Background
Cervical cancer results from uncontrolled growth of malignant cells started within the uterine cervix and is one of the most common malignancies in women worldwide [
1‐
3]. Although this disease is almost preventable with routine genetic screening and vaccination, more than 80% of cervical cancers with a majority in the advanced stage are currently found in developing countries including China, leading to a high risk of recurrence and poor survival [
2,
4]. Thus, there is a compelling need to explore novel therapeutic interventions for this disease.
Emerging evidence suggests that targeting cancer cell metabolism is a promising therapeutic approach in human cancers. AMP-activated protein kinase (AMPK) is a known cellular metabolic sensor and plays an important role in the control of energy homeostasis in response to external stresses [
5‐
8]. Recent studies have documented that pharmacological activation of AMPK is able to block cancer cell growth in various human cancers [
8‐
11]. Indeed, we have previously reported that pharmaceutical AMPK activators such as AICAR (ATP-dependent) and A23187 (ATP-independent) could suppress cervical cancer cell growth in the presence or absence of LKB1, an upstream kinase of AMPK [
10]. We also proposed mechanistic evidence showing that metformin, AICAR and A23187 suppress cervical cancer cell growth through reducing DVL3, a positive effector of Wnt/β-catenin signaling cascade which has been shown to be constitutively active during cervical cancer development [
12]. Yet, it is still believed that there are other molecular mechanisms by which these pharmaceutical AMPK activators suppress cancer cell growth. The understanding of these mechanisms will assist in exploring better therapeutic regimes when using these drugs.
Forkhead Box M1 (FOXM1) is a member of the Forkhead Box transcription factors which is essential for cell proliferation and apoptosis in the development and function of many organs [
13‐
17]. We previously reported that aberrant upregulation of FOXM1 is associated with the progression and development of human cervical squamous cell carcinoma (SCC) [
18]. Biochemical and functional studies confirmed that FOXM1 is critically involved in cervical cancer cell growth through upregulating cyclin B1, cyclin D1 and cdc25B and downregulating p27 and p21 expressions. These findings suggest that FOXM1 plays a vital role in cervical cancer cell growth and oncogenesis.
In this study, we reported that the activated AMPK inhibits the cell growth by reducing FOXM1 expression in human cervical cancer cells upon treatments with hypoxia, glucose deprivation and pharmaceutical AMPK activators. We provided both biochemical and functional evidence to support our findings that the repression of FOXM1 expression of AMPK is dependent on the AKT/FOXO3a/FOXM1 signaling cascade.
Methods
Cell lines and reagents
Cervical cancer cell lines HeLa, CaSki, C33A and SiHa (American Type Culture Collection, Rockville, Md., USA) (cell line authentication was done by in-house STR DNA profiling analysis) were employed in this study. They were maintained in Dulbecco’s Modified Eagle Medium (DMEM) (Invitrogen, Carlsbad, CA) supplemented with 10% (v/v) fetal bovine serum (Gibco), 100 units/ml penicillin/streptomycin (Gibco) at 37°C in an incubator with humidified atmosphere of 5% CO2 and 95% air. AMPK activators AICAR, A23187 and metformin and AKT inhibitor LY294002 were obtained from Tocris Bioscience (Bristol, UK). FOXM1 inhibitor Thiostrepton was purchased from Calbiochem (La Jolla, CA, USA).
Plasmids and cell transfection
To study the effects of enforced FOXM1 expression, the FOXM1c-expressing plasmid pcDNA3–FOXM1c was used because the c isoform has higher transactivating activity and is expressed dominantly in cells as well as tissues. Whereas, the pcDNA3 empty vector was used in mock transfections as control. Besides, the vector-based shRNA plasmid pTER–FOXM1 was used to knockdown endogenous FOXM1. All of these plasmids had been described previously [
18]. As controls in knockdown assays, the p-super GFP and pcDNA3 vectors were used in mock transactions. To knockdown human FOXO3a, the TriFECTa RNAi Kit which contains three siRNAs targeting human FOXO3a was purchased from IDT (Integrated DNA Technologies, Inc., Iowa, USA). Cell transfection was carried out using LipofectAMINETM 2000 (Invitrogen) according to the manufacturer’s instructions. Expression patterns were analyzed by Western blotting. The parental vector pEGFP-C1 was used as empty vector control.
Cell proliferation assay
Cell proliferation kit (XTT) (Roche, Basel, Switzerland) was used to measure cell viability according to the manufacturer’s protocol. Three independent experiments were performed in triplicates.
RNA extraction and quantitative reverse transcriptase–PCR (Q-PCR)
According to the instruction of the manufacturer, total RNA was extracted using TRIzol reagent (Invitrogen). Complementary DNA (cDNA) was subsequently synthesized using a reverse transcription reagent kit (Applied Biosystems, Foster City). The expression level of FOXM1 was then evaluated by q-PCR in an ABI PRISM™ 7500 system (Applied Biosystems) using Taqman® Gene expression Assays; human FOXM1 (Assay ID: Hs00153543_m1). The human 18S rRNA (Assay ID: Hs99999901_m1) was used as an internal control.
Western blot analysis
Proteins in cell lysates were separated by 10% SDS-PAGE and transferred to polyvinylidene-difluoride (PVDF) membranes. The membranes were blotted with 5% skimmed milk and subsequently probed overnight at 4°C with primary antibodies specific for p-AMPKα, AMPKα, p-AKT, AKT, p-FOXO3a, FOXO3a (Cell Signaling, Beverly, MA, USA), FOXM1 (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) and β-actin (Sigma-Aldrich, St. Louis, MO, USA) and then incubated with horseradish peroxidase conjugated goat anti-rabbit or anti-mouse secondary antibody (Amersham, Uppsala, Sweden). Immunodetection was performed with enhanced chemiluminescent reagent solution (AmershamTM ECLTM) and visualized using medical X-ray film.
Data analysis
Student’s t test was applied to the data analysis. All data were expressed as mean ± SEM. P-value of less than 0.05 was considered as significant.
Discussion
Recent studies have suggested that targeting cancer cell metabolism is an alternative therapeutic approach in cancer treatment. AMPK is a pivotal energy sensor governing normal and cancer cell metabolism. Our previous research has shown that pharmaceutical AMPK activators are able to repress cervical cancer cell growth through targeting DVL3 in the Wnt/β-catenin signaling pathway [
12]. In this study, we report another molecular mechanism by which AMPK can retard cervical cancer cell growth: inhibition of FOXM1 function via AMPK/AKT/FOXO3a signaling. We demonstrated that AMPK activated by either micro-environmental stresses or pharmaceutical AMPK activators could reduce FOXM1 expression through blocking the AKT/FOXO3a signaling pathway, that in turn impaired cervical cancer cell growth.
The Forkhead box transcription factor FOXM1 regulates a number of key cell cycle regulators that control the G
1 to S and the G
2 to M transitions [
21‐
26]. Accumulating evidences have shown that the upregulation of FOXM1 is often involved in the development of various human cancers [
27‐
32]. We previously reported that there is a progressive increase in FOXM1 level in the progression of human cervical cancer [
18]. The inhibition of FOXM1 by genetic or pharmaceutical approach significantly impairs tumor growth of this cancer
in vitro and
in vivo[
18,
25,
27,
33], suggesting that targeting FOXM1 is an appealing and potential approach for anticancer therapeutics. On the other hand, we and others have proved that activation of AMPK is able to inhibit the growth of various human cancers including cervical cancer [
10,
34‐
36]. The pharmacological activation of AMPK using AICAR or metformin has been shown to inhibit cell growth and induce cell apoptosis of a wide spectrum of cancer cells through modulation of p53 [
37], p27 [
38,
39], or p21 [
18,
40], or DVL3 in Wnt/β-catenin signaling in cervical cancer [
12]. Herein, we demonstrated that the activation of AMPK by various AMPK activators or hypoxia and glucose deprivation stresses induces a remarkable reduction of FOXM1 which in turn leads to a remarkable decrease of cervical cancer cell growth in both HPV positive (Caski, Hela and SiHa) and HPV negative (C33A) cell lines. On the other hand, ectopic expression of FOXM1c could counteract the suppressive effect of activated AMPK. These findings indicate that FOXM1 is a key oncogenic factor associated with cervical cancer cell growth, while activated AMPK inhibits cervical cancer cell growth through downregulation of endogenous FOXM1.
In fact, we demonstrated that reduction of FOXM1 occurred at both the mRNA but and protein levels in cervical cancer cells. This is suggestive of a transcriptional suppression of FOXM1 by its upstream effectors. Previous studies have reported that FOXM1 is transcriptionally suppressed by FOXO3a, which is a critical downstream effector of the PI3K/AKT/FOXO signaling pathway [
19,
20]. For example, it has been reported that FOXO3a represses estrogen receptors α (ERα) activity in breast cancer cells through an alternative mechanism by which FOXO3a interacts and downregulates the expression of FOXM1 [
29,
41]. These evidences imply that the reduction of FOXM1 at both the mRNA and protein levels is due to the presence of FOXO3a in cervical cancer cells. In fact, our data using siRNA-mediated FOXO3a knockdown showed that FOXO3a expression is required for FOXM1 reduction upon activation of AMPK.
How do activated AMPK leads to FOXO3a accumulating in the nucleus and blocking the transcription of
FOXM1 mRNA? FOXO3a, which belongs to the class O of Forkhead/winged helix box (FOXO) transcription factors, is a key tumor suppressor involved in different cellular processes [
42]. FOXO3a is modified by phosphorylation, acetylation and ubiquitination, which in turn affect its nuclear/cytoplasm shuttling, transcriptional activity and stability [
43‐
46]. It is known that the PI3K/AKT signaling is the main regulatory pathway of FOXO3a [
44,
47,
48]. When PI3K/AKT signaling is activated, FOXO3a is not only inactivated and phosphorylated at Thr32, Ser253 and Ser315 residues, but is also exported out from the nucleus to the cytoplasm where it is ubiquitinated and subjected to proteasome-dependent degradation [
43,
48]. Therefore, nuclear FOXO3a functions as transcriptional regulator, whereas cytoplasmic FOXO3a is considered inactive [
46]. On the other hand, AKT is a signaling kinase known to be inactivated by activated AMPK [
49,
50]. In our study, treatment of either AMPK activator (metformin) or PI3K/AKT inhibitor (LY294002) showed significant inhibition of p-AKT and a remarkable reduction of p-FOXO3a (Ser253), an AKT-specific phosphorylation site, suggesting that suppression of FOXO3a is reduced. As a result, FOXO3a will be more nuclear-localized and activated to inhibit
FOXM1 mRNA expression in cervical cancer cells.
Aforementioned, AMPK activation can commonly inhibit FOXM1 expression in cervical cancer cells. However, whether there exists a feedback loop on the activity of AMPK is still unknown. To test this notion, cervical cancer cells were treated with the FOXM1 inhibitor thiostrepton to investigate the effect on AMPK activation. Results showed that treatment of thiostrepton only reduced expression of FOXM1 but not the activity of AMPK. In addition, depletion of endogenous FOXM1 using shRNAs gave similar findings as the treatment of thiostrepton, implying that FOXM1 is acting downstream of AMPK without any feedback regulation.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
MY and DC designed research; MY, DC and VL performed the experiments; DC, VL and KY contributed new reagents-analytic tools; MY, DC and HN analyzed and interpreted data; MY and DC wrote the manuscript. All authors were involved in editing the manuscript and had final approval of the submitted and published versions.