Genetic and epigenetic profiling of complex chromosomal rearrangements in a Sonic Hedgehog Medulloblastoma sample from a patient with Li Fraumeni syndrome

Rene (German Cancer Research Center - DKFZ) began his presentation by describing the motivation behind his work; with a clear emphasis on disentangling the effects of structural variations and epigenetic changes that drive medulloblastoma formation. To do this, long nanopore reads formed the cornerstone of his work.

Rene then provided some context to his work, and discussed Li-Fraumeni syndrome or SHH-Medulloblastoma. He described that TP53 mutations underpin the syndrome, which often culminates in multiple cancers throughout a lifetime. The specific patient used in their study, whom harboured the archetypal TP53 mutation, went on to develop a specific brain cancer - Sonic Hedgehog Medulloblastoma. Rene explained that such medulloblastomas, are typically associated with large-scale chromosomal rearrangements

Rene then discussed the available data sets they had from the single patient with SHH-MB. He had Oxford Nanopore WGS data, and WGS short-read data from three samples: the primary tumour, another from a tumour relapse, and a third from a germline sample. For the nanopore dataset, the three samples had been sequenced across a total of four PromethION flow cells resulting in a coverage of around 15 – 30X. In addition, short-read RNA seq data was also available, along with methylation array from the 450 k platform.

Rene proceeded to delve into the genomic analyses he carried out. Starting off with the copy number profile from the short-read data, Rene highlighted some of the striking aberrations present in the patients genome, including a  loss of heterozygosity in chromosome 3 and a drastically rearranged chromosome 7. To facilitate further analysis Rene and his team integrated long nanopore read and short-read data for haplotype/chromosome-arm phasing. He discussed the stages, in which he started by calling variants from short reads and phased the variants using long nanopore reads with the tool WhatsHap. Then to further phase the haplotype blocks generated from WhatsHap,population phasing was performed with the shapeit tool. He explained that this gives quite long haplotype blocks. Next, where allelic imbalances were observed, switching errors could be corrected by matching the allele specific copy number state of neighbouring haplotype blocks.

Rene then went on to explain, they used long nanopore reads to call large structural variations from the primary tumour data. He found a significant number of intra and inter-chromosomal rearrangements, 180 and 127 respectively. Perhaps more unexpectedly, was the discovery of rearrangements between different chromosomes. When zooming in on these inter-chromosomal rearrangements, there appeared to be two main clusters, the most obvious was between chromosome 11 and chromosome 17.

Another observation made by Rene was the high copy number of short somatic templated insertions, which was initially flagged up by further interrogating the short-read data. He explained that there is a large body of evidence that implicates these templated insertions with repair mechanisms of DSBs, an event which seems to be prevalent in the patients sample. To obtain more granular detail, he decided to use the long nanopore reads to see where these templated insertions were located in the genome. He found what he described as several bursts of templated insertions. By further analysing the data, he speculated that the repair mechanism driving these templated insertions seem to be going back and fourth of the forward and reverse strands of the same template, but also hopping between the template sequences.

Rene spoke about his findings of chromosome 11 and 17 in more detail. He and his team believed they had discovered a novel chromosome – combined from segments of chromosome 11 and 17. Using the long-read data, they succeeded in assembling 1.7Mb of the novel chromosome, containing 57 genes in total, of which 22 were protein coding. Rene and his team further validated their hypothesis of a novel chromosome by using FISH, and found signals of chromosome 11 and 17 in much closer proximity in the medulloblastoma tissue when compared to the healthy donor tissue.

Next, Rene moved on to the epigenetic facets of his experiments. He analysed the raw nanopore signals for 5-methylcytosine at CpG sites in all three tissues using Nanopolish. These were subsequently validated against the 450K array data, which demonstrated a very high correlation between the two methods. Rene stated that he looked for regions of differential methylation using the tool pycoMeth and discussed in detail how it works. He and his team used NanoEpiseg in order to segment the methylome based on read-level methylation calls. Rene also developed his own file format - MetH5 Format, which efficiently stores and retrieves methylation predictions without losing the read-level information provided by nanopore sequencing. Rene then went on to explain how segmentation occurs using the Nanoepiseg model.

Comparing the primary tumour and relapse sample, Rene discovered 370 differentially methylated regions, 101 of which overlapped with promoters of various protein coding genes and 32 were cerebellum-specific enhancers. Rene also stated that around half of the intervals they found using nanopore sequencing would not have been detectable by common methylation arrays since there were no probes for these regions. He also used the same method to call allele-specific methylation using nanopore phased reads – by comparing regions that are methylated in only one haplotype. Another interesting affect he observed in the novel chromosome was large-scale demethylation, which was absent in the normal haplotype. He stated that the demethylation spanned a 200 kb region containing the genes STK33 and TRIM66.

Next, Rene pointed out some of his discoveries from analysing gene expression data. Using short-read RNA-seq data, Rene tried to see if allele-specific methylation or allele-specific copy number in the primary tumour would explain allele-specific expression. He briefly noted his findings. Finally, Rene spoke about validating gene fusions using SVs called from long reads. He touched upon the process, explaining that they could disentangle how the gene fusions may have arisen.

In conclusion, Rene combined long- and short-read data for chromosome arm phasing, and constructed targeted assemblies from long reads to help resolve structural variants in a medulloblastoma sample. In addition, nanopore sequencing allowed for comprehensive methylome profiling, and was used to validate and understand some complex gene fusions.