Integrative genomics approach identifies molecular features associated with early-stage ovarian carcinoma histotypes

Hanna Engqvist,Anikó Kovács,Peter Larsson,Per Karlsson,Jana Biermann,Elisabeth Werner Rönnerman,Toshima Z Parris,Khalil Helou,Karin Sundfeldt

doi:10.1038/s41598-020-64794-8

Hanna Engqvist, Anikó Kovács + Show 7 more

Open Access

https://doi.org/10.1038/s41598-020-64794-8

Copy DOI

Abstract

Ovarian cancer comprises multiple subtypes (clear-cell (CCC), endometrioid (EC), high-grade serous (HGSC), low-grade serous (LGSC), and mucinous carcinomas (MC)) with differing molecular and clinical behavior. However, robust histotype-specific biomarkers for clinical use have yet to be identified. Here, we utilized a multi-omics approach to identify novel histotype-specific genetic markers associated with ovarian carcinoma histotypes (CCC, EC, HGSC, and MC) using DNA methylation, DNA copy number alteration and RNA sequencing data for 96 primary invasive early-stage (stage I and II) ovarian carcinomas. More specifically, the DNA methylation analysis revealed hypermethylation for CCC in comparison with the other histotypes. Moreover, copy number imbalances and novel chromothripsis-like rearrangements (n = 64) were identified in ovarian carcinoma, with the highest number of chromothripsis-like patterns in HGSC. For the 1000 most variable transcripts, underexpression was most prominent for all histotypes in comparison with normal ovarian samples. Overall, the integrative approach identified 46 putative oncogenes (overexpressed, hypomethylated and DNA gain) and three putative tumor suppressor genes (underexpressed, hypermethylated and DNA loss) when comparing the different histotypes. In conclusion, the current study provides novel insights into molecular features associated with early-stage ovarian carcinoma that may improve patient stratification and subclassification of the histotypes.

Highlights

Ovarian cancer comprises multiple subtypes (clear-cell (CCC), endometrioid (EC), high-grade serous (HGSC), low-grade serous (LGSC), and mucinous carcinomas (MC)) with differing molecular and clinical behavior
We performed a comprehensive genome- and transcriptome-wide analysis integrating DNA methylation, copy number alterations (CNAs) and RNA sequencing (RNA-seq) data for 96 primary invasive early-stage ovarian carcinoma samples characterized as CCC, EC, HGSC and MC
Highly methylated CpG sites were more frequently found in the gene body, 3′ untranslated regions (3′ UTR) and intergenic regions (IGR), as well as in CpG shelves and open sea (Fig. 1b)

Summary

Introduction

Ovarian cancer comprises multiple subtypes (clear-cell (CCC), endometrioid (EC), high-grade serous (HGSC), low-grade serous (LGSC), and mucinous carcinomas (MC)) with differing molecular and clinical behavior. We utilized a multi-omics approach to identify novel histotype-specific genetic markers associated with ovarian carcinoma histotypes (CCC, EC, HGSC, and MC) using DNA methylation, DNA copy number alteration and RNA sequencing data for 96 primary invasive early-stage (stage I and II) ovarian carcinomas. We performed a comprehensive genome- and transcriptome-wide analysis integrating DNA methylation, CNA and RNA sequencing (RNA-seq) data for 96 primary invasive early-stage (stage I and II) ovarian carcinoma samples characterized as CCC, EC, HGSC and MC. Based on the assumption that the genetic profiles of early-stage tumors are generally less complex compared to the later stages, we chose to only include early-stage ovarian carcinoma to enable the classification of early events in ovarian carcinoma tumorigenesis This may permit the identification of specific genomic alterations related to ovarian carcinoma. We provide an extensive overview of the genome, methylome, and transcriptome for early-stage ovarian carcinomas, thereby identifying putative genetic markers for ovarian carcinoma, such as oncogenes and tumor suppressor genes

Methods

Results

Conclusion