EACR25-1713

Comprehensive analysis of somatic structural variations in Malignant pleural mesothelioma (MPM) using long-read sequencing

M. Alhazmi1, A. Dawson2, A. Nakas3, C. Poile2, J. Dzialo2, A. Bzura2, K. Kutywayo2,3, C. Tufarelli1, D. Fennell2,3, E. Hollox1
1University of Leicester, Department of Genetics, Genomics and Cancer Sciences, Leicester, United Kingdom
2University of Leicester, Leicester Cancer Research Centre, Leicester, United Kingdom
3University Hospitals of Leicester NHS Trust, Leicester, United Kingdom
Introduction:

Malignant pleural mesotheliomas (MPMs) harbour extensive somatic genome structural variations. Molecular targeted therapy is lacking. Identifying somatic driver mutations in MPM is critical in understanding the development of mesothelioma and developing new therapy strategies. Long-read DNA sequencing using Oxford Nanopore Technology (ONT) can detect chromosomal rearrangements, shorter insertion/deletions, collectively known as structural variants (SVs) compared with short-read whole exome sequencing. ONT also directly profiles the methylome, allowing assessment of epigenetic silencing. In this proof-of-principle study, we aim to identify and validate somatic structural variants using long read sequencing in three MPMs using matched blood samples as a reference, and comparing with matched WGS short-read DNA sequencing. We also examine the matched transcriptomes data.

Material and method:

Three MPM tumours with matched blood DNA samples were sequenced at high coverage using ONT PromethION. Structural variants were identified using minimap2 and Severus, then compared to variants identified by Illumina whole genome sequencing using BWA and Manta. The segmentation copy number alteration was detected by CNVkit. The gene expression was detected by using Salmon. Finally, the affected genes were examined their impact on survival time using TCGA data.

Result and discussion:

The three MPM tumours were sequenced to between 38x and 42x coverage, with a median read length between 20 kb and 28 kb, and a maximum read length between 580 kb and 248 kb. The most common SV type was translocation (35.7%), followed by deletion (27.8%), insertion (21.1%), inversion (0.11%), and duplication (0.11%). We identified the double hits of MPM driver genes such as NF2, LATS2, BAP1 using both whole genome sequencing Illumina and whole genome sequencing ONT. On the other hand, Severus identified between 61-156 novel SVs not reported by whole genome sequencing Illumina. These SVs affected multiple COSMIC cancer-associated genes including LRP1B, BIRC6, EXT1, NBEA, GPC5, LPP, ROBO2, JAZF1, CDH1, FLCN, SPECC1, and EPHA7. Additional, potential cancer genes were previously reported in other cancers such as GPC6 and PRKN. Furthermore, some of these SVs were involved in complex rearrangement on same haplotype. Gene expression levels of GPC6 and PRNK were associated with survival time.

Conclusion:

Long read sequencing using Oxford nanopore technology identifies novel structural variants in MPM, missed by short-read sequencing, including potential complex events such as chromoplexy. For future studies, the ONT could be applied to larger cohorts of MPM patients to identify the double hit and complex rearrangements during MPM evolution, and to support new precision medicine strategies.