Elevated MiR-222-3p Promotes Proliferation and Invasion of Endometrial Carcinoma via Targeting ERα

MicroRNAs play key roles in tumor proliferation and invasion. Here we show distinct expression of miR-222-3p between ERα-positive and ERα-negative endometrial carcinoma (EC) cell lines and primary tumors, and investigation of its relationship with ERα and other clinical parameters. In vitro, the function of miR-222-3p was examined in RL95-2 and AN3CA cell lines. MiR-222-3p expression was negatively correlated with ERα. Over-expressed miR-222-3p in RL95-2 cells promoted cell proliferation, enhanced invasiveness and induced a G1 to S phase shift in cell cycle. Furthermore, the miR-222-3p inhibitor decreased the activity of AN3CA cells to proliferate and invade. In vivo, down-regulated miR-222-3p of AN3CA cells inhibited EC tumor growth in a mouse xenograft model. Additionally, miR-222-3p increased raloxifene resistance through suppressing ERα expression in EC cells. In conclusion, miR-222-3p plays a significant role in the regulation of ERα expression and could be potential targets for restoring ERα expression and responding to antiestrogen therapy in a subset of ECs.


Introduction
EC is the most common malignancy of the female genital system, with estimated 8,190 deaths in the USA in 2013 [1]. EC is often associated with excessive estrogen exposure, and often coexist with, or is preceded by endometrial hyperplasia [2]. In China, the incidence of EC has been increasing with a shift to younger population due to multiple factors, such as obesity and lifestyle changes [3,4]. Upon initial diagnosis, as 28% of the patients already have regional or distant metastasis [5]. However, the etiology of EC remains unclear. Current treatments for EC (including surgery, chemotherapy, and radiation therapy) always produce significant side effects. To date, our knowledge of this disease is still quite limited. Therefore, further elucidation of the molecular mechanisms during endometrial cancer is urgently required.
Estrogen is a classical etiological factor for EC and the prooncogenic effect of estrogen is mediated primarily by estrogen receptor alpha (ERa) activation of target genes that promote cell proliferation or decrease apoptosis [6]. ERa is an important marker for prognosis and is predictive of response to endocrine therapy in patients with EC [7]. The most common basis for determining risk of recurrent disease is the categorization into type I and II endometrial carcinoma [8]. Type I tumors are most frequent, characterized by endometrioid histology, low histologic grade, low Federation Internationale des Gynaecologistes et Obstetristes (FIGO) stage, and favorable prognosis. In contrast, type II tumors are characterized by nonendometrioid histology, high histologic grade and stage, and a tendency to recur, even when treated at early stage [9]. The molecular basis for type I and II cancers is only partially understood. Hyperestrogenic risk factors and positivity for ERa are common in type I in contrast to type II cancers, and ERa status is reported to be a prognostic marker in endometrial cancer [10,11]. Decreased or absent expression of ERa is associated with extensive invasion, disease progression and poor prognosis [12]. Understanding the basis for ERa discrepancies between EC and normal endometrium will hopefully improve the understanding of endometrial carcinogenesis and facilitate research and development of novel therapy.
MicroRNAs (miRNAs) are small noncoding RNAs that silence of their cognate target genes by either degrading mRNAs or inhibiting their translation [13]. As such, miRNAs are implicated in the regulation of a variety of cellular processes, including stemness and metastasis; also, miRNAs could function as either oncogenes or tumor suppressors [14][15][16]. MiR-222-3p overexpression facilitate the growth, metastasis and invasion of a variety of malignant tumors, including breast cancer [17,18], lung cancer [19], colorectal carcinoma [20], and melanoma [21] via genetic or epigenetic mechanisms.
In our previous study using microRNAs microarray, we found that miR-222-3p significant overexpression in ERa-negative EC cells (vs. ERa-positive EC cells). Here, we represent a comprehensive analysis of miR-222-3p expression in atypical hyperplasia, clinical EC tumor samples and normal endometrium. We investigated the regulatory effects of miR-222-3p on ERa, and explored the potential therapeutic value of miR-222-3p in EC cell lines through functional analysis. These findings provided new insights into the invasive mechanisms in EC, and encouraged exploring miR-222-3p as a target for intervention.  [22]. The local ethics committee approved the research project. Informed consent for the experimental use of the surgically removed samples was obtained from all patients. None of the patients received hormone therapy, radiotherapy, or chemotherapy prior to the sample collection. Following excision, tissue samples were immediately snap-frozen in liquid nitrogen and stored at 280uC until RNA extraction. Clinical and pathological data are presented in Table 1.

Immunoblot analysis
For each of three independent experiments, 60-mg total protein extract was separated on 10% SDS-PAGE gels and transferred to PVDF membrane. The levels of ERa expression were evaluated by using a rabbit monoclonal anti-ERa antibody (Epitomics, 1:2000 dilution). As a loading control, b-actin expression levels were measured using mouse monoclonal anti-actin antibody (Proteintech Group Inc., 1:2000 dilution). The secondary horseradish peroxidase-conjugated antibodies (Santa Cruz, goat antirabbit, 1:1000 dilution; Proteintech Group Inc., goat anti-mouse, 1:4000 dilution) were detected using ECL Plus Western blotting detection reagents (Amersham Biosciences). Bands were quantified with ImageJ 1.34 software.

Cell invasion assay
Cell invasion was evaluated using transwell chamber assay (Millipore, Billerica, MA, USA) according to the manufacturer's instruction. For invasion assay, totally 5610 4 of RL95-2 and AN3CA cells were seeded on an 8-mm pore size transwell insert coated with extracellular matrix (ECM) (1:6) (BD Biosciences). After incubated at 37uC for 48 h, the cells adherent to the upper surface of the filter were removed using a cotton applicator, then stained with crystal violet, and the values obtained were calculated by averaging the total numbers of cells from triplicate determinations.

Construction of reporter plasmids and luciferase assays
The 39 UTR of the human ERa gene was PCR amplified using the following primers: 59-TCAGAG-cctattgttggatattgaatgacagacaatcttatgtagcaaagattatgcctgaaaagggatcc-GC-39 (forward) and 39-GGCCGC-ggatcccttttcaggcataatctttgctacataagattgtctgtcattcaatatccaacaatagg-C-59 (reverse) and cloned between the Xho1 and Not1 sites of the psiCHECK 2 Vector (Promega, Madison, WI, USA), giving rise to the p39UTR-ERa plasmid. The construct was used to generate, by inverse PCR, the p39UTRmut-ERa plasmid

Tumorigenicity assays in nude mice
The Ethic Committee for Animal Experimentation of Shanghai Jiaotong University approved all experimental protocols. A total of 5610 6 cells suspended in 100 ml 16 PBS were injected into subcutaneously to the interscapular area in 4-week old female BALB/C athymic nude mice. A group of mice (n = 4) received AN3CA cells (1610 7 ) transfected with LV-has-miR-222-3p-down (miR-222-3p inhibitor in lentivirus vector; LV-miR-222i). Another negative control group received AN3CA cells (1610 7 ) transfected with LV-has-miR-222-3p-down NC (lentivirus vector alone; LV-miR-222i NC). The size of tumors was measured weekly for 4 weeks. Mice were sacrificed at 30 days post-injection. Tumors were excised and measured. Tumor volume (cm 3 ) was calculated by using the following formula: (the longest diameter)6(the shortest diameter) 2 60.5.

Statistical analysis
All data are expressed as mean 6 standard deviation (SD). Statistical evaluation of the data was performed with one-way ANOVA. Pair-wise comparisons were conducted by Student's t test. The P value less than 0.05 were considered statistically significant. All analyses in the study were evaluated with SPSS version 17.0 software (Chicago, IL, USA).
Furthermore, cell cycle analyses indicated that RL95-2 cells overexpressing miR-222-3p had a significant increase in S phase population, as compared with miR-222m NC cells, with a concomitant decrease of the G1 portion (Fig. 3A). On the contrary, inhibiting miR-222-3p resulted in an accumulation of cells in G0/G1 phase and a decrease in S phase in AN3CA cells (Fig. 3A).
Furthermore, representative H&E staining was shown in Fig. 5E. The expression of Ki67, a measure for tumor cell proliferation, was decreased by LV-miR-222i in xenograft tumor (Fig. 5E). IHC results also showed that miR-222-3p knockdown lightly increased PTEN and TIMP3 expression in vivo (Fig. S1). Additionally, ERa protein expression in the xenograft was not affected by LV-miR-222i (Fig. 5E).

Discussion
EC remains a major cause of cancer-related morbidity and mortality among women [23]. Worldwide, it is estimated that nearly 200,000 new cases of endometrial cancer are diagnosed annually [24]. In North America and Europe, EC is the most common gynecologic malignancy [25].
Estrogen is a classic etiological factor for endometrial tumorigenesis [26]. Despite the normal and beneficial physiological actions of endogenous estrogen in women, abnormally high estrogen levels are associated with the increased incidence of certain types of cancer, especially those of the breast and endometrium [6]. The binding of estrogen to ERs induces conformational changes in protein structure that allow receptor dimerization and interaction with coactivators. The pro-oncogenic effect of estrogen is mediated primarily by ERa activation of target genes that promote cell proliferation or decrease apoptosis [6].
In EC, deregulated ERa caused by genomic or epigenetic aberrations was a prevalent phenomenon, which reduced the expression of ERa and associated with high stage and poor differentiation [11,27]. Recent studies indicated that miRNAs are atypically expressed in virtually all cancers, including ECs [28]. Leivonen reported that five ERa-regulating miRNAs (e.g. miR-18a, miR-18b, miR-193b, miR-302c, and miR-206) directly targeted ERa in the 39UTR [29]. Dampening of the ERa signaling by let-7 miRNAs inhibited cell proliferation and subsequently triggered the cell apoptotic process in MCF7 cells [30]. Other studies have demonstrated that miR-22 overexpression leads to a reduction of ERa level, at least in part by inducing mRNA degradation, and compromises estrogen signaling, as exemplified by its inhibitory impact on the ERa-dependent proliferation of breast cancer cells [31,32]. In breast cancer, miR-222-3p directly repressed ERa and knockdown of miR-222-3p sensitized MDA-MB-468 cells to tamoxifen-induced cell growth arrest and apoptosis [17].
In this study, we tested 75 cases of EC samples, and demonstrated up-regulation of miR-222-3p in ERa-negative EC tissues. Also, miR-222-3p overexpression is correlated to higher grades, later stages and more nodal metastasis. Furthermore, miR-222-3p expression was significantly higher in ERa-negative cells, AN3CA and KLE, than in those of ERa-positive cells. By ectopic expressing miR-222-3p, the potential of cell proliferation and invasion was apparently enhanced in RL95-2 cells. In vivo, dampening of miR-222-3p could significantly inhibit tumor growth. Unexpectedly, we did not detect significant upregulation of ERa protein upon LV-miR-222i treatment.
In a previous report, Zhao et al. have demonstrated that ERa is suppressed by miR-221 and miR-222-3p [17]. In the current study, we found that miR-222-3p also inhibits ERa expression in EC cell lines. Unlike Zhao's study, we found that miR-222-3p inhibited ERa expression at both protein and mRNA level in EC  analyzed by qRT-PCR and western blot method, respectively. ERa levels decreased when miR-222-3p was upregulated in response to the miR-222m in RL95-2 cells, whereas the reverse was observed for ERa expression when miR-222-3p was knocked down in AN3CA cells. (D) The expression of pS2, cyclin D1 and PR was down regulated after miR-222m transfection in RL95-2 cell lines. Oppositely, in AN3CA cells, these ERa downstream genes were increased with miR-222-3p dampened. * P,0.05, ** P,0.01, *** P,0.001, **** P,0.0001. doi:10.1371/journal.pone.0087563.g004 cells. The downstream genes of ERa, including PR, cyclinD1 and pS2, were also inhibited after miR-222-3p ectopic expression in our study.
Since RL95-2 cells enhanced invasive potential after miR-222-3p upregulated, we found that expression of MMP-2 and MMP-9 was increased. In several malignancies, MMPs have been linked to aggressive behavior, and gelatinases (MMP-2 and MMP-9) in particular are prognostic factors in EC [39,40]. These results indicated that miR-222-3p could enhance invasive potential of ECs via promoting MMP-2 and MMP-9 secretion.
Besides the oncogenic role of miR-222-3p in vitro, tumor formation assay confirmed that decreased miR-222-3p expression could inhibit the proliferation of EC cells in a mouse xenograft model. At 1 week after injection, there were tumors in the interscapular area of mice; from 2 weeks, volume of tumors in LV-miR-222i was much smaller than LV-miR-222i NC, and in the following weeks the differences became much apparent. At 4 weeks, volume of tumors in LV-miR-222i NC was nearly 100 fold than that of LV-miR-222i. Higher proliferation in cells treatment with LV-miR-222i NC was also evident in immunohistochemical staining of Ki67. Moreover, the expression of PTEN and TIMP3 was increased in LV-miR-222i lightly. Interestingly, we did not detected apparent upregulation of ERa protein upon LV-miR-222i treatment. Given that many other mechanisms like single nucleotide polymorphism [41] and promoter hypermethylation [42] were involved, we hypothesize miR-222-3p overexpression was one of the reasons for ERa loss in AN3CA cells. Taken together, these results indicate that miR-222-3p is a crucial oncogene and may be an important determinant of ERa status in EC.
Except these routine therapies, SERMs (selective oestrogenreceptor modulators) were another choice for EC patients, which could bind the ER and modulate ER-mediated gene transcription [43]. In general, patients with ERa-positive respond favorably to SERMs; however, loss of ERa often showed SERMs resistance. As miR-222-3p directly inhibited ERa protein expression, we then further explored whether alterations of miR-222-3p have effects on cellular reaction to raloxifene (a SERM already in clinical use) in EC cells. We found that increased miR-222-3p induced resistance to raloxifene in RL95-2 cells, while down-regulation of miR-222-3p restored sensitivity of AN3CA cells to raloxifene via promoting notably cell apoptosis. Our results demonstrated that miR-222-3p overexpression was a novel mechanism for raloxifene resistance in EC patients.
In summary, our findings confirmed proto-oncogenic role of miR-222-3p. MiR-222-3p was overexpressed in ERa-negative EC tumors and was associated with high grade, late stage and nodal metastasis. Up-regulating miR-222-3p promoted cell proliferation, enhanced invasiveness and induced a G1 to S phase transition. Down-regulated miR-222-3p of AN3CA cells inhibited EC tumor growth in a mouse xenograft model. High level of miR-222-3p was a mechanism for raloxifene resistance in EC therapy. Therefore, miR-222-3p could serve as potential therapeutic targets for a subset of ERa-negative ECs and might be developed as a biomarker for EC.  in tumor tissues, as compared with LV-miR-222i NC transfected group. All experiments were repeated at least three times. (TIF)