Overcoming cisplatin resistance in osteosarcoma through the miR-199a-modulated inhibition of HIF-1α

Abstract Dysregulation of miRNAs has been shown to contribute to multiple tumorigenic processes, as well as to correlate with tumour progression and prognosis. miR-199a has been shown to be dysregulated in multiple tumour types. However, the association between miR-199a and the chemoresistance features of osteosarcoma are not well understood, the target genes for miR-199a and the regulatory mechanisms are also unknown. In the present study, we demonstrated that miR-199a is expressed at low levels in osteosarcoma cells and patient samples. By the selection and establishment of cisplatin resistant osteosarcoma cell line, we observed a correlation between miR-199a and cisplatin resistance in osteosarcoma cells: resistant cells exhibit attenuated miR-199a expressions and exogenous overexpression of miR-199a sensitizes osteosarcoma cells to cisplatin. Moreover, we identified HIF-1α as a direct target for miR-199a. Intriguingly, cisplatin resistant osteosarcoma cells display significantly elevated HIF-1α expression under hypoxia. We report here overexpression of miR-199a resensitizes cisplatin resistant cells to cisplatin through inhibition of HIF-1α in vitro and in vivo. Finally, by analysing the clinical osteosarcoma patient samples, we demonstrate a reverse correlation between miR-199a and HIF-1α mRNAs. Our study will provide mechanisms for the miRNA-mediated anticancer therapy and miR-199a may be considered a promising therapeutic agent for osteosarcoma patients who fail to respond to conventional chemotherapy.


Introduction
miRNAs are a class of small non-coding, single-stranded endogenous RNA fragments that repress target protein translation by base pairing to the 3 -untranslated mRNA region [1,2]. miR-NAs have emerged as important regulators involved in multiple cellular processes such as cell proliferation, apoptosis and autophagy [3,4]. The effects of miRNAs on tumorigenesis and cancer progression have been revealed that miRNA expressions are significantly up-or down-regulated in tumour tissues, resulting in a prominent diagnostic and prognostic tool [5,6]. Moreover, miR-NAs have been reported to be involved in drug resistance. One study showed that miR-125b is correlated with Taxol-resistant breast cancer cells and inhibition of miR-125b could resensitize Taxol-resistant cells [7]. Another study illustrated that miR-34a overexpression results in attenuated chemoresistance to the camptothecin by targeting SIRT1 gene [8]. Therefore, elucidating the critical functions of miRNAs in tumour progression is important to discover novel antitumour agents.
Osteosarcoma which arises from mesenchymal cells is the most common form of malignant bone tumour and occurs predominantly in adolescents and young adults [9]. Currently, a combination of therapies that include surgery, chemotherapy and radiation therapy is the most common therapeutic plan for osteosarcoma [10]. However, patients with recurrent or metastatic osteosarcomas still have very poor prognosis [11]. Among the antitumour agents, although cisplatin is one of the most effective drugs against osteosarcoma, outcomes remain poor for most patients who developed chemoresistance [12]. Thus, understanding the molecular mechanisms involved in the chemoresistance of osteosarcoma cells is a critical stage to improve the treatments.
miR-199a has been described to be deregulated after cisplatin treatment in cancer cell lines [13], suggesting that it may be a therapeutic target. However, the molecular mechanisms underlying this process are not well understood and the functions of miR-199a in osteosarcoma have not been clearly illustrated. In the present study, we focus on the roles of miR-199a in the cisplatin sensitivity in osteosarcoma. Moreover, the downstream target of miR-199a will be identified and we will investigate whether overexpression of miR-199a could sensitize cisplatin resistant osteosarcoma cells. We supposed that miR-199a could be a promising therapeutic agent in osteosarcoma.

Materials and methods
A total of 20 osteosarcoma patient samples and their matched adjacent normal tissues were obtained from osteosarcoma patients at the Department of Orthopedic, First People's Hospital in Kashgar area (Xinjiang, People's Republic of China). The study was approved by the ethics committee of Department of Orthopedic, First People's Hospital in Kashgar area (Xinjiang, People's Republic of China). Written informed consents were obtained from all the patients. Tissue samples were collected at surgery, immediately frozen in liquid nitrogen and stored until analysis in the present study.
The human osteosarcoma cell lines (MG-63, U-2OS and SaoS-2) were purchased from the American Type Cell Culture Collection (Manassas, VA, U.S.A.). Human normal osteoblast cell line NHOST, hFOB and HOB were purchased from the Cell Bank of Chinese Academy of Sciences (Shanghai, China) and grown in Dulbecco's Modified Eagle's Medium (DMEM, Gibco, U.S.A.) supplemented with 10% FBS (Sigma, U.S.A.), 100 units/ml of penicillin streptomycin (Invitrogen, Carlsbad, CA). Cells were cultured in a humidified incubator at 37 • C with 5% CO 2 and 95% air. Hypoxia was performed in hypoxic chamber that was filled with gas mixture of 95% N 2 and 5% CO 2 for 24 h, then returned to regular 37 • C, 5% CO 2 and 95% air and regular cell culture medium.

Antibodies and reagents
Rabbit monoclonal antibody anti-HIF-1α was purchased from Cell Signaling (#14179, Danvers, MA, U.S.A.); mouse monoclonal antibody anti-β-actin was purchased from Cell Signaling (#3700, Danvers, MA, U.S.A.). Cisplatin was purchased from Sigma-Aldrich (Shanghai, China) and stored as 20 mM solution in ddH 2 O at -20 • C and diluted with DMEM medium prior to use.

Selection of cisplatin resistant cell line
The selection of cisplatin resistant osteosarcoma cells was performed according to a recent report [14]. Briefly, SaoS-2 parental cells were treated with increased concentrations of cisplatin from 0.5-25 μM for the selection of survival cells. The treatments continued for 3 months and survival clones were pooled for the following assays in the present study.

Real-time PCR
Total RNA was extracted from osteosarcoma cells by the TRIzol method as recently described [14]. The reverse transcription reaction was performed to reverse transcribe RNA into cDNA using an oligo (dT) primer from the miScript II RT Kit (Qiagen, Hilden, Germany). A total of 50 ng cDNA of each samples was added up to 25 ml per reaction. qPCR assays were carried out using a StepOnePlus sequence detection system. The cycling conditions were as follows: 10 min of polymerase activation at 95 • C, followed by 40 cycles at 95 • C for 15 s and 60 • C for 60 s. U6 was used as an internal control. For detection of the HIF-1α mRNA, PCR amplification consisted of 40 cycles of denaturation at 95 • C for 10 s, annealing at 55 • C for 10 s and extension at 72 • C for 20 s using the following primers: HIF-1A, sense: 5-GAAAGCGCAAGTCTTCAAAG-3, antisense: 5-TGGGTAGGAGATGGAGATGC-3; GAPDH, sense: 5-AATCCCATCACCATCTTCCA-3, antisense: 5-TGGACTCCACGACGTACTCA-3. Experiments were repeated three times using the 2 -C T method.

miRNAs and plasmid DNA transfection
miRNA mimics (miR-199a mimic) and negative control miRNA mimics were purchased from GenePharma (Shanghai, China). Lipofectamine 2000 transfection reagent (Invitrogen) was used for the transfection of miRNAs. miR-199a or negative control mimic was transfected at 25 nM. Forty-eight hours after transfection, cells were collected for the downstream experiments. The expression of miR-199a was detected by real-time PCR. HIF-1α overexpression vector was transfected using the Lipofectamine 2000 transfection reagent (Invitrogen). Forty-eight hours after transfection, cells were collected for the downstream experiments. The experiments were repeated three times.

Luciferase assay
The luciferase assay was performed according to a previous report [13], Briefly, cells were seeded in a 24-well plate at a density of 5 × 10 4 /well and cultured overnight. The pGL3-luciferase reporter gene plasmids pGL3-HIF-1A-3 -UTR (WT), pGL3-HIF-1A-3 -UTR (mutant) or the control-luciferase plasmid were cotransfected into the cells with the control mimic or miR-199a mimic using Lipofectamine 2000 Reagent (Invitrogen) according to the manufacturer's protocol. Dual-luciferase activity assays were assayed 48 h after transfection.

Cell viability assay
Cell viability was evaluated by MTT. Briefly, 1 × 10 4 cells were plated in 96-well tissue culture plates for overnight. Cells were then exposed to drugs under normoxic or hypoxic conditions. After that, MTT (Sigma, U.S.A.) was added directly into cells and the generated formazan was dissolved in DMSO and the absorbance was recorded. All the experiments were performed in triplicate and repeated three times.

Western blot
The cellular lysates were prepared by RIPA buffer (Thermofisher Scientific Inc., Waltham, MA). Protein concentration was determined using the Bradford assay (Thermofisher Scientific Inc., Waltham, MA). Proteins were resolved on SDS/PAGE (10% gel) and transferred to immobilon PVDF membranes. Membranes were blocked with 4% BSA for 1 h at room temperature and incubated with the following primary antibodies overnight at 4 • C: mouse monoclonal anti-β-actin, (Santa Cruz Biotechnology) or rabbit polyclonal anti-HIF-1α (Santa Cruz Biotechnology). After washing completely in TBS with 0.05% Tween 20 (TBS-Tween), the blots were subsequently incubated with a donkey antirabbit or donkey antimouse peroxidase-conjugated secondary antibody for 1 h at room temperature. The blots were visualized by chemiluminescence methods using the Pierce ECL Western Blotting Substrate (Thermofisher Scientific Inc., U.S.A.).

Mouse xenograft models
The xenograft experiments were performed as previously described [15]. Osteosarcoma cells were injected subcutaneously into the right flanks of the nude mice. Three weeks after injection, the subcutaneous tumour size had reached a tumour volume of approximately 100 mm 3

Statistics
The values given are mean + − S.E.M. The significance of difference between the experimental groups and controls was assessed by Student's t test. The difference was considered significant with the P-value was less than 0.05.

miR-199a displays tumour suppressor roles in osteosarcoma cells
To investigate the potential roles of miR-199a in osteosarcoma, we compared the expressions of miR-199a in normal osteoblast cells and osteosarcoma cells. As we expected, all three osteosarcoma cell lines showed significantly down-regulated miR-199a levels compared with non-tumorous osteoblast cells ( Figure 1A), suggesting miR-199a might be a tumour suppressor in osteosarcoma. To explore whether miR-199a is down-regulated in osteosarcoma tumour tissues, we compared the miR-199a expression levels in 20 osteosarcomas and their adjacent normal tissues. Consistently, we observed miR-199a levels were significantly down-regulated in 20 cases of osteosarcoma samples compared with normal tissues by TaqMan real-time PCR ( Figure 1B). Taken together, our results detected miR-199a expressed lower levels in osteosarcoma, suggesting a tumour suppressor role of miR-199a.

Overexpression of miR-199a suppresses cell proliferation and motility
We next investigated whether exogenous overexpression of miR-199a could affect the osteosarcoma cellular processes. To assess the effect of miR-199a on the cell growth and motility of osteosarcoma cells, we transfected miR-199a mimics into two osteosarcoma cell lines, SaoS-2 and MG63 ( Figure 2A). Figure 2B showed in both Saos-2 and MG63 cell lines, cells with miR-199a overexpressing demonstrated significantly suppressed proliferation after 72 h compared with cells with control miRNAs transfection. Moreover, we observed a significant decrease in the number of cell migrations in SaoS-2 cells transfected with miR-199a after 36-h scratch. (Figure 2C), indicating miR-199a might be selected as a target to treat osteosarcoma.

Cisplatin treatments inhibits miR-199a and cisplatin resistant osteosarcoma cells show down-regulated miR-199a
As we described above, cisplatin remains the second most commonly used chemotherapy for high-grade osteosarcoma [10,11]. Therefore, we focused on the functions of miR-199a in the modulation of cisplatin sensitivity in osteosarcoma. To determine whether miR-199a levels are correlated with chemotherapeutic response, we identified the expressions of miR-199a before and after cisplatin-based chemotherapy. We found that miR-199a expression levels were significantly decreased in SaoS-2 or MG63 cells after gradient cisplatin treatments ( Figure 3A), suggesting that miR-199a involves the modulation of cisplatin sensitivity in osteosarcoma cells.
To assess the directly regulatory effects of miR-199a on cisplatin sensitivity in osteosarcoma cells, we established cisplatin resistant cell line from the SaoS    of miR-199a was down-regulated in SaoS-2 cisplatin resistant cells, ( Figure 3C). These data indicate that inhibition of miR-199a may be applied to the cisplatin-based chemotherapy.

HIF-1α is correlated with the miR-199a-molulated cisplatin sensitivity
Since, a previous study demonstrated that HIF-1α is the master regulator during hypoxia response, which leads to cancer cells' chemoresistance via provoking multiple adaptive responses [17]. To investigate the underlying mechanisms for the miR-199a-modulated cisplatin sensitivity in osteosarcoma cells, we compared the expressions of HIF-1α in SaoS-2 parental and cisplatin resistant cells. It was interesting to find that HIF-1α is significantly up-regulated in cisplatin resistant cells at both protein ( Figure 4A) and mRNA ( Figure 4B) levels, intriguing us to explore the correlation of HIF-1α and miR-199a in the regulation of cisplatin sensitivities of osteosarcoma cells. As we expected, HIF-1α was induced under hypoxia in SoaS-2 and MG-63 cells ( Figure 4C). In contrast, miR-199a was suppressed under hypoxia ( Figure 4D). These data illustrated the miR-199a-modulated cisplatin sensitivity is correlated with HIF1-α expression.

HIF-1α is a direct target of miR-199a in osteosarcoma cells
We next investigated whether there is direct regulation between miR-199a and HIF-1α in osteosarcoma cells. By analysing miRNA target prediction public databases (TargetScan), we noticed that the 3 -UTR mRNA of HIF-1α contains a highly conserved binding site for miR-199a ( Figure 5A). MG-63 and SaoS-2 cells were transfected with control miRNAs or miR-199a mimic. Results showed the expressions of HIF-1α were inhibited by exogenous overexpression of HIF-1α under hypoxia ( Figure 5B). To investigate whether miR-199a could directly target on 3 -UTR of HIF-1α mRNA, sequential replacement of seven base pair region from position 31 to 37 of HIF-1α 3 -UTR was performed to produce the mutant vector. SoaS-2 and MG63 cells were cotransfected with a vector containing pMIR reporter-luciferase fused with original sequence or predicted binding site mutant of the 3 -UTR of HIF-1α mRNA and miR-199a mimics or control miRNAs. Fluorescence intensity was detected at 24 h post-transfection. The results showed that the luciferase activity of the luciferase reporter gene with the 3 -UTR of HIF-1α mRNA was significantly decreased by 66.7%, but that there was no difference in luciferase activity in the cotransfection of vector containing mutant of the 3 -UTR of HIF-1α mRNA and miR-199a mimics ( Figure 5C). These results indicated that HIF-1α expression was significantly inhibited by miR-199a through the direct binding to the 3 -UTR region of HIF-1α mRNA.

Overexpression of miR-199a sensitizes cisplatin resistant osteosarcoma cells through inhibition of HIF-1α in vitro and in vivo
To study whether targeting miR-199a could resensitize cisplatin resistant osteosarcoma cells to cisplatin, we transfected SaoS-2 cisplatin resistant cells with control miRNAs or miR-199a mimics then treated these cells with cisplatin at 0, 6, 12 and 24 μM. Cell viability was assayed to show that cisplatin treatments markedly inhibited viabilities of cells with miR-199a-mimics transfection compared with control miRNAs transfection ( Figure 6A). Forced expression of miR-199a into Saos-2 cisplatin resistant cells decreased the cisplatin IC 50 from 31.51 to 8.85 μM. To assess whether the miR-199a modulated sensitization of cisplatin resistant cells was through the inhibition of HIF-1α, we transfected HIF-1α overexpression vector into miR-199a pre-transfected SaoS-2 cisplatin resistant cells to restore the HIF-1α expression to the original level. As we expected, restoration of HIF-1α expression in miR-199a overexpressing cells led to a significant tolerance to cisplatin compared with the transfection with control vector in miR-199a overexpressing cisplatin resistant cells ( Figure 6B), confirming that overexpression of miR-199a sensitizes cisplatin resistant osteosarcoma cells to cisplatin by the inhibition of HIF-1α.
Based on the potent anticisplatin resistance activity of miR-199a observed in in vitro assays, we investigated the roles of miR-199a in the regulation of chemosensitivity in an in vivo model. The antitumour growth potency was investigated in a subcutaneous mouse model with inoculation of SaoS-2 cells with or without transfection of miR-199a mimics into the mammary fat pads of 6-week-old nude mice. Consistent with the in vitro results, the miR-199a mimic treated mice exhibited a significant reduction in tumour growth ( Figure 7A). Moreover, the expression of HIF-1α was down-regulated in miR-199a mimics transfected cells derived tumours compared with control miRNAs transfection ( Figure 7B). To test whether miR-199a could promote the sensitivity of cisplatin resistant osteosarcoma cells, we treated tumour xenografts with cisplatin (3 mg/kg, two times/week) or PBS to mice with injection of SaoS-2 cisplatin resistant cells with or without miR-199a transfection once tumours were 100 mm 3 in size. Our results demonstrated that compared with the control group and miR-199a mimic alone group, the combination of miR-199a transfection plus cisplatin treatment synergistically inhibited cisplatin resistant cell derived tumour growth ( Figure 7C). Remarkably, the expression of HIF-1α was also down-regulated in miR-199a mimics transfected cells derived tumours compared with control miRNAs transfection in xenograft tumours ( Figure 7D). In summary, our xenograft mouse model revealed that miR-199a could sensitize cisplatin resistant osteosarcoma cells through inhibition of HIF-1α in vivo.

Reverse correlation between miR-199a and HIF-1α in osteosarcoma patient samples
The above results demonstrated that HIF-1α is a major target of miR-199a in osteosarcoma cell lines and in vivo model, we next investigated the correlation between miR-199a and HIF-1α expressions in osteosarcoma tissues. We examined HIF-1α mRNA expressions in 45 osteosarcoma specimens with immunohistochemical staining. Representative images of HIF-1α expressions in Figure 8A showed that HIF-1α was down-regulated in miR-199a high expressed osteosarcoma tissues and highly expressed in miR-199a low level osteosarcoma tissues. Statistically, of the 17 osteosarcoma cases with elevated miR-199a, 14 (82.36%) of them had low levels of HIF-1α, and 25 of 28 (89.29%) cases with down-regulated miR-199a presented high levels of HIF-1α ( Figure 8B). These findings illustrated that miR-199a regulates HIF-1α expression in clinical osteosarcoma patient specimens.

Discussion
As we discussed above, miR-199a has been reported to be deregulated and negatively correlated with multiple cancer types. In the present study, we found that miR-199a levels were significantly decreased in osteosarcoma cancer cell lines and patients. Moreover, forced expression of miR-199a suppressed osteosarcoma cell proliferation, consistently with the previous study [18]. Since an increasing number of osteosarcoma patients develop resistance to chemotherapy drugs, we next focused on cisplatin sensitivity of osteosarcoma cells.
miRNAs are found to be novel modulators of cisplatin sensitivity. Galluzzi et al. [19] reported that miR-181a could enhance the cisplatin-triggered cell death in A549 cells and miR-630 could reduce the cisplatin sensitivity in the  same cells. However, the precise roles of miR-199a in cisplatin resistance in osteosarcoma cells have not been well elucidated. A previous study illustrated that cisplatin treatments decreased miR-199a levels in human hepatocellular carcinoma cell lines [13]. In addition, forced expression of miR-199a could enhance the cisplatin sensitivity by activating autophagy pathway [13], suggesting that miR-199a negatively regulates cisplatin resistance. We demonstrated that cisplatin treatments down-regulated miR-199a expressions in osteosarcoma cells. By establishment of cisplatin resistant cell line originating from SaoS-2 cells, we detected that miR-199a was down-regulated in osteosarcoma cisplatin resistant cells, intriguing us to explore the mechanisms for the miR-199a-modulated cisplatin sensitivity.
HIF-1α is the main transcription factor responsible for the cellular adaptation to hypoxia [20]. Moreover, it has been widely studied that hypoxia commonly exists in solid tumours, leading to cancer cells chemoresistance via activation of adaptive pathways [20]. Other than our above results, some studies have reported that hypoxia can activate autophagy and the hypoxia-induced autophagy may contribute to cisplatin resistance [21]. Our data revealed HIF-1α was up-regulated in cisplatin resistant osteosarcoma cells, indicating that HIF-1α contributes to cisplatin sensitivity. Importantly, we identified HIF-1α as a direct target of miR-199a and overexpression of miR-199a sensitized cisplatin resistant cells in vitro and in vivo. In addition, we investigated whether the miR-199a-modulated cisplatin sensitization was through the direct inhibition of HIF-1α with the restoration of HIF-1α expression in miR-199a overexpressing cells. Here, we show that the rescue of HIF-1α in miR-199a overexpressing osteosarcoma cells enhanced resistance to cisplatin, suggesting an miR-199a-HIF-1α-cisplatin sensitivity axis in osteosarcoma. Currently, the detailed mechanisms for the HIF-1α-induced cisplatin resistance under hypoxia is still under our investigation.
Taken together, the present study indicates that miR-199a contributes to the reversal of cisplatin resistance by blocking the expression of HIF-1α in cisplatin resistant osteosarcoma cancer cells. During this process, HIF-1α is a direct target gene of miR-199a. According to the present study, we predict that miR-199a may be a potential therapeutic target for cisplatin-resistant osteosarcoma tumours. Our future studies will focus on whether additional molecular mechanisms exist for the HIF-1α-induced cisplatin resistance in an in vivo model. This model may be useful for future treatment schedules for overcoming cisplatin resistant osteosarcoma patients.