Integrated analysis of optical mapping and whole-genome sequencing reveals intratumoral genetic heterogeneity in metastatic lung squamous cell carcinoma

Abstract

Background: Intratumoral heterogeneity is a crucial factor to the outcome of patients and resistance to therapies, in which structural variants play an indispensable but undiscovered role. Methods:We performed an integrated analysis of optical mapping and whole-genome sequencing on a primary tumor (PT) and matched metastases including lymph node metastasis (LNM) and tumor thrombus in the pulmonary vein (TPV). Single nucleotide variants, indels and structural variants were analyzed to reveal intratumoral genetic heterogeneity among tumor cells in different sites.Results: Our results demonstrated there were less nonsynonymous somatic variants shared with PT in LNM than in TPV, while there were more structural variants shared with PT in LNM than in TPV. More private variants and its affected genes associated with tumorigenesis and progression were identified in TPV than in LNM. It should be noticed that optical mapping detected an average of 77.1% (74.5-78.5%) large structural variants (>5,000 bp) not detected by whole-genome sequencing and identified several structural variants private to metastases.Conclusions: Our study does demonstrate structural variants, especially large structural variants play a crucial role in intratumoral genetic heterogeneity and optical mapping could make up for the deficiency of whole-genome sequencing to identify structural variants.

Identification of Somatic Structural Variants in Solid Tumors By Optical Genome Mapping

“…Some prior studies have begun to demonstrate the utility of Bionano UHMW DNA isolation protocols in solid tissue tumor analysis. These include studies of lung squamous cell carcinoma and metastatic prostate carcinoma [7,34,35]. This current report demonstrates the utility of the a DNA isolation protocol and SV analysis in a wide variety of solid tissue types, and expands the feasibility of such analysis for previously unused human tissue types.…”

“…For these four tongue tumor samples, we identified an average of 1474 total SVs per sample. Filtering these SVs using the Rare Variant Analysis pipeline for SVs not found in the Bionano control database yielded an average of 72 total SVs per sample, consisting of 11 insertions (range 9-15), 31 deletions (range 11-47), 3 inversions (range 1-6), 14 duplications (range 2-23), and 14 translocations (6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19). Filtering against the variants found in the corresponding blood samples returned an average of 58 total SVs per sample, consisting of 10 insertions (range 9-10), 20 deletions (range 7-35), 2 inversions (range 0-4), 13 duplications (range 4-24), and 14 translocations (range 6-19) (Figure 3, upper panel).…”

“…For instance, recurrent translocations, such as the Philadelphia chromosome, can activate oncogenes but at the same time reveal avenues for implementing or developing effective targeted drug therapies [1][2][3][4]. Likewise, SV identification plays an increasingly important role in cancer diagnosis and prognosis [5,6], and SVs have been shown to play a crucial role in intratumoral genetic heterogeneity [7]. Therefore, SV identification and analysis is important to understanding oncogenesis and tumor behavior.…”

Identification of Somatic Structural Variants in Solid Tumors by Optical Genome Mapping

“…For instance, recurrent translocations, such as the Philadelphia chromosome, can activate oncogenes but at the same time reveal avenues for implementing or developing effective targeted drug therapies [ 1 , 2 , 3 , 4 ]. Likewise, SV identification plays an increasingly important role in cancer diagnosis and prognosis [ 5 , 6 ], and SVs have been shown to play a crucial role in intra-tumoral genetic heterogeneity [ 7 ]. Therefore, SV identification and analysis are important to understanding oncogenesis and tumor behavior.…”

“…Despite its success in visualizing SVs in liquid tumors and cell lines, OGM has not yet seen widespread application in solid tissue tumors, due primarily to the difficulty of obtaining high-quality, high-molecular-weight DNA from solid tumor samples. Nonetheless, previous work has shown the feasibility of high-quality high-molecular-weight DNA isolation and analysis using earlier workflow iterations [ 33 ], and recent feasibility studies have shown the importance of OGM application to solid tumor analysis [ 7 , 34 , 35 ]. Peng et al demonstrated large SVs not detected by WGS implicated in metastatic lung squamous cell carcinoma [ 7 ], and Jaratlerdiri et al and Crumbaker et al similarly found SVs impacting oncogenic and tumor-suppressing genes not identified by NGS or WGS alone in prostate cancer [ 34 , 35 ].…”