BISCUIT: an efficient, standards-compliant tool suite for simultaneous genetic and epigenetic inference in bulk and single-cell studies.

Data from both bulk and single-cell whole-genome DNA methylation experiments are under-utilized in many ways. This is attributable to inefficient mapping of methylation sequencing reads, routinely discarded genetic information, and neglected read-level epigenetic and genetic linkage information. We introduce the BISulfite-seq Command line User Interface Toolkit (BISCUIT) and its companion R/Bioconductor package, biscuiteer, for simultaneous extraction of genetic and epigenetic information from bulk and single-cell DNA methylation sequencing. BISCUIT's performance, flexibility and standards-compliant output allow large, complex experimental designs to be characterized on clinical timescales. BISCUIT is particularly suited for processing data from single-cell DNA methylation assays, with its excellent scalability, efficiency, and ability to greatly enhance mappability, a key challenge for single-cell studies. We also introduce the epiBED format for single-molecule analysis of coupled epigenetic and genetic information, facilitating the study of cellular and tissue heterogeneity from DNA methylation sequencing.

Comprehensive molecular and clinical characterization of NUP98 fusions in pediatric acute myeloid leukemia.

NUP98 fusions comprise a family of rare recurrent alterations in AML, associated with adverse outcomes. In order to define the underlying biology and clinical implications of this family of fusions, we performed comprehensive transcriptome, epigenome, and immunophenotypic profiling of 2,235 children and young adults with AML and identified 160 NUP98 rearrangements (7.2%), including 108 NUP98-NSD1 (4.8%), 32 NUP98-KDM5A (1.4%) and 20 NUP98-X cases (0.9%) with 13 different fusion partners. Fusion partners defined disease characteristics and biology; patients with NUP98-NSD1 or NUP98-KDM5A had distinct immunophenotypic, transcriptomic, and epigenomic profiles. Unlike the two most prevalent NUP98 fusions, NUP98-X variants are typically not cryptic. Furthermore, NUP98-X cases are associated with WT1 mutations, and have epigenomic profiles that resemble either NUP98-NSD1 or NUP98-KDM5A. Cooperating FLT3-ITD and WT1 mutations define NUP98-NSD1, and chromosome 13 aberrations are highly enriched in NUP98-KDM5A. Importantly, we demonstrate that NUP98 fusions portend dismal overall survival, with the noteworthy exception of patients bearing abnormal chromosome 13 (clinicaltrials gov. Identifiers: NCT00002798, NCT00070174, NCT00372593, NCT01371981).

Integrated DNA Copy Number and Expression Profiling Identifies IGF1R as a Prognostic Biomarker in Pediatric Osteosarcoma.

Osteosarcoma is a primary malignant bone tumor arising from bone-forming mesenchymal cells in children and adolescents. Despite efforts to understand the biology of the disease and identify novel therapeutics, the survival of osteosarcoma patients remains dismal. We have concurrently profiled the copy number and gene expression of 226 osteosarcoma samples as part of the Strategic Partnering to Evaluate Cancer Signatures (SPECS) initiative. Our results demonstrate the heterogeneous landscape of osteosarcoma in younger populations by showing the presence of genome-wide copy number abnormalities occurring both recurrently among samples and in a high frequency. Insulin growth factor receptor 1 (IGF1R) is a receptor tyrosine kinase which binds IGF1 and IGF2 to activate downstream pathways involved in cell apoptosis and proliferation. We identify prevalent amplification of IGF1R corresponding with increased gene expression in patients with poor survival outcomes. Our results substantiate previously tenuously associated copy number abnormalities identified in smaller datasets (13q34+, 20p13+, 4q35-, 20q13.33-), and indicate the significance of high fibroblast growth factor receptor 2 (FGFR2) expression in distinguishing patients with poor prognosis. FGFR2 is involved in cellular proliferation processes such as division, growth and angiogenesis. In summary, our findings demonstrate the prognostic significance of several genes associated with osteosarcoma pathogenesis.

Integrative genomic analysis of matched primary and metastatic pediatric osteosarcoma.

Despite being the most common childhood bone tumor, the genomic characterization of osteosarcoma remains incomplete. In particular, very few osteosarcoma metastases have been sequenced to date, critical to better understand mechanisms of progression and evolution in this tumor. We performed an integrated whole genome and exome sequencing analysis of paired primary and metastatic pediatric osteosarcoma specimens to identify recurrent genomic alterations. Sequencing of 13 osteosarcoma patients including 13 primary, 10 metastatic, and 3 locally recurring tumors revealed a highly heterogeneous mutational landscape, including cases of hypermutation and microsatellite instability positivity, but with virtually no recurrent alterations except for mutations involving the tumor suppressor genes RB1 and TP53. At the germline level, we detected alterations in multiple cancer related genes in the majority of the cohort, including those potentially disrupting DNA damage response pathways. Metastases retained only a minimal number of short variants from their corresponding primary tumors, while copy number alterations showed higher conservation. One recurrently amplified gene, KDR, was highly expressed in advanced cases and associated with poor prognosis. © 2019 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

SeSAMe: reducing artifactual detection of DNA methylation by Infinium BeadChips in genomic deletions.

We report a new class of artifacts in DNA methylation measurements from Illumina HumanMethylation450 and MethylationEPIC arrays. These artifacts reflect failed hybridization to target DNA, often due to germline or somatic deletions and manifest as incorrectly reported intermediate methylation. The artifacts often survive existing preprocessing pipelines, masquerade as epigenetic alterations and can confound discoveries in epigenome-wide association studies and studies of methylation-quantitative trait loci. We implement a solution, P-value with out-of-band (OOB) array hybridization (pOOBAH), in the R package SeSAMe. Our method effectively masks deleted and hyperpolymorphic regions, reducing or eliminating spurious reports of epigenetic silencing at oft-deleted tumor suppressor genes such as CDKN2A and RB1 in cases with somatic deletions. Furthermore, our method substantially decreases technical variation whilst retaining biological variation, both within and across HM450 and EPIC platform measurements. SeSAMe provides a light-weight, modular DNA methylation data analysis suite, with a performant implementation suitable for efficient analysis of thousands of samples.

GLI3 repressor determines Hedgehog pathway activation and is required for response to SMO antagonist glasdegib in AML.

The Hedgehog (Hh) signaling pathway is activated in many cancers and is a promising target for therapeutic development. Deletions in the receptor Patched (PTCH) or activating mutations in Smoothened (SMO) have been reported in basal cell carcinoma and medulloblastoma, but are largely absent in most tumor types. Therefore, the mechanism of pathway activation in most cancers, including hematological malignancies, remains unknown. In normal tissues, Hh pathway activation via PTCH/SMO causes an increase in the downstream transcriptional activator GLI1 and a decrease in the GLI3 transcriptional repressor (GLI3R). In this article, we confirm that the Hh pathway is active in acute myeloid leukemia (AML), however, this activity is largely independent of SMO. Epigenetic and gene expression analysis of The Cancer Genome Atlas AML data set reveals that GLI3 expression is silenced in most AML patient samples, and the GLI3 locus is abnormally methylated. We show that GLI3R is required for the therapeutic effect of SMO antagonists in AML samples and restoration of GLI3R suppresses the growth of AML. We additionally demonstrate that GLI3R represses AML growth by downregulating AKT expression. In summary, this study provides the first evidence that GLI3R plays an essential role in SMO-independent Hh signaling in AML, and suggests that GLI3R could serve as a potential biomarker for patient selection in SMO antagonist clinical trials. Furthermore, these data support rational combinations of hypomethylating agents with SMO antagonists in clinical trials.

Preprocessing, normalization and integration of the Illumina HumanMethylationEPIC array with minfi.

Summary: The minfi package is widely used for analyzing Illumina DNA methylation array data. Here we describe modifications to the minfi package required to support the HumanMethylationEPIC ('EPIC') array from Illumina. We discuss methods for the joint analysis and normalization of data from the HumanMethylation450 ('450k') and EPIC platforms. We introduce the single-sample Noob ( ssNoob ) method, a normalization procedure suitable for incremental preprocessing of individual methylation arrays and conclude that this method should be used when integrating data from multiple generations of Infinium methylation arrays. We show how to use reference 450k datasets to estimate cell type composition of samples on EPIC arrays. The cumulative effect of these updates is to ensure that minfi provides the tools to best integrate existing and forthcoming Illumina methylation array data. Availability and Implementation: The minfi package version 1.19.12 or higher is available for all platforms from the Bioconductor project. Contact: khansen@jhsph.edu. Supplementary information: Supplementary data are available at Bioinformatics online.

Reprogramming of the human intestinal epigenome by surgical tissue transposition.

Extracellular cues play critical roles in the establishment of the epigenome during development and may also contribute to epigenetic perturbations found in disease states. The direct role of the local tissue environment on the post-development human epigenome, however, remains unclear due to limitations in studies of human subjects. Here, we use an isogenic human ileal neobladder surgical model and compare global DNA methylation levels of intestinal epithelial cells pre- and post-neobladder construction using the Infinium HumanMethylation450 BeadChip. Our study is the first to quantify the effect of environmental cues on the human epigenome and show that the local tissue environment directly modulates DNA methylation patterns in normal differentiated cells in vivo. In the neobladder, the intestinal epithelial cells lose their tissue-specific epigenetic landscape in a time-dependent manner following the tissue's exposure to a bladder environment. We find that de novo methylation of many intestine-specific enhancers occurs at the rate of 0.41% per month (P < 0.01, Pearson = 0.71), while demethylation of primarily non-intestine-specific transcribed regions occurs at the rate of -0.37% per month (P < 0.01, Pearson = -0.57). The dynamic resetting of the DNA methylome in the neobladder not only implicates local environmental cues in the shaping and maintenance of the epigenome but also illustrates an unexpected cross-talk between the epigenome and the cellular environment.

Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia.

BACKGROUND: Many mutations that contribute to the pathogenesis of acute myeloid leukemia (AML) are undefined. The relationships between patterns of mutations and epigenetic phenotypes are not yet clear. METHODS: We analyzed the genomes of 200 clinically annotated adult cases of de novo AML, using either whole-genome sequencing (50 cases) or whole-exome sequencing (150 cases), along with RNA and microRNA sequencing and DNA-methylation analysis. RESULTS: AML genomes have fewer mutations than most other adult cancers, with an average of only 13 mutations found in genes. Of these, an average of 5 are in genes that are recurrently mutated in AML. A total of 23 genes were significantly mutated, and another 237 were mutated in two or more samples. Nearly all samples had at least 1 nonsynonymous mutation in one of nine categories of genes that are almost certainly relevant for pathogenesis, including transcription-factor fusions (18% of cases), the gene encoding nucleophosmin (NPM1) (27%), tumor-suppressor genes (16%), DNA-methylation-related genes (44%), signaling genes (59%), chromatin-modifying genes (30%), myeloid transcription-factor genes (22%), cohesin-complex genes (13%), and spliceosome-complex genes (14%). Patterns of cooperation and mutual exclusivity suggested strong biologic relationships among several of the genes and categories. CONCLUSIONS: We identified at least one potential driver mutation in nearly all AML samples and found that a complex interplay of genetic events contributes to AML pathogenesis in individual patients. The databases from this study are widely available to serve as a foundation for further investigations of AML pathogenesis, classification, and risk stratification. (Funded by the National Institutes of Health.).

Insulin-like Growth Factor 2 Gene Expression Molecularly Differentiates Pleuropulmonary Blastoma and Embryonal Rhabdomyosarcoma.

The sarcomatous element in pleuropulmonary blastoma (PPB) is often histologically indistinguishable from embryonal rhabdomyosarcoma (ERMS). A diagnosis of PPB is often made after definitive surgical resection based on pathologic features, most notably the presence of hamartomatous pulmonary elements. Samples from seven PPB patients were obtained from the rhabdomyosarcomatous portion of the tumor by macrodissection. Representative ERMS tumor tissue was selected from 21 ERMS patient samples. Formalin-fixed paraffin-embedded tissue scrolls from each sample were analyzed using the Affymetrix Human Exon arrays. All PPB patients and 7 of 21 ERMS patients were 3 years old and younger. Twenty transcripts (10 annotated, 10 noncoding RNAs) were significantly differentially expressed in ERMS when compared with PPB samples. Insulin-like growth factor 2 (IGF2) was uniformly overexpressed in ERMS (19/21>400) but was expressed at low levels in PPB (P<0.001). Two ERMS cases that had low level IGF2 expression were 3 years and younger of age. No other differences between the 2 approached this degree of significance, despite a common rhabdomyogenic phenotype in the sarcomatous areas of PPB. PPB, unlike most ERMS, appears not to be driven by autocrine IGF2 signaling.

Low-level processing of Illumina Infinium DNA Methylation BeadArrays.

We propose a novel approach to background correction for Infinium HumanMethylation data to account for technical variation in background fluorescence signal. Our approach capitalizes on a new use for the Infinium I design bead types to measure non-specific fluorescence in the colour channel opposite of their design (Cy3/Cy5). This provides tens of thousands of features for measuring background instead of the much smaller number of negative control probes on the platforms (n = 32 for HumanMethylation27 and n = 614 for HumanMethylation450, respectively). We compare the performance of our methods with existing approaches, using technical replicates of both mixture samples and biological samples, and demonstrate that within- and between-platform artefacts can be substantially reduced, with concomitant improvement in sensitivity, by the proposed methods.

CelFiE-ISH: a probabilistic model for multi-cell type deconvolution from single-molecule DNA methylation haplotypes.

Deconvolution methods infer quantitative cell type estimates from bulk measurement of mixed samples including blood and tissue. DNA methylation sequencing measures multiple CpGs per read, but few existing deconvolution methods leverage this within-read information. We develop CelFiE-ISH, which extends an existing method (CelFiE) to use within-read haplotype information. CelFiE-ISH outperforms CelFiE and other existing methods, achieving 30% better accuracy and more sensitive detection of rare cell types. We also demonstrate the importance of marker selection and of tailoring markers for haplotype-aware methods. While here we use gold-standard short-read sequencing data, haplotype-aware methods will be well-suited for long-read sequencing.

EGFR-dependent endocytosis of Wnt9a and Fzd9b promotes ß-catenin signaling during hematopoietic stem cell development in zebrafish.

Cell-to-cell communication through secreted Wnt ligands that bind to members of the Frizzled (Fzd) family of transmembrane receptors is critical for development and homeostasis. Wnt9a signals through Fzd9b, the co-receptor LRP5 or LRP6 (LRP5/6), and the epidermal growth factor receptor (EGFR) to promote early proliferation of zebrafish and human hematopoietic stem cells during development. Here, we developed fluorescently labeled, biologically active Wnt9a and Fzd9b fusion proteins to demonstrate that EGFR-dependent endocytosis of the ligand-receptor complex was required for signaling. In human cells, the Wnt9a-Fzd9b complex was rapidly endocytosed and trafficked through early and late endosomes, lysosomes, and the endoplasmic reticulum. Using small-molecule inhibitors and genetic and knockdown approaches, we found that Wnt9a-Fzd9b endocytosis required EGFR-mediated phosphorylation of the Fzd9b tail, caveolin, and the scaffolding protein EGFR protein substrate 15 (EPS15). LRP5/6 and the downstream signaling component AXIN were required for Wnt9a-Fzd9b signaling but not for endocytosis. Knockdown or loss of EPS15 impaired hematopoietic stem cell development in zebrafish. Other Wnt ligands do not require endocytosis for signaling activity, implying that specific modes of endocytosis and trafficking may represent a method by which Wnt-Fzd specificity is established.

EZH2-driven immune evasion defines high-risk pediatric AML with t(16;21) FUS::ERG gene fusion.

Despite decades of research, acute myeloid leukemia (AML) remains a remarkably lethal malignancy. While pediatric AML (pAML) carries a more favorable prognosis than adult AML, the past 25 years of large clinical trials have produced few improvements in pAML survival. Nowhere is this more evident than in patients carrying a t(16;21)(p11;q22) translocation, which yields the FUS::ERG fusion transcript. Patients with FUS::ERG-positive AML are often primary refractory, and most responders quickly relapse. In COG clinical trials, allogeneic stem cell transplantation was of no benefit to FUS::ERG pAML patients; 100% of transplanted patients succumbed to their disease. Expression of major histocompatibility complex (MHC) class I & II and costimulatory molecules is absent at diagnosis in FUS::ERG AML, mirroring the epigenetic mechanism of post-transplant relapse seen in adult AML and its associated dismal outcomes. Here we show that this class-defining immune-repressive phenotype is driven by overexpression of the EZH2 histone lysine methyltransferase in vitro and in multiple clinical cohorts. We show that treatment with the FDA-approved EZH2 inhibitor tazemetostat along with IFN-γ reverses this phenotype, re-establishes MHC presentation, and severely impairs the viability of FUS::ERG AML cells. EZH2 inhibitors may thus provide the first targeted therapeutic option for patients with this high-risk subtype of pAML, with particular benefit as a bridge to successful allogeneic stem cell transplantation.

Validation of a genomic classifier that predicts metastasis following radical prostatectomy in an at risk patient population.

PURPOSE: Patients with locally advanced prostate cancer after radical prostatectomy are candidates for secondary therapy. However, this higher risk population is heterogeneous. Many cases do not metastasize even when conservatively managed. Given the limited specificity of pathological features to predict metastasis, newer risk prediction models are needed. We report a validation study of a genomic classifier that predicts metastasis after radical prostatectomy in a high risk population. MATERIALS AND METHODS: A case-cohort design was used to sample 1,010 patients after radical prostatectomy at high risk for recurrence who were treated from 2000 to 2006. Patients had preoperative prostate specific antigen greater than 20 ng/ml, Gleason 8 or greater, pT3b or a Mayo Clinic nomogram score of 10 or greater. Patients with metastasis at diagnosis or any prior treatment for prostate cancer were excluded from analysis. A 20% random sampling created a subcohort that included all patients with metastasis. We generated 22-marker genomic classifier scores for 219 patients with available genomic data. ROC and decision curves, competing risk and weighted regression models were used to assess genomic classifier performance. RESULTS: The genomic classifier AUC was 0.79 for predicting 5-year metastasis after radical prostatectomy. Decision curves showed that the genomic classifier net benefit exceeded that of clinical only models. The genomic classifier was the predominant predictor of metastasis on multivariable analysis. The cumulative incidence of metastasis 5 years after radical prostatectomy was 2.4%, 6.0% and 22.5% in patients with low (60%), intermediate (21%) and high (19%) genomic classifier scores, respectively (p<0.001). CONCLUSIONS: Results indicate that genomic information from the primary tumor can identify patients with adverse pathological features who are most at risk for metastasis and potentially lethal prostate cancer.

bamSliceR: cross-cohort variant and allelic bias analysis for rare variants and rare diseases.

Rare diseases and conditions create unique challenges for genetic epidemiologists precisely because cases and samples are scarce. In recent years, whole-genome and whole-transcriptome sequencing (WGS/WTS) have eased the study of rare genetic variants. Paired WGS and WTS data are ideal, but logistical and financial constraints often preclude generating paired WGS and WTS data. Thus, many databases contain a patchwork of specimens with either WGS or WTS data, but only a minority of samples have both. The NCI Genomic Data Commons facilitates controlled access to genomic and transcriptomic data for thousands of subjects, many with unpaired sequencing results. Local reanalysis of expressed variants across whole transcriptomes requires significant data storage, compute, and expertise. We developed the bamSliceR package to facilitate swift transition from aligned sequence reads to expressed variant characterization. bamSliceR leverages the NCI Genomic Data Commons API to query genomic sub-regions of aligned sequence reads from specimens identified through the robust Bioconductor ecosystem. We demonstrate how population-scale targeted genomic analysis can be completed using orders of magnitude fewer resources in this fashion, with minimal compute burden. We demonstrate pilot results from bamSliceR for the TARGET pediatric AML and BEAT-AML projects, where identification of rare but recurrent somatic variants directly yields biologically testable hypotheses. bamSliceR and its documentation are freely available on GitHub at https://github.com/trichelab/bamSliceR.

Developmental priming of cancer susceptibility.

DNA mutations are necessary drivers of cancer, yet only a small subset of mutated cells go on to cause the disease. To date, the mechanisms that determine which rare subset of cells transform and initiate tumorigenesis remain unclear. Here, we take advantage of a unique model of intrinsic developmental heterogeneity (Trim28+/D9) and demonstrate that stochastic early life epigenetic variation can trigger distinct cancer-susceptibility 'states' in adulthood. We show that these developmentally primed states are characterized by differential methylation patterns at typically silenced heterochromatin, and that these epigenetic signatures are detectable as early as 10 days of age. The differentially methylated loci are enriched for genes with known oncogenic potential. These same genes are frequently mutated in human cancers, and their dysregulation correlates with poor prognosis. These results provide proof-of-concept that intrinsic developmental heterogeneity can prime individual, life-long cancer risk.

Long Noncoding RNA Expression Independently Predicts Outcome in Pediatric Acute Myeloid Leukemia.

PURPOSE: Optimized strategies for risk classification are essential to tailor therapy for patients with biologically distinctive disease. Risk classification in pediatric acute myeloid leukemia (pAML) relies on detection of translocations and gene mutations. Long noncoding RNA (lncRNA) transcripts have been shown to associate with and mediate malignant phenotypes in acute myeloid leukemia (AML) but have not been comprehensively evaluated in pAML. METHODS: To identify lncRNA transcripts associated with outcomes, we evaluated the annotated lncRNA landscape by transcript sequencing of 1,298 pediatric and 96 adult AML specimens. Upregulated lncRNAs identified in the pAML training set were used to establish a regularized Cox regression model of event-free survival (EFS), yielding a 37 lncRNA signature (lncScore). Discretized lncScores were correlated with initial and postinduction treatment outcomes using Cox proportional hazards models in validation sets. Predictive model performance was compared with standard stratification methods by concordance analysis. RESULTS: Training set cases with positive lncScores had 5-year EFS and overall survival rates of 26.7% and 42.7%, respectively, compared with 56.9% and 76.3% with negative lncScores (hazard ratio, 2.48 and 3.16; P < .001). Pediatric validation cohorts and an adult AML group yielded comparable results in magnitude and significance. lncScore remained independently prognostic in multivariable models, including key factors used in preinduction and postinduction risk stratification. Subgroup analysis suggested that lncScores provide additional outcome information in heterogeneous subgroups currently classified as indeterminate risk. Concordance analysis showed that lncScore adds to overall classification accuracy with at least comparable predictive performance to current stratification methods that rely on multiple assays. CONCLUSION: Inclusion of the lncScore enhances predictive power of traditional cytogenetic and mutation-defined stratification in pAML with potential, as a single assay, to replace these complex stratification schemes with comparable predictive accuracy.

Evolutionary Landscape of SOX Genes to Inform Genotype-to-Phenotype Relationships.

The SOX transcription factor family is pivotal in controlling aspects of development. To identify genotype-phenotype relationships of SOX proteins, we performed a non-biased study of SOX using 1890 open-reading frame and 6667 amino acid sequences in combination with structural dynamics to interpret 3999 gnomAD, 485 ClinVar, 1174 Geno2MP, and 4313 COSMIC human variants. We identified, within the HMG (High Mobility Group)- box, twenty-seven amino acids with changes in multiple SOX proteins annotated to clinical pathologies. These sites were screened through Geno2MP medical phenotypes, revealing novel SOX15 R104G associated with musculature abnormality and SOX8 R159G with intellectual disability. Within gnomAD, SOX18 E137K (rs201931544), found within the HMG box of ~0.8% of Latinx individuals, is associated with seizures and neurological complications, potentially through blood-brain barrier alterations. A total of 56 highly conserved variants were found at sites outside the HMG-box, including several within the SOX2 HMG-box-flanking region with neurological associations, several in the SOX9 dimerization region associated with Campomelic Dysplasia, SOX14 K88R (rs199932938) flanking the HMG box associated with cardiovascular complications within European populations, and SOX7 A379V (rs143587868) within an SOXF conserved far C-terminal domain heterozygous in 0.716% of African individuals with associated eye phenotypes. This SOX data compilation builds a robust genotype-to-phenotype association for a gene family through more robust ortholog data integration.