Preparation of Duplex Sequencing Libraries for Archival Paraffin-Embedded Tissue Samples Using Single-Strand-Specific Nuclease P1.

DNA from formalin-fixed paraffin-embedded (FFPE) tissues, which are frequently utilized in cancer research, is significantly affected by chemical degradation. It was suggested that approaches that are based on duplex sequencing can significantly improve the accuracy of mutation detection in FFPE-derived DNA. However, the original duplex sequencing method cannot be utilized for the analysis of formalin-fixed paraffin-embedded (FFPE) tissues, as FFPE DNA contains an excessive number of damaged bases, and these lesions are converted to false double-strand nucleotide substitutions during polymerase-driven DNA end repair process. To resolve this drawback, we replaced DNA polymerase by a single strand-specific nuclease P1. Nuclease P1 was shown to efficiently remove RNA from DNA preparations, to fragment the FFPE-derived DNA and to remove 5'/3'-overhangs. To assess the performance of duplex sequencing-based methods in FFPE-derived DNA, we constructed the Bottleneck Sequencing System (BotSeqS) libraries from five colorectal carcinomas (CRCs) using either DNA polymerase or nuclease P1. As expected, the number of identified mutations was approximately an order of magnitude higher in libraries prepared with DNA polymerase vs. nuclease P1 (626 ± 167/Mb vs. 75 ± 37/Mb, paired t-test p-value 0.003). Furthermore, the use of nuclease P1 but not polymerase-driven DNA end repair allowed a reliable discrimination between CRC tumors with and without hypermutator phenotypes. The utility of newly developed modification was validated in the collection of 17 CRCs and 5 adjacent normal tissues. Nuclease P1 can be recommended for the use in duplex sequencing library preparation from FFPE-derived DNA.

Comprehensive evaluation of the test for 5'-/3'-end mRNA unbalanced expression as a screening tool for ALK and ROS1 fusions in lung cancer.

BACKGROUND: Despite the progress in the development of next-generation sequencing (NGS), diagnostic PCR assays remain to be utilized in clinical routine due to their simplicity and low cost. Tests for 5'-/3'-end mRNA unbalanced expression can be used for variant-independent detection of translocations, however, many technical aspects of this methodology require additional investigations. METHODS: Known ALK/ROS1 fusions and 5'-/3'-end unbalanced expression were analyzed in 2009 EGFR mutation-negative non-small cell lung cancer (NSCLC) samples with RT-PCR tests, which were optimized for the use with FFPE-derived RNA. RESULTS: Variant-specific PCR tests for 4 common ALK and 15 common ROS1 translocations detected 115 (5.7%) and 44 (2.2%) rearrangements, respectively. Virtually all samples with common ALK fusions demonstrated some level of 5'/3' mRNA ends unbalanced expression, and 8 additional NSCLCs with rare ALK fusions were further identified by PCR or NGS among 48 cases selected based on ALK expression measurements. Interestingly, NSCLCs with unbalanced 5'-/3'-end ALK expression but without identified ALK translocations had elevated frequency of RAS mutations (21/40, 53%) suggesting the role of RAS activation in the alternative splicing of ALK gene. In contrast to ALK, only a minority of ROS1 translocation-positive cases demonstrated unbalanced gene expression, with both 5'- and 3'-end mRNA expression being elevated in most of the samples with translocations. Surprisingly, high ROS1 expression level was also found to be characteristic for NSCLCs with activating mutations in other tyrosine kinases such as EGFR, ALK, or MET. CONCLUSIONS: Comprehensive ALK analysis can be performed by the test for 5'-/3'-end unbalanced expression with minimal risk of missing an ALK rearrangement. In contrast, the use of the test for 5'-/3'-end unbalanced expression for the detection of ROS1 fusions is complicated; hence, the utilization of variant-specific PCR assays for ROS1 testing is preferable.