Single cell DNA sequencing reveals punctuated and gradual clonal evolution in hepatocellular carcinoma

Abstract Background & Aims Copy number alterations (CNAs), elicited by genome instability, is a major source of intratumor heterogeneity. How CNAs evolve in hepatocellular carcinoma (HCC) remains unknown. Methods We performed single cell DNA sequencing (scDNA-seq) on 1275 cells isolated from 10 HCC patients, ploidy-resolved scDNA-seq of 356 cells from 1 additional patient, and single cell RNA sequencing (scRNA-seq) on 27344 cells from 3 additional patients. Three statistical fitting models were compared to investigate the CNA accumulation pattern. Results Cells in the tumor were categorized into three subpopulations: euploid, pseudoeuploid, and aneuploid. Our scDNA-seq analysis revealed that CNA accumulation followed a dual-phase copy number evolution (DPCNE) model, i.e. a punctuated phase followed by a gradual phase. Patients exhibited prolonged gradual phase showed higher intratumor heterogeneity and worse disease-free survival. Integrating bulk RNA-seq of 17 HCC patients, published datasets of 1196 liver tumors, and immunohistochemical staining of 202 HCC tumors, we showed that high expression of CAD, a gene involved in pyrimidine synthesis, was correlated with rapid tumorigenesis and reduced survival. The DPCNE model was validated by our scRNA-seq data and published scDNA-seq datasets of other cancer types. Furthermore, ploidy-resolved scDNA-seq revealed the common clonal origin of diploid- and polyploid-aneuploid cells, suggesting that polyploid tumor cells were generated by whole genome doubling of diploid tumor cells. Conclusions Our work revealed a novel DPCNE model, showed HCC with longer gradual phase were more severe, identified CAD as a promising biomarker for early recurrence of HCC, and supported the diploid origin of polyploid HCC.