Systematic discovery of gene fusions in pediatric cancer by integrating RNA-seq and WGS.

BACKGROUND: Gene fusions are important cancer drivers in pediatric cancer and their accurate detection is essential for diagnosis and treatment. Clinical decision-making requires high confidence and precision of detection. Recent developments show RNA sequencing (RNA-seq) is promising for genome-wide detection of fusion products but hindered by many false positives that require extensive manual curation and impede discovery of pathogenic fusions. METHODS: We developed Fusion-sq to overcome existing disadvantages of detecting gene fusions. Fusion-sq integrates and "fuses" evidence from RNA-seq and whole genome sequencing (WGS) using intron-exon gene structure to identify tumor-specific protein coding gene fusions. Fusion-sq was then applied to the data generated from a pediatric pan-cancer cohort of 128 patients by WGS and RNA sequencing. RESULTS: In a pediatric pan-cancer cohort of 128 patients, we identified 155 high confidence tumor-specific gene fusions and their underlying structural variants (SVs). This includes all clinically relevant fusions known to be present in this cohort (30 patients). Fusion-sq distinguishes healthy-occurring from tumor-specific fusions and resolves fusions in amplified regions and copy number unstable genomes. A high gene fusion burden is associated with copy number instability. We identified 27 potentially pathogenic fusions involving oncogenes or tumor-suppressor genes characterized by underlying SVs, in some cases leading to expression changes indicative of activating or disruptive effects. CONCLUSIONS: Our results indicate how clinically relevant and potentially pathogenic gene fusions can be identified and their functional effects investigated by combining WGS and RNA-seq. Integrating RNA fusion predictions with underlying SVs advances fusion detection beyond extensive manual filtering. Taken together, we developed a method for identifying candidate gene fusions that is suitable for precision oncology applications. Our method provides multi-omics evidence for assessing the pathogenicity of tumor-specific gene fusions for future clinical decision making.

Elevated enhancer-oncogene contacts and higher oncogene expression levels by recurrent CTCF inactivating mutations in acute T cell leukemia.

Monoallelic inactivation of CCCTC-binding factor (CTCF) in human cancer drives altered methylated genomic states, altered CTCF occupancy at promoter and enhancer regions, and deregulated global gene expression. In patients with T cell acute lymphoblastic leukemia (T-ALL), we find that acquired monoallelic CTCF-inactivating events drive subtle and local genomic effects in nearly half of t(5; 14) (q35; q32.2) rearranged patients, especially when CTCF-binding sites are preserved in between the BCL11B enhancer and the TLX3 oncogene. These solitary intervening sites insulate TLX3 from the enhancer by inducing competitive looping to multiple binding sites near the TLX3 promoter. Reduced CTCF levels or deletion of the intervening CTCF site abrogates enhancer insulation by weakening competitive looping while favoring TLX3 promoter to BCL11B enhancer looping, which elevates oncogene expression levels and leukemia burden.

Molecular Characterization Reveals Subclasses of 1q Gain in Intermediate Risk Wilms Tumors.

Chromosomal alterations have recurrently been identified in Wilms tumors (WTs) and some are associated with poor prognosis. Gain of 1q (1q+) is of special interest given its high prevalence and is currently actively studied for its prognostic value. However, the underlying mutational mechanisms and functional effects remain unknown. In a national unbiased cohort of 30 primary WTs, we integrated somatic SNVs, CNs and SVs with expression data and distinguished four clusters characterized by affected biological processes: muscle differentiation, immune system, kidney development and proliferation. Combined genome-wide CN and SV profiles showed that tumors profoundly differ in both their types of 1q+ and genomic stability and can be grouped into WTs with co-occurring 1p-/1q+, multiple chromosomal gains or CN neutral tumors. We identified 1q+ in eight tumors that differ in mutational mechanisms, subsequent rearrangements and genomic contexts. Moreover, 1q+ tumors were present in all four expression clusters reflecting activation of various biological processes, and individual tumors overexpress different genes on 1q. In conclusion, by integrating CNs, SVs and gene expression, we identified subgroups of 1q+ tumors reflecting differences in the functional effect of 1q gain, indicating that expression data is likely needed for further risk stratification of 1q+ WTs.

Implementation of paediatric precision oncology into clinical practice: The Individualized Therapies for Children with cancer program 'iTHER'.

iTHER is a Dutch prospective national precision oncology program aiming to define tumour molecular profiles in children and adolescents with primary very high-risk, relapsed, or refractory paediatric tumours. Between April 2017 and April 2021, 302 samples from 253 patients were included. Comprehensive molecular profiling including low-coverage whole genome sequencing (lcWGS), whole exome sequencing (WES), RNA sequencing (RNA-seq), Affymetrix, and/or 850k methylation profiling was successfully performed for 226 samples with at least 20% tumour content. Germline pathogenic variants were identified in 16% of patients (35/219), of which 22 variants were judged causative for a cancer predisposition syndrome. At least one somatic alteration was detected in 204 (90.3%), and 185 (81.9%) were considered druggable, with clinical priority very high (6.1%), high (21.3%), moderate (26.0%), intermediate (36.1%), and borderline (10.5%) priority. iTHER led to revision or refinement of diagnosis in 8 patients (3.5%). Temporal heterogeneity was observed in paired samples of 15 patients, indicating the value of sequential analyses. Of 137 patients with follow-up beyond twelve months, 21 molecularly matched treatments were applied in 19 patients (13.9%), with clinical benefit in few. Most relevant barriers to not applying targeted therapies included poor performance status, as well as limited access to drugs within clinical trial. iTHER demonstrates the feasibility of comprehensive molecular profiling across all ages, tumour types and stages in paediatric cancers, informing of diagnostic, prognostic, and targetable alterations as well as reportable germline variants. Therefore, WES and RNA-seq is nowadays standard clinical care at the Princess Máxima Center for all children with cancer, including patients at primary diagnosis. Improved access to innovative treatments within biology-driven combination trials is required to ultimately improve survival.

Mesenchymal tumor organoid models recapitulate rhabdomyosarcoma subtypes.

Rhabdomyosarcomas (RMS) are mesenchyme-derived tumors and the most common childhood soft tissue sarcomas. Treatment is intense, with a nevertheless poor prognosis for high-risk patients. Discovery of new therapies would benefit from additional preclinical models. Here, we describe the generation of a collection of 19 pediatric RMS tumor organoid (tumoroid) models (success rate of 41%) comprising all major subtypes. For aggressive tumors, tumoroid models can often be established within 4-8 weeks, indicating the feasibility of personalized drug screening. Molecular, genetic, and histological characterization show that the models closely resemble the original tumors, with genetic stability over extended culture periods of up to 6 months. Importantly, drug screening reflects established sensitivities and the models can be modified by CRISPR/Cas9 with TP53 knockout in an embryonal RMS model resulting in replicative stress drug sensitivity. Tumors of mesenchymal origin can therefore be used to generate organoid models, relevant for a variety of preclinical and clinical research questions.

Improved Gene Fusion Detection in Childhood Cancer Diagnostics Using RNA Sequencing.

PURPOSE: Gene fusions play a significant role in cancer etiology, making their detection crucial for accurate diagnosis, prognosis, and determining therapeutic targets. Current diagnostic methods largely focus on either targeted or low-resolution genome-wide techniques, which may be unable to capture rare events or both fusion partners. We investigate if RNA sequencing can overcome current limitations with traditional diagnostic techniques to identify gene fusion events. METHODS: We first performed RNA sequencing on a validation cohort of 24 samples with a known gene fusion event, after which a prospective pan-pediatric cancer cohort (n = 244) was tested by RNA sequencing in parallel to existing diagnostic procedures. This cohort included hematologic malignancies, tumors of the CNS, solid tumors, and suspected neoplastic samples. All samples were processed in the routine diagnostic workflow and analyzed for gene fusions using standard-of-care methods and RNA sequencing. RESULTS: We identified a clinically relevant gene fusion in 83 of 244 cases in the prospective cohort. Sixty fusions were detected by both routine diagnostic techniques and RNA sequencing, and one fusion was detected only in routine diagnostics, but an additional 24 fusions were detected solely by RNA sequencing. RNA sequencing, therefore, increased the diagnostic yield by 38%-39%. In addition, RNA sequencing identified both gene partners involved in the gene fusion, in contrast to most routine techniques. For two patients, the newly identified fusion by RNA sequencing resulted in treatment with targeted agents. CONCLUSION: We show that RNA sequencing is sufficiently robust for gene fusion detection in routine diagnostics of childhood cancers and can make a difference in treatment decisions.

A systematic analysis of genetic interactions and their underlying biology in childhood cancer.

Childhood cancer is a major cause of child death in developed countries. Genetic interactions between mutated genes play an important role in cancer development. They can be detected by searching for pairs of mutated genes that co-occur more (or less) often than expected. Co-occurrence suggests a cooperative role in cancer development, while mutual exclusivity points to synthetic lethality, a phenomenon of interest in cancer treatment research. Little is known about genetic interactions in childhood cancer. We apply a statistical pipeline to detect genetic interactions in a combined dataset comprising over 2,500 tumors from 23 cancer types. The resulting genetic interaction map of childhood cancers comprises 15 co-occurring and 27 mutually exclusive candidates. The biological explanation of most candidates points to either tumor subtype, pathway epistasis or cooperation while synthetic lethality plays a much smaller role. Thus, other explanations beyond synthetic lethality should be considered when interpreting genetic interaction test results.

Structural variant detection in cancer genomes: computational challenges and perspectives for precision oncology.

Cancer is generally characterized by acquired genomic aberrations in a broad spectrum of types and sizes, ranging from single nucleotide variants to structural variants (SVs). At least 30% of cancers have a known pathogenic SV used in diagnosis or treatment stratification. However, research into the role of SVs in cancer has been limited due to difficulties in detection. Biological and computational challenges confound SV detection in cancer samples, including intratumor heterogeneity, polyploidy, and distinguishing tumor-specific SVs from germline and somatic variants present in healthy cells. Classification of tumor-specific SVs is challenging due to inconsistencies in detected breakpoints, derived variant types and biological complexity of some rearrangements. Full-spectrum SV detection with high recall and precision requires integration of multiple algorithms and sequencing technologies to rescue variants that are difficult to resolve through individual methods. Here, we explore current strategies for integrating SV callsets and to enable the use of tumor-specific SVs in precision oncology.

Suboptimal Global Transcriptional Response Increases the Harmful Effects of Loss-of-Function Mutations.

The fitness impact of loss-of-function mutations is generally assumed to reflect the loss of specific molecular functions associated with the perturbed gene. Here, we propose that rewiring of the transcriptome upon deleterious gene inactivation is frequently nonspecific and mimics stereotypic responses to external environmental change. Consequently, transcriptional response to gene deletion could be suboptimal and incur an extra fitness cost. Analysis of the transcriptomes of ∼1,500 single-gene deletion Saccharomyces cerevisiae strains supported this scenario. First, most transcriptomic changes are not specific to the deleted gene but are rather triggered by perturbations in functionally diverse genes. Second, gene deletions that alter the expression of dosage-sensitive genes are especially harmful. Third, by elevating the expression level of downregulated genes, we could experimentally mitigate the fitness defect of gene deletions. Our work shows that rewiring of genomic expression upon gene inactivation shapes the harmful effects of mutations.

Genome-wide off-rates reveal how DNA binding dynamics shape transcription factor function.

Protein-DNA interactions are dynamic, and these dynamics are an important aspect of chromatin-associated processes such as transcription or replication. Due to a lack of methods to study on- and off-rates across entire genomes, protein-DNA interaction dynamics have not been studied extensively. Here, we determine in vivo off-rates for the Saccharomyces cerevisiae chromatin organizing factor Abf1, at 191 sites simultaneously across the yeast genome. Average Abf1 residence times span a wide range, varying between 4.2 and 33 min. Sites with different off-rates are associated with different functional characteristics. This includes their transcriptional dependency on Abf1, nucleosome positioning and the size of the nucleosome-free region, as well as the ability to roadblock RNA polymerase II for termination. The results show how off-rates contribute to transcription factor function and that DIVORSEQ (Determining In Vivo Off-Rates by SEQuencing) is a meaningful way of investigating protein-DNA binding dynamics genome-wide.

An Optimized Chromatin Immunoprecipitation Protocol for Quantification of Protein-DNA Interactions.

Transcription factors are important regulators of cell fate and function. Knowledge about where transcription factors are bound in the genome is crucial for understanding their function. A common method to study protein-DNA interactions is chromatin immunoprecipitation (ChIP). Here, we present a revised ChIP protocol to determine protein-DNA interactions for the yeast Saccharomyces cerevisiae. We optimized several aspects of the procedure, including cross-linking and quenching, cell lysis, and immunoprecipitation steps. This protocol facilitates sensitive and reproducible quantitation of protein-DNA interactions. For complete details on the use and execution of this protocol, please refer to (de Jonge et al., 2019).

An organoid biobank for childhood kidney cancers that captures disease and tissue heterogeneity.

Kidney tumours are among the most common solid tumours in children, comprising distinct subtypes differing in many aspects, including cell-of-origin, genetics, and pathology. Pre-clinical cell models capturing the disease heterogeneity are currently lacking. Here, we describe the first paediatric cancer organoid biobank. It contains tumour and matching normal kidney organoids from over 50 children with different subtypes of kidney cancer, including Wilms tumours, malignant rhabdoid tumours, renal cell carcinomas, and congenital mesoblastic nephromas. Paediatric kidney tumour organoids retain key properties of native tumours, useful for revealing patient-specific drug sensitivities. Using single cell RNA-sequencing and high resolution 3D imaging, we further demonstrate that organoid cultures derived from Wilms tumours consist of multiple different cell types, including epithelial, stromal and blastemal-like cells. Our organoid biobank captures the heterogeneity of paediatric kidney tumours, providing a representative collection of well-characterised models for basic cancer research, drug-screening and personalised medicine.

DNA Methylation Profiling Identifies Distinct Clusters in Angiosarcomas.

PURPOSE: DNA methylation profiling has previously uncovered biologically and clinically meaningful subgroups within many tumor types, but was not yet performed in angiosarcoma. Angiosarcoma is a rare sarcoma with very heterogeneous clinical presentations, which may be based on differences in biological background. In this exploratory study, DNA methylation profiling of 36 primary angiosarcoma samples from visceral, deep soft tissue, radiation-induced, and UV-induced localizations was performed. EXPERIMENTAL DESIGN: Primary angiosarcoma formalin-fixed paraffin-embedded samples from visceral, soft tissue, radiation-induced, and UV-induced origin were collected from a nationwide search for angiosarcoma in the Netherlands. DNA was extracted for methylation profiling with the Illumina Infinium MethylationEPIC array. Quality control assessment and unsupervised hierarchical clustering were performed. Copy-number profiles were generated and analyzed for chromosomal stability. Clinical data were obtained from the Netherlands Cancer Registry. RESULTS: DNA methylation profiling by unsupervised hierarchical clustering of 36 angiosarcoma samples (6 visceral, 5 soft tissue, 14 radiation-induced, 11 UV-induced) revealed two main clusters (A and B), which were divided into four subclusters. The clusters largely corresponded with clinical subtypes, showing enrichment of UV-induced cases in cluster A1 and radiation-induced cases in cluster A2. Visceral and soft tissue cases almost exclusively fell into cluster B. Cluster A showed significantly increased chromosomal instability and better overall survival (22 vs. 6 months, P = 0.046) compared with cluster B. CONCLUSIONS: In this novel methylation profiling study, we demonstrated for the first time four different angiosarcoma clusters. These clusters correlated with clinical subtype, overall survival, and chromosomal stability.

Kaposiform hemangioendothelioma and tufted angioma - (epi)genetic analysis including genome-wide methylation profiling.

Kaposiform hemangioendothelioma (KHE) is a locally aggressive vascular condition of childhood and is clinicopathologically related to tufted angioma (TA), a benign skin lesion. Due to their rarity molecular data are scarce. We investigated 7 KHE and 3 TA by comprehensive mutational analysis and genome-wide methylation profiling and compared the clustering, also with vascular malformations. Lesions were from 7 females and 3 males. The age range was 2 months to 9 years with a median of 10 months. KHEs arose in the soft tissue of the thigh (n = 2), retroperitoneum (n = 1), thoracal/abdominal (n = 1), supraclavicular (n = 1) and neck (n = 1). One patient presented with multiple lesions without further information. Two patients developed a Kasabach-Merritt phenomenon. TAs originated in the skin of the shoulder (n = 2) and nose/forehead (n = 1). Of the 5 KHEs and 2 TAs investigated by DNA sequencing, one TA showed a hot spot mutation in NRAS, and one KHE a mutation in RAD50. Unsupervised hierarchical clustering analysis indicated a common methylation pattern of KHEs and TAs, which separated from the homogeneous methylation pattern of vascular malformations. In conclusion, methylation profiling provides further evidence for KHEs and TAs potentially forming a spectrum of one entity. Using next generation sequencing, heterogeneous mutations were found in a subset of cases (2/7) without the presence of GNA14 mutations, previously reported in KHE and TA.

A framework for exhaustive modelling of genetic interaction patterns using Petri nets.

MOTIVATION: Genetic interaction (GI) patterns are characterized by the phenotypes of interacting single and double mutated gene pairs. Uncovering the regulatory mechanisms of GIs would provide a better understanding of their role in biological processes, diseases and drug response. Computational analyses can provide insights into the underpinning mechanisms of GIs. RESULTS: In this study, we present a framework for exhaustive modelling of GI patterns using Petri nets (PN). Four-node models were defined and generated on three levels with restrictions, to enable an exhaustive approach. Simulations suggest ∼5 million models of GIs. Generalizing these we propose putative mechanisms for the GI patterns, inversion and suppression. We demonstrate that exhaustive PN modelling enables reasoning about mechanisms of GIs when only the phenotypes of gene pairs are known. The framework can be applied to other GI or genetic regulatory datasets. AVAILABILITY AND IMPLEMENTATION: The framework is available at http://www.ibi.vu.nl/programs/ExhMod. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

The ability of transcription factors to differentially regulate gene expression is a crucial component of the mechanism underlying inversion, a frequently observed genetic interaction pattern.

Genetic interactions, a phenomenon whereby combinations of mutations lead to unexpected effects, reflect how cellular processes are wired and play an important role in complex genetic diseases. Understanding the molecular basis of genetic interactions is crucial for deciphering pathway organization as well as understanding the relationship between genetic variation and disease. Several hypothetical molecular mechanisms have been linked to different genetic interaction types. However, differences in genetic interaction patterns and their underlying mechanisms have not yet been compared systematically between different functional gene classes. Here, differences in the occurrence and types of genetic interactions are compared for two classes, gene-specific transcription factors (GSTFs) and signaling genes (kinases and phosphatases). Genome-wide gene expression data for 63 single and double deletion mutants in baker's yeast reveals that the two most common genetic interaction patterns are buffering and inversion. Buffering is typically associated with redundancy and is well understood. In inversion, genes show opposite behavior in the double mutant compared to the corresponding single mutants. The underlying mechanism is poorly understood. Although both classes show buffering and inversion patterns, the prevalence of inversion is much stronger in GSTFs. To decipher potential mechanisms, a Petri Net modeling approach was employed, where genes are represented as nodes and relationships between genes as edges. This allowed over 9 million possible three and four node models to be exhaustively enumerated. The models show that a quantitative difference in interaction strength is a strict requirement for obtaining inversion. In addition, this difference is frequently accompanied with a second gene that shows buffering. Taken together, these results provide a mechanistic explanation for inversion. Furthermore, the ability of transcription factors to differentially regulate expression of their targets provides a likely explanation why inversion is more prevalent for GSTFs compared to kinases and phosphatases.

Next-generation phenotyping using computer vision algorithms in rare genomic neurodevelopmental disorders.

PURPOSE: The interpretation of genetic variants after genome-wide analysis is complex in heterogeneous disorders such as intellectual disability (ID). We investigate whether algorithms can be used to detect if a facial gestalt is present for three novel ID syndromes and if these techniques can help interpret variants of uncertain significance. METHODS: Facial features were extracted from photos of ID patients harboring a pathogenic variant in three novel ID genes (PACS1, PPM1D, and PHIP) using algorithms that model human facial dysmorphism, and facial recognition. The resulting features were combined into a hybrid model to compare the three cohorts against a background ID population. RESULTS: We validated our model using images from 71 individuals with Koolen-de Vries syndrome, and then show that facial gestalts are present for individuals with a pathogenic variant in PACS1 (p = 8 × 10-4), PPM1D (p = 4.65 × 10-2), and PHIP (p = 6.3 × 10-3). Moreover, two individuals with a de novo missense variant of uncertain significance in PHIP have significant similarity to the expected facial phenotype of PHIP patients (p < 1.52 × 10-2). CONCLUSION: Our results show that analysis of facial photos can be used to detect previously unknown facial gestalts for novel ID syndromes, which will facilitate both clinical and molecular diagnosis of rare and novel syndromes.

The clinical implementation of copy number detection in the age of next-generation sequencing.

INTRODUCTION: The role of copy number variants (CNVs) in disease is now well established. In parallel NGS technologies, such as long-read technologies, there is continual development and data analysis methods continue to be refined. Clinical exome sequencing data is now a reality for many diagnostic laboratories in both congenital genetics and oncology. This provides the ability to detect and report both SNVs and structural variants, including CNVs, using a single assay for a wide range of patient cohorts. Areas covered: Currently, whole-genome sequencing is mainly restricted to research applications and clinical utility studies. Furthermore, detecting the full-size spectrum of CNVs as well as somatic events remains difficult for both exome and whole-genome sequencing. As a result, the full extent of genomic variants in an individual's genome is still largely unknown. Recently, new sequencing technologies have been introduced which maintain the long-range genomic context, aiding the detection of CNVs and structural variants. Expert commentary: The development of long-read sequencing promises to resolve many CNV and SV detection issues but is yet to become established. The current challenge for clinical CNV detection is how to fully exploit all the data which is generated by high throughput sequencing technologies.

Growth condition dependency is the major cause of non-responsiveness upon genetic perturbation.

Investigating the role and interplay between individual proteins in biological processes is often performed by assessing the functional consequences of gene inactivation or removal. Depending on the sensitivity of the assay used for determining phenotype, between 66% (growth) and 53% (gene expression) of Saccharomyces cerevisiae gene deletion strains show no defect when analyzed under a single condition. Although it is well known that this non-responsive behavior is caused by different types of redundancy mechanisms or by growth condition/cell type dependency, it is not known what the relative contribution of these different causes is. Understanding the underlying causes of and their relative contribution to non-responsive behavior upon genetic perturbation is extremely important for designing efficient strategies aimed at elucidating gene function and unraveling complex cellular systems. Here, we provide a systematic classification of the underlying causes of and their relative contribution to non-responsive behavior upon gene deletion. The overall contribution of redundancy to non-responsive behavior is estimated at 29%, of which approximately 17% is due to homology-based redundancy and 12% is due to pathway-based redundancy. The major determinant of non-responsiveness is condition dependency (71%). For approximately 14% of protein complexes, just-in-time assembly can be put forward as a potential mechanistic explanation for how proteins can be regulated in a condition dependent manner. Taken together, the results underscore the large contribution of growth condition requirement to non-responsive behavior, which needs to be taken into account for strategies aimed at determining gene function. The classification provided here, can also be further harnessed in systematic analyses of complex cellular systems.

Molecular mechanisms that distinguish TFIID housekeeping from regulatable SAGA promoters.

An important distinction is frequently made between constitutively expressed housekeeping genes versus regulated genes. Although generally characterized by different DNA elements, chromatin architecture and cofactors, it is not known to what degree promoter classes strictly follow regulatability rules and which molecular mechanisms dictate such differences. We show that SAGA-dominated/TATA-box promoters are more responsive to changes in the amount of activator, even compared to TFIID/TATA-like promoters that depend on the same activator Hsf1. Regulatability is therefore an inherent property of promoter class. Further analyses show that SAGA/TATA-box promoters are more dynamic because TATA-binding protein recruitment through SAGA is susceptible to removal by Mot1. In addition, the nucleosome configuration upon activator depletion shifts on SAGA/TATA-box promoters and seems less amenable to preinitiation complex formation. The results explain the fundamental difference between housekeeping and regulatable genes, revealing an additional facet of combinatorial control: an activator can elicit a different response dependent on core promoter class.