Identification of molecular pattern and prognostic risk model based on ligand-receptor pairs in liver cancer

Abstract

Introduction: The tumor microenvironment of hepatocellular carcinoma is composed of multiple cells, and the interactive communication between cells drives tumor progression and characterizes the tumor. Communication between cells is mainly achieved through signal transduction between receptor ligands, and the rise of single-cell technology has made it possible to analyze the communication network between cells.

Methods: We applied a train of bioinformatic techniques and in vitro experiments. We analyzed the composition of the microenvironment of liver cancer by combining single-cell sequencing data and transcriptome sequencing data from liver cancer to construct molecular typing and risk models for LRs. Then, we analyzed association of it with prognosis, mutation, KEGG, tumor microenvironment (TME), immune infiltration, tumor mutational burden (TMB) and drug sensitivity in liver cancer. qPCR and was used to identify SLC1A5 expression in LIHC cell lines and CCK8, transwell and cell colony formation were performed to validate the function of SLC1A5. Meanwhile, we also performed polarization of macrophages.

Results: In this experiment, we found that liver cancer tissues are rich in immune and mesenchymal cells, and there is extensive signaling between individual cells, so we constructed molecular typing and risk models for LRs. Combining clinical data revealed significant differences in clinical characteristics, prognosis and mutated genes between the molecular typing of receptor-ligand pairs, as well as in sensitivity to drugs; similarly, there were significant prognostic differences between the risk models. There were also notable differences in activated signaling pathways, infiltrating immune cells and immune subtypes. Subsequently, we used siRNA to knock down SLC1A5 in hepatocellular carcinoma cells and found that cell proliferation, migration and invasion were diminished.

Conclusions: In conclusion, our LRs model may become a marker to guide clinical treatment and prognosis.

Keywords: TME; ligand-receptor; liver cancer; molecular pattern; risk model.