Over-expressions of AMPK subunits in ovarian carcinomas with significant clinical implications

BACKGROUND: AMP-activated protein kinase (AMPK) has recently been considered as a potential target for cancer therapy. However, the expression status of various subunits of the heterotrimeric AMPK in human cancers is rarely reported. We decided to determine their expressions in ovarian carcinomas and their relationships with the disease. METHODS: Expressions and locations of the AMPK-α1, -α2, -β1, -β2, -γ1 and -γ2 were detected by quantitative PCR (Q-PCR) and immunohistochemical staining (IHC). Their expression levels in ovarian tumors were compared with normal controls and also correlated with clinicopathological parameters. RESULTS: Except AMPK-α1, expressions of the other five AMPK subunits are significantly higher in ovarian carcinomas as determined by Q-PCR. Although IHC detection of AMPK-γ1 and -γ2 were not successful, over-expressions of AMPK-α2, -β1, and -β2 were further confirmed by IHC. Over-expressions of various AMPK subunits occurred independently and were mainly detected in the cytoplasm. Interestingly, AMPK-α2 and -β1 were also detected in the nucleus and cell membrane, respectively. Clinical correlation analyses indicate that expressions of different AMPK subunits are associated with different subtypes of carcinoma. High expression of AMPK-α2 is significantly associated with endometrioid carcinomas. On the other hand, high expressions of AMPK-β and -γ subunits are associated with mucinous and serous carcinomas, respectively. Furthermore, high expressions of AMPK-β1 and -γ2 are also associated with early and late stages of disease, respectively. Finally, patients with high expression of AMPK-α2 had better prognosis. CONCLUSIONS: Aberrant expressions of AMPK subunits may play important roles in ovarian carcinogenesis. Each AMPK subunit may have its own function other than just a component of the AMPK molecule. Correlations with clinical parameters suggest that expressions of AMPK subunits have different clinical implications in ovarian cancer development.