Single-cell RNA sequencing reveals unique expression pattern of APOBEC3 enzymes in HPV-associated cancers
Introduction
APOBEC is a class of enzymes that, upregulated in response to viral infection, clears exogenous viral infections primarily by massively catalyzing the cytidine to uracil mutations in viral genomes. Recent studies showed that APOBEC-specific mutations are enriched in many types of cancers and were associated with treatment resistance and prognosis. Interestingly, APOBEC3A and APOBEC3B were observed consistently upregulated in HPV-positive oropharyngeal and cervical cancers, and APOBEC-specific mutations are enriched in both viral and tumor genomes. Given such a scenario, the role of these enzymes is elusive in primarily promoting or impeding carcinogenesis. With the improvement of transcriptome sequencing from bulk to a single-cell resolution, we could have a better understanding of the expression pattern and unique features of APOBEC3A and APOBEC3B to decipher this puzzle.
Methods
We collected a single-cell RNA sequencing dataset of head and neck cancers with known HPV status (GSE182227). Scanpy and Scran were used to preprocess and normalize data. A copy number signal-based mixture strategies were performed to infer the malignant epithelial cells. A non-negative matrix factorization algorithm was applied to characterize the feature of malignant cells.
Results
A total of 70,031 cells from GSE182227 passed the quality control and were used for further analysis. Epithelial cells were assigned based on cell-specific markers (Keratin 14, EpCAM), and further reassigned into substructures - basal layer, spinous layer, and granular layer cells. APOBEC3A is enriched in the granular layer, while APOBEC3B is widely expressed in all three layers. Malignant cells were identified, and their cellular status was characterized. Most of the APOBEC3A high-expression cells were in epithelial senescence status, and over half of APOBEC3B high-expression cells were in cell cycling status. By analyzing the co-expression pattern of APOBECs and HPV gene, APOBEC3A was enriched in HPV gene-negative cells and APOBEC3B was enriched in HPV gene-positive cells.
Conclusion
APOBEC3A and APOBEC3B have different expression patterns in HPV-positive head and neck cancers. APOBEC3A tends to be overexpressed in well-differentiated epithelial cells undergoing senescence. APOBEC3A was associated with viral gene silence, while APOBEC3B had a good consistency with HPV gene expression.