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

To recruit tissues for research study, patient’s consent was obtained before surgery at Queen Mary Hospital, Hong Kong. The use of the clinical specimens was approved by the local Institutional Review Board. Tumors containing more than 70% tumor cells were only used in the present study. A total of 76 tumor samples surgically resected from primary ovarian cancer patients and 53 normal ovary samples from benign diseases (including 5 borderline cases) were randomly selected for this study. The histological subtypes and disease stages of the tumors were classified according to International Federation of Gynecology and Obstetrics (FIGO) criteria. Clinical data of patients were retrieved from the records kept at the Department of Obstetrics & Gynecology, Queen Mary Hospital, Hong Kong.

Total RNA from the ovarian tissues was prepared using the TRIzol reagent (Invitrogen). Quality and quantity of RNA were determined by agarose electrophoresis and spectrophotometry, respectively. The cDNA was then prepared using the Taqman reverse transcription kit according to the manufacturer’s instructions (Applied Biosystems, Foster City, CA). The gene expression assay kits were also purchased from Applied Biosystems. The primers and probes used to detect and quantify the expressions of AMPK subunits were AMPK-α1 (Assay ID: Hs00178893_m1); AMPK-α2 (Hs00178903_m1); AMPK-β1 (Hs00272166_m1); AMPK-β2 (Hs00271294_m1); AMPK-γ1 (Hs00176952_m1); and AMPK-γ2 (Hs00211903_m1).

The real-time quantitative PCR (Q-PCR) was performed in an ABI 7500 system (Applied Biosystems). The relative expression level of each gene was normalized against the endogenous β-actin (Product no: 4326315E) control. The GAPDH (Assay ID: Hs99999905_m1) control was also used to further verify the relative expression level of each gene. The relative quantity of the RNA expression was calculated using the comparative CT method with the 7500 System SDS software [10].

Immunohistochemical (IHC) staining for the six AMPK subunits was performed separately on an ovarian cancer tissue array (OVC1021) (Pantomics Inc, San Francisco, CA). The tissue array contains 97 cancer cases and 5 normal/benign cases. The antibodies against the six AMPK subunits were all purchased from Cell Signaling Technology, Inc., USA. They were anti-AMPK-α1 (Cat. no.: 2795); anti-AMPK-α2 (2757); anti-AMPK-β1 (4182); anti-AMPK-β2 (4148); anti-AMPK-γ1 (4187); and anti-AMPK-γ2 (2536). In addition, the anti-phospho-AMPK-α (2535) was also use to detect the phospho-AMPK-α. The details of the IHC staining procedures have been published elsewhere [11, 12]. The percentages of positively stained cells in tumors and normal epithelia were assessed. The proportions of positive cells were ranged from 10 to 100%, while the intensity of staining was scored as 0 (negative), 1 (very weak), 2 (weak), 3 (moderate), 4 (intense), and 5 (very intense) in the most strongly stained tumor area. The immunoreactivity score for each case was taken as percentage of positive cells multiplied by the intensity of staining.

The cellular locations of AMPK-α2, -β1 and -γ2 were examined in the ovarian cancer cells, A2780CP and SKOV3. The pCMV6–AMPK-α2–GFP, pCMV6–AMPK-β1–GFP and pCMV6–AMPK-γ2–GFP tagged plasmids (OriGene Technologies) were used. The analytical procedure has already been reported in a previous study [12]. The fluorescence signals were captured through confocal microscopy.

Statistical analysis was performed using Chi-squared test to evaluate the relationship between gene expression (low/high) and clinicopathological parameters, including age, grade, stage, histological subtype and recurrence. The nonparametric Mann–Whitney test was used to determine the significance on the difference in the distribution of gene expression in cancer, borderline and normal samples. Kaplan-Meier and the log rank test were used for survival analyses (both overall survival; from diagnosis to death of any cause and disease-free survival; from diagnosis to relapse). The SPSS 15.0 software was used to carry out the statistical analyses. All P values reported were two-sided and P < 0.05 was considered as statistically significant.

Ovarian tissues, including tumor and normal tissues, were used in the current study. The relative levels of the six AMPK subunits including α1, α2, β1, β2, γ1, and γ2 in ovarian tissues were measured using real-time Q-PCR. Expression of the AMPK-γ3 subunit was not conducted because it is not expressed in ovarian tissues.

The mRNA levels of all AMPK subunits, except AMPK-α1, were statistically significantly higher in ovarian carcinomas than those of the normal controls (Figure 1 & Table 1). Over-expressions of different AMPK subunits independently occurred and were not associated with one another.

To further confirm the over-expressions of AMPK subunits in ovarian carcinomas, IHC staining of each AMPK subunit was carried out separately on an ovarian tissue array slide.

IHC was successful using the antibodies against AMPK-α1, -α2, -β1, and -β2. However, IHC using antibodies against AMPK-γ1 and -γ2 subunits was not successful, and the protein levels of the two γ subunits cannot be determined. Consistent with Q-PCR results, AMPK-α2, -β1, and -β2 subunits were over-expressed in ovarian carcinomas compared with the normal controls (Figure 2), whereas AMPK-α1 expression generally had no significant difference between cancer samples and normal controls.

Concerning the cellular locations of the AMPK subunits, all four subunits were detected in the cytoplasm. In addition, AMPK-α2 and -β1 were also detected in the nucleus and cell membrane, respectively (Figure 2). The active form of AMPK, the phospho-AMPK-α was only detected in the cytoplasm (Figure 3).

To confirm further the cellular locations of the AMPK subunits detected in IHC, confocal microscopy was used to localize the AMPK subunits in ovarian cancer cells with separately enforced expression of the subunits in A2780CP cell (Figure 4). Expression of AMPK-α2 was detected in both cytoplasm and nucleus. Expression of AMPK-β1 was observed in the cytoplasm as well as in the cellular membrane. Thus, the expressions of AMPK-α2 and AMPK-β1 detected either by IHC or confocal microscopy match with each other. Although IHC was not successful in detecting AMPK-γ2, expressions of AMPK-γ2 were evenly detected in the cytoplasm, membrane, and nucleus of the A2780CP cell (Figure 4). The cellular locations of AMPK-β1 and -γ2 in another ovarian cancer cell SKOV3 were shown in Additional file 1: Figure S1. These results indicate that the expressions of AMPK subunits are quite distinct from one another.

After determining the expression levels of the various AMPK subunits by Q-PCR, the relationships between the expression levels of the various AMPK subunits and clinicopathological data were analyzed. The ovarian carcinoma samples were divided into two groups based on the relative expression levels of the AMPK subunits, that is, greater than or smaller than/equal to two-folds of the mean of the relative expression levels of the normal samples. The two groups of samples were then correlated with the clinicopathological parameters of the samples used in the current study. Table 2 summarizes the results of all correlations. The details of the correlation analyses are shown in Additional file 2: Table S1 to Table S7.

The high expression of the AMPK-α2 subunit was found to be associated with endometrioid carcinomas (Additional file 2: Table S2). On the other hand, the high expressions of the AMPK-β and -γ subunits were found to be associated with mucinous (Additional file 2: Table S3 and Table S4) and serous carcinomas (Additional file 2: Table S5 and Table S6), respectively. Clear cell carcinomas had no association with any of the high expressions of the AMPK subunits (Additional file 2: Table S1 to Table S6). Since the antibodies for detecting AMPK-γ subunits were not good enough for IHC, we can only present the IHC staining quantifications of the AMPK-α and β subunits. The results are shown in Additional file 2: Table S8-Table S11. However, since the numbers of samples of the various histological subtypes in the tissue various greatly, it is not easy to match the results between Q-PCR and IHC staining. Nevertheless, the results determined by Q-PCR and IHC staining match one another in the grade and stage of diseases as described below.

The expressions of various AMPK subunits were also correlated with the stages of the disease. High expressions of AMPK-β1 (Additional file 2: Table S3) and AMPK-γ2 (Additional file 2: Table S6) were found in the early and late stages of the disease, respectively. The expressions of other AMPK subunits had no correlation with the stages of the disease. The correlation between AMPK-β1 expression and the stage of the disease was further confirmed by IHC results (Additional file 2: Table S10).

The high expression of AMPK-γ2 was associated with low-grade tumors (Additional file 2: Table S6). However, this result cannot be further confirmed because IHC staining of AMPK-γ subunits was not successful. The expressions of other AMPK subunits had no association with the grades of tumors.

Patients with high expressions of AMPK-α2 had less disease recurrence rate (Additional file 2: Table S2) and better overall and disease-free survival rate (Figure 5). This finding is consistent with the results described in another report [13].