Supporting data for “Clusterin as a functional downstream target in Sox9-expressing hepatocellular carcinoma”.

Hepatocellular carcinoma (HCC) is one of the deadliest cancers across the globe. Patients are usually diagnosed at advanced stage. Presence of microvascular metastasis, oversized tumor burden and insufficient liver remnant reduce the feasibility to surgical and local ablative treatment modalities. Besides, high propensity of tumor recurrence and metastasis increase the level of challenge in eliminating HCC. As a matter of fact, targeted therapy confers survival benefits to HCC patients in late phase. In this regard, elucidating the genetic events that underlies HCC is of the utmost importance to supply potential biomarkers for HCC detection and druggable targets for curative purpose.

As a transcription factor, Sox9 masters embryogenesis and sexual development. Emerging evidence illustrated that Sox9 overexpression facilitates tumor initiation and progression in multiples solid cancers. Sox9 overexpression is associated with aggressive behavior. Of note, Sox9 is a stem cell factor which endows stemness properties in cancers. Based on our previous study, knocking down Sox9 inhibited cell growth, cell migration and invasion, but enhanced chemosensitivity. Apart from this, decline of Sox9 expression suppressed the expression of stemness markers and tumorsphere formation ability. The oncogenic potential of Sox9 was mediated by Wnt signaling pathway via transcription regulation of Frizzled-7. To further dissect the molecular mechanism underlying Sox9-expression HCC, we tried to identify the molecular target from cancer secretome regulated by Sox9. Furthermore, its functional role and clinical significance were evaluated.

To this end, the HCC secretome upon Sox9 silencing in HCC cell lines was examined by mass spectrometry. Our results revealed Clusterin (CLU) was a commonly downregulated in the conditioned medium (CM) from Sox9 knockdown clones compared with the control. A high serum CLU level was correlated with advanced HCC stages. Moreover, CLU could better differentiate tumor stages and predict clinical outcomes when used in combination with serum alpha-fetoprotein (AFP) level. After investigating the clinical features of CLU, we further explored the oncogenic potential of Sox9-CLU axis using in vitro assays. Through cell migration and invasion assays, CM from Sox9 knockdown clones retarded cell migration ability and invasiveness. Conversely, addition of CLU recombinant proteins enhanced cell motility in the transwell assays. Enhanced cell migration and invasion induced by Sox9 overexpression could be abrogated by the use of neutralizing CLU antibody which reduced extracellular CLU expression. By predictive analysis, Sox9 putative binding sites were identified in the promoter region of CLU. The binding interaction was validation by chromatin immunoprecipitation (ChIP) assay.

In summary, findings from this study suggests that CLU is a downstream effector in the secretome of Sox9-expressing HCC. Targeting secreted CLU may present a novel therapeutic strategy to improve the disease outcome of Sox9-expressing HCC.