Abstract
BACKGROUND: Liver hepatocellular carcinoma (LIHC) is a leading cause of cancer-related mortality, with an immunosuppressive tumor microenvironment (TME) contributing to therapeutic resistance. Although neutrophils are recognized as key regulators of LIHC progression, their functional heterogeneity and metabolic drivers are not yet fully understood. METHODS: We integrated bulk RNA sequencing (RNA-seq) data from The Cancer Genome Atlas (TCGA) and the Gene Expression Omnibus (GEO) database (GSE39791) alongside scRNA-seq data from GSE149614 and GSE290925. Neutrophils were annotated based on specific marker genes (FCGR3B, CSF3R) and classified into three metabolic states: high arginine state (HAS), intermediate arginine state (DTAS), and low arginine state (LAS) using arginine metabolism-related gene sets. Differentiation trajectories were reconstructed via CytoTRACE and monocle2. Intercellular communication was analyzed using CellChat, while machine learning, incorporating seven different algorithms, was applied to identify key regulatory genes. RESULTS: scRNA-seq analysis revealed three distinct neutrophil subgroups: high (HAS), intermediate (DTAS), and low (LAS) arginine metabolism states. The proportion of LAS neutrophils was significantly enriched in tumor tissues compared to normal tissues (p < 0.001). Trajectory analysis indicated that LAS neutrophils exhibited a less differentiated state. From this landscape, ATP11B and PADI4 were identified as key genes, with PADI4 expression being approximately 3-fold higher in HAS compared to LAS neutrophils. Functional studies demonstrated that silencing PADI4 in LIHC cell lines inhibited cell proliferation by approximately 50% at 96 hours, increased apoptosis by 2-fold, and reduced cell invasion by 50%. CONCLUSIONS: Arginine metabolism shapes neutrophil polarization in the LIHC TME. Targeting metabolic pathways may provide new therapeutic strategies to modulate the immune landscape and improve patient outcomes.