Identification of prognostic claudins signature in hepatocellular carcinoma from a hepatocyte differentiation model.

Background: Loss of terminal differentiation markers and gain of stem cell-like properties are a major hallmark of cancer malignant progression. Identification of novel biomarkers representing tumor developmental progeny and predictive of patients' prognosis would greatly benefit clinical cancer management.

Methods: Human embryonic stem cells were induced to differentiate into hepatocytes along hepatic lineages. Transcriptomic data from different liver developmental stages were analyzed combining with the RNA-seq data from The Cancer Genome Atlas (TCGA) project. Kaplan-Meier survival analysis and Cox regression analyses were used to analyze the clinical significance in HCC patients.

Results: A shifted expression pattern of claudin (CLDN) family genes were identified to be closely associated with liver development and tumor progression. Claudins with hepatic features were found to be significantly down-regulated and predicted better prognosis in HCC patients. Conversely, another set of claudins with embryonic stem cell features were found to be significantly up-regulated and predicted worse prognosis in HCC patients. A claudin signature score system was further established by combining the two sets of claudin genes. The newly established claudins signature could robustly predict HCC patients' prognosis in the training, testing, and independent validation cohorts.

Conclusions: In the present study, we developed a novel embryonic developmental claudins signature to monitor the extent of tumor dedifferentiation in HCC from an in vitro hepatocyte differentiation model. The claudins signature might present a great potential in predicting prognostic significance in HCC as cell surface biomarkers, and provide novel therapeutic targets for precision oncology further in the clinic.