DEPDC1B Promotes Melanoma Angiogenesis and Metastasis through Sequestration of Ubiquitin Ligase CDC16 to Stabilize Secreted SCUBE3.

The ability of melanoma to acquire metastasis through the induction of angiogenesis is one of the major causes of skin cancer death. Here, it is found that high transcript levels of DEP domain containing 1B (DEPDC1B) in cutaneous melanomas are significantly associated with a poor prognosis. Tissue microarray analysis indicates that DEPDC1B expression is positively correlated with SOX10 in the different stages of melanoma. Consistently, DEPDC1B is both required and sufficient for melanoma growth, metastasis, angiogenesis, and functions as a direct downstream target of SOX10 to partly mediate its oncogenic activity. In contrast to other tumor types, the DEPDC1B-mediated enhancement of melanoma metastatic potential is not dependent on the activities of RHO GTPase signaling and canonical Wnt signaling, but is acquired through secretion of signal peptide, CUB domain and EGF like domain containing 3 (SCUBE3), which is crucial for promoting angiogenesis in vitro and in vivo. Mechanistically, DEPDC1B regulates SCUBE3 protein stability through the competitive association with ubiquitin ligase cell division cycle 16 (CDC16) to prevent SCUBE3 from undergoing degradation via the ubiquitin-proteasome pathway. Importantly, expression of SOX10, DEPDC1B, and SCUBE3 are positively correlated with microvessel density in the advanced stage of melanomas. In conclusion, it is revealed that a SOX10-DEPDC1B-SCUBE3 regulatory axis promotes melanoma angiogenesis and metastasis, which suggests that targeting secreted SCUBE3 can be a therapeutic strategy against metastatic melanoma.

FSTL1 Secreted by Activated Fibroblasts Promotes Hepatocellular Carcinoma Metastasis and Stemness.

The tumor microenvironment plays a critical role in maintaining the immature phenotype of tumor-initiating cells (TIC) to promote cancer. Hepatocellular carcinoma (HCC) is a unique disease in that it develops in the setting of fibrosis and cirrhosis. This pathologic state commonly shows an enrichment of stromal myofibroblasts, which constitute the bulk of the tumor microenvironment and contribute to disease progression. Follistatin-like 1 (FSTL1) has been widely reported as a proinflammatory mediator in different fibrosis-related and inflammatory diseases. Here we show FSTL1 expression to be closely correlated with activated fibroblasts and to be elevated in regenerative, fibrotic, and disease liver states in various mouse models. Consistently, FSTL1 lineage cells gave rise to myofibroblasts in a CCL4-induced hepatic fibrosis mouse model. Clinically, high FSTL1 in fibroblast activation protein-positive (FAP+) fibroblasts were significantly correlated with more advanced tumors in patients with HCC. Although FSTL1 was expressed in primary fibroblasts derived from patients with HCC, it was barely detectable in HCC cell lines. Functional investigations revealed that treatment of HCC cells and patient-derived 3D organoids with recombinant FSTL1 or with conditioned medium collected from hepatic stellate cells or from cells overexpressing FSTL1 could promote HCC growth and metastasis. FSTL1 bound to TLR4 receptor, resulting in activation of AKT/mTOR/4EBP1 signaling. In a preclinical mouse model, blockade of FSTL1 mitigated HCC malignancy and metastasis, sensitized HCC tumors to sorafenib, prolonged survival, and eradicated the TIC subset. Collectively, these data suggest that FSTL1 may serve as an important novel diagnostic/prognostic biomarker and therapeutic target in HCC. SIGNIFICANCE: This study shows that FSTL1 secreted by activated fibroblasts in the liver microenvironment augments hepatocellular carcinoma malignancy, providing a potential new strategy to improve treatment of this aggressive disease.