DNA damage response-related signatures characterize the immune landscape and predict the prognosis of HCC via integrating single-cell and bulk RNA-sequencing.

Background: The occurrence and progression of hepatocellular carcinoma (HCC) are significantly affected by DNA damage response (DDR). Exploring DDR-related biomarkers can help predict the prognosis and immune characteristics of HCC.

Methods: First, the single-cell RNA sequencing (scRNA-seq) dataset GSE242889 was processed and performed manual annotation. Then we found the marker genes of DDR-active subgroups based on "AUCell" algorithm. The "Limma" R package was used to identify differentially expressed genes (DEGs) between tumor and normal samples of HCC. The risk prognostic model was constructed by filtering genes using univariate Cox and LASSO regression analyses. Finally, the signatures were analyzed for immune infiltration, gene mutation, and drug sensitivity. Last but not least, KPNA2, which had the largest coefficient in our model was validated by experiments including western blot, MTT, colony formation and γ-H2AX assays.

Results: We constructed a prognostic model based on 5 DDR marker genes including KIF2C, CDC20, KPNA2, UBE2S and ADH1B for HCC. We also proved that the model had an excellent performance in both training and validation cohorts. Patients in the high-risk group had a poorer prognosis, different immune features, gene mutation frequency, immunotherapy response and drug sensitivity compared with the low-risk group. Besides, our experimental results proved that KPNA2 was up-regulated in liver cancer cells than in hepatocytes. More importantly, the knockdown of KPNA2 significantly inhibited cell variability, proliferation and promoted DNA damage.

Conclusions: We innovatively integrated scRNA-seq and bulk RNA sequencing to construct the DDR-related prognostic model. Our model could effectively predict the prognosis, immune landscape and therapy response of HCC.