ASPO Abstracts

Authors: Bauer MA, Chiang TC, Hsu PC, Su LJ

Category: Early Detection & Risk Prediction
Conference Year: 2021

Abstract Body:
Purpose This pilot study examines whether DNA methylation in samples collected before the breast cancer diagnosis is predictive of cancer in a rural cohort population. Methods DNA samples extracted from saliva were randomly selected from 51 breast cancer cases and 83 non-cases in the Arkansas Rural Community Health study cohort, where 26,347 women between the ages of 18 and 100 were recruited throughout 75 counties in Arkansas. All samples were collected at the time of baseline recruitment between 2007 and 2013. Subject records were linked with the Arkansas Central Cancer Registry for newly diagnosed breast cancer at least one year after initial study enrollment. The R packages ChAMP and CpGAssoc were used to identify significantly differentially methylated positions (DMPs) and regions (DMRs). Gene Set Enrichment Analysis (GSEA) was performed using the online tool NetworkAnalyst, and the KEGG and Reactome pathway databases. The Functional Epigenetic Modules (FEM) algorithm was used to identify differentially methylated gene networks. Results The study samples included 74 European Americans (28 breast cancer cases) and 60 American Americans (23 cases) aged 20 to 79 years at baseline. The differential analysis identified 1,077 significant DMPs and 37 DMRs associated with breast cancer status after multiple testing correction. Eight DMRs were mapped to genes that were found to be significant to probe position level and included LGALS1 (Galectin 1) and WT1 (Wilms tumor 1). Both are protein-coding genes involved in cell growth and differentiation and have been associated with breast cancer. We also identified 3 FEMs centered around CRMP1, CCL5, and CCDC85B. Notably, CCL5 (chemokine C- C motif ligand 5) has been implicated in breast cancer metastasis. GSEA identified "CREB phosphorylation through the activation of Ras" and "Ras activation upon Ca2+ influx through NMDA receptor" using the Reactome database and "Calcium signaling pathway" and "cAMP signaling pathway" using the KEGG pathways database. Conclusions We observed DNA methylation differences between samples that later developed breast cancer vs those that had not up until the time point of analysis. We further identified several differentially methylated genes that play a role in the oncogenic process.

Keywords: breast cancer, DNA methylation, saliva, rural, nested case-control

Return to Abstract Listing