Abstract Viral infection plays a significant role in the development and evolution of many cancers. It is estimated that between 13% to 18% of all new cancer cases are caused by infection. It is known that viral infection can induce transcriptional changes in cancer, including altering the expression of a seven-membered DNA editing enzyme family known as APOBEC3 or A3. These cytosine deaminases have been implicated in infectious diseases and cancer for their ability to induce cytidine to uridine (CU) mutations in both viral and host DNA. Indeed, A3-associated mutational signatures SBS2 and SBS13 are present in more than 70% of cancer types and around 50% of all cancer genomes. However, little is known about the interplay between viral infection and A3 dysregulation in cancer. Accurate profiling of A3 expressions and their dysregulations caused by viral infections in cancer has historically been challenging using bulk RNA-seq. To address this, we leverage existing single-cell RNA-seq (scRNA-seq) data sets and our custom single cell bioinformatics pipeline, which uses multiple machine learning models for cell type annotation and cancer cell risk prediction, statistical methods for detection of mutational signatures at the single cell resolution, and custom k-mer based viral reference alignment for viral read detection in order to identify the source of A3 and viral reads across individual cell types within the breast cancer tumor microenvironment. Our studies using this pipeline have provided unparalleled resolution within the tumor microenvironment required to address questions about the association between viral presence and A3 dysregulation in various cancers. Previous analysis of head and neck tumors (GEO: GSE173468) allowed our lab to identify the presence of HPV16 viral reads, which were associated with a significant increase in A3B expression in infected basal cells of the tumor. Our recent analysis of breast cancer tumors (GEO: GSE16529 and GSE243526) has given us critical insights into cell type-specific A3 expression dynamics, indicating that the source of A3A expression was solely found coming from macrophage cells in the tumor and that high-risk cancer cells had primarily high expression of A3B and lacked A3A expression. In this study, we analyzed publicly available scRNA-seq datasets from two normal tissues (NTs), three primary tumors (PTs), and three tamoxifen-treated recurrent tumors (RTs) (GEO: GSE240112) in order to investigate the association between the presence of viral reads, A3 expression, and mutational signatures SBS2/13. We are currently optimizing our pipeline to further improve our viral read identification and association with specific cell types within each tumor microenvironment. This work is supported by the Texas Biomedical Research Institute’s Postdoctoral Forum Fellowship Grant. Citation Format: J. D. Lehle, M. Soleimanpour, N. Haghjoo, D. Ebrahimi. Single cell mutational fingerprints and the viral culprits: uncovering APOBEC3 dysregulation in breast cancer abstract. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PS2-12-19.
Lehle et al. (Tue,) studied this question.