Reconstructing Chromothriptic Chromosomes in Oesophageal Adenocarcinomas
The epigenetic landscape is regulated by a myriad of factors. This regulation ranges from functional compartmentalisation of genomic sequences into topologically associating domains, to chromosome looping, to short-range promoter-enhancer interactions. The underlying genome sequence contributes to this regulation, likely at a variety of scales, however the extent of this contribution is not fully understood. Chromothripsis is a localised catastrophic genome shattering event that can be used to study how the underlying genomic sequence affects this higher order structuring. Since chromothripsis tends to affect only one of the two alleles, in every cell a direct comparison can be made between the wild-type chromosome and the chromothriptic chromosome. The wild-type chromosome represents the genome sequence and structure before reshuffling and the chromothriptic derivative chromosome can be used to query the direct effects of this reshuffling.
Chromothripsis has been seen in up to 32% of cases of oesophageal adenocarcinomas. Therefore, patient-derived oesophageal adenocarcinoma organoids with evidence of chromothripsis restricted to one allele were used to better understand how the genome is regulated. Complex regions of structural variation between alleles in cancer genomes coupled with subclonal variants means haplotype-aware de novo assemblies are essential for contiguous cancer genome assemblies. Our method takes haplotype blocks and assigns PacBio circular consensus sequencing reads to the appropriate allele using B-allele frequencies of single nucleotide polymorphisms and presence of structural variants. The chromosomes are then assembled separately and scaffolded using Hi-C reads, which we also haplotype resolve. This produces contiguous assemblies, even on chromosomes with over 900 structural rearrangements compared to the reference genome. This methodology has been used to reconstruct chromothriptic derivative chromosomes and the associated wild-type chromosomes in five organoid models, as well as other chromosomes with complex rearrangements. All types of structural variant have been reconstructed, other than tandem duplications which are collapsed by current assembly tools.
With these cancer-specific reference assemblies, the epigenome of the chromothriptic and wild-type chromosomes can be profiled. Hi-C chromosome capture has been used to study topologically associated domains; ATAC-seq to study chromatin accessibility; ChIP-seq to identify CTCF binding and histone modifications (H3K27me3, H3K4me3, H3K27ac) and Iso-seq to phase long read transcripts to their respective chromosomes. There are widespread differences between the chromothriptic and wild-type chromosomes for each epigenetic mark. This indicates that the shattering of the chromosome has dramatic consequences for gene regulation, far beyond what we see when comparing two wild-type alleles of the same chromosome. It highlights that, while underlying genome sequence has a fundamental role in gene regulation, the epigenetic context of that sequence also has a profound impact. The work done to assemble these chromosomes allows for unprecedented insight into the regulatory impact of structural variation.