Progestin resistance is the main obstacle to successful conservative therapy in young endometrial cancer patients. To investigate the molecular events that lead to progestin resistance and to find a possible way to reverse progestin resistance in endometrial cancer, we established a progestin-resistant Ishikawa cell line by long-term progestin treatment to downregulate progesterone receptor (PR) expression. Both medoxyprogesterone acetate (MPA) and LY294002, a phosphatidylinositol 3-kinase (PI3K) inhibitor, were assayed for their effects on the proliferation of progestin-sensitive and progestin-resistant cancer cells, respectively. The MPA inhibited the PI3K/Akt pathway and suppressed cell proliferation in progestin-sensitive Ishikawa cells, but activated the PI3K/Akt pathway and had no effect on cell proliferation in progestin-resistant Ishikawa cells or HEC-1A cells. Inhibiting the PI3K/Akt pathway by LY294002 upregulated PR expression and diminished cell growth, especially in progestin-resistant endometrial cancer cells. In vivo endometrial cancer xenograft studies in nude mice also showed that inhibiting the PI3K/Akt pathway reversed progestin resistance in endometrial cancer. Our results indicate that activation of the PI3K/Akt pathway by progestin without PR mediation plays an important role in progestin resistance to endometrial cancer cells. In addition, inhibiting the PI3K/Akt pathway might reverse progestin resistance in endometrial cancer. (Cancer Sci 2011; 102: 557–564)
Endometrial cancer is the most common cancer in the female reproductive system and up to 75–85% of cases are adenocarcinoma (type I endometrial cancer).(1–3) Progestins, especially synthetic progestins such as medroxyprogesterone acetate (MPA), were applied in the conservative endocrine treatment to young patients of clinical stage I, grade 1 endometrial adenocarcinoma in order to preserve their fertility, as well as in palliative treatment to advanced-stage patients.(4–7) To date, more than 30% of patients with early stage endometrial cancers did not respond to progestin due to de novo or acquired progestin resistance during progestin treatment, despite the different drugs or regimens used.(6,8–11) The mechanism of progestin resistance is still unknown. Hysterectomy is the only choice for this group of patients to reduce the risk of possible cancer invasion. Currently, continuous efforts are being made to find a way to reverse progestin resistance and to improve the effect of fertility-sparing treatment for endometrial cancer. Although our previous studies have shown that long-term progestin treatment to Ishikawa endometrial cancer cells could downregulate progesterone receptor (PR) expression and induce progestin resistance in Ishikawa cells,(12–14) its causative molecular mechanism is still unclear.
Progestin is suggested to inhibit endometrial cancer cell proliferation and to reverse their malignant biological behavior by binding to its nuclear PR. However, the administration of progestins to endometrial cancer cells without PR expression does not have an inhibitory effect; instead, it stimulates cell proliferation. This implies that progestin might function through another pathway different to the classical PR one.(15–19) It was reported in other studies that progestin would activate the PI3K/Akt pathway independent of PR in breast cancer cells, and activation of the PI3K/Akt pathway by growth factors such as insulin-like growth factor (IGF) would result in downregulation of PR in breast cancer cells.(5,20–22) Therefore, in this paper, we showed that progestin might also activate the PI3K/Akt pathway in endometrial cancers independent of PR, and that inhibiting the PI3K/Akt pathway could reverse the progestin resistance in endometrial cancer.
Materials and Methods
Cells cultures. Human endometrial carcinoma cell line, Ishikawa cells (European Collection of Cells Cultures), were grown in Dulbecco’s modified Eagle’s Medium (DMEM; Invitrogen, Carlsbad, CA, USA) and HEC-1A cells (Cells Bank of the Chinese Academy of Science) were grown in McCoy’s 5A Medium (Invitrogen). Both culture mediums contained 10% fetal bovine serum (FBS; Gibco) and 50 mg/mL gentamycin.
Induction of progestin-resistant endometrial cancer cell lines. Briefly, Ishikawa cells were maintained in DMEM supplemented with 10% FBS and a kind of synthetic progestin MPA (Sigma, St. Louis, MO, USA) for 10 months. The concentration of MPA was increased by 2.5 μM every 4 weeks until it reached 10 μM. The medium containing MPA was changed every 2 or 3 days. When the surviving cells grew to a high density but were still less than confluent, they were subcultured by 0.02% EDTA and 0.25% trypsin prepared in Hank’s balanced salt solution. This culture method produced a subline refractory to the growth-suppressive effects of MPA. The subline was considered to be progestin-resistant (progestin-R) Ishikawa cells and was thereafter maintained in 10 μM MPA.
Sulforhodamine B (SRB) assay. Cells were incubated in phenol red-free and serum-free medium with a different concentration (0–20 μM) of MPA for 24–96 h. Prior to the administration of MPA, the cultures were pre-treated with specific PI3K inhibitor LY294002 (Sigma) for 1 h. Ten micromolar MPA was used because this dose showed the maximal inhibition effect on Ishikawa cell growth in vitro. At the indicated times, the medium was replaced with 10% (wt/vol) trichloroacetic acid and stained with SRB (Sigma) for 30 min. The dye was then washed off with 1% (vol/vol) acetic acid. The protein-bound dye was then dissolved in 10 mM Tris base solution for the optical density (OD570nm) value. Each experiment was conducted in triplicate and repeated at least three times. The MPA and LY294002 were dissolved in DMSO and prepared in culture media with a final DMSO concentration not exceeding 0.1%.
RNA extraction and real-time PCR. Cells were plated at a density of 1.5 × 106 per 6-cm-diameter dish, and total RNA was extracted from the cells after the treatment by using Trizol reagent (Sigma). Total RNA (2 μg) was reverse transcribed into cDNA in a 20 μL reaction system according to the protocol of the PrimeScriptTMRT reagent Kit (TaKaRa, Shiga, Japan). The PCR primers were synthesized by Invitrogen Bioengineering Corporation. The upstream primer 5′- GCC CTA TCT CAA CTA CCT GAG G -3′ and downstream primer 5′- GCG GAT TTT ATC AAC GAT GCA G -3′ were used to specificity amplify PR gene. Relative mRNA of PR was normalized to β-actin.
Real-time PCR analysis was conducted using the Applied Biosystems 7000 Sequence Detection System. The real-time PCR conditions were as follows: 50°C for 2 min followed by 95°C for 10 min, 40 cycles at 95°C for 15 s, and 60°C for 1 min. All real-time experiments were carried out in triplicate and at the end of the PCR the calculated cycle threshold (CT) values were exported to the SPSS (Chicago, IL, USA) software program for analysis.
Western blot analysis. Cells were plated at a density of 1.5 × 106 per 6-cm-diameter dish and collected after the indicated treatment. The collected cells were lysed in ice-cold cell lysis buffer (Tris–Cl [pH = 8.0] 50 mmol/L, NaCl 150 mmol/L, NaN 30.02%, SDS 0.1%, NP 40.1%, NaTDC 0.5%, EDTA 1 mmol/L) for plasmosin extraction or a Nuclear Extraction Kit (Sigma) for nucleoprotein. The protein extracts were separated by SDS-PAGE and transferred to polyvinylidene difluoride membranes. The membranes were incubated with antibodies against pAkt, total Akt, PR (Cell Signaling Technology, Danvers, MA, USA) overnight at 4°C, rinsed and incubated with optimal secondary antibody. Immunoreactive bands were visualized using the ECL detection system (Pierce, Rockford, IL, USA). β-actin (for plasmosin) (Kangchen, Shanghai, China) and Lamin B1 (for nucleoprotein) (Santa Cruz Biotechnology, Santa Cruz, CA, USA) were used as the loading controls.
SiRNA, plasmid constructs and transient transfections. The hPR siRNA (Genechem Biological, Shanghai, China) was transfected into Ishikawa cells to knockdown PR expression. The hPR expression plasmids were generous gifts from Dr DP McDonnell (Duke University, Durham, NC, USA). Briefly, the day before transfection, cells were plated in six-well plates and grown for 1 day when the cells had reached 80–95% confluence. Transient transfections were performed using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s protocol.
Antitumor effects of MPA and/or LY294002 in vivo. Female BALB/c mice (Shanghai Cancer Institute) used in the present study were treated according to the protocols approved by the ethical committee of Fudan University. Mice at the age of 4–6 weeks (n = 60) were injected subcutaneously with human Ishikawa cells, progestin-R Ishikawa cells or HEC-1A cells (5 × 106 in 100 μL PBS) in the right and left hind limbs. Seven days after injection, the mice of each type were then randomly divided into four groups (n = 5 mice/group) to receive either vehicle (control), MPA (100 mg/kg bodyweight), LY294002 (25 mg/kg bodyweight) or MPA plus LY294002 intraperitoneally.(23,24) The MPA was injected three times per week (nine times in total) and LY294002 twice a week (six times in total). The control mice were injected with the same volume of vector only. The mice were killed at day 28 after tumor cell transplantation. Tumor size was measured with calipers weekly, and tumor volume was calculated using the following formula: tumor volume (mm3) = (tumor length [mm] × tumor width [mm]2)/2. Bodyweights were also measured to assess the side-effects.
Statistical analysis. The data are presented as mean ± standard deviation. The statistical significance of the results was assessed by multiple anova followed by Scheffé’s post hoc tests using SPSS 11.5 software with P < 0.05 considered significant. The general linear model for repeated measures was used to test drug interaction.
Characterization of progestin sensitivity of endometrial cancer cells lines. The progestin-R Ishikawa cell subline was established by long-term MPA induction.(12) Medoxyprogesterone acetate significantly inhibited the proliferation of parental, but not progestin-R Ishikawa cells (Fig. 1a). Medoxyprogesterone acetate in various doses was then administered to three cells lines: Ishikawa cells, progestin-R Ishikawa cells and HEC-1A cells, respectively. The results showed that MPA inhibited cell growth in a dose-dependent manner in the Ishikawa cells, and the growth inhibition rates were 17.8%, 22.3% and 32.9% when treated with 1, 10 and 20 μM MPA, respectively, for 72 h (Fig. 1b). However, this inhibitory effect was not found in the progestin-R Ishikawa cells, and only slight inhibition was observed in the HEC-1A cells treated with 20 μM MPA (Fig. 1b).
Both progestin-R Ishikawa cells and HEC-1A cells were found to have significantly lower PR mRNA or protein than Ishikawa cells (Fig. 1c,d). This indicates that long-term progestin treatment could cause PR downregulation in endometrial cancer cells, and the PR level might play a key role in determining sensitivity of endometrial cancer cells to progestin.
Differential effects of progestin on the phosphatidylinositol 3-kinase pathway in progestin-sensitive and progestin-resistant endometrial cancer cells. To investigate whether the PI3K/Akt pathway was activated by progestins, we assayed the phosphorylation of Akt in response to MPA administration in three endometrial cancer cell lines, that is, Ishikawa, progestin-R Ishikawa and HEC-1A cells that exhibit different PR expression levels, as discussed above. Interestingly, MPA caused a reduction of Akt phosphorylation in the Ishikawa cells in a dose-dependent manner, while it induced Akt phosphorylation in the other two types of cells (Fig. 2). Quantification analysis showed that a 30-min MPA treatment at concentrations of 0.1, 1, 10 and 20 μM resulted in reductions of Akt phosphorylation by 19.2%, 49.3%, 51.6% and 37.8%, respectively, in the Ishikawa cells, but caused increases of Akt phosphorylation by 27.5%, 27.8%, 57.0% and 98.0%, respectively, in the progestin-R Ishikawa cells, and 34.5%, 30.3%, 70.9% and 41.4%, respectively, in the HEC-1A cells (Fig. 2).
In our investigation, the inhibition of Akt phosphorylation in the Ishikawa cells was with peak inhibition effect at 60 min. Induction of Akt phosphorylation by 10 μM of MPA treatment was obvious at 15 and 30 min in the progestin-R Ishikawa cells and the HEC-1A cells, respectively (Fig. 3).
Progestin activates the PI3K/Akt pathway in PR knocked-down endometrial cancer cells. To address whether MPA activates Akt phosphorylation through PR, we knocked down the PR expression in the Ishikawa cells using siRNA (Fig. 4). The PR siRNA successfully reduced PR mRNA by 93.1% in the Ishikawa cells (Fig. 4a). Unlike the inhibition in the PR-positive Ishikawa cells, Akt phosphorylation was increased by 169.3% in the Ishikawa cells with knocked-down PR after 30 min 10 μM MPA treatment, compared with the RNAi-neg control (Fig. 4c). Progestin activated the PI3K/Akt pathway when PR was knocked down in the Ishikawa cells. Then, MPA in various doses was administered to the PR knocked-down Ishikawa cells and we found that unlike the parental progestin-sensitive Ishikawa cells, the PR knocked-down Ishikawa cells turned to be progestin resistant (Fig. 4d).
We then transfected pcDNA PR plasmid to overexpress PR in progestin-R Ishikawa cells and HEC-1A cells (Fig. 4b). Phosphorylation of Akt in the Ishikawa cells and the HEC-1A cells, which are progestin resistant and PR overexpressed, was reduced by 17.4%and 61.2%, respectively, after 30 min of treatment with 10 μM MPA (Fig. 4c). These results confirmed that progestin activated Akt phosphorylation in a PR-independent manner in the endometrial cancer cells with very low or no PR expression, but inhibited Akt phosphorylation in a PR-dependent manner in PR-overexpressed endometrial cancer cells. We then found MPA inhibited cell growth in a dose-dependent manner in PR-overexpressed progestin-R.
Ishikawa cells and HEC-1A cells, and the growth inhibition rates were 12.8%, 26.3%, 36.4% and 14.1%, 18.7%, 28.6% when treated with 1, 10 and 20 μM MPA, respectively, for 72 h (Fig. 4d). The PR-overexpressed endometrial cancer cells appeared to regain their sensitivity to progestin.
Inhibition of the PI3K/Akt pathway restores PR expression in progestin-resistant cancer cells. Previous studies have shown that phosphorylation of Akt was involved in the regulation of PR expression in breast cancer cells.(12,14,22) We then investigated whether phosphorylation of Akt was also involved in the regulation of PR expression in endometrial cancer cells using PI3K specific inhibitor LY294002. The PR expression in progestin-R Ishikawa cells was measured at 24, 48, 72 and 96 h of treatment with 10 μM MPA in the presence of 1 μM of LY294002. It was found that the PR expression in the progestin-R Ishikawa cells was upregulated in a time-dependent manner and it increased by 180% (P < 0.05) compared with the control group at 96 h of treatment (Fig. 5a).
Inhibition of the PI3K/Akt pathway diminishes cell growth especially in progestin-resistant endometrial cancer cells. The IC50 values of LY294002 were 68.9, 4.16 and 7.59 μM in parental Ishikawa, progestin-R Ishikawa and HEC-1A cells, respectively (Fig. 5b), which meant the PI3K/Akt pathway was a much more important survival pathway in the progestin-resistant cells than in the progestin-sensitive endometrial cancer cells.
We further investigated whether inhibition of the PI3K/Akt pathway by LY294002 could restore progestin sensitivity in the progestin resistant endometrial cancer cells. The Ishikawa cells, progestin-R Ishikawa cells and HEC-1A cells were treated with 10 μM of MPA and/or 1 μM, 5 μM of LY294002 for 72 h. The combined treatment of 10 μM MPA and 1 μM LY294002, 10 μM MPA and 5 μM LY294002 induced a further 33.7%, 49.5%, 48.1% and 44.5%, 70.8%, 60.9% reduction of the proliferation of Ishikawa, progestin-R Ishikawa and HEC-1A cells, respectively (Fig. 5c). We used a general linear model to test the drug group data and found the combination of treatment of MPA and LY294002 enhanced the antitumor effect, especially on progestin-resistant endometrial cancer cells (P < 0.05), but not so significantly on the progestin-sensitive endometrial cancer cells, which indicated that inhibition of the PI3K/Akt pathway diminished cell growth, especially in the progestin-resistant endometrial cancer cells.
Inhibition of the PI3K/Akt pathway reversed progestin resistance in endometrial cancer in vivo. In the MPA-treated group, the tumor volume of Ishikawa xenograft was reduced by 29.5% compared with that of the control group, which was treated by vehicle only. No difference in tumor volume was found in either the progestin-R Ishikawa xenografts or the HEC-1A xenografts treated with MPA alone. Consistent with our in vitro study, LY294002 inhibited the growth of all three endometrial cancer cell xenografts, and the combination of LY294002 and MPA exhibited the most significant antitumor effects on both the progestin-sensitive and progestin-resistant endometrial cancer xenografts. The tumor volume of Ishikawa, progestin-R Ishikawa and HEC-1A xenografts treated with LY294002 alone was reduced by 17.8%, 36.3% and 28.9%, respectively (Table 1). Strikingly, in the group on the combinational treatment, the tumor volume in the Ishikawa, progestin-R Ishikawa and HEC-1A xenografts was significantly reduced by 44.9%, 58.7% and 38.2%, respectively (P < 0.05, Table 1, Fig. 6a,b). Accordingly, tumor weight in the Ishikawa, progestin-R Ishikawa and HEC-1A xenografts was reduced by 42.3%, 61.5% and 48.9%, respectively, after the combined treatment of MPA and LY294002 (P < 0.05, Table 1, Fig. 6c).
Table 1. Evaluation of neoplasm growth in mice treated with MPA, LY (LY294002) or a combination at the point of death (mean ± SD)
Groups (n = 5)
MPA (100 mg/kg)
LY (25 mg/kg)
MPA (100 mg/kg) + LY (25 mg/kg)
The tumor weights and volumes at the end of therapy on day 28 of the in vivo experiments are shown. Values represent the means (±SD) of n = 5. *P < 0.05 compared with the control group. MPA, medoxyprogesterone acetate.
Tumor volume (mm3)
469.62 ± 75.91
331 ± 37.33*
385.33 ± 90.52
258.6 ± 51.66*
Tumor weight (g)
1.127 ± 0.139
0.903 ± 0.057*
0.957 ± 0.050
0.650 ± 0.177*
Tumor volume (mm3)
412.03 ± 82.64
431.55 ± 63.23
262.30 ± 61.18*
170.07 ± 66.60*
Tumor weight (g)
0.617 ± 0.019
0.725 ± 0.066
0.403 ± 0.059*
0.238 ± 0.042*
Tumor volume (mm3)
605.82 ± 75.33
576.81 ± 101.01
430.85 ± 73.03*
374.13 ± 29.25*
Tumor weight (g)
1.154 ± 0.268
1.100 ± 0.108
0.908 ± 0.112
0.590 ± 0.100*
It has long been understood that progestins, which play an important role in endometrial cancer therapy by binding to nuclear PR could inhibit cell proliferation and induce apoptosis as well as benign differentiation. Progesterone receptor has been traditionally considered to act via the regulation of transcriptional processes involving nuclear translocation and binding to the progesterone response elements, and ultimately leading to the regulation of gene expression. However, novel non-transcriptional mechanisms of signal transduction of hormones in therapeutic resistance cancer cells have been identified.(10,25–30) We further showed that MPA treatment stimulated the proliferation of endometrial cancer cells with a low level of PR expression, which is consistent with the in vivo study by Hanekamp.(12,31) These results provide evidence that the proliferation effect of progestins on progestin-resistant endometrial cancer cells might be mediated by a non-transcriptional pathway other than the PR pathway.
Ubiquitous regulatory cascades such as mitogen-activated protein kinases, the phosphatidylinositol 3-OH kinase and tyrosine kinases are modulated through non-transcriptional mechanisms by steroid hormones, and it has recently been shown that Akt activation induced endocrine resistance in metastatic breast cancer and non-small-cell lung cancer.(22,27,32) A rapid increase in Akt phosphorylation was also seen with MPA treatment in progestin-resistant endometrial cancer cells in the present study (Figs 2,3). Upon MPA treatment, Akt activation was reduced in progestin-sensitive endometrial cancer cells but increased in progestin-R endometrial cancer cells (Figs 2,3). The Akt activation was not likely attributable to PR mediation because the progestin-resistant endometrial cancer cells showed very low or no expression of PR (Fig. 1c,d). Thus, we postulate that activation of the PI3K-Akt pathway by progestin without PR mediation plays an important role in progestin resistance to endometrial cancer cells.(27,33–36)
Progestin resistance is a multi-factorial phenomenon involving downregulation of PR, deregulation of the apoptotic pathways, abnormal Cox-2 and HER-2 expression, systemic insulin resistance and upregulation of the non-transcriptional mechanisms involving PI3K/Akt pathway activation.(15,16,27,37–40) Knocking down PR in MPA-treated progestin-sensitive Ishikawa cells promoted Akt activation and induced progestin resistance, whereas overexpressing PR in progestin-resistant endometrial cancer cells reduced Akt activation and regained their sensitivity to progestin (Fig. 4). According to these results, we hypothesize that both PR-mediated classical and the non-genomic PI3K/Akt pathway of progestin exist independently in endometrial cancer cells. The proliferation-inhibiting effect of progestin mediated by the classic PR pathway was obvious in cells of high PR expression, while the proliferation-promotion effect of progestin mediated by the non-genomic PI3K/Akt pathway was efficient in cells of low or no PR expression.
Without the inhibitory effect of the classical PR pathway, the survival of progestin-resistant endometrial cancer cells largely depended on the PI3K/Akt pathway. The PI3 kinase inhibitor LY294002 inhibits PI3 kinase-dependent Akt phosphorylation and kinase activity, and greatly potentiates chemotherapy-induced apoptosis in cells with high Akt levels, but not in cells with low Akt levels.(32,41) Thus, the antitumor activity of LY294002 is more significant in progestin-R endometrial cancer cells than in progestin-sensitive endometrial cancer cells.
The results of the present study strongly suggest that the PR level should be taken as a crucial criterion by physicians in evaluating whether their patient needs progestin treatment.(42,43) Loss of PR gene expression has been attributed to loss of heterozygosity, loss of ER function and methylation of a CpG island in the PR promoter.(44) Our previous findings provide implications for another theory, in which potent growth factor signaling, especially the highly activated TGF-EGFR signaling, contributes to PR regulation.(12) The results are consistent with other key areas of study on HER2 and insulin-like growth factor-I and PR regulation in breast cancer.(22,45,46) Long-term progestin treatment might downregulate PR expression through utilizing the proteasome and inhibiting PR gene transcription, thereby inducing progestin resistance in PR-positive endometrial cancers cells. But in PR-negative endometrial cancer cells, the regulation of PR expression is mostly mediated through a PI3K-dependent pathway. In the presence of LY294002, MPA treatment restored PR expression in progestin-R Ishikawa cells in the present study, which indicates that the PI3K/Akt pathway might be the downstream pathway responsible for the growth factor regulation of PR in progestin-resistant cancer cells.
Our result indicates that inhibiting the PI3K/Akt pathway might not only get strong antitumor activity, but also restore PR expression and activate the classical PR pathway in progestin-R cancer cells. Thus, progestin-R endometrial cancer cells, which possibly have highly active growth factor signaling and PI3K/Akt signaling, respond better to the combination group of MPA and LY294002 than the progestin-sensitive endometrial cancer cells in vitro and in vivo (Figs 5,6). Inhibiting the PI3K/Akt pathway could reverse progestin resistance in endometrial cancer.
In conclusion, the present study suggests that low or absent PR in endometrial cancer cells is associated with a poor response to progesterone therapy. Long-term progestin treatment downregulates PR expression in progestin-sensitive endometrial cancer cells and leads to progestin resistance. In addition, progestin could activate the PI3K/Akt pathway independent of PR mediation in PR-negative or low-expression endometrial cancer cells. Inhibition of the PI3K/Akt pathway could restore PR expression and progestin sensitivity in progestin-resistant endometrial cancer. The combination treatment of progestin and PI3K inhibitor might be an effective conservative regimen for endometrial cancer patients who unfortunately developed progestin resistance but hoped to preserve their ability to remain fertile.
This study was supported by the National Natural Science Foundation of China (NSFC No. 30900901); Shanghai Leading Academic Discipline Project, Project no. B117; Science and Technology Commission of Shanghai Municipality, Project no. 09QA1400800; the Youth Research Program of Shanghai Health Bureau, Project no. 2009Y015; and the Open Project of Biomedical Research Institute, Fudan University, Project no. IBS0834.