Search

DMC identity and annotation for all clusters. This is part 3/3 (file split by columns due to upload limitations)

DMC identity and annotation for all clusters. This is part 1/3 (file split by columns due to upload limitations)

The supplementary methods contain a more detailed description of the analytic and experimental approach beyond the scope of the main methods section. Supplementary Figure S1. Ability of specific genomic lesions to predict epigenetic clustering. Supplementary Figure S2. ERRBS coverage. Supplementary Figure S3. Differential methylation at promoters and active enhancers. Supplementary Figure S4. DMC distribution. Supplementary Figure S5. Transcription factor binding site analysis at differentially methylated active enhancers in IDH2, DNMT3A, and IDH1/DNMT3A AMLs.

DMC identity and annotation for IDH2, DNMT3A, and IDH1/DNMT3A cohorts.

DMC identity and annotation for all clusters. This is part 2/3 (file split by columns due to upload limitations)

Supplementary Table 1 from A Transcriptional Profiling Study of CCAAT/Enhancer Binding Protein Targets Identifies Hepatocyte Nuclear Factor 3β as a Novel Tumor Suppressor in Lung Cancer

We showed previously that CCAAT/enhancer binding protein α (C/EBPα), a tissue-specific transcription factor, is a candidate tumor suppressor in lung cancer. In the present study, we have performed a transcriptional profiling study of C/EBPα target genes using an inducible cell line system. This study led to the identification of hepatocyte nuclear factor 3β (HNF3β), a transcription factor known to play a role in airway differentiation, as a downstream target of C/EBPα. We found down-regulation of HNF3β expression in a large proportion of lung cancer cell lines examined and identified two novel mutants of HNF3β, as well as hypermethylation of the HNF3β promoter. We also developed a tetracycline-inducible cell line model to study the cellular consequences of HNF3β expression. Conditional expression of HNF3β led to significant growth reduction, proliferation arrest, apoptosis, and loss of clonogenic ability, suggesting additionally that HNF3β is a novel tumor suppressor in lung cancer. This is the first study to show genetic abnormalities of lung-specific differentiation pathways in the development of lung cancer.

B-cell lymphoma 6 (BCL6) is a transcription repressor and proto-oncogene that plays a crucial role in the innate and adaptive immune system and lymphoid neoplasms. However, its role in myeloid malignancies remains unclear. Here, we explored the role of BCL6 in acute myeloid leukemia (AML). BCL6 was expressed at variable and often high levels in AML cell lines and primary AML samples. AMLs with higher levels of BCL6 were generally sensitive to treatment with BCL6 inhibitors, with the exception of those with monocytic differentiation. Gene expression profiling of AML cells treated with a BCL6 inhibitor revealed induction of BCL6-repressed target genes and transcriptional programs linked to DNA damage checkpoints and downregulation of stem cell genes. Ex vivo treatment of primary AML cells with BCL6 inhibitors induced apoptosis and decreased colony-forming capacity, which correlated with the levels of BCL6 expression. Importantly, inhibition or knockdown of BCL6 in primary AML cells resulted in a significant reduction of leukemia-initiating capacity in mice, suggesting ablation of leukemia repopulating cell functionality. In contrast, BCL6 knockout or inhibition did not suppress the function of normal hematopoietic stem cells. Treatment with cytarabine further induced BCL6 expression, and the levels of BCL6 induction were correlated with resistance to cytarabine. Treatment of AML patient-derived xenografts with BCL6 inhibitor plus cytarabine suggested enhanced antileukemia activity with this combination. Hence, pharmacologic inhibition of BCL6 might provide a novel therapeutic strategy for ablation of leukemia-repopulating cells and increased responsiveness to chemotherapy. Key Points: • Primary human AML cells with higher levels of BCL6 expression displayed sensitivity when exposed to BCL6 inhibitors. • BCL6 induction contributed to cytarabine resistance, and inhibition of BCL6 in AML cells decreased leukemia stem cell activity.

Chromosomal rearrangements are a frequent cause of oncogene deregulation in human malignancies. Overexpression of EVI1 is found in a subgroup of acute myeloid leukemia (AML) with 3q26 chromosomal rearrangements, which is often therapy resistant. In AMLs harboring a t(3;8)(q26;q24), we observed the translocation of a MYC super-enhancer (MYC SE) to the EVI1 locus. We generated an in vitro model mimicking a patient-based t(3;8)(q26;q24) using CRISPR-Cas9 technology and demonstrated hyperactivation of EVI1 by the hijacked MYC SE. This MYC SE contains multiple enhancer modules, of which only one recruits transcription factors active in early hematopoiesis. This enhancer module is critical for EVI1 overexpression as well as enhancer-promoter interaction. Multiple CTCF binding regions in the MYC SE facilitate this enhancer-promoter interaction, which also involves a CTCF binding site upstream of the EVI1 promoter. We hypothesize that this CTCF site acts as an enhancer-docking site in t(3;8) AML. Genomic analyses of other 3q26-rearranged AML patient cells point to a common mechanism by which EVI1 uses this docking site to hijack enhancers active in early hematopoiesis., Chromosome rearrangements can be a cause of altered oncogene expression in cancer, such as a 3q26 translocation in some acute myeloid leukemias (AML) that leads to overexpression of EVI1. Here the authors engineer this rearrangement in a cell line and show that EVI1 overexpression is a result of ‘enhancer hijacking’ of the MYC superenhancer, which is facilitated by CTCF-mediated loops.

The presence of minimal residual disease (MRD) is widely recognized as a powerful predictor of therapeutic outcome in acute myeloid leukemia (AML), but methods of measurement and quantification of MRD in AML are not yet standardized in clinical practice. There is an urgent, unmet need for robust and sensitive assays that can be readily adopted as real-time tools for disease monitoring. NPM1 frameshift mutations are an established MRD marker present in half of patients with cytogenetically normal AML. However, detection is complicated by the existence of hundreds of potential frameshift insertions, clonal heterogeneity, and absence of sequence information when the NPM1 mutation is identified using capillary electrophoresis. Thus, some patients are ineligible for NPM1 MRD monitoring. Furthermore, a subset of patients with NPM1-mutated AML will have false-negative MRD results because of clonal evolution. To simplify and improve MRD testing for NPM1, we present a novel digital PCR technique composed of massively multiplex pools of insertion-specific primers that selectively detect mutated but not wild-type NPM1. By measuring reaction end points using digital PCR technology, the resulting single assay enables sensitive and specific quantification of most NPM1 exon 12 mutations in a manner that is robust to clonal heterogeneity, does not require NPM1 sequence information, and obviates the need for maintenance of hundreds of type-specific assays and associated plasmid standards.

Cancer-associated mutations in genes encoding RNA splicing factors (SFs) commonly occur in leukemias, as well as in a variety of solid tumors, and confer dependence on wild-type splicing. These observations have led to clinical efforts to directly inhibit the spliceosome in patients with refractory leukemias. Here, we identify that inhibiting symmetric or asymmetric dimethylation of arginine, mediated by PRMT5 and type I protein arginine methyltransferases (PRMTs), respectively, reduces splicing fidelity and results in preferential killing of SF-mutant leukemias over wild-type counterparts. These data identify genetic subsets of cancer most likely to respond to PRMT inhibition, synergistic effects of combined PRMT5 and type I PRMT inhibition, and a mechanistic basis for the therapeutic efficacy of PRMT inhibition in cancer.

Publisher's Note: There is an Inside Blood Commentary on this article in this issue.

Introduction Transcriptional deregulation is a central event in the development of acute myeloid leukemia (AML), with most mutations occurring in genes related to transcription, chromatin regulation and DNA methylation. Furthermore, alterations involving cis-regulatory elements have been shown to play a critical role in aberrant gene expression in AML. Genetic variation in cis-regulatory regions usually involves a single allele, which results in differential expression of the two alleles. This phenomenon, termed allele-specific expression (ASE), is therefore an accurate marker for cis-regulatory variation (Pastinen, 2010). We propose that a systematic study of genes with aberrant ASE in AML may uncover aberrantly expressed genes caused by abnormalities in cis-regulatory elements. Therefore we aim to 1) chart the landscape of ASE in AML, 2) establish a link between relevant ASE events and AML subtypes, and 3) investigate the mechanisms driving ASE. Methods We performed whole exome sequencing (WES) and RNA-seq on leukemic blasts from 168 de novo AML patients, representing all major subtypes of the disease. Combining both datasets, we assessed ASE in every gene with informative (non-homozygous) single nucleotide variants (SNVs). Results Patients had a median of 37 genes with ASE, several of which were recurrently detected across multiple patients. To shorten the gene list we selected for this study genes known to be involved either in cancer or in myeloid development. The gene most commonly found to show ASE (53/140 cases with SNVs) was GATA2, which encodes a transcription factor crucial for proliferation and maintenance of hematopoietic stem cells with a known involvement in AML. Interestingly, integration with molecularly defined classification of AML revealed that all cases (n=17) with biallelic CEBPA mutations exhibited GATA2 ASE (p-value = 6.00·10-7, Fisher's test). Biallelic CEBPA mutations (CEBPA DM) identify an AML subtype with favorable clinical outcome and frequently co-occur with GATA2 mutations (Greif PA, 2012), pointing to a functional connection between these two genes. Indeed, 44% of the cases in our cohort exhibited a GATA2 mutation, and 27% carried a second, subclonal mutation in the same gene. Importantly, in cases where a GATA2 mutation was found, the mutant allele was always preferentially expressed. These findings were validated in the TCGA dataset, where all four CEBPA DM patients with informative SNVs in GATA2 exhibited GATA2 ASE. Although GATA2 ASE was present in other AML subtypes, none of these subtypes showed a significant association with this finding. Patients with a t(8;21) rearrangement (n=5), which represses CEBPA expression, did not exhibit GATA2 ASE, and we only observed GATA2 ASE in 4 out of 8 CEBPA silenced leukemias (Wouters BJ, 2007). Altogether, this demonstrates the uniqueness of the 1-to-1 relationship between CEBPA DM and GATA2 ASE, and excludes a causative role for inactive CEBPA protein in mediating mono-allelic expression of GATA2. The average expression of GATA2 in CEBPA DM patients was comparable to other AMLs, even in cases with monoallelic GATA2 expression. This suggests that a) ASE was achieved by repression of one allele rather than dramatically increased expression of the other, b) there was a compensation of the non-repressed allele. DNA methylation analysis of the GATA2 promoter did not reveal methylation-mediated gene silencing of the repressed allele. The long-distance +77 kb GATA2 enhancer appears to be involved in ASE, as RNA read-through levels at the enhancer were significantly different in CEBPA DM AMLs (p-value < 10-4, Wald test) in an allele-specific manner. The involvement of the enhancer was further confirmed by differences in H3K27ac levels between the two alleles. Conclusions An unbiased screen of 168 de novo AML cases revealed that all patients (n=17) with CEBPA biallelic mutations display GATA2 ASE. GATA2 mutations were found in 8 of the 17 cases, always in the allele that is preferentially expressed. Since GATA2 ASE is present in all CEBPA DM and GATA2 mutations only in a fraction, we hypothesize that GATA2 ASE is acquired first and mutations are only selected if they occur in the expressed allele. Moreover, given that other subgroups with CEBPA abnormalities do not show a similar pattern, we propose that ASE of GATA2 is not a consequence of CEBPA mutations, but rather a requirement for the development of AML in these patients. Disclosures No relevant conflicts of interest to declare.

We performed cytosine methylation sequencing on genetically diverse patients with acute myeloid leukemia (AML) and found leukemic DNA methylation patterning is primarily driven by nonpromoter regulatory elements and CpG shores. Enhancers displayed stronger differential methylation than promoters, consisting predominantly of hypomethylation. AMLs with dominant hypermethylation featured greater epigenetic disruption of promoters, whereas those with dominant hypomethylation displayed greater disruption of distal and intronic regions. Mutations in IDH and DNMT3A had opposing and mutually exclusive effects on the epigenome. Notably, co-occurrence of both mutations resulted in epigenetic antagonism, with most CpGs affected by either mutation alone no longer affected in double-mutant AMLs. Importantly, this epigenetic antagonism precedes malignant transformation and can be observed in preleukemic LSK cells from Idh2R140Q or Dnmt3aR882H single-mutant and Idh2R140Q/Dnmt3aR882H double-mutant mice. Notably, IDH/DNMT3A double-mutant AMLs manifested upregulation of a RAS signaling signature and displayed unique sensitivity to MEK inhibition ex vivo as compared with AMLs with either single mutation. Significance: AML is biologically heterogeneous with subtypes characterized by specific genetic and epigenetic abnormalities. Comprehensive DNA methylation profiling revealed that differential methylation of nonpromoter regulatory elements is a driver of epigenetic identity, that gene mutations can be context-dependent, and that co-occurrence of mutations in epigenetic modifiers can result in epigenetic antagonism. Cancer Discov; 7(8); 868–83. ©2017 AACR. This article is highlighted in the In This Issue feature, p. 783

C/EBPs are a family of transcription factors that regulate growth. control and differentiation of various tissues. We found that C/EBP gamma is highly upregulated in a subset of acute Myeloid leukemia (AML) samples characterized by C/EBP alpha hypermethylation/silencing Similarly, C/EBP gamma was upregulated in murine hematopoietic stem/progenitor cells lacking C/EBP alpha, as C/EBP alpha mediates C/EBP gamma suppression. Studies in myeloid cells demonstrated that CEBP6 overexpression blocked neutrophilic differentiation. Further, downregulation of Cebpg in murine Cebpa-deficient stem/progenitor cells or in human CEBPA-silenced AML samples restored granulocytic differentiation hi addition, treatment of these leukemias with demethylating agents restored the. C/EBP alpha-C/EBP gamma balance and upregulated the expression of myeloid differentiation markers. Our-results indicate that C/EBP gamma mediates the myeloid differentiation arrest induced by C/EBP alpha deficiency and that targeting the C/EBP alpha-C/EBP gamma axis rescues neutrophilic differentiation in this unique subset of AMLs.

We evaluated concurrent gene mutations, clinical outcome, and gene expression signatures of CCAAT/enhancer binding protein alpha (CEBPA) double mutations (CEBPA(dm)) versus single mutations (CEBPA(sm)) in 1182 cytogenetically normal acute myeloid leukemia (AML) patients (16-60 years of age). We identified 151 (12.8%) patients with CEBPA mutations (91 CEBPA(dm) and 60 CEBPA(sm)). The incidence of germline mutations was 7% (5 of 71), including 3 C-terminal mutations. CEBPA(dm) patients had a lower frequency of concur-rent mutations than CEBPA(sm) patients (P < .0001). Both, groups were associated with a favorable outcome compared with CEBPA(wt) (5-year overall survival [OS] 63% and 56% vs 39%; P < .0001 and P = .05, respectively). However, in multivariable analysis only CEBPA(dm) was a prognostic factor for favorable OS outcome (hazard ratio [HR] 0.36, P < .0001; event-free survival, HR 0.41, P < .0001; relapse-free survival, HR 0.55, P = .001). Outcome in CEBPA(sm) is dominated by concurrent NPM1 and/or FLT3 internal tandem duplication mutations. Unsupervised and supervised GEP analyses showed that CEBPA(dm) AML (n = 42), but not CEBPA(sm) AML (n = 18), expressed a unique gene signature. A 25-probe set prediction signature for CEBPA(dm) AML showed 100% sensitivity and specificity. Based on these findings, we propose that CEBPA(dm) should be clearly defined from CEBPA(sm) AML and considered as a separate entity in the classification of AML. (Blood. 2011;117(8):2469-2475)

Acute myeloid leukemia (AML) represents a highly heterogeneous myeloid stem cell disorder classified based on various genetic defects. Besides genetic alterations, epigenetic changes are recognized as an additional mechanism contributing to leukemogenesis, but insight into the latter process remains minor. Using a combination of Methyl-CpG-Immunoprecipitation (MCIp-chip) and MALDI-TOF analysis of bisulfite-treated DNA in a cohort of 196 AML patients we previously demonstrated that (cyto)genetically defined AML subtypes, including CBFB-MYH11, AML-ETO, NPM1-mut, CEBPA-mut or IDH1/2-mut subtypes, express specific DNA-methylation profiles (Gebhard et al, Leukemia, 2018). A fraction of AML patients (5/196) displayed a unique abnormal hypermethylation profile that was completely distinct from any other AML subtype. These patients present immature leukemia (FAB M0, M1) with various chromosomal aberrations but very few mutations (e.g. no IDH1/2, KRAS, DNMT3A) that might explain the CpG island methylator phenotype (CIMP) phenotype. The CIMP patients showed high resemblance with a recently reported CEBPA methylated subgroup (Wouters et al, 2007 and Figueroa et al, 2009), which we confirmed by MCIp-chip and MALDI-TOF analysis. To explore the whole range of epigenetic alterations in the CIMP-AML patients we performed in-depth global DNA methylation and gene expression analyses (MCIp-seq and RNA-seq) in 45 AML and 12 CIMP patients from both studies. Principle component analysis and t-distributed stochastic neighbor embedding (t-SNE) revealed that CIMP patients express a unique DNA-methylation and gene-expression signature that separated them from all other AMLs. We could discriminate promoter methylation from non-promoter methylation by selecting MCIp-seq peaks within 3kb around TSS. Promoter hypermethylation was highly associated with repression of genes (PCC = -0.053, p-value = 0.00075). Hypermethylation of non-promoter regions was more strongly associated with upregulation of genes (PCC = 0.046, p-value = 4.613e-06). Interestingly, differentially methylated regions also showed a positive association with myeloid lineage CTCF binding sites (27% vs 18% expected, p-value < 2.2e-16 in a chi-square test of independence). Methylation of CTCF sites causes loss of CTCF binding, which has been reported to disrupt boundaries between so-called topologically associated domains (TADs), allowing enhancers located in a particular TAD to become accessible to genes in adjacent TADs and affect their transcription. Whether this is the case is under investigation. In this study we particularly focused on the role of hypermethylation of promoters in CIMP-AMLs. Promoters of many transcriptional regulators that are involved in the differentiation of myeloid lineages of which several are frequently mutated in AML were hypermethylated and repressed, including CEBPA, CEBPD, IRF8, GATA2, KLF4, MITF or MAFB. Notably, HMGA2, a critical regulator of myeloid progenitor expansion, exhibited the largest degree of CIMP promoter hypermethylation compared to the other AMLs, accompanied by a reduction in gene expression. Moreover, multiple members of the HOXB family and KLF1 (erythroid differentiation) were methylated and repressed as well. In addition, these patients frequently showed hypermethylation of many chromatin factors (e.g. LMNA, CHD7 or TET2). Hypermethylation of the TET2 promoter could result in a loss of maintenance DNA demethylation and therefore successive hypermethylation at CpG islands. We carried out regulome-capture-bisulfite sequencing on CIMP-AMLs compared to other AML samples and normal blood cell controls and confirmed methylation of the same transcription and chromatin factor promoters. We conclude that these leukemias represent very primitive HSCPs which are blocked in differentiation into multiple hematopoietic lineages, due to the absence of regulators of these lineages. Although the underlying cause for the extreme hypermethylation signature is still subject to ongoing studies, the consequence of promoter hypermethylation is silencing of key lineage regulators causing the differentiation arrest in these cells. We argue that these patients may particularly benefit from therapies that revert DNA methylation. Disclosures Ehninger: Cellex Gesellschaft fuer Zellgewinnung mbH: Employment, Equity Ownership; GEMoaB Monoclonals GmbH: Employment, Equity Ownership; Bayer: Research Funding. Thiede:AgenDix: Other: Ownership; Novartis: Honoraria, Research Funding.

We examined the gene expression profiles of two independent cohorts of patients with acute myeloid leukemia [n=247 and n=214 (younger than or equal to 60 years)] to study the applicability of gene expression profiling as a single assay in prediction of acute myeloid leukemia-specific molecular subtypes. The favorable cytogenetic acute myeloid leukemia subtypes, i.e., acute myeloid leukemia with t(8;21), t(15;17) or inv(16), were predicted with maximum accuracy (positive and negative predictive value: 100%). Mutations in NPM1 and CEBPA were predicted less accurately (positive predictive value: 66% and 100%, and negative predictive value: 99% and 97% respectively). Various other characteristic molecular acute myeloid leukemia subtypes, i.e., mutant FLT3 and RAS, abnormalities involving 11q23, -5/5q-, -7/7q-, abnormalities involving 3q (abn3q) and t(9;22), could not be correctly predicted using gene expression profiling. In conclusion, gene expression profiling allows accurate prediction of certain acute myeloid leukemia subtypes, e.g. those characterized by expression of chimeric transcription factors. However, detection of mutations affecting signaling molecules and numerical abnormalities still requires alternative molecular methods.

The past decade has shown a marked increase in the use of high-throughput assays in clinical research into human cancer, including acute myeloid leukemia (AML). In particular, genome-wide gene expression profiling (GEP) using DNA microarrays has been extensively used for improved understanding of the diagnosis, prognosis, and pathobiology of this heterogeneous disease. This review discusses the progress that has been made, places the technologic limitations in perspective, and highlights promising future avenues

Mutations in CCAAT/enhancer binding protein α (CEBPA) are seen in 5% to 14% of acute myeloid leukemia (AML) and have been associated with a favorable clinical outcome. Most AMLs with CEBPA mutations simultaneously carry 2 mutations (CEBPAdouble-mut), usually biallelic, whereas single heterozygous mutations (CEBPAsingle-mut) are less frequently seen. Using denaturing high-performance liquid chromatography and nucleotide sequencing, we identified among a cohort of 598 newly diagnosed AMLs a subset of 41 CEBPA mutant cases (28 CEBPAdouble-mut and 13 CEBPAsingle-mut cases). CEBPAdouble-mut associated with a unique gene expression profile as well as favorable overall and event-free survival, retained in multivariable analysis that included cytogenetic risk, FLT3-ITD and NPM1 mutation, white blood cell count, and age. In contrast, CEBPAsingle-mut AMLs did not express a discriminating signature and could not be distinguished from wild-type cases as regards clinical outcome. These results demonstrate significant underlying heterogeneity within CEBPA mutation-positive AML with prognostic relevance.

Gene expression profiling of acute myeloid leukemia (AML) allows the discovery of previously unrecognized molecular entities. Here, we identified a specific subgroup of AML, defined by an expression profile resembling that of AMLs with mutations in the myeloid transcription factor CCAAT/enhancer-binding protein alpha (C/EBPα), while lacking such mutations. We found that in these leukemias, the CEBPA gene was silenced, which was associated with frequent promoter hypermethylation. The leukemias phenotypically showed aberrant expression of T-cell genes, of which CD7 was most consistent. We identified 2 mechanisms that may contribute to this phenotype. First, absence of Cebpa led to up-regulation of specific T-cell transcripts (ie, Cd7 and Lck) in hematopoietic stem cells isolated from conditional Cebpa knockout mice. Second, the enhanced expression of TRIB2, which we identify here as a direct target of the T-cell commitment factor NOTCH1, suggested aberrantly activated Notch signaling. Putatively activating NOTCH1 mutations were found in several specimens of the newly identified subgroup, while a large set of control AMLs was mutation negative. A gene expression prediction signature allowed the detection of similar cases of leukemia in independent series of AML.

We showed previously that CCAAT/enhancer binding protein α (C/EBPα), a tissue-specific transcription factor, is a candidate tumor suppressor in lung cancer. In the present study, we have performed a transcriptional profiling study of C/EBPα target genes using an inducible cell line system. This study led to the identification of hepatocyte nuclear factor 3β (HNF3β), a transcription factor known to play a role in airway differentiation, as a downstream target of C/EBPα. We found down-regulation of HNF3β expression in a large proportion of lung cancer cell lines examined and identified two novel mutants of HNF3β, as well as hypermethylation of the HNF3β promoter. We also developed a tetracycline-inducible cell line model to study the cellular consequences of HNF3β expression. Conditional expression of HNF3β led to significant growth reduction, proliferation arrest, apoptosis, and loss of clonogenic ability, suggesting additionally that HNF3β is a novel tumor suppressor in lung cancer. This is the first study to show genetic abnormalities of lung-specific differentiation pathways in the development of lung cancer.

Segmental uniparental disomy as a recurrent mechanism for homozygous CEBPA mutations in acute myeloid leukemia

Tiling arrays are useful for exploring local functions of regions of the genome in an unbiased fashion. The exact determination of those genomic regions based on tiling-array data, e.g., generated by means of hybridization with immunopreciptated DNA-fragments to the arrays is a challenge. Many different statistical methodologies have been developed to find biological relevant regions-of-interest (ROI) by using the quantitative signal intensity of each probe. We previously developed a method called Hypergeometric Analysis of Tiling arrays (HAT) for the analysis of tiling-array data, but it is developed such that it can also be used to study data derived by genome-wide deep sequencing approaches. Here we applied HAT to analyze two publicly available tiling-array data sets. After the detection of statistically significant ROI, these are often used in additional analysis for hypothesis testing. We therefore discuss, by using the results of the tiling-array experiment, pathway and motif analyses.

Numerous molecular markers have been recently discovered as potential prognostic factors in acute myeloid leukemia (AML). It has become of critical importance to thoroughly evaluate their interrelationships and relative prognostic importance. Gene expression profiling was conducted in a well-characterized cohort of 439 AML patients (age < 60 years) to determine expression levels of EVI1, WT1, BCL2, ABCB1, BAALC, FLT3, CD34, INDO, ERG and MN1. A variety of AML-specific mutations were evaluated, that is, FLT3, NPM1, N-RAS, K-RAS, IDH1, IDH2, and CEBPADM/SM (double/single). Univariable survival analysis shows that (1) patients with FLT3ITD mutations have inferior overall survival (OS) and event-free survival (EFS), whereas CEBPADM and NPM1 mutations indicate favorable OS and EFS in intermediate-risk AML, and (2) high transcript levels of BAALC, CD34, MN1, EVl1, and ERG predict inferior OS and EFS. In multivariable survival analysis, CD34, ERG, and CEBPADM remain significant. Using survival tree and regression methodologies, we show that CEBPADM, CD34, and IDH2 mutations are capable of separating the intermediate group into 2 AML subgroups with highly distinctive survival characteristics (OS at 60 months: 51.9% vs 14.9%). The integrated statistical approach demonstrates that from the multitude of biomarkers a greatly condensed subset can be selected for improved stratification of intermediate-risk AML.

Acute myeloid leukemia is a heterogeneous disease from the molecular and biologic standpoints, and even patients with a specific gene expression profile may present clinical and molecular heterogeneity. We studied the epigenetic profiles of a cohort of patients who shared a common gene expression profile but differed in that only half of them harbored mutations of the CEBPA locus, whereas the rest presented with silencing of this gene and coexpression of certain T-cell markers. DNA methylation studies revealed that these 2 groups of patients could be readily segregated in an unsupervised fashion based on their DNA methylation profiles alone. Furthermore, CEBPA silencing was associated with the presence of an aberrant DNA hypermethylation signature, which was not present in the CEBPA mutant group. This aberrant hypermethylation occurred more frequently at sites within CpG islands. CEBPA-silenced leukemias also displayed marked hypermethylation compared with normal CD34+ hematopoietic cells, whereas CEBPA mutant cases showed only mild changes in DNA methylation compared with these normal progenitors. Biologically, CEBPA-silenced leukemias presented with a decreased response to myeloid growth factors in vitro.

Minimally differentiated acute myeloid leukemia (AML-M0) is defined by immature morphology and expression of early hematologic markers. By gene expression profiling (GEP) and subsequent unsupervised analysis of 35 AML-M0 samples and 253 previously reported AML cases, we demonstrate that AML-M0 cases express a unique signature that is largely separated from other molecular subtypes. Hematologic transcription regulators such as CEBPA, CEBPD, and ETV6, and the differentiation associated gene MPO appeared strongly down-regulated, in line with the primitive state of this leukemia. AML-M0 frequently carries loss-of-function RUNX1 mutation. Unsupervised analyses revealed a subdivision between AML-M0 cases with and without RUNX1 mutations. RUNX1 mutant AML-M0 samples showed a distinct up-regulation of B cell-related genes such as members of the B-cell receptor complex, transcription regulators RUNX3, ETS2, IRF8, or PRDM1, and major histocompatibility complex class II genes. Importantly, prediction with high accuracy of the AML-M0 subtype and prediction of patients carrying RUNX1 mutation within this subtype were possible based on the expression level of only a few transcripts. We propose that RUNX1 mutations in this AML subgroup cause lineage infidelity, leading to aberrant coexpression of myeloid and B-lymphoid genes. Furthermore, our results imply that AML-M0, although originally determined by morphology, constitutes a leukemia subgroup. (Blood. 2009; 114: 3001-3007)

In this concise overview, we discuss recent findings concerning a distinct subgroup of acute myeloid/T-lymphoid leukemia. We describe how we identified these leukemias in multiple cohorts of acute myeloid leukemia (AML) using a combination of gene expression profiling and additional analytic approaches, and how we obtained insight in possible mechanisms leading to their phenotype.

Tribbles homolog 2 (Trib2) was identified as a downregulated transcript in leukemic cells undergoing growth arrest. To investigate the effects of Trib2 in hematopoietic progenitors, mice were reconstituted with hematopoietic stem cells retrovirally expressing Trib2. Trib2-transduced bone marrow cells exhibited a growth advantage ex vivo and readily established factor-dependent cell lines. In vivo, Trib2-reconstituted mice uniformly developed fatal transplantable acute myelogenous leukemia (AML). In mechanistic studies, we found that Trib2 associated with and inhibited C/EBPalpha. Furthermore, Trib2 expression was elevated in a subset of human AML patient samples. Together, our data identify Trib2 as an oncogene that induces AML through a mechanism involving inactivation of C/EBPalpha.

Mutations in CEBPA , the gene encoding the transcription factor CCAAT/enhancer binding protein alpha (C/EBPalpha), have been reported in multiple studies, and are found in approximately 8% of patients with acute myeloid leukemia (AML).[1][1],[2][2] Specific regions of the gene tend to be most

Abstract 554 Acute myeloid leukemia (AML) is a heterogeneous disease of different subtypes characterized by distinct cytogenetic or molecular abnormalities. We recently identified a previously unrecognized subtype with a unique epigenetic feature, i.e. silencing of the gene that encodes CCAAT-enhancer binding protein alpha (CEBPα) by DNA hypermethylation (denoted as the CEBPαsilenced group). The leukemic blast cells of these patients express myeloid as well as T-lymphoid markers. Moreover, gene expression and DNA methylation profiling put these leukemias in between AML and T-lymphoid leukemia (T-ALL). CEBPα is a transcription regulator that is essential for normal neutrophil development. We hypothesize that methylation and consequently silencing of the gene encoding CEBPα is abnormal and an important hit in the transformation of this unique leukemia subtype. The mechanism by which CEBPα silencing plays a role in transformation of cells with myeloid/T-lymphoid features is the aim of our study. We carried out gene expression profiling of the CEBPαsilenced group (n=10) and compared gene expression levels to normal CD34+ bone marrow cells (n=11) and the remaining AML group (n=506) using a three-way ANOVA and a post-hoc test (tukey-kramer method). We detected 686 differentially expressed genes with P < 0.05 after multiple testing. Of these 686 genes, 288 were up- and 401 down-regulated in the CEBPαsilenced group. We next asked the question which of those genes might be bona fide CEBPα targets and whether expression has been altered as the result of CEBPα silencing. We transduced estrogen-inducible C/EBPα in 32D cells and carried out ChIP-on-chip analysis using ER specific antibodies. The analysis yielded a collection of 529 CEBPα target genes that are significant enriched for C/EBPα binding (P < 0.05) using Hypergeometric Analysis of Tiling-arrays (HAT). We hypothesized that the direct target genes of CEBPα, derived from the 32D model system, were also present among deregulated genes in CEBPαsilenced human AMLs. We therefore overlaid the detected direct binding targets of CEBPα in 32D cells, with the differentially expressed genes in the CEBPαsilenced group and identified 49 overlapping genes (P=1×10−7) as putative direct targets. Among these 49 genes, 25 were down-regulated and 24 were upregulated in the primary CEBPαsilenced AML group. Both groups of genes were highly enriched for the CEBPα consensus binding sequence, i.e. 16/25 and 20/24 promoter regions respectively. The 25 downregulated genes, among which ACSL1, MYCT1 or Slc7a11, represent targets that are most likely normally activated by CEBPα, but are not transcribed in CEBPαsilenced human AMLs due to the absence of CEBPα. Among the genes that were upregulated in CEBPαsilenced leukemias are BCL2, CCR9, CEBPG or CD47. These putative target genes are repressed in CEBPα expressing cells and activated when CEBPα is silenced. This observation suggests that the transcription factor CEBPα may also acts as a repressor of gene transcription. We therefore studied the expression of two of those genes, i.e. CEBPG and CCR9 in Lin-Sca+Kit+ (LSK) bone marrow hematopoietic stem cells from wild type versus conditional Mx-Cre/CEBPα knock-out mice. Similar to what we observed in human CEBPαsilenced leukemias, we found that these two genes were switched off in LSK cells from CEBPα knock-out animals. Ingenuity pathway analysis of the 24 upregulated genes detected high enrichment (P < 0.001) of pathways involved with T-cell development. Our data clearly suggest that CEBPα may act as repressor of T-cell related genes through direct promoter interaction. We therefore propose that silencing of CEBPα by promoter hypermethylation is one of the transforming events that driving towards mixed myeloid/T-lymphoid leukemia. Disclosures: No relevant conflicts of interest to declare.

Background Tiling-arrays are applicable to multiple types of biological research questions. Due to its advantages (high sensitivity, resolution, unbiased), the technology is often employed in genome-wide investigations. A major challenge in the analysis of tiling-array data is to define regions-of-interest, i.e., contiguous probes with increased signal intensity (as a result of hybridization of labeled DNA) in a region. Currently, no standard criteria are available to define these regions-of-interest as there is no single probe intensity cut-off level, different regions-of-interest can contain various numbers of probes, and can vary in genomic width. Furthermore, the chromosomal distance between neighboring probes can vary across the genome among different arrays. Results We have developed Hypergeometric Analysis of Tiling-arrays (HAT), and first evaluated its performance for tiling-array datasets from a Chromatin Immunoprecipitation study on chip (ChIP-on-chip) for the identification of genome-wide DNA binding profiles of transcription factor Cebpa (used for method comparison). Using this assay, we can refine the detection of regions-of-interest by illustrating that regions detected by HAT are more highly enriched for expected motifs in comparison with an alternative detection method (MAT). Subsequently, data from a retroviral insertional mutagenesis screen were used to examine the performance of HAT among different applications of tiling-array datasets. In both studies, detected regions-of-interest have been validated with (q)PCR. Conclusions We demonstrate that HAT has increased specificity for analysis of tiling-array data in comparison with the alternative method, and that it accurately detects regions-of-interest in two different applications of tiling-arrays. HAT has several advantages over previous methods: i) as there is no single cut-off level for probe-intensity, HAT can detect regions-of-interest at various thresholds, ii) it can detect regions-of-interest of any size, iii) it is independent of probe-resolution across the genome, and across tiling-array platforms and iv) it employs a single user defined parameter: the significance level. Regions-of-interest are detected by computing the hypergeometric-probability, while controlling the Family Wise Error. Furthermore, the method does not require experimental replicates, common regions-of-interest are indicated, a sequence-of-interest can be examined for every detected region-of-interest, and flanking genes can be reported.

Response With great interest we have read the letter by Terriou and coworkers, who have performed an extensive survey of CEBPA CpG promoter hypermethylation in T-ALL as well as in a selection of immature acute myeloid leukemia (AML) cases. This study relates to our previous work in which we

Acute myeloid leukemia (AML) is a heterogeneous disease characterized by various cytogenetic and molecular abnormalities, some of which can be used as prognostic markers. Mutations in the transcription factor CCAAT/enhancer binding protein alpha (CEBPA) occur in 5–10% of AML and have consistently been associated with a favorable outcome. Three types of mutations have been described: N-terminal out-of-frame mutations, inframe mutations in the basic leucine zipper (bZIP) region, and a small variable group of remaining aberrations. Most CEBPA mutant AML cases carry two mutations, usually on different alleles (double mutant). However, there are also cases that only express a single heterozygous mutation, and thus retain a wild type allele (single mutant). It is not known whether single and double CEBPA mutations should be considered of equal biological and/or clinical importance. We applied dHPLC WAVE technology in combination with nucleotide sequencing of the entire CEBPA gene in a cohort of 598 cases of adult de novo AML. After exclusion of previously described polymorphisms, we identified 41 cases (6.9%) with at least one mutation. Of these cases, 28 carried double mutations, i.e. two different heterozygous mutations or one homozygous mutation, whereas the remaining cases carried a single heterozygous mutation. To investigate whether CEBPA mutations were associated with specific transcriptional signatures, we examined genome-wide gene expression (GEP) data of 525/598 AMLs, including 38/41 CEBPA mutant cases. Class prediction of total CEBPA mutation status based on GEP data resulted in a relatively large number of false negatives in cross-validation using the PAM algorithm (sensitivity 68%, specificity 99%). Strikingly however, all these missed cases appeared to be of the single mutant group, while the double mutants were recognized with high accuracy. In agreement, unsupervised cluster analysis of the 525 AMLs led to distinct grouping of cases with double mutations, while cases with a single heterozygous mutation did not. These observations suggested that double and single CEBPA mutant AMLs represent distinct biological entities. We next assessed the clinical relevance of this finding. In concordance with previous studies, total CEBPA mutation status associated with favorable overall survival (OS) and event-free survival (EFS) (P=0.023 and P=0.042, log rank test), which was maintained in multivariable Cox’s proportional hazards models with cytogenetic risk group, FLT3-ITD and NPM1 mutation status, age and white blood cell count (hazard ratio [HR] 0.47, 95% confidence interval [CI] 0.29–0.77; P=0.002 and HR 0.52, 95% CI 0.33–0.82; P=0.004). Surprisingly, when the double and single CEBPA mutant cases were separately analyzed, only the double mutants showed a highly favorable outcome, while the single mutants could not be distinguished from CEBPA wild type AMLs (P=0.003 versus P=0.51 (OS) and P=0.004 versus P=0.18 (EFS)). In multivariable analysis, CEBPA double mutation status remained associated with favorable outcome (OS HR 0.31, 95% CI 0.16–0.59; P

Acute Myeloid Leukemia (AML) is a heterogeneous disease from the molecular and biological standpoints. In order to resolve some of this complexity, a recent microarray-based expression profiling study segregated cohorts of patients with common gene signatures. One of these signatures was associated with alterations of the CCAAT/enhancer-binding protein alpha (CEBPA) gene. Among these patients, a subset harbored CEBPA mutations, while the remainder failed to express CEBPA, which in a number of cases correlated with hypermethylation of its promoter. This latter subgroup of leukemias with silenced CEBPA presented with significant biological differences compared to CEBPA mutant patients, including expression of T-cell markers and activating mutations of NOTCH1 (Wouters BJ et. al., PMID:17671232). Since our preliminary data show that DNA methylation profiling is extremely accurate in identifying distinct biological phenotypes in AML and other tumors, we wondered whether genome-wide epigenetic analysis would identify the biological difference between these patients. In order to determine the DNA methylation profiles of these patients we performed HELP (HpaII tiny fragment Enrichment by Ligation-mediated PCR), a quantitative genome-wide DNA methylation method, using a 400,000 feature cutom-desinged microarray, representing 24,000 gene promoters with 50-mer oligonucleotides. We studied the complete previously identified cluster of AML cases presenting with a CEBPA expression signature, and compared and contrasted the DNA methylation profiles of cases carrying the CEBPA mutation and those presenting with CEBPA silencing. Remarkably, unsupervised (unbiased) clustering of DNA methylation profiles revealed that samples were readily segregated into two groups that overlapped perfectly with the presence or absence of the CEBPA mutation, indicating the presence of underlying genome-wide DNA methylation differences between these two groups. We next performed supervised analysis of the samples to compare the DNA methylation profiles of CEBPA mutated vs. CEBPA non-mutated samples. The analysis was performed using a moderated T test, and 291 genes promoters were identified as differentially methylated between the two groups at a significance level of p

Background High-throughput experiments, such as with DNA microarrays, typically result in hundreds of genes potentially relevant to the process under study, rendering the interpretation of these experiments problematic. Here, we propose and evaluate an approach to find functional associations between large numbers of genes and other biomedical concepts from free-text literature. For each gene, a profile of related concepts is constructed that summarizes the context in which the gene is mentioned in literature. We assign a weight to each concept in the profile based on a likelihood ratio measure. Gene concept profiles can then be clustered to find related genes and other concepts. Results The experimental validation was done in two steps. We first applied our method on a controlled test set. After this proved to be successful the datasets from two DNA microarray experiments were analyzed in the same way and the results were evaluated by domain experts. The first dataset was a gene-expression profile that characterizes the cancer cells of a group of acute myeloid leukemia patients. For this group of patients the biological background of the cancer cells is largely unknown. Using our methodology we found an association of these cells to monocytes, which agreed with other experimental evidence. The second data set consisted of differentially expressed genes following androgen receptor stimulation in a prostate cancer cell line. Based on the analysis we put forward a hypothesis about the biological processes induced in these studied cells: secretory lysosomes are involved in the production of prostatic fluid and their development and/or secretion are androgen-regulated processes. Conclusion Our method can be used to analyze DNA microarray datasets based on information explicitly and implicitly available in the literature. We provide a publicly available tool, dubbed Anni, for this purpose.

Trib2 is a member of the Trib family of serine/threonine kinase-like proteins (Trib1, Trib2, Trib3), whose function in hematopoiesis is not documented. To investigate the effects of Trib2 in hematopoietic progenitors, mice were reconstituted with hematopoietic stem cells retrovirally expressing Trib2. Trib2-reconstituted mice uniformly developed fatal transplantable acute myelogenous leukemia (AML), with a median survival of 179 days. Retroviral Trib2 expression in hematopoietic stem cells perturbed myeloid development, enhanced progenitor proliferation, and directly inhibited the function of C/EBPα, a critical transcription factor that is frequently dysregulated in AML. Furthermore, an analysis of microarray data generated from 285 AML patient samples identified elevated Trib2 expression in a distinct subset of patients in a cluster with a high frequency of C/EBPα mutations. In mechanistic studies, we found that Trib2 associated with and led to the proteasomal-dependent degradation of C/EBPα. To determine the structural requirements for Trib2 to inhibit C/EBPα, and to induce AML, we have initiated a structure/function analysis of Trib2. Together, our data identify Trib2 as a novel oncogene that induces AML through a mechanism involving inactivation of C/EBPα. The identification of Trib2 as potent leukemogen points to new pathogenic mechanisms and possible therapeutic opportunities in this aggressive cancer, which is not currently curable in the majority of patients.

Recently, gene expression profiling has allowed the identification of specific molecular signatures of established as well as novel subgroups of acute myeloid leukemia (AML). CEBPA mutations are observed in ~10% of AML. In a gene expression profiling study of 285 AML cases, we previously found two distinct clusters consisting of almost all specimens with CEBPA mutations, i.e. cluster #15 and cluster #4. In cluster #4, only 9 out of 15 cases were CEBPA mutant. Thus, this cluster contained 6 patients with a highly similar expression signature, but without CEBPA mutations. We wished to investigate whether these cases represent a previously unidentified subgroup of leukemias. All 6 cases appeared morphologically immature (FAB-M0 or M1) and did not carry any of the common genetic aberrations. Supervised array expression analysis revealed CEBPA to be among the most highly downregulated genes in these 6 cases, which was associated with CEBPA CpG promoter hypermethylation in 4/6. None of the CEBPA mutant cases or a random selection of 10 AML cases from other clusters exhibited CEBPA promoter methylation. Strikingly, several T-cell related genes, e.g. CD7, CD3D, IL7R and T-cell receptor delta locus, appeared significantly overexpressed in these cases. Multicolor flowcytometry demonstrated high expression of CD34 and combinations of myeloid (e.g. CD13, CD33, CD117 and/or myeloperoxidase) as well as T-lymphoid (CD7, CD3, TdT) lineage markers on all six. Expression of T-cell markers correlated with T-cell receptor rearrangements in 4/6 cases. Other highly expressed genes in these biphenotypic acute leukemias included NOTCH1 and its recently identified target gene TRIB2. Sequencing analysis demonstrated putatively activating NOTCH1 mutations in three out of six samples, as described previously in T-cell acute lymphoblastic leukemia (T-ALL). The remaining 3 cases and 25 control AMLs, including all CEBPA mutant cases, did not demonstrate any NOTCH1 mutation. Finally, based on this first cohort of patients, we determined the minimal number of probe sets accurately predicting cluster #4 biphenotypic cases, and identified eight probe sets, including those for T-cell receptor delta/alpha locus. Applying this predictor set to an independent new cohort of 272 AML cases, an identical subgroup of another 6 AML patients with frequent NOTCH1 mutations and CEBPA promoter hypermethylation could be identified. Available clinical data suggested an adverse prognosis of this type of leukemia. In conclusion, we identified a subgroup of biphenotypic acute leukemias characterized by a CEBPA mutant gene expression profile, frequent CEBPA promoter hypermethylation, and frequent NOTCH1 mutations. To our knowledge, this is the first report to describe NOTCH1 mutations in leukemia other than T-ALL.