Cancer‐Associated Fibroblasts: Orchestrating the Crosstalk Between Liver Cancer Cells and Neutrophils Through the Cardiotrophin‐Like Cytokine Factor 1–Mediated Chemokine (C‐X‐C motif) Ligand 6/TGF‐β Axis

Liver cancer (hepatocellular carcinoma, HCC) is one of the deadliest diseases and the 4th leading cause of cancer mortality in the world. To prolong the survival of HCC patients, immense efforts have been made to target the intrinsic key signaling pathways regulating the survival of HCC cells. However, the behaviors of cancer cells within the tumor are critically regulated by stromal cells, including fibroblasts, immune cells and endothelial cells, within the tumor microenvironment (TME). Cancer-associated fibroblasts (CAFs) are one of the major cell populations within the TME, and activated fibroblasts are the predominant form of CAFs found in several malignancies. Specifically, in HCC, more than 80% of cases develop within the context of cirrhosis, which is always accompanied by an enrichment of activated fibroblasts due to chronic inflammation.1 Therefore, it is crucial to understand the potential crosstalk between CAFs and HCC cells.


SEE ARTICLE ON PAGE 1717
L iver cancer (hepatocellular carcinoma [HCC]) is one of the deadliest diseases and is the fourth leading cause of cancer mortality in the world. To prolong the survival of patients with HCC, immense efforts have been made to target the intrinsic key signaling pathways regulating the survival of HCC cells. However, the behaviors of cancer cells within the tumor are critically regulated by stromal cells, including fibroblasts, immune cells, and endothelial cells, within the tumor microenvironment (TME). Cancer-associated fibroblasts (CAFs) are one of the major cell populations within the TME, and activated fibroblasts are the predominant form of CAFs found in several malignancies. Specifically, in HCC, more than 80% of cases develop within the context of cirrhosis, which is always accompanied by an enrichment of activated fibroblasts due to chronic inflammation. (1) Therefore, it is crucial to understand the potential crosstalk between CAFs and HCC cells.
In 2011, Mazzocca et al. (2) provided the first evidence, to our knowledge, showing the crosstalk between CAFs and HCC cells. First, they demonstrated the molecular and functional differences between peritumoral fibroblasts (PTFs) and CAFs. Furthermore, they identified tumor-derived lysophosphatidic acid to be crucial in inducing the differentiation of PTFs, which in turn accelerates HCC progression. Evidence that CAF-derived hepatocyte growth factor is capable of regulating the plasticity of liver cancer stem cells (CSCs) through the c-Met/FRA1/HEY1 and AP1/KRT19 signaling pathways and bears clinical significance has also begun to emerge. (3,4) Recently, CAFs were found to recruit immune cells within the TME, including neutrophils and macrophages. For example, CAF-derived endosialin promoted macrophage recruitment and polarization, resulting in HCC progression. (5) In addition, CAFs induced PDL1 + tumor-associated neutrophils (TANs) and impaired T-cell function through the IL6/signal transducer and activator of transcription 3 (STAT3) pathway. (6) Although crosstalk between CAFs/HCC cells and CAFs/immune cells has been reported, the nature of the intricate crosstalk among CAFs, HCC cells, and other cellular components within the TME remains largely unknown, particularly during different HCC stages, which significantly impedes efforts to develop effective cancer therapeutics.
In this issue of Hepatology, Song et al. (7) reported comprehensive cytokine-mediated crosstalk among CAFs, HCC cells, and TANs, augmenting cancer stemness and TAN recruitment in HCC. The authors reported for the first time, to our knowledge, that CAFs isolated from HCC at advanced stages showed a greater tumorpromoting effect in vivo than those isolated from early stages. By comparing the expression profiles between PTFs and CAFs, cardiotrophin-like cytokine factor 1 (CLCF1) was found to be preferentially expressed and secreted by CAFs, and its levels were correlated with advanced HCC stages. Clinically, high levels of CLCF1 expression were correlated with aggressive tumor behaviors and poor patient survival. CLCF1 was also identified to be preferentially secreted by lung CAFs with clinical significance. (8) Upon gene set enrichment analysis, CLCF1 expression showed a negative correlation with stem-cell differentiation in a cohort of HCC samples. Along with these data, the authors found that CAF-derived CLCF1 enhanced cancer stemness in HCC, and the effect was mitigated by repressing the expression of CNTFR in HCC cells. Additionally, this study uncovered a mechanism underlying the CSCenhancing role of CAF-derived CLCF1 in HCC. As an inflammatory cytokine, CLCF1 was found to upregulate two key cytokines, chemokine (C-X-C motif ) ligand 6 (CXCL6) and TGFβ, through the Akt/ extracellular signal-regulated kinase 1/2 (ERK1/2)/ STAT3 signaling axis. Mechanistically, CXCL6 was found to regulate cancer stemness through the transcriptional regulation of stemness-associated genes, including the SRY (sex determining region Y)-box 2 gene through E2F1, whereas TGFβ affected its CSC-enhancing effect through the p38 pathway. Interestingly, the suppression of STAT3 fully abolished the CSC-enhancing effect of CLCF1, whereas the addition of CXCL6 and TGFβ only partially restored this suppressive effect, suggesting that STAT3 is the major effector of CLCF1mediated cancer stemness in HCC. This result is in line with a previous report showing the critical involvement of STAT3 in the regulation of liver CSCs in HCC. (9) As TANs are one of the most infiltrated immune cells within the TME of HCC, it is possible that there is potential crosstalk between CAFs and TANs. Another interesting result is that Song et al. dissected the interactions among CAFs, HCC, and TANs, which are mediated through a cytokine network. Data from the Cancer Genome Atlas database showed that CXCL6 was significantly correlated with CD15, the cell surface marker of neutrophils, whereas TGFβ displayed a strong association with matrix metalloproteinase 9, a functional marker of N2-polarized neutrophils. Based on this interesting finding, the authors performed a functional study and found that recombinant CXCL6 and TGFβ function to recruit and induce the "N2" polarization of TANs, respectively. This result, together with the data showing the enhanced recruitment of TANs upon the treatment of HCC cells with either CLCF1 or a CAF supernatant, suggests the role of CAF-derived CLCF1 in recruiting and inducing the N2 polarization of TANs through the induced secretion of CXCL6 and TGFβ in HCC cells. However, because TGFβ also induced the activation, maturation, and differentiation of other immune cells, including natural killer cells, macrophages, dendritic cells, and T cells, there is also a possibility that other immune cells may also be involved in the CLCL1-mediated signaling cascade.
Additionally, this study provided mechanistic insight into how HCC cells educate CAFs to enhance their potential to regulate cancer stemness and the recruitment of TANs. Song et al. showed that ERK1/2 phosphorylation was associated with CLCF1 expression and that their levels in CAFs were higher when isolated from late-stage HCC tissues than they were when isolated from early-stage HCC tissues. Furthermore, they found that ERK1/2 signaling in CAFs is required for CLCF1 expression in CAFs, which is activated by HCC-secreted CXCL6 and TGFβ in a paracrine manner. Using an in vivo mouse model, they were able to confirm the functional role of CLCF1 in coordinating the crosstalk among CAFs, HCC cells, and TANs through an ERK1/2-induced CLCF1/CNTFR positive feedback loop. In a cohort of HCC clinical samples, they showed the role of the CLCF1-CXCL6/TGFβ axis in the regulation of cancer stemness, and the recruitment of N2 TANs was clinically relevant, which contributes to the poor prognosis of patients with HCC.
In summary, this study provides a CLCF1-driven CXCL6/TGFβ/ERK1/2 positive feedback loop involving crosstalk among different cellular components, including CAFs, HCC cells, and TANs (Fig. 1). Accordingly, this study opens a therapeutic avenue for targeting HCC by the selective blockade of CLCF1/CNTFR or ERK1/2 signaling, warranting a thorough investigation.