Poster ASHG 2022

Single-cell identity and state coupled with genome-wide SNV and CNV in primary breast cancer with ResolveOME™ profiling.

J.S. Zawistowski¹; I. Salas-Gonzalez¹, T.V. Morozova¹, T. Tate¹, K.A. Kennedy¹, D.M. Arvapalli¹, S.D. Velivela¹, J.G. Blackinton¹, J.R. Marks², E.S. Hwang², G.L. Harton, V.J. Weigman¹, J.A.A. West¹

¹BioSkryb Genomics, Inc., Durham, NC, USA

²Department of Surgery, Duke University Medical Center, Durham, NC, USA

The molecular events governing the transition from ductal carcinoma in situ (DCIS) to invasive breast cancer are still being elucidated, whereby precise definition of these events has the potential provide a therapeutic window of intervention. To simultaneously expose both genomic and transcriptomic underpinnings in primary breast cancer samples and to ascertain intratumoral heterogeneity we utilized ResolveOME to profile single cells from tumor biopsies of DCIS/invasive ductal carcinoma (IDC). While earlier single-cell methods have importantly unified assessment of copy number variation (CNV) and transcriptomics, they do not yield complete and uniform genome-wide coverage for single nucleotide-level data, made possible with ResolveOME’s employment of primary template-directed amplification. As input into ResolveOME, we stratified singulated mastectomy samples by epithelial cell adhesion molecule (EpCAM) surface marker expression with FACS. The genomic arm of ResolveOME followed by analysis with BaseJumper™ software cataloged genome-wide single nucleotide variation (SNV) in 24 single cells expressing either high or low levels of EpCAM, including the identification of oncogenic PIK3CA N345K, while identifying cooccurring DCIS/IDC prototypical chromosomal loss of 11q, 13q and 16q/17p harboring tumor suppressor loci. Concurrently, the transcriptomic arm of ResolveOME enabled the calling of cell identity with the Human Cell Atlas transcriptional database, revealing monocytic and fibroblastic infiltration in the biopsy samples in addition to the expected cells of epithelial identity. The coupling of SNV data and transcriptome data critically unveiled phenotypic cell state heterogeneity, whereby an epithelial cell with both PIK3CA N345K and associated chromosomal losses manifested with a stemness/fibroblastic gene expression signature characteristic of the EpCAM low clade of patient cells. We are currently expanding ResolveOME profiling to additional DCIS/IDC patient tumors to comprehensively define driver and regulatory SNV while simultaneously distinguishing between infiltration of non-epithelial cell types from instances of epithelial morphing of physiological cell state within a biopsy. Furthermore, we are defining cell lineage at both the CNV and SNV level as additional single cells are sequenced for each patient sample. These data collectively highlight the molecular complexity and heterogeneity even among a small number of biopsied cells, and underscore the criticality of the interplay of DNA/RNA tiers of information when positing oncogenic mechanisms.