Integration of Hi-C and Nanopore Sequencing for Structural Variant Analysis in AML with a Complex Karyotype: (Chromothripsis)²

Background. Acute myeloid leukemia (AML) with a complex karyotype (CK-AML) is an AML subtype with a still dismal outcome despite recent therapeutic advances. The prognosis is even worse when the underlying structural variants (SVs) lead to an extremely complex pattern of rearrangements, called chromothripsis, with a median overall survival of only 120 days. Except for the presence of inactivating TP53 aberrations in about 70% of all AML-CK cases, the pathogenesis is poorly understood. To gain novel insights into the molecular mechanisms underlying CK-AML reliable high precision SV delineation is needed, which so far has been a major limitation in cancer research.

Aim. We developed a SV detection pipeline by integrating Oxford Nanopore Technology (ONT) based whole genome sequencing (WGS) and Hi-C sequencing. This pipeline generated precise characterization of SVs for which the impact on gene expression and the emergence of novel fusion genes was studied by RNA-seq and ONT transcriptome sequencing.

Patients and Methods. We applied our WGS and Hi-C SV detection pipeline to a cohort of 11 AML-CK cases. Nanopore DNA Sequencing was performed until a genomic coverage >10x per patient was reached. The samples of 9 patients were also subjected to Nanopore cDNA sequencing for fusion gene analysis and Illumina based RNA-seq for transcript quantification. As controls for Hi-C and Illumina RNA sequencing, CD34+ hematopoietic stem cell enriched samples from five healthy donors were used.

Results. Our SV detection pipeline enabled us to fully reconstruct the derivate chromosome structure even of very complex, chromothriptic rearrangements in CK-AML. This enabled us to identify features of chromothripsis, that could previously not be detected using conventual technologies. We found local clustering of breakpoints in three of the patients with up to 31 Inversions and Translocations located in a genomic region of just 2.7 kb. These breakpoints were present in the Hi-C as well as in our Nanopore SV dataset. Our SV pipeline also showed that in these highly clustered regions, the very small rejoined fragments (in many cases less than 1 kb in size) often showed an elevated copy number (CN) state, i.e. small amplifications. We termed this newly discovered phenomenon chromothripsis-in-chromothripsis or (chromothripsis)². The precise knowledge about these breakpoints, which were validated by two different technologies, enabled us to study the pathogenesis of CK-AML at a so far unprecedented resolution. Fusion transcripts could be very precisely mapped and the impact of the breakpoints and CN changes on gene expression levels could be validated, thereby indicating functional relevance of the respective aberrations.

Conclusions. The combination of Hi-C and long-read sequencing for SV detection proved to be a powerful tool for precise SV detection. Our SV pipeline allowed us to discover a new level of complexity in chromothripsis. Application of this pipeline to leukemias as well as other types of cancer can improve the precision of SV detection, thereby raising new opportunities for functional interpretation of complex genomic aberrations of pathogenic relevance.