Friend or foe? The elusive role of hepatic stellate cells in liver cancer

Abstract

Liver fibrosis is a substantial risk factor for the development and progression of liver cancer, which includes hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (iCCA). Studies utilizing cell fate mapping and single-cell transcriptomics techniques have identified quiescent perisinusoidal hepatic stellate cells (HSCs) as the primary source of activated collagen-producing HSCs and liver cancer-associated fibroblasts (CAFs) in HCC and liver metastasis, complemented in iCCA by contributions from portal fibroblasts. At the same time, integrative computational analysis of single-cell, single-nucleus and spatial RNA sequencing data have revealed marked heterogeneity among HSCs and CAFs, with distinct subpopulations displaying unique gene expression signatures and functions. Some of these subpopulations have divergent roles in promoting or inhibiting liver fibrogenesis and carcinogenesis. In this Review, we discuss the dual roles of HSC subpopulations in liver fibrogenesis and their contribution to liver cancer promotion, progression and metastasis. We review the transcriptomic and functional similarities between HSC and CAF subpopulations, highlighting the pathways that either promote or prevent fibrosis and cancer, and the immunological landscape from which these pathways emerge. Insights from ongoing studies will yield novel strategies for developing biomarkers, assessing prognosis and generating new therapies for both HCC and iCCA prevention and treatment.

Figure 1.. A model of HSC and CAF plasticity upon activation based on single-cell analyses.

Quiescent HSCs are the major source of activated HSCs/myofibroblasts in fibrosis and CAFs in both primary and metastatic liver cancer scRNA-Seq analyses have demonstrated the transition from quiescent HSCs to several activated HSC and CAF phenotypes. Furthermore, in silico analyses describe the transcriptional differentiation of activated HSCs into subtypes that include proliferative (pHSC), inflammatory (iHSC), contractile and migrative (cmHSC), and fibrogenic myofibroblasts (myHSC) phenotypes. Other subtypes expressing microvasculature genes are depicted within the subpopulation of vascular HSCs (vHSC), yet their origin(s) is (are) unclear. A de-activated HSC (dHSC) subpopulation appears during regression of liver fibrosis, characterized by an intermediary gene expression signature between quiescent and activated HSCs. However, full transdifferentiation of this subtype back to the quiescent phenotype has not been described. Similarly, CAFs display a transcriptional transition from an inflammatory (iCAF) to a fibrogenic phenotype (myCAF), with intermediary subpopulations such as vascular (vCAF) and antigen presenting CAFs (apCAFs). Due to the remarkable plasticity of HSCs and CAFs during liver fibrogenesis and carcinogenesis, their distinct subpopulations show complementary or ambiguous functions in response to specific chronic inflammatory and tumor microenvironments.

Figure 2.. Roles of CAF subpopulations in the liver tumor microenvironment.

scRNA-Seq analyses have revealed significant heterogeneity of CAF subpopulation in HCC, iCCA and liver metastasis. The main CAF subtypes described in liver cancer are myofibroblastic (myCAF), inflammatory (iCAF) and vascular (vCAF). In HCC, HSCs expressing type 1 collagen (myHSCs) increase liver stiffness, which promotes proliferation of tumor cells. Conversely, HSCs expressing cytokines and growth factors (iHSC) suppress HCC growth via hepatocyte growth factor (HGF) and its receptor MET. In iCCA, myCAFs produce hyaluran synthase 2, the enzyme responsible for hyaluronic acid (HA) production, promoting tumor growth and progression. On the other hand, type 1 collagen produced by the myCAF subpopulation contributes to liver stiffness, but without effects on tumor growth. In human iCCA, the vCAF subpopulation promotes tumor growth via the interleukin (IL)-6/IL-6R axis. In liver metastasis, type 1 collagen derived from myCAFs suppresses tumor growth by mechanically restraining the tumor, whereas HA promotes tumor growth. In both iCCA and metastases, iCAFs promote tumor growth via the HGF-MET axis. Arrows indicate pro-tumorigenic effects and inhibitory arrows indicate suppression of tumor growth.

Figure 3.. Hepatic stellate cell subpopulations are potential targets for anti-fibrotic and anti-tumor therapies in NASH.

Activated HSCs have elevated expression of alpha V integrin and TGFβ receptor, which can be targeted to reduce fibrosis. As proof of principle, a small molecule CWHM12 pharmacologically blocks alpha V integrin to attenuate fibrosis, and TGFβ receptor signaling can also be locally inhibited by targeting caveolin, hyaluronic acid synthase 2, CD147, hydrogen peroxide inducible clone 5, or galectin-1. After prolonged activation, HSCs can senesce and secrete the SASP component IL-33 via the gasdermin D pore, which promotes tumor development. Disulfiram reduces tumor burden in mice by inhibiting the gasdermin D pore. Senescent HSCs can be depleted by senolytic anti-uPAR CAR T cells to reduce liver fibrosis. HSC-derived CAFs might be depleted by targeting fibroblast activation protein (FAP) via DNA vaccines, CAR T cell therapy or oncolytic viruses, to potentially reduce hepatic fibrosis and tumor burden.