion, dimethyl-H3K9 seems to be less distinct in specifying silencing for the genes tested. Conclusion This study demonstrates in L1210 leukemia cells a diverse relationship between histone modifications and DNA methylation in the maintenance of gene silencing. Acetyl-H3K9 shows an inverse relationship between DNA methylation and histone acetylation in regulating gene silencing as expected. However, dimethyl-H3K9 seems to be less distinct in relation to promoter methylation. Meanwhile, a combination of epigenomic tools is of help in understanding the heterogeneity of epigenetic regulation, which may further our vision accumulated from single-gene studies.

The epigenetic control of gene transcription in cancer has been the theme of many recent studies and therapeutic approaches. Carcinogenesis is frequently associated with hypermethylation and consequent down-regulation of genes that prevent cancer, e.g., those that control cell proliferation and apoptosis. We used the demethylating drug zebularine to induce changes in DNA methylation, then examined patterns of gene expression using cDNA array analysis and Restriction Landmark Genomic Scanning followed by RNase protection assay and reverse transcription-PCR to confirm the results. Microarray studies revealed that many genes were epigenetically regulated by methylation. We concluded that methylation decreased the expression of, or silenced, several genes, contributing to the growth and survival of multiple myeloma cells. For example, a number of genes (BAD, BAK, BIK, and BAX) involved in apoptosis were found to be suppressed by methylation. Sequenced methylation-regulated DNA fragments identified by Restriction Landmark Genomic Scanning were found to contain CpG islands, and some corresponded to promoters of genes that were regulated by methylation. We also observed that after the removal of the demethylating drug, the addition of interleukin 6 restored CpG methylation and re-established previously silenced gene patterns, thus implicating a novel role of interleukin 6 in processes regulating epigenetic gene repression and carcinogenesis.

We developed a mouse model of acute T/natural killer (NK) acute lymphoblastic leukemia (ALL) that is always preceded by polyclonal lymphocyte expansion to determine how aberrant promoter DNA methylation and consequent gene silencing might be contributing to leukemic transformation. We demonstrated the non-random nature of methylation in the transformation from benign polyclonal expansion to leukemia, and identified a novel tumor suppressor gene silenced in the majority of human leukemia but not solid tumors. To this, we utilized restriction landmark genomic scanning (RLGS) on eight IL-15 transgenic (tg) mice with T/NK ALL, and performed RLGS on an additional five spleens from IL-15tg mice without ALL, but with polyclonal T/NK expansion, and on four spleens from wild type (wt) mice. A total of 2447 RLGS fragments with good resolution on each RLGS profile were analyzed. Only 1–2 variable changes were detected in either wt or polyclonal T/NK expansion controls. In contrast, the eight spleens with clonal T/NK leukemia had 45 to 209 changes (1.8 to 8.5% of total fragments) consistent with aberrant DNA methylation. The association of RLGS fragment losses or reduced intensity with leukemic transformation versus polyclonal expansion was highly significant (P