Massively parallel functional profiling prioritises breast cancer risk variants and implicates CCDC88C in ER-positive breast cancer.
TL;DR
This work identified 709 variants mapping to 140 risk regions, that are associated with significant variation between REF and ALT alleles, that may influence breast cancer risk through expression of CCDC88C, which in turn could impact prognosis.
Abstract
Genome wide association studies (GWAS), combined with fine-mapping have identified 196 independent signals associated with breast cancer risk. Deciphering the functional basis of these associations can inform our understanding of the biology and aetiology of breast cancer. Decoding GWAS risk associations is challenging due to linkage disequilibrium between variants and because most variants map to non-coding regions, influencing breast cancer risk via cis-regulatory mechanisms that modulate the expression of target genes. To prioritise variants with allele-specific regulatory activity at breast cancer risk loci, we carried out a lentivirus-based massively parallel reporter assay (lentiMPRA) to screen 5,116 credible causal variants across these signals. We identified 709 variants mapping to 140 risk regions, that are associated with significant variation between REF and ALT alleles. A follow-up investigation at 14q32.11 revealed rs7153397 may influence breast cancer risk through expression of CCDC88C, which in turn could impact prognosis. These findings provide a prioritised set of functional variants for downstream analyses, advancing our understanding of breast cancer risk mechanisms.