WYMM Tour: Munich

Applying whole exome sequencing (WES) has revolutionized the field of diagnostics of rare diseases. However, although most of the rare diseases are caused by genetic alterations, only about 50% of the patients can be solved by WES. Thus, as a next step, short read genome sequencing (srGS) was implemented in some European diagnostic centers. I will show the advantage to apply srGS first, but will also show its limitation and how long read genome sequencing (lrGS) will overcome some of the unsolved cases.

Epigenetic analyses have significantly advanced human brain tumour classification, starting around 2014 with childhood cancers that are histologically similar but represent a highly diverse biological spectrum. Reproducible diagnoses are vital for therapy personalisation and optimisation. The epigenetic diagnostic principle was extended through the development of a brain tumour classifier released in 2017 by the German Cancer Research Centre (DKFZ, Heidelberg). While the "WHO Classification of Tumours" listed epigenetic features in the context of only very few newly discovered brain tumour types in 2016, the 2021 edition has grown to almost twice the number of entities and recommends epigenetic testing for a significant proportion of tumour types. Functional tumour epigenetics are still to be characterised in depth;nevertheless, epigenomic patterns are highly indicative of cellular lineages which can be exploited diagnostically, e.g., through unsupervised and supervised machine learning approaches. In parallel to DNA methylation signatures, larger-scale epigenetic analyses determine genome-wide copy number profiles that - for many tumour types - are of diagnostic, prognostic, or predictive relevance. We developed the freely accessible open-source framework EpiDiP/NanoDiP ("Epigenetic Digital Pathology" - epidip.org / "Nanopore Digital Pathology"). This online tool is based on UMAP dimension reduction and places every anonymously uploaded microarray-derived methylation pattern in the context of currently approx. 20'000 annotated reference datasets, comprising neoplastic and non-neoplastic tissues. Its companion offline tool NanoDiP analogously enables diagnostics and research by methylation- and copy number profiling within hours through nanopore sequencing (ONT) on cost-effective, robust, small-scale, portable embedded computers. NanoDiP enables almost any pathology laboratory to engage in epigenetic workup of cancer specimens. Our reference data collection covers large fractions of human cancer types. The fully automated sequencing (R9), methylation, and copy number calling processes have been accredited as a laboratory-developed test (LDT) in Switzerland: NanoDiP does not only cover data analysis but also controls sequencing through the MinKNOW API. Hence, it can safely be operated by laboratory personnel without informatics knowledge.

Structural rearrangements in cancer can lead to the aberrant expression of an oncogene, by bringing an enhancer in the gene's vicinity. Typically, the oncogene's promoter is methylated in normal tissue and loses methylation as a result of the enhancer hijacking, but only on the rearranged allele. The long reads provided by nanopore sequencing often coveral several SNPs and can therefore be phased to the paternal and maternal haplotypes. In addition, nanopore sequencing provides methylation information at CpG sites. Therefore, nanopore sequencing can uncover allele-specific methylation, in particular at promoters of genes which hijack an enhancer. We verified in the leukemic GDM1 cell line, which is known to aberrantly express MNX1 because of an enhancer hijacking, that the MNX1 promoter is only unmethylated in the rearranged allele. We are now applying nanopore sequencing to acute myeloid leukemia samples with complex karyotypes, in the hope to discover new genes activated by enhancer hijacking,

Chromatin accessibility in GDM1 cell line was determined using nanoNOMe and nanoPia5, a technique developed in-house for this master thesis. Accessible chromatin regions were encoded by 5mC or 6mA methylation of GpC or adenine nucleotides on the DNA. Comparative analysis with ATAC sequencing at regulatory regions revealed increased chromatin accessibility, particularly in promoter regions with 43% of all adenines methylated by nanoPia5 protocol. The DiMeLo protocol allowed the targeted insertion of methylations on DNA at histones H3K27ac or H3K9me3 to determine their genomic location in GDM1 cells. In K562 enhancer regions, the protocol allowed targeted 6mA methylations at H3K27ac positions, with an increase of 8-15% of methylated adenines in these regions. Comparing H3K27ac localization with RNA expression of genes showed a correlation with a Spearman correlation coefficient of 0.604. Regions with increased signals in ACT and DiMeLo sequencing were identified using peak calling algorithms. Segments with increased 6mA signal in DiMeLo showed higher signals of H3K27ac localisation in ACTseq. Analysis of H3K9me3 included comparison of the signal distribution at the chromosome level and examination of 6mA accumulation in heterochromatin regions, demonstrating the abbility of DiMeLo to also determine the localisation of proteins in chromatin inaccessible regions. Some limitations of the methylation callers used in this work, and the difficulties in specifically distinguishing between the 6mA and 5mC signals in sequencing, were highlighted in control experiments with λ DNA by analysis of methylated motifs and at the sequence level.

Genetic diagnostics of repeat expansion and contraction disorders, including hereditary ataxias and facioscapulohumeral muscular dystrophy (FSHD), respectively, present significant challenges due to their phenotypic overlap and genetic complexity. Traditional methods lack precision and fail to capture crucial epigenetic markers. To overcome these limitations, we implemented nanopore Cas9-targeted sequencing for repeat expansions, enabling precise analysis of 10 repeat loci associated with hereditary ataxias. This method allowed for parallel repeat length, sequence, and methylation detection. Application of this approach to 100 undiagnosed ataxia patients revealed causative repeat expansions in 28 individuals, with RFC1 biallelic expansions emerging as the most common etiology. Moreover, identifying a novel repeat motif underscores the diagnostic potential of nanopore sequencing in uncovering previously unrecognized genetic variants. In the context of repeat contractions in FSHD, conventional diagnostics often fail to capture the full spectrum of genetic and epigenetic factors associated with this disease. To that end, we employed Cas9-targeted nanopore sequencing to characterize permissive haplotypes, repeat size, methylation profiles, and structural variants in FSHD patients. As proof of concept, we clarified the unexpectedly high average distal methylation in a paucisymptomatic patient homozygous for a permissive haplotype, harboring a repeat contraction with two repeat units on one allele. The ONT data revealed the presence of the contraction as a mosaic, causing an increase in the average percent methylation. In another patient, resolving a complex hybrid allele revealed the presence of a hypermethylated rare 4qC166H haplotype, which makes FSHD unlikely as the reason for the patient’s symptoms. Overall, our work showcases the power of nanopore sequencing in providing comprehensive insights into the genetic and epigenetic landscape of repeat expansion and contraction disorders and in uncovering novel disease mechanisms.

Virally encoded RNA structures influence the timing and efficiency of critical processes such as replication, protein synthesis, and evasion of immune detection, thereby determining the virus's ability to infect, proliferate, and persist within the host. Current technologies used to interrogate RNA structure and function in situ predominantly rely on ensemble analyses, which can mask the diversity and complexity inherent in RNA structures.
In response to this challenge, we apply long-read and single-molecule nanopore sequencing for a more precise understanding of RNA dynamics. We recently developed Nano-DMS-MaP for isoform-resolved RNA structure analysis, which we used to demonstrate how differently folded HIV-1 transcripts affect their function. We are also working toward single-molecule RNA structural analysis. As an essential step towards this goal, we have now implemented real-time multiplexing