H2AK119Ub1 and H3K27Me3 in molecular staging for survival prediction of patients with pancreatic ductal adenocarcinoma.

Oncotarget

PubMedID: 25431952

Chen S, Chen J, Zhan Q, Zhu Y, Chen H, Deng X, Hou Z, Shen B, Chen Y, Peng C. H2AK119Ub1 and H3K27Me3 in molecular staging for survival prediction of patients with pancreatic ductal adenocarcinoma. Oncotarget. 2014;5(21):10421-33.
Polycomb group (PcG) proteins Ring1B and EZH2, which have been characterized as catalyzing the two epigenetic modifications H2AK119 monoubiquitination (H2AK119Ub1) and H3K27 trimethylation (H3K27Me3), are well-known epigenetic silencers implicated in embryonic development and tumorigenesis. However, the status of polycomb-associated histone modifications and their clinical implications in pancreatic cancer remain unclear. Here, we performed immunohistochemistry on tissue microarrays (TMAs) containing 80 pairs of human pancreatic cancer specimens to assess the expression levels of Ring1B, H2AK119Ub1, EZH2, and H3K27Me3 in tumors. More than 50% of the tumor cells showed a high expression of H2AK119Ub1, Ring1B, and EZH2, whereas more than 50% of the tumor cells showed a low level of H3K27Me3. Different expression patterns of H2AK119Ub1 and H3K27Me3 in tumors were negatively correlated (r = -0. 247, P = 0. 027). Both H2AK119Ub1 and H3K27Me3 independently predicted the clinical prognosis. In particular, a combinatorial pattern of elevated H2AK119Ub1 and decreased H3K27Me3 in tumors was significantly correlated with a poorer prognosis. Furthermore, compared to the tumor, lymph node, metastasis (TNM) staging system, histone modifications can discriminate the survival difference more accurately, especially for patients with stage I or stage II tumors. Simultaneous silencing of Ring1B and EZH2 via shRNA depleted H2AK119Ub1 and H3K27Me3 in the pancreatic cancer cells PanC1 and AsPC1, enhanced HOX gene derepression, and inhibited tumor cell growth in vitro and in tumor xenograft models. These results demonstrated that H2AK119Ub1 and H3K27Me3 cooperate in tumors and are associated with the clinical prognosis in combinatorial patterns. We have proposed that epigenetic modifications may serve as discriminatory biomarkers for molecular staging of pancreatic cancer.