Chromatin remodeling complexes play essential roles in metazoan development through widespread control of gene expression, but the precise molecular mechanisms by which they do this in vivo remain ill defined. Using an inducible system with fine temporal resolution, we show that the nucleosome remodeling and deacetylation (NuRD) complex controls chromatin architecture and the protein binding repertoire at regulatory regions during cell state transitions. This is primarily exerted through its nucleosome remodeling activity while deacetylation at H3K27 follows changes in gene expression. Additionally, NuRD activity influences association of RNA polymerase II at transcription start sites and subsequent nascent transcript production, thereby guiding the establishment of lineage-appropriate transcriptional programs. These findings provide a detailed molecular picture of genome-wide modulation of lineage-specific transcription by an essential chromatin remodeling complex as well as insight into the orchestration of molecular events involved in transcriptional transitions in vivo. VIDEO

Tumourigenic transformation of normal cells into cancer typically involves several steps resulting in acquisition of unlimited growth potential, evasion of apoptosis and non-responsiveness to growth inhibitory signals. Both genetic and epigenetic changes can contribute to cancer development and progression. Given the vast genetic heterogeneity of human cancers and difficulty to monitor cancer-initiating events in vivo, the precise relationship between acquisition of genetic mutations and the temporal progression of epigenetic alterations in transformed cells is largely unclear. Here, we use an in vitro model system to investigate the contribution of cellular immortality and oncogenic transformation of primary human cells to epigenetic reprogramming of DNA methylation and gene expression. Our data demonstrate that extension of replicative life span of the cells is sufficient to induce accumulation of DNA methylation at gene promoters and large-scale changes in gene expression in a time-dependent manner. In contrast, continuous expression of cooperating oncogenes in immortalized cells, although essential for anchorage-independent growth and evasion of apoptosis, does not affect de novo DNA methylation at promoters and induces subtle expression changes. Taken together, these observations imply that cellular immortality promotes epigenetic adaptation to highly proliferative state, whereas transforming oncogenes confer additional properties to transformed human cells.

Haemoglobinopathies are the commonest monogenic diseases worldwide and are caused by variants in the globin gene clusters. With over 2400 variants detected to date, their interpretation using the American College of Medical Genetics and Genomics (ACMG)/Association for Molecular Pathology (AMP) guidelines is challenging and computational evidence can provide valuable input about their functional annotation. While many in silico predictors have already been developed, their performance varies for different genes and diseases. In this study, we evaluate 31 in silico predictors using a dataset of 1627 variants in HBA1, HBA2, and HBB. By varying the decision threshold for each tool, we analyse their performance (a) as binary classifiers of pathogenicity and (b) by using different non-overlapping pathogenic and benign thresholds for their optimal use in the ACMG/AMP framework. Our results show that CADD, Eigen-PC, and REVEL are the overall top performers, with the former reaching moderate strength level for pathogenic prediction. Eigen-PC and REVEL achieve the highest accuracies for missense variants, while CADD is also a reliable predictor of non-missense variants. Moreover, SpliceAI is the top performing splicing predictor, reaching strong level of evidence, while GERP++ and phyloP are the most accurate conservation tools. This study provides evidence about the optimal use of computational tools in globin gene clusters under the ACMG/AMP framework.

Several types of haemoglobinopathies are caused by copy number variants (CNVs). While diagnosis is often based on haematological and biochemical parameters, a definitive diagnosis requires molecular DNA analysis. In some cases, the molecular characterisation of large deletions/duplications is challenging and inconclusive and often requires the use of specific diagnostic procedures, such as multiplex ligation-dependent probe amplification (MLPA). Herein, we collected and comprehensively analysed all known CNVs associated with haemoglobinopathies. The dataset of 291 CNVs was retrieved from the IthaGenes database and was further manually annotated to specify genomic locations, breakpoints and MLPA probes relevant for each CNV. We developed IthaCNVs, a publicly available and easy-to-use online tool that can facilitate the diagnosis of rare and diagnostically challenging haemoglobinopathy cases attributed to CNVs. Importantly, it facilitates the filtering of available entries based on the type of breakpoint information, on specific chromosomal and locus positions, on MLPA probes, and on affected gene(s). IthaCNVs brings together manually curated information about CNV genomic locations, functional effects, and information that can facilitate CNV characterisation through MLPA. It can help laboratory staff and clinicians confirm suspected diagnosis of CNVs based on molecular DNA screening and analysis.

Introduction: Haemoglobinopathies are the commonest monogenic diseases worldwide and are caused by variants in the globin gene clusters. With over 2400 variants detected to date, their interpretation using the ACMG/AMP guidelines is challenging, with computational evidence able to provide valuable input about their functional annotation. While many in silico predictors have already been developed, their performance varies for different genes and diseases. Materials and Methods: We evaluate 31 in silico predictors using a dataset of 1627 variants in HBA1, HBA2, and HBB. Through varying the decision threshold for each tool, we analyse their performance (a) as binary classifiers of pathogenicity, and (b) using different nonoverlapping pathogenic and benign thresholds for their optimal use in the ACMG/AMP framework. Results: CADD, Eigen-PC, and REVEL are the overall top performers, with the former reaching moderate strength level for pathogenic prediction. Eigen-PC and REVEL achieve the highest accuracies for missense variants, while CADD is also a reliable predictor of nonmissense variants. Moreover, SpliceAl is the top performing splicing predictor, reaching strong level of evidence, while GERP++ and phyloP are the most accurate conservation tools. Discussion: This study provides evidence about the optimal use of computational evidence in globin gene clusters under the ACMG/AMP framework.