Single-cell and bulk transcriptome analysis unveils a ligand-receptor-based signature for prognostication and reveals that TREM1 controls the malignant behaviors of hepatocellular carcinoma

Abstract

The transcriptional heterogeneity and cellular ecosystem diversity of HCC await further exploration. Single-cell and bulk RNA sequencing data from HCC cells are analyzed to generate a LASSO model for HCC prognostication. CCK-8, scratch assay, flow cytometry, and ROS assays are used to validate how TREM1 may affect HCC cell biological behaviors in vitro. qPCR, western blot analysis, immunohistochemistry, and flow cytometry are applied in a xenograft model to test the effects of TREM1 knockdown on carcinogenesis and the tumor microenvironment. A single-cell atlas of the multicellular ecosystem comprising 13 cell types in HCC is constructed. On the basis of ligand-receptor marker genes specifically extracted from the cell populations, a prognostic model is defined and subsequently validated in additional clinical cohorts. For the first time, a heterogeneous immune microenvironment is observed between low- and high-risk patients, primarily involving macrophages, CD4+ T cells, M1 macrophages, and regulatory T (Treg) cells. Sufficient evidence validates the positive effects of TREM1 on HCC cell proliferation, migration, and apoptosis. Additionally, TREM1 positively modulates the levels of the proinflammatory cytokines IL-1β, TNF-α, and MCP-1. TREM1 downregulation alters the proportions of M1 macrophages and Tregs in the tumor tissue from our HCC xenograft model. Eventually, the Nrf2/Keap1 signaling pathway, which is related to oxidative stress, is shown to be a key pathway downstream of TREM1 downregulation. In summary, we construct a novel prognostic model for HCC on the basis of ligand-receptor marker genes and investigate the role of TREM1 in HCC progression and its impact on the TME.

1. Introduction

Hepatocellular carcinoma (HCC) is a highly aggressive malignancy with a rising incidence globally. The prognosis is still dismal, with an overall five-year survival rate of 12%–15% [1]. In the past few years, remarkable progress has been made in treatment regimens: antiangiogenic multikinase agents (sorafenib [2] and lenvatinib [3]) have gained approval as first-line systemic therapeutic options for unresectable HCCs; other multikinase inhibitors as well as immune checkpoint blockade agents against PD-1/PD-L1 have been approved as second-line treatments [4]. Nevertheless, unresectable HCC patients still face increased unmet medical needs and unsatisfactory survival outcomes.

A tumor is an extremely complex ecosystem, defined by spatiotemporal relationships between heterogeneous cell populations composed of malignant, immune, and stromal cell types [5]. Hence, characterizing the landscape of the HCC multicellular ecosystem and critical components linked to tumor progression and immunotherapy is essential. Using bulk transcriptome approaches, prior research has revealed that each HCC tumor has its own personalized expression profile containing diverse transcriptional programs [6]. Despite the progress in bioinformatics, deconvolution algorithms are unable to analyze rare cell populations and cell-to-cell interplay, among other methods.

Single-cell RNA sequencing (scRNA-seq) has been proven to be an efficient tool for characterizing expression data across numerous cells simultaneously, which enables the generation of integrated profiles of diverse cell types in tumors under different biological states or conditions [7]. Recently, scRNA-seq research has provided unique insights into many aspects of HCC biology. For example, scRNA-seq reveals the immunosuppressive landscape and tumor heterogeneity of HBV-related HCC [8]. Single-cell analysis has revealed that proliferative Prom1+ tumor-propagating cells, as well as their dynamic cellular transitions, are involved in HCC development [9]. Through single-cell transcriptomic profiling, the landscape of intratumoral heterogeneity and stemness-associated subpopulations has been revealed in HCC [10]. In the present study, we integrated single-cell and bulk transcriptome analyses to reveal a novel ligand-receptor-based prognostic model for HCC and revealed the role of TREM1 in controlling the malignant behaviors of HCC cells both in vitro and in vivo.