Genetic deletion of the chemokine Cxcl10 diminish HCC burden in a murine liver cancer model by modulation of tumor stroma

Background and aims:

HCC progression is linked to the dynamic process of neovascularization, chronic liver inflammation and the crosstalk between tumor cells, peritumoral parenchymal cells as well as non-parenchymal cells e.g. infiltrating immune cells. The chemokine CXCL10 is implicated in these processes and plays a key role during acute and chronic liver injury in men and mice. Here we aimed to investigate the role of CXCL10in HCC progression.

Methods:

We applied a fibrosis-triggered cancer mouse model encompassing a single i.p. injection of N-Diethylnitrosamin (DEN, 14 d after birth) and weekly i.p. injections of low dose carbon tetrachloride (CCl4, week 4 – 26) to Cxcl10-/- and wild-type (WT). Tumor burden was assessed after 26 weeks. Tumor and surrounding tissue was characterized by histological staining for proliferation, cell death/apoptosis and vascularization marker (Ki67, TUNEL, cleaved caspase 3 and VEGF-R2/KDR). Leukocyte subpopulations were measured by multicolor flow cytometry (FACS) and immunofluorescence (IF)/-histological (IHC) staining (e.g. CD3, CD4, CD8) with specific focus on the HCC-related T cell immune response. In vitro, the tumor cell line Hepa1.6 was stimulated with CXCL10 and proliferation was determined by BrDU incorporation.

Results:

CXCL10 expression is significantly increased in tumor tissue as compared to control tissue (qPCR and IHC). Genetic deletion of Cxcl10 in a murine liver cancer model led to a decreased tumor burden (tumor number and size, P < 0.05) compared to WT mice, which was linked to a slight reduction of Ki67+ tumor cells (P < 0.07). In vitro, the stimulation of the tumor cell line/Hepa 1 – 6 with CXCL10 did not alter cell proliferation implicating no direct effects of CXCL10 on tumor growth. As CXC chemokines are implicated in hallmarks like angiogenesis and immune cell recruitment, we hypothesized that CXCL10 exert its pro-tumorigenic effects via modulation of the tumor microenvironment. Indeed, CXCL10-/- tumors showed a lower amount of VEGF-R2/KDR+ area (IF, P < 0.01) and microvessel density (P < 0.05) (= neovascularization). Moreover, IHC staining and multicolor flow cytometry analysis revealed an accumulation of CD3+ T cells (P < 0.001) in the tumor stroma tissue of DEN/CCl4 treated Cxcl10-/- mice, with enhanced stromal accumulation of CD4+ T helper cells and CD8+ cytotoxic T cells in DEN/CCl4 treated Cxcl10-/- mice in comparison to DEN/CCl4 treated WT mice (CD4+ cells P < 0.01, CD8+ cells P < 0.05). Of note, the boost of T cells was notassociated with a compensational increase of CXCL9 – the other ligand of the CXCL10 receptor CXCR3 – in Cxcl10-/- mice.

Conclusions:

Our results uncover the pro-tumorigenic function of the chemokine CXCL10 in a murine liver cancer model. CXCL10 stimulates tumor neovascularization and suppresses the accumulation of T cell subsets with potential anti-tumoral properties. Hence, these data implicate CXCL10-directed tumor microenvironment modification as novel targets for HCC treatment.