Biology subjects

Kemmeren, P.

Publications and source records attributed to Kemmeren, P..

11 recordsLinked to original sources

bioRxiv2025

A pathway-informed mutual exclusivity framework to detect genetic interactions in pediatric cancer

BackgroundThe exponential increase of sequenced cancer genomes has enabled the in-silico study of genetic interactions in tumors across many cancer types - particularly synthetic lethality, where two gene alterations lead to cell death - and identify new candidate therapeutic targets. This rise is primarily present in adult cancer, while in-silico investigation of genetic interactions remains challenging in pediatric cancer especially in tumors of low incidence. Consequently, this underscores the need for specialized approaches to advance our understanding of genetic interactions in pediatric oncology. MethodsHere, we present the pathway-informed genetic interaction framework (PIGI) that employs mutual exclusivity and co-occurrence testing and leverages biological pathways to infer candidate genetic interactions. Pathways facilitate the detection of hidden biology by grouping genes in functional units, as well as alleviate key confounders of these analyses: pathway epistasis and cancer subtypes - thereby highlighting genes of greater interest. ResultsPIGI detected candidate genetic interactions by assessing pathway mutual exclusivity and co-occurrence in two primary pediatric cancer datasets, DKFZ and TARGET. PIGI detected, from high significance pathway relationships, 35 mutually exclusive and 2 co-occurring mutated gene pairs. The already known ME gene pair of TP53-DROSHA is detected in the much smaller collection of the DKFZ Wilms tumors. Over half of the identified gene pairs represent new discoveries that have not been previously described in the literature. Four of them could be promising candidates for synthetic lethal genetic interactions. ConclusionThese findings highlight the benefits of genetic interactions inference by exploring a different aspect of pediatric cancer data through pathways and propose new gene pairs for follow-up synthetic lethality experimentation.

bioinformatics↗

bioRxiv2025

Cell surface proteomics of patient-derived malignant rhabdoid tumor organoids identifies ROBO1 as potential CAR T cell target for pediatric solid tumors.

Malignant rhabdoid tumors are highly aggressive pediatric malignancies with limited treatment options and poor outcomes. To expand therapeutic options, we employed mass spectrometry-based cell surface proteomics on malignant rhabdoid tumor and patient-matched normal kidney organoids to identify potential CAR T cell targets. Integrating these findings with transcriptomics and protein expression data, we revealed ROBO1 as a promising target, showing strong and uniform expression across malignant rhabdoid and other pediatric tumors. ROBO1-targeted CAR T cells displayed potent anti-tumor activity in vitro, effectively eliminating tumor cells in co-cultures with organoids from malignant rhabdoid tumors, rhabdomyosarcoma, and neuroblastoma. In vivo, ROBO1 CAR T cells infiltrated tumors, induced potent tumor regression and significantly increased survival in malignant rhabdoid tumor-bearing mice. These findings establish ROBO1 as a compelling therapeutic target for CAR T cell therapy and offer a promising approach to address the critical need for effective treatments in high-risk pediatric solid tumors. Statement of significanceThis study identifies ROBO1 as a promising CAR T cell target for pediatric solid tumors. Using patient-derived tumor organoids and tissues, we demonstrate strong ROBO1 expression across several tumor entities and robust anti-tumor efficacy of ROBO1-targeted CAR T cells in vitro and in vivo, underscoring its potential for clinical translation.

cancer biology↗

bioRxiv2023

Complex structural variation is prevalent and highly pathogenic in pediatric solid tumors

BackgroundIn pediatric cancer, structural variants (SVs) and copy number alterations can contribute to cancer initiation and progression, and hence aid diagnosis and treatment stratification. The few studies into complex rearrangements have found associations with tumor aggressiveness or poor outcome. Yet, their prevalence and biological relevance across pediatric solid tumors remains unknown. ResultsIn a cohort of 120 primary tumors, we systematically characterized patterns of extrachromosomal DNA, chromoplexy and chromothripsis across five pediatric solid cancer types: neuroblastoma, Ewing sarcoma, Wilms tumor, hepatoblastoma and rhabdomyosarcoma. Complex SVs were identified in 56 tumors (47%) and different classes occurred across multiple cancer types. Recurrently mutated regions tend to be cancer-type specific and overlap with cancer genes, suggesting that selection contributes to shaping the SV landscape. In total, we identified potentially pathogenic complex SVs in 42 tumors that affect cancer driver genes or result in unfavorable chromosomal alterations. Half of which were known drivers, e.g. MYCN amplifications due to ecDNA and EWSR1::FLI1 fusions due to chromoplexy. Recurrent novel candidate complex events include chromoplexy in WT1 in Wilms tumors, focal chromothripsis with 1p loss in hepatoblastomas and complex MDM2 amplifications in rhabdomyosarcomas. ConclusionsComplex SVs are prevalent and pathogenic in pediatric solid tumors. They represent a type of genomic variation which currently remains unexplored. Moreover, carrying complex SVs seems to be associated with adverse clinical events. Our study highlights the potential for complex SVs to be incorporated in risk stratification or exploited for targeted treatments.

cancer biology↗

bioRxiv2022

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. For 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. We identified 1q+ in eight tumors that differ in mutational mechanisms, subsequent rearrangements and genomic contexts. 1q+ tumors were present in all four expression clusters and individual tumors overexpress different genes on 1q. Through 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.

cancer biology↗

bioRxiv2022

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 pediatric RMS tumor organoid (tumoroid) models comprising all major subtypes. For aggressive tumors, tumoroid models can often be established within four to eight 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 six 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.

cancer biology↗

bioRxiv2021

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

BackgroundGene 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. ResultsWe developed Fusion-sq to detect tumor-specific gene fusions by integrating and "fusing" evidence from RNA-seq and whole genome sequencing (WGS) using intron-exon gene structure. In a pediatric pan-cancer cohort of 130 patients, we identified 165 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 characterised by underlying SVs or expression changes indicative of activating or disruptive effects. ConclusionsOur 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 fusions that is suitable for precision oncology applications. Our method provides multi-omics evidence for assessing the pathogenicity of tumor-specific fusions for future clinical decision making.

genomics↗

bioRxiv2020

A comprehensive map of genetic interactions in childhood cancer reveals multiple underlying biological mechanisms

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 mechanisms underlying most candidates are 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 results of genetic interaction tests.

cancer biology↗

bioRxiv2020

Trecode: a FAIR eco-system for the analysis and archiving of omics data in a combined diagnostic and research setting

MotivationThe increase in speed, reliability and cost-effectiveness of high-throughput sequencing has led to the widespread clinical application of genome (WGS), exome (WXS) and transcriptome analysis. WXS and RNA sequencing is now being implemented as standard of care for patients and for patients included in clinical studies. To keep track of sample relationships and analyses, a platform is needed that can unify metadata for diverse sequencing strategies with sample metadata whilst supporting automated and reproducible analyses. In essence ensuring that analysis is conducted consistently, and data is Findable, Accessible, Interoperable and Reusable (FAIR). ResultsWe present "Trecode", a framework that records both clinical and research sample (meta) data and manages computational genome analysis workflows executed for both settings. Thereby achieving tight integration between analyses results and sample metadata. With complete, consistent and FAIR (meta) data management in a single platform, stacked bioinformatic analyses are performed automatically and tracked by the database ensuring data provenance, reproducibility and reusability which is key in worldwide collaborative translational research. Availability and implementationThe Trecode data model, codebooks, NGS workflows and client programs are currently being cleared from local compute infrastructure dependencies and will become publicly available in spring 2021. Contactp.kemmeren@prinsesmaximacentrum.nl

bioinformatics↗

bioRxiv2020

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 organising factor Abf1, at 191 sites simultaneously across the yeast genome. Average Abf1 residence times span a wide-range, varying between 4.5 and 37 minutes. 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.

molecular biology↗

bioRxiv2019

An extensively optimized chromatin immunoprecipitation protocol for quantitatively comparable and robust results

Chromatin immunoprecipitation (ChIP) is a commonly used technique to investigate which parts of a genome are bound by a particular protein. The result of ChIP is often interpreted in a binary manner: bound or not bound. Due to this focus, ChIP protocols frequently lack the ability to quantitatively compare samples with each other, for example in a time series or under different growth conditions. Here, using the yeast S. cerevisiae transcription factors Cbf1, Abf1, Reb1, Mcm1 and Sum1, we optimized the five major steps of a commonly used ChIP protocol: cross-linking, quenching, cell lysis, fragmentation and immunoprecipitation. Quenching with glycine is inefficient and can lead to large degrees of variability, an issue that is resolved by using tris(hydroxymethyl)aminomethane (Tris). Another source of variability is degradation of the protein of interest during the procedure. Enzymatic cell lysis with zymolyase can lead to extensive protein degradation, which is greatly reduced by mechanical lysis through bead beating. Degradation also occurs during sonication of chromatin, affecting large proteins in particular. An optimal mix of protease inhibitors and cross-linking with a higher percentage of formaldehyde reduces the extent of this degradation. Finally we also show that the immunoprecipitation step itself can be greatly improved with magnetic beads and optimized incubation/washing steps. The study results in a highly optimized protocol, which is shorter, easier to perform and has a stronger, more reproducible signal with less background. This protocol is presented in detail. In addition, the results highlight the greatest sources of variability in many other protocols, showing which steps are important to focus on for reproducible and quantitatively comparable ChIP experiments.

molecular biology↗

bioRxiv2018

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 putative 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 bakers 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.\n\nAuthor SummaryThe relationship between genotype and phenotype is one of the major challenges in biology. While many previous studies have identified genes involved in complex genetic diseases, there is still a gap between genotype and phenotype. One of the difficulties in filling this gap has been attributed to genetic interactions. Large-scale studies have revealed that genetic interactions are widespread in model organisms such as bakers yeast. Several molecular mechanisms have been proposed for different genetic interaction types. However, differences in occurrence and underlying molecular mechanism of genetic interactions have not yet been compared between gene classes of different function. Here, we compared genetic interaction patterns identified using gene expression profiling for two classes of genes: gene specific transcription factors and signaling related genes. We modelled all possible molecular networks to unravel putative molecular differences underlying different genetic interaction patterns. Our study proposes a new mechanistic explanation for a certain genetic interaction pattern that is more strongly associated with transcription factors compared to signaling related genes. Overall, our findings and the computational methodologies implemented here can be valuable for understanding the molecular mechanisms underlying genetic interactions.

genetics↗