Bulk and single-cell RNA sequencing identify prognostic signatures related to FGFBP2+ NK cell in hepatocellular carcinoma.

Background: Hepatocellular carcinoma (HCC) is a highly aggressive malignancy. As a specific immune cell subpopulation, FGFBP2⁺ NK cells play a crucial part in immune surveillance of HCC progression. This study set out to identify prognostic signature related to FGFBP2⁺ NK cell in HCC.

Methods: Bulk and scRNA-seq data were derived from the public databases. The single cell atlas of HCC and heterogeneity of natural killer (NK) cells were delineated by "Seurat" package. Pseudo-time trajectory of FGFBP2⁺ NK cell was constructed by "Monocle2" package. Cell-cell interactions were analyzed by "CellChat" package. Prognostic signature was screened to develop a RiskScore model, and the prediction robustness was verified. Immune cell infiltration and immunotherapy response were assessed between different risk groups. Drug sensitivity was predicted by "oncoPredict" package. The expressions of the prognosis gene signature were detected by in vitro test utilizing HCC cells. The effects of key genes on the proliferative, migratory and invasive capacity of HCC cells were assessed by EdU assay, wound healing and Transwell assay.

Results: The proportion of NK cell in HCC samples was markedly decreased than that in healthy samples. NK cell was further divided into three cell subpopulations, and FGFBP2⁺ NK cell was associated with the prognosis of HCC patients. Pseudo-time trajectory analysis of FGFBP2⁺ NK cell revealed two differential expression gene clusters. FGFBP2⁺ NK cell exhibited extensive intercellular communication in HCC. Further, eight prognostic signatures were identified, including six "risk" genes (UBE2F, AHSA1, PTP4A2, CDKN2D, FTL, RGS2) and two "protective" genes (KLF2, GZMH). RiskScore model was established with good prognostic prediction performance. In comparison to low-risk group, high-risk group had poorer prognosis, lower immune cell infiltration, and higher TIDE score. Moreover, 16 drugs showed significant correlation with RiskScore. Additionally, the expressions of GZMH was downregulated while FTL, PTP4A2, UBE2F, CDKN2D, RGS2, and AHSA1 were up-regulated in HCC cells. FTL and PTP4A2 silencing could suppress the proliferation, migration and invasion abilities of HCC cells.

Conclusion: This study identified eight prognostic gene signatures related to FGFBP2⁺ NK cell in HCC, which may serve as potential therapeutic targets for HCC.