Abstract 1142P
Background
Duplex sequencing is an extremely accurate NGS approach, which allows correcting errors caused by DNA chemical degradation, PCR amplification or wrong sequencer readings. FFPE-derived DNA, which is frequently utilized in cancer research, is significantly affected by chemical degradation. The adjustment of duplex sequencing-based methods to the requirements for the analysis FFPE tissues is of high potential importance.
Methods
We applied BotSeqS, a duplex sequencing-based method developed by M.L. Hoang et al. (2016) to quantify rare somatic mutations in normal tissues, for the study of mutational rate in tumors. The use of DNA polymerase for blunting DNA ends may introduce errors to duplex sequencing data; therefore, we attempted to substitute this enzyme with a single strand-specific nuclease P1, which is also capable of fragmenting FFPE-derived DNA. In total, about 0.4 Mb of random duplex sequences per tumor was obtained, and polymorphisms with population frequency of more than 0.01 by GNOMAD v3.1.1 database were filtered out.
Results
As we expected, libraries, prepared using standard DNA fragmentation and end repair methods contained significantly more duplex-supported mutations than libraries, prepared with nuclease P1 (599±140/Mb vs. 71±38/Mb, p = 0.0007). Using nuclease P1-based method, we quantified number of mutations in colorectal tumors with presumably impaired DNA repair or error-correction systems: MSI-positive tumors (n=4), tumors with biallelic MUTYH mutations (n=4), tumors with mutations in POLD1 (n=3) and POLE (n=1) genes. With our method, high number of mutations (69-172/Mb) was found in all MSI-positive tumors and also in the tumor with S297F mutation in POLE gene (107/Mb). However, we failed to differentiate between MUTYH-positive (6-24/Mb) or POLD1-positive (16-29/Mb) cases and colorectal carcinomas with no such defects (16-30/Mb, n=5), probably due to the lack of information regarding germline status of all identified mutations.
Conclusions
Nuclease P1 can be recommended for the use in duplex sequencing library preparation from FFPE-derived DNA. The developed method can help to identify tumors with very high mutational burden.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
The authors.
Funding
Russian Scientific Fund, grant number 19-15-00312.
Disclosure
All authors have declared no conflicts of interest.