LAUSR

Hepatocellular carcinoma (HCC) is the fourth leading cause of cancerrelated deaths worldwide with few effective drug treatments. Recent combination therapies inhibiting immune checkpoints and angiogenesis showed better overall and progressionfree survival outcomes than the protein kinase inhibitor sorafenib in patients with unresectable HCC, suggesting targeting tumour microenvironment (TME) is a promising strategy for the treatment of HCC. However, how the TME is developed and regulated is still poorly known. Thus, understanding the pathogenesis of HCC TME would be essential in order to develop more effective HCC therapies. In Gut, Fu et al have demonstrated that microRNA223 (miR223) is a key regulator of HCC TME by modulating tumour hypoxic microenvironment, immunosuppression and angiogenesis. miRNAs are a class of small (~19 to 24 nucleotides in length) endogenous RNAs that play critical roles in controlling the development, progression and prognosis of liver disease. miR223 is a welldocumented myeloidenriched miRNA that is expressed at the highest levels in neutrophils, followed by macrophages, but present at low levels in hepatocytes. Relevantly, miR223 acts as an important feedback inhibitor to ameliorate liver inflammation by attenuating the expression of numerous inflammatory genes. In this new study, Fu et al have revealed another important function of miR223, that is, to negatively regulate the expression of programmed cell death 1 (PD1) in T cells and programmed cell death ligand 1 (PDL1) in macrophages in two preclinical mouse models of HCC induced by injection of diethylnitrosamine or by orthotopic HCC cell implantation together with chronic CCl4 injection. They found that the expression of PD1/PDL1 in immune cells was markedly upregulated in these two mouse models of HCC, and this upregulation was more pronounced in miR223 knockout mice. Because the activation of the PD1/PDL1 pathway is an important mechanism for tumours to escape from immune surveillance, the upregulated PD1/PDL1 axis may eventually contribute to the increased susceptibility of miR223 knockout mice to HCC development and progression, although this possibility needs to be more thoroughly addressed. Interestingly, this study also revealed that miR223 did not directly inhibit PD1 and PDL1 expression in immune cells; rather it indirectly attenuated their expression by downregulating HCC hypoxiainducible factor 1α (HIF1α), a key protein that controls HCC responses to hypoxia as well as PD1 and PDL1 levels. 5 The authors have further provided mechanistic insights showing that miR223 directly targets and inhibits HIF1α expression in HCC cells and subsequently limits hypoxia/HIF1α-driven angiogenesis and immunosuppression. HIF1α, which is upregulated in HCCs, is known to stimulate the expression of CD39 and CD73, the two key ectonucleotidases that have been proven to drive tumour immunosuppression. Elevated CD39 and CD73 generate more adenosine in the tumour environment, and the elevated adenosine subsequently induces PD1 and PDL1 in immune cells, resulting in immunosuppression. All of these effects are suppressed by miR223 due to the direct targeting and downregulation of HIF1α mRNA by this miRNA in HCC cells. miR223 is expressed at low levels in hepatocytes and is further downregulated in HCCs, while myeloid cells, especially neutrophils, express high levels of miR223. Because HCCs are often associated with significant intrahepatic infiltration of myeloid cells such as neutrophils, and myeloid cells can transfer miR223 into hepatocytes via extracellular vesicles (EVs), the authors speculate that infiltrated myeloid cells may transfer miR223 into HCC and subsequently inhibit HIF1α, thus modulating the TME. However, although this tenet is plausible, more studies are required to confirm this mechanism. Interestingly, in vitro evidence showing that HCC cells are able to take up neutrophilderived EVs and miR223 was provided in this study. The authors previously performed bone marrow transplantation experiments, and their data suggested that immune cellderived miR223 can be directly transferred into hepatocytes in vivo in mouse models of nonalcoholic fatty liver disease. Future studies using bone marrow transplantation or cellspecific miR223 knockout mice are required to confirm the transfer of myeloid cellderived miR223 into HCCs in vivo in mouse models of liver cancer. Finally, Fu et al have also tested the therapeutic potential of miR223 for the treatment of HCC in two preclinical HCC models with chronic liver inflammation. It was found that gene delivery of miR223 in hepatocytes ameliorates HCC tumour angiogenesis, hypoxia, PD1/PDL1 expression and tumour growth. Given the marked reduction of miR223 in HCC, restoration of miR223 in tumour tissues could improve the efficacy of the current combination therapy for HCC with PD1/ PDL1 inhibitors and antiangiogenic regimens. MiRNAbased therapeutic strategies are being actively evaluated for the treatment of human diseases including cancer. miR223 could be a potential therapeutic target to be clinically tested for the treatment of HCC, particularly if tumourspecific miRNA delivery strategies can be implemented. More importantly, this study also implied the novel role of myeloid cells in orchestrating TME in HCC, controlling immune escape and angiogenesis, which may provide new perspectives for cancer treatment in the future. Nevertheless, given the increasingly recognised impact of nonalcoholic steatohepatitis (NASH) in hepatocarcinogenesis, it will be interesting to revisit the role of miR223 in mouse models of NASHassociated HCC that capture the particularities of liver tumour development in this background.