WYMM Tour: Leiden

Long read nanopore sequencing provides a paradigm shift for human diagnostics. The use of long reads may not only provide improved diagnostic yields for the detection of developmental or rare disorders, but may also streamline the diagnostic process by integrating methylome information. For the first time, it may become possible to map hitherto unmappable regions, regions which may harbor hitherto undiscoverd genes. In addition, the use of long reads may extend beyond rare diseases. We have been exploring its use for liquid biopsies, prenatal diagnosis and single cell analyses. I will showcase, where and how we can expect diagnostic implementation of this technology when validated.

Abstract: Epigenetic changes at the level of methylation, histone modifications or non-coding RNA can lead to divergent medically relevant phenotypes. Nanopore sequencing leverages the ability to simultaneously phase large blocks of the human genome with long-reads and capture the all-context methylation status of cytosine. Here we characterize the causative genetic and epigenetic variations of one subject with epi-cblC and of several subjects with FSHD.

Faithful epigenetic regulation of the genome is critical for normal development. Disruption to these processes can result in faulty DNA methylation patterns, a phenomenon that is increasingly being recognized for human developmental disease. These so-called “episignatures” have been implemented in clinical diagnostics. However, the impact of abnormal DNA methylation, which can easily be detected in patient blood, on molecular and developmental processes is not fully understood. The major challenge now, is to link (ab)normal DNA methylation patterns to disease phenotype. We utilize genome-wide and adaptive sampling Oxford Nanopore long read sequencing approaches to map DNA methylation patterns of complex genomic regions in patient material and mouse models carrying pathogenic mutations in epigenetic regulators. Our recent work uncovered that global aberrant DNA hypomethylation can translate to localized changes in chromatin state and gene expression at gene clusters. This emphasizes the importance of considering tissue and cell type-specific effects when studying disorders where germline mutations in epigenetic regulators are correlated with genome-wide defects in DNA methylation.

Next-generation sequencing has revolutionized our understanding of genetic variation and its impact on health and disease. However, traditional short-read bulk sequencing methods often fail to capture complex structural variants and are unable to detect the intricate diversity within individual cells. To address this limitation, we have developed a single-cell multiomic approach that combines the power of cellular barcoding from 10x Genomics and long-read nanopore sequencing from Oxford Nanopore Technologies. Our method enables the simultaneous analysis of the genome, transcriptome, and chromatin accessibility of thousands of individual cells. By generating high-quality, long-read sequencing data, we can accurately identify structural variants, gene fusions, alternative splicing, and other complex genetic alterations that are often missed by short-read sequencing. These information-rich datasets allow us to better understand genome function and discern genetic and non-genetic contributions to clonal evolution during health and disease.

Performing and reporting genetic test results within days, can significantly alter clinical decisions, guide therapeutic strategies, and increase patient positive outcomes. This is especially true in a pediatric and neonatal critical care settings, where each hour can make a difference. Nonetheless, conventional testing often requires multiple methods to reach a diagnosis leading to turnarounds extending over weeks. Ultrarapid genetic testing with nanopore sequencing is recently emerging as a transformative tool in a critical care setting thanks to the streamlined workflow, speed, and real-time data generation. In this study, we aimed at assessing the potential future applicability of Nanopore sequencing for ultrarapid genetic diagnostics, as a single test able to detect at once multiple variant types, DNA modifications, and phasing information. Our results confirm the potential of this technology to revolutionize the diagnostic testing for ultrarapid testing purposes.

Abstract: Nanopore sequencing is a powerful technique with numerous advantages for clinical genetics applications. In a single run, it can detect and analyse structural variants, indels, and SNVs, phase alleles, and study modified bases in both coding and non-coding regions of the genome, including single and repeated regions. These capabilities enable precise detection and a deeper understanding of patients’ genetic variants, with the potential to refine diagnoses and genetic counselling. Adaptive Sampling enhances this process by easily increasing coverage and precision in specific regions of interest (e.g. repeats, differentially methylated regions, targeted loci). We will present various technical points and clinical cases.

Despite the implementation of massive parallel sequencing as standard of care, about half of patients with developmental disorders (DD) remain without genetic diagnosis. Structural and epigenetic variants have long been known to be involved in the pathogenesis of DD but remain challenging to map due to technological limitations. However, with dropping costs and increasing accuracy of long read sequencing (LRS) platforms, the concomitant assessment of single nucleotide, structural and epigenetic variation becomes feasible. Methods: To unravel hitherto unidentified causes of intellectual disability and/or multiple congenital anomalies, we performed whole genome nanopore sequencing in 25 patient-parent trios without molecular diagnosis after short read exome or genome sequencing. We developed an analytical pipeline and built a population reference set to assess structural variants (SV), comparing the hg38 and T2T human reference genomes. In addition, we validated phased methylation detection looking at imprinted regions as well as X-inactivation and explored the detection of genome-wide episignatures. Results: We identify around 25.000 SV per individual. Approximately 100 of these SV are classified as de novo. However, only 0.4 potential true de novo SV are retained upon manual curation and population filtering. A mean of 1.4 X-linked inherited SV are identified in males. Using the T2T reference genome, we identified 3 de novo deletions that had been missed using hg38. Some of the identified SV disrupt regulatory sequences of DD associated genes. We were able to demonstrate the concurrent detection of episignatures and underlying genomic variants. The haplotyped methylome is now being scrutinized to investigate potential pathogenic mechanisms of non-coding variants. Conclusion: LRS enables the detection of few, but potential disease-causing, hitherto unidentified SV. LRS also opens the door to concurrent assessment of genetic and epigenetic variation, a major advantage for clinical research.

Long-read whole genome sequencing (LR-WGS) could identify all possibly pathogenic DNA variants in genetic disorders. However, DNA analysis alone is often insufficient to interpret the effect of DNA variation, and by extension whether it could cause disease. To improve diagnostic yield, we use transcriptome analysis in a diagnostic setting in undiagnosed disorders (Dekker et al., 2023, Am J Hum Genet). However, interpretation of the results may remain challenging. As combining RNAseq and LR-WGS could increase our understanding of complex variants, we here used LR-WGS and RNA-sequencing (Oxford Nanopore Technologies, ONT) in unresolved genetic disorders. We selected undiagnosed individuals, some awaiting diagnosis for more than 15 years, where previous RNAseq analysis indicated strong clues regarding the genetic cause—such as transcript isoform switching, fusion transcripts, transcriptional read-through and transposon insertion—, but further routine diagnostic DNA analysis proved insufficient to identify the precise causative variant. For example, in an undiagnosed individual with a clinical phenotype matching Treacher Collins syndrome, we identified a switch in TCOF1 transcript isoform from one allele. Further DNA analysis including optical genome mapping identified a possible breakpoint in TCOF1, but the underlying event could not fully be resolved. LR-WGS subsequently identified a ~3.5 kb mobile element insertion (SVA) in the TCOF1 intronic sequence and together with the observed effect in the RNAseq data, this provided a genetic signature with the potential for diagnosis. In all, our results support the idea that ONT LR-WGS can be rapidly implemented for undiagnosed genetic disease and stress the need for complementary analysis such as RNA-sequencing, necessary to interpret the clinical impact of complex DNA variants.