Tumor Suppressive Role of miR-342-5p and miR-491-5p in Human Osteosarcoma Cells

Osteosarcomas are the most common type of malignant bone tumor. These tumors are characterized by the synthesis of an osteoid matrix. Current treatments are based on surgery and combination chemotherapy. However, for metastatic or recurrent tumors, chemotherapy is generally ineffective, and osteosarcomas are sometimes unresectable. Thus, the use of microRNAs (miRNAs) may represent an attractive alternative for the development of new therapies. Using high-throughput functional screening based on impedancemetry, we previously selected five miRNAs with potential chemosensitizing or antiproliferative effects on chondrosarcoma cells. We validated the tumor-suppressive activity of miR-491-5p and miR-342-5p in three chondrosarcoma cell lines. Here, we carried out individual functional validation of these five miRNAs in three osteosarcoma cell lines used as controls to evaluate their specificity of action on another type of bone sarcoma. The cytotoxic effects of miR-491-5p and miR-342-5p were also confirmed in osteosarcoma cells. Both miRNAs induced apoptosis. They increased Bcl-2 homologous antagonist killer (Bak) protein expression and directly targeted Bcl-2 lymphoma-extra large (Bcl-xL). MiR-342-5p also decreased B-cell lymphoma-2 (Bcl-2) protein expression, and miR-491-5p decreased that of Epidermal Growth Factor Receptor (EGFR). MiR-342-5p and miR-491-5p show tumor-suppressive activity in osteosarcomas. This study also confirms the potential of Bcl-xL as a therapeutic target in osteosarcomas.


Introduction
Osteosarcomas are the most common type of bone tumor [1][2][3]. They primarily affect children and adolescents. These tumors mainly develop in the metaphyseal region of long bones, such as the distal femur, the proximal tibia and the proximal humerus [4,5]. Osteosarcomas are bone-forming malignant tumors. They are classified into several histological subtypes: central, intramedullary and surface tumors, with subtypes in each group [5]. Conventional osteosarcomas represent 80% of all primary osteosarcomas. Conventional treatment of osteosarcomas consists of surgery in combination with chemotherapy, leading to an approximately 70% chance of patient remission after 5 years [6]. Nevertheless, the survival of patients with metastatic or recurrent tumors has remained unchanged over the past 40 years, with an overall survival rate of approximately 20% at 5 years [7]. At the time of diagnosis, 20% of patients have pulmonary metastases, but it is estimated that We first evaluated the antimetabolic effects of miR-149-5p, miR-342-5p, miR-491-5p, miR-541-5p and miR-625-5p alone in the presence of a sublethal dose of cisplatin (CDDP) to test their chemosensitizing potential ( Figure 1).
In the HOS cell line, miR-491-5p significantly decreased the metabolic activity compared with miR-Ctrl, both without CDDP (2.2-fold decrease, p < 0.001) and with CDDP (1.9-fold decrease, p < 0.01) ( Figure 1A). MiR-342-5p also induced a significant decrease in the metabolic activity in both conditions (1.9-fold, p < 0.01). The other miRNAs did not significantly modulate the metabolic activity of HOS cells.
In the MG-63 cell line, only miR-491-5p and miR-342-5p significantly lowered the metabolic activity of the cells ( Figure 1B). MiR-491-5p induced a 4-fold and 4.9-fold decrease (p < 0.001) in metabolic activity compared with miR-Ctrl and miR-Ctrl/CDDP-treated cells, respectively, with no significant difference between the two treatments. MiR-342-5p decreased metabolic activity by 2.9-fold in both conditions (p < 0.001 and p < 0.01 respectively). After 48 h, cells were treated for 24 h with 0.5 µg/mL CDDP. The metabolic activity was measured 72 h after transfection. Data are presented as mean optical density (OD) ± SD of four independent experiments. Statistically significant differences between miR-Ctrl and miRNA-treated cells were determined using one-way ANOVA (***: p < 0.001, **: p < 0.01, *: p < 0.05).
Overall, these data demonstrate the antimetabolic effects of only miR-491-5p and miR-342-5p on all three osteosarcoma cell lines and of miR-625-5p only on the SaOS-2 cell line. Moreover, none of these miRNAs showed a significant chemosensitizing effect to CDDP on the metabolism of osteosarcoma cell lines.
2.2. MiR-342-5p and MiR-491-5p Have Cytotoxic Effects on Osteosarcoma Cell Lines Next, we studied the potential cytotoxic effect of these five miRNAs on the three osteosarcoma cell lines ( Figure 2).
Only miR-491-5p and miR-342-5p induced significant cytotoxic effects in MG-63 cells (approximately 3-fold increase with or without CDDP compared with miR-Ctrl) ( Figure 2B). The effect of miR-625-5p was not statistically significant in the presence or absence of CDDP (1.6-fold and 2.3-fold increases, respectively).
In the SaOS-2 cell line, only miR-342-5p significantly triggered cytotoxic effects with and without CDDP (2.5-fold and 2-fold increases relative to the respective miR-Ctrl, p < 0.05), with no significant difference between the two conditions ( Figure 2C). MiR-491-5p enhanced cytotoxicity by approximately 2-fold in both conditions, but this effect was only significant (p < 0.01) in combination with CDDP.
Only miR-491-5p and miR-342-5p showed significant cytotoxic effects on the three osteosarcoma cell lines. The combination of CDDP and miRNAs did not seem to improve the effects of miRNAs.

MiR-342-5p and MiR-491-5p Induce Cell Death in Osteosarcoma Cell Lines
We then analyzed cell cycle progression in the three osteosarcoma cell lines in the presence or absence of a sublethal dose of CDDP ( Figure 3).
CDDP increased the percentage of events in the S and G2/M phases, whereas it decreased the percentage of events in the G0/G1 phase in HOS cells ( Figure 3A). CDDP enhanced the percentage of events only in the S phase in MG-63 and SaOS-2 cells (Figure 3B,C). MiR-491-5p and miR-342-5p alone did not induce a cell cycle blockade (no stalling in S or G2/M phases), but they increased the percentage of cells in the sub-G1 phase, an indicator of the induction of cell death.
In the HOS cell line, miR-491-5p increased the percentage of sub-G1 events with and without CDDP (by 2.4-fold and 2.9-fold relative to the respective miR-Ctrl ( Figure 3A). This effect was significant only with miR-491-5p used alone (p < 0.05). MiR-342-5p induced a more significant increase in cell death, which increased by 5.6-fold and 7.5-fold with and without CDDP, respectively (p < 0.001). The proportions of cells in the sub-G1 phase were 4.8%, 11.3% and 26.7% without CCDP and 4.3%, 12.3% and 32.2% with CCDP for miR-Ctrl, miR-491-5p and miR-342-5p, respectively. Both miRNAs decreased cellular density 72 h post-transfection ( Figure S1, Supplementary Materials). Moreover, we observed more cellular debris and condensed and/or fragmented chromatin, typical features of apoptotic cells ( Figure S1).  Figure 1. The cytotoxicity of miRNAs was measured 72 h after transfection. Data are presented as the mean relative luciferase units (RLU) ± SD of four independent experiments. Statistically significant differences between miR-Ctrl and miRNA-treated cells were determined using one-way ANOVA (***: p < 0.001, **: p < 0.01, *: p < 0.05).
In MG-63 cells, miR-491-5p and miR-342-5p induced a significant increase (p < 0.001) in the number of sub-G1 events ( Figure 3B). For miR-491-5p, we observed a 3.2-fold increase without CDDP and a 3.6-fold increase with CDDP. MiR-342-5p increased cell death by 4.3-fold when used alone and by 4-fold in the presence of CDDP. The proportions of the cells in the sub-G1 phase were 17.4%, 56.4% and 75.3% without CCDP and 18%, 64.6% and 72.5% with CCDP for miR-Ctrl, miR-491-5p and miR-342-5p, respectively. A decrease in confluency and a high amount of cellular debris and apoptotic bodies were observed after the transfection of both miRNAs ( Figure S2, Supplementary Materials).
In the SaOS-2 cell line, miR-491-5p enhanced the percentage of cells in the sub-G1 phase by 3-fold with and without CDDP (p < 0.001 and p < 0.01) compared with their respective miR-Ctrl ( Figure 3C). MiR-342-5p also increased the number of sub-G1 events in both conditions (by 2.6-fold alone; by 2.8-fold with CDDP; p < 0.01). The proportions of cells in the sub-G1 phase were 16%, 48.8% and 41.8% without CCDP and 16.9%, 50% and 47.4% with CCDP for miR-Ctrl, miR-491-5p and miR-342-5p, respectively. This induction of cell death by both miRNAs was confirmed with microscopic observations at 72 h posttransfection ( Figure S3, Supplementary Materials).
Overall, miR-342-5p and miR-491-5p functioned as apoptomiRs in HOS, MG-63 and SaOS-2 osteosarcoma cell lines. These miRNAs caused cell death at the same magnitude with and without CDDP, suggesting that they do not enhance the effect of CDDP.  Figure 1. Flow cytometry was carried out 72 h after transfection to analyze cell cycle phase distribution. The histograms show the analysis of five independent experiments for HOS cells and four independent experiments for MG-63 and SaOS-2 cells (mean ± SD with the different phases of the cycle). Statistically significant differences in the percentage of sub-G1 events between miR-Ctrl and miRNA-treated cells are presented. They were determined using one-way ANOVA (***: p < 0.001, **: p < 0.01, *: p < 0.05).

MiR-342-5p and MiR-491-5p Activate the Apoptosis Pathway in Osteosarcoma Cells
We then investigated the induction of apoptosis by the two miRNAs used alone in the three osteosarcoma cell lines. MiR-491-5p and miR-342-5p led to poly(ADP-Ribose) Polymerase (PARP) cleavage in the three cell lines, with better cleavage detected with miR-491-5p compared with miR-342-5p in SaOS-2 cells ( Figure 4A). In addition, both miRNAs also enhanced the cleavage of caspase-3 ( Figure 4B) and induced caspase-3/7 activities ( Figure 4C) to different extents, depending on the cell line. Representative blots of three independent Western blots are shown. Quantification of the bands of interest, normalized to β-tubulin or GAPDH signals and to miR-Ctrl, is shown below each blot. (C) Caspase-3/7 activity was measured in real time for 72 h after transfection as described in Section 4. The graphs show the analysis of fluorescence of four independent experiments performed in triplicate (mean relative fluorescent units (RFUs) ± SD). Statistically significant differences between miR-Ctrl and miRNA-treated cells were determined using two-way ANOVA (***: p < 0.001, **: p < 0.01).
Because PARP cleavage by activated caspases and cleavage or activation of caspase-3/7 are characteristic of apoptosis, these results demonstrated that miR-491-5p and miR-342-5p induce apoptosis in osteosarcoma cells. MiR-342-5p appears to be the best apoptomiR in the HOS cell line, whereas miR-491-5p appears to be the best one in the SaOS-2 cell line. Both miRNAs were efficient in the MG-63 cell line, with miR-342-5p showing a better effect. These results are consistent with our previous results on cell death ( Figure 3).
Both miRNAs potentially target members of the anti-apoptotic Bcl-2 family. Bcl-xL was validated as their direct target in the SW1353 chondrosarcoma cell line in our previous study [16]. Bcl-2 was also validated as a direct target of miR-342-5p in SW1353 cells. Mcl-1 was also a predicted target for both miRNAs [16]. We therefore studied their expression to evaluate their contribution to the induction of apoptosis by both miRNAs ( Figure 4B,D,E).
Bcl-2 protein expression was decreased only upon miR-342-5p transfection ( Figure 4B). The greatest inhibition was observed in the HOS cell line, with a 2-fold decrease compared with miR-Ctrl. MiR-342-5p downregulated Bcl-2 protein expression in the other two cell lines, but to a lesser extent (by 1.4-fold). We did not observe inhibition with miR-491-5p overexpression in any of the three cell lines, suggesting that Bcl-2 is not a target for this miR.
In summary, the inhibition of the expression of Bcl-xL and Bcl-2 may be responsible for the pro-apoptotic effect of miR-342-5p on osteosarcoma cells, and that of Bcl-xL may be involved in the pro-apoptotic effect of miR-491-5p.
A putative binding site for miR-342-5p is located between nucleotides 818 and 824 in BCL2-3 UTR ( Figure 5A). Transfection of miR-342-5p reduced the luciferase activity of the WT-BCL2-3 UTR reporter vector by 17% compared with miR-Ctrl ( Figure 5B). By contrast, co-transfection with anti-miR-342-5p increased luciferase reporter activity (by 20%). A mutation in the BCL2-3 UTR binding site prevented the inhibitory effect of miR-342-5p. Nevertheless, the inhibitory effect of miR-342-5p on BCL2-3 UTR was not statistically significant. Therefore, we cannot unambiguously confirm that BCL2 mRNA is a direct target of miR-342-5p in HOS cells. Computational analysis predicted two potential binding sites for miR-342-5p in BCL2L1-3 UTR at positions 679-686 and 1407-1413. Several constructs were generated with two individual mutations in these sites (MUT1 and MUT2, respectively) and a construct with both mutations (MUT1/MUT2) ( Figure 5C).

MiR-342-5p and MiR-491-5p Regulate the Expression of Numerous Proteins Related to Apoptosis or Proliferation
Next, we studied the expression of some of their validated or predicted targets involved in cell survival or cell death.
Transfection of miR-342-5p did not significantly downregulate the expression of the cyclin D1 protein in any of the three osteosarcoma cell lines ( Figure 6A). The inhibition did not exceed 1.2-fold in MG-63 cells. We did not observe the inhibition of cyclin D1 protein expression by miR-491-5p in the three osteosarcoma cell lines.
MiR-491-5p inhibited EGFR protein expression by 1.6-fold in HOS cells and by 2.5-fold in MG-63 and SaOS-2 cells ( Figure 6B). By contrast, miR-342-5p had no significant inhibitory effect on EGFR in any of the three osteosarcoma cell lines.
In the three osteosarcoma cell lines, miR-342-5p induced a strong decrease in procaspase-9 protein expression (47 kDa; by 3.3-fold in HOS and MG-63 cells and by 2.5-fold in SaOS-2 cells) ( Figure 6C). This decrease was not due to the induction of large cleaved caspase-9 fragments (35 and 37 kDa). MiR-491-5p did not modulate pro-caspase-9 protein expression nor induce protein cleavage.
In summary, miR-491-5p inhibited the expression of EGFR, a protein involved in cell survival, and it enhanced the expression of Bak, a pro-apoptotic protein in osteosarcoma cells. MiR-342-5p also increased Bak protein expression, but at the same time, it decreased pro-caspase-9 protein expression, a caspase involved in the apoptosis pathway.

Discussion
Osteosarcomas represent the most frequent forms of bone sarcomas, with an annual prevalence of 1 in 100,000. Today, osteosarcomas are generally well-treated tumors due to the effectiveness of combination chemotherapy associated with surgery. However, for patients with metastases or recurrence, the survival rate is much lower. As a consequence, treatment seems to have a low success rate, which has not improved for many years. It is therefore essential to develop new tools for the diagnosis and therapy for this type of tumor.
For miR-541-5p, miR-625-5p and miR-149-5p, we did not observe any significant cytotoxic effect after 72 h incubation. However, for miR-625-5p, we detected a slight increase in the cytotoxic effect on the three cell lines, as well as significant downregulation in the metabolic activity of the SaOS-2 cell line. It is possible that a longer incubation period is required to observe cell death effects or chemosensitizing effects for these miRNAs. Indeed, no significant effect of miR-149-5p overexpression was observed on SaOS-2 and U-2OS proliferation after 3 days of incubation in a MTT assay, whereas antiproliferative effect was observed at 4 days [21]. Chemosensitization of miR-625-5p to CDDP was observed in an MTT assay after 48 h of incubation with CDDP in two multidrug-resistant gastric cancer cell lines, whereas we performed the XTT assay after 24 h in osteosarcoma cell lines [18]. In our previous study performed on chondrosarcoma cell lines [16], only miR-342-5p and miR-491-5p clearly showed significant antimetabolic and cytotoxic effects, as in the present study on osteosarcoma. We then studied the effects of miR-342-5p and miR-491-5p on the cell cycle. Both miRNAs enhanced the number of sub-G1 events, reflected by the presence of cellular debris, a hallmark of apoptosis, in the three cell lines. MiR-342-5p was the most effective inducer of cell death in the HOS cell line. In MG-63 cells, both miRNAs induced an equivalent increase in the number of sub-G1 events, with a slightly better effect for miR-342-5p, and vice versa in SaOS-2 cells.
Even though we did not previously observe chemosensitizing effects of miR-491-5p and miR-342-5p in chondrosarcoma cells, we evaluated the ability of these miRNAs to enhance the sensitivity of the three osteosarcoma cell lines to CDDP. Neither of the two miR-NAs was able to increase sensitivity to CDDP. However, we chose to use a sublethal dose of CDDP (1.65 µM) to only induce cell cycle arrest in the S and/or G2/M phases for the three cell lines, rather than doses close to the inhibitory concentrations to reach 50% reduction in cell viability (IC 50 ), as has been used by others. In the literature, IC 50 values for the three cell lines are very variable, depending on the incubation time and the viability assays used (from 3.5 µM to 78 µM for HOS cells, 1.65 to 26 µM for MG-63 cells and 4.9 µM to 46 µM for SaOS-2 cells) [22][23][24]. MG-63 seemed to be the least resistant. We checked whether these concentrations of CDDP induced more cell death to validate the relative sensitivity of the cell lines to a CDDP treatment alone ( Figure S4, Supplementary Materials). Indeed, elevated concentrations of CDDP increased the percentage of cells in the sub-G1 phase rather than cause a cell cycle blockade in the S and G2-M phases for the HOS cell line and in the S phase for the MG-63 and SaOS-2 cell lines. We hypothesized that cell drug resistance was too high with sublethal doses of CDDP to see a chemosensitizing effect of miRNAs. Moreover, after the transfection of miRNA, we incubated the cells only for 24 h with CDDP versus 48 h in most of the previous studies, as already mentioned for miR-625-5p [18]. Another study demonstrated that the overexpression of miR-491-5p was able to sensitize osteosarcoma cell lines (MG-63 and U2OS) after incubation with 10 µM CDDP for 48 h [25]. Therefore, the chosen kinetics and the sublethal dose of CDDP may not be adequate to reveal any chemosensitizing effect of miRNAs, as in our previous study [16]. Other experiments are required to explore this hypothesis.
The analysis of the expression of proteins such as PARP and caspase-3 allowed us to confirm that apoptosis is involved in miRNA-induced cell death. MiR-342-5p and miR-491-5p induced PARP cleavage in the three osteosarcoma cell lines. We also revealed an enhancement of caspase-3/7 activity, although to different extents, with both miRNAs (Figure 7). Because miR-Ctrl also increased caspase-3/7 activity in HOS and SaOS-2 cells, the results appeared to be less significant than in MG-63 cells. We also observed PARP cleavage for the miR-Ctrl condition in Western blots with HOS and SaOS-2 cells. However, we clearly demonstrated that miR-342-5p is the best apoptosis inducer in the HOS cell line, whereas miR-491-5p is the best inducer in the SaOS-2 cell line. Both miRNAs significantly induced caspase-3/7 activity in the MG-63 cell line, with miR-342-5p showing a stronger effect. These results correlate with the data obtained using flow cytometry.
MiR-491-5p has previously been studied in osteosarcoma. One study showed that this miRNA is downregulated in osteosarcoma tissues and cell lines [26]. Another study demonstrated that the level of miR-491-5p decreases in the serum of osteosarcoma patients and correlates with increased metastasis, chemoresistance and thus a lower survival rate [25]. The overexpression of miR-491-5p has been shown to restore the chemosensitivity of osteosarcoma cells to CDDP and to directly target CRYAB encoding crystallin alpha B, a protein that inhibits apoptosis and contributes to intracellular architecture [25]. The latter study showed that miR-491-5p exerts its tumor-suppressing activity by directly targeting CRYAB. Other direct targets of miR-491-5p have been identified in osteosarcoma cells, such as FOXP4 and PKM2 (encoding forkhead-box P4 and pyruvate kinase muscle 2, respectively) [26,27]. Their inhibition increases apoptosis or suppresses cell proliferation of osteosarcoma cells. Nevertheless, the mechanisms involved in the expression of miR-491-5p and its biological role in osteosarcomas still need to be explored. In cancer, constitutive activation of EGFR is associated with cell proliferation, survival and migration. Expression of EGFR has frequently been observed in high-grade osteosarcomas. EGFR does not appear to be a major driver for osteosarcoma cell growth, but it may contribute to starvation and chemotherapy resistance [28]. EGFR mRNA is a validated target of miR-491-5p in other types of cancers, such as glioblastomas and ovarian carcinomas [17,29]. As previously shown in chondrosarcoma cell lines [16], miR-491-5p decreased the level of EGFR expression in osteosarcoma cell lines, which may contribute to a decrease in the cell proliferation rate, as shown in Figure 7.
Several studies have shown that overexpression of miR-491-5p directly inhibits the expression of the Bcl-xL protein in various cancers [17,30,31]. We also previously demonstrated this direct inhibition in SW1353 chondrosarcoma cells [16]. In the present study, we confirmed that miR-491-5p induced cell death in the HOS cell line by directly targeting BCL2L1 mRNA and consequently decreasing Bcl-xL protein expression. By contrast, although the anti-apoptotic member Mcl-1 may be a putative target of miR-491-5p, this miRNA did not influence Mcl-1 expression in any of the three osteosarcoma cell lines investigated. On the other hand, overexpression of miR-491-5p increased the expression level of the pro-apoptotic protein Bak. Inhibition of Bcl-xL expression and overexpression of Bak may therefore contribute to its apoptotic activity in osteosarcoma cells (Figure 7), as in chondrosarcoma cells [16].
Only one study has investigated the effect of miR-342-5p in osteosarcomas. It directly targeted the Wnt Family member 7b (Wnt7b) oncogene, inhibited tumor growth, migration and invasion and restored cell sensitivity to doxorubicin [32]. However, in other studies, including ours, miR-342-5p has been shown to be a tumor suppressor in neuroblastomas, colon cancers, ovarian cancer cells and chondrosarcomas [16,[33][34][35]. Although EGFR and MCL1 mRNA may be potential targets of miR-342-5p, our data did not show significant inhibition of their protein expression in the three osteosarcoma cell lines analyzed. CCND1 mRNA is a validated target of miR-342-5p in neuroblastomas [33], but as in chondrosarcoma cell lines [16], we did not observe downregulated cyclin D1 expression in any of the three osteosarcoma cell lines.
MiR-342-5p decreased Bcl-2 protein expression (from 1.4 to 2-fold) in the three osteosarcoma cell lines. Although we validated BCL2 mRNA as a direct target of miR-342-5p in our previous study on chondrosarcomas [16], here, we did not validate BCL2 mRNA as a real direct target of miR-342-5p. Transfection of miR-342-5p decreased the luciferase activity of the WT-BCL2-3 UTR vector in HOS cells by 1.2-fold, but it was not statistically significant. We suspect that the pro-apoptotic effects of miR-342-5p are linked to Bcl-xL inhibition rather than to Bcl-2 inhibition in osteosarcoma cells. In effect, miR-342-5p downregulated Bcl-xL protein expression from 2.5 to 5-fold in osteosarcoma cell lines and significantly decreased the activity of the WT-BCL2L1-3 UTR vector by 4.2-fold in HOS cells. As shown for neuroblastomas [33], we validated BCL2L1 mRNA as a direct target of miR-342-5p. Moreover, as in our previous study on chondrosarcomas [16], we also clearly identified its binding site at position 679-686 of BCL2L1-3 UTR. Overexpression of Bak also contributes to the apoptotic effects of miR-342-5p (Figure 7).
We showed that miR-342-5p also had anti-apoptotic effects on the three osteosarcoma cell lines by reducing pro-caspase-9 protein expression (by approximately 3-fold). We did not observe the induction of caspase-9 cleavage by either miRNA, but, unlike other caspases, caspase-9 showed complete activity in its uncleaved form [36]. These results are consistent with our previous study on chondrosarcoma cells [16] and with another study on cardiomyocytes [36]. In osteosarcoma cells, miR-342-5p clearly induced apoptosis by inducing caspase-3/7 activity and PARP cleavage but could also have some anti-apoptotic potential (Figure 7).
In osteosarcomas, overexpression of Bcl-2 and Bcl-xL has been reported to be associated with poor prognosis [37,38]. Metastatic or recurrent osteosarcomas have a high expression of Bcl-xL, which is correlated with a low probability of survival, mainly due to resistance to chemotherapy [38]. This correlation between Bcl-xL expression and chemoresistance in cancers is very common and has been demonstrated in a panel of 60 cell lines [39]. Additionally, a study showed that pharmacological inhibition of Bcl-xL improves the sensitivity of osteosarcomas to doxorubicin [40]. Here, miR-491-5p and miR-342-5p targeted BCL2L1 mRNA and induced apoptosis. We thus confirmed the potential of Bcl-xL as a relevant therapeutic target in osteosarcomas.

Metabolic Activity Analysis
HOS, MG-63 and SaOS-2 cells were seeded onto 96-well microplates at a density of 3.2 × 10 3 cells/well in triplicate. Cells were grown for 24 h before transfection of 20 nM miRNA. Treatment with CDDP was initiated 48 h later, as described above. Cell metabolic activity was assessed 72 h after transfection. XTT assay (11465015001, Roche, Basel, Switzerland) was performed as previously described [16]. Absorbance was measured after 1 h of incubation with a microplate reader (Spark 10M, Tecan Lyon, France) at 450 nm with a reference wavelength at 600 nm.

Cytotoxicity Assay
Cells were cultured under the same conditions as for the XTT assays. Seventy-two hours after transfection of miRNA, cytotoxicity was evaluated with a bioluminescence cytotoxicity assay kit (CA4680, Interchim, Montluçon, France). Adenylate kinase detection was carried out in the supernatant as previously described [16]. Luminescence was measured with a microplate reader (Spark 10M, Tecan Lyon, France).

Analysis of Nuclear Morphology
Morphological characterization of apoptotic cells by nuclear staining with DAPI (sc-24941, Santa Cruz Biotechnology, Dallas, TX, USA) was performed after cyto-centrifugation as previously described [16]. Cell confluence and nuclear staining were monitored using a fluorescence microscope (Nikon Eclipse Ti; Nikon Instruments Inc., Melville, NY, USA).

Statistical Analysis
Experiments were independently repeated at least three times. Values are reported as means ± SD. Statistical analyses were performed using one-way ANOVA corrected for multiple comparisons using Dunnett's test or two-way ANOVA corrected for multiple comparisons using Sidak's test. Alternatively, statistical significance was assessed by two-tailed unpaired Student's t-test with Welch's correction to analyze statistical differences within representative experiments performed in triplicate. Data were analyzed with GraphPad Prism 7 software (San Diego, CA, USA): *** p < 0.001, ** p < 0.01 and * p < 0.05.

Conclusions
We clearly demonstrated the antimetabolic, cytotoxic and antiproliferative effects of miR-342-5p and miR-491-5p on three osteosarcoma cell lines and also their pro-apoptotic effects. Unfortunately, they failed to induce chemosensitivity to CDDP in our experimental conditions, as observed in previous impedancemetry experiments. We also showed that miR-342-5p and miR-491-5p directly target BCL2L1 mRNA. Moreover, miR-491-5p inhibited the protein expression of EGFR and increased that of Bak. MiR-342-5p also increased the protein expression of Bak and decreased that of Bcl-2 and Bcl-xL but, at the same time, decreased pro-caspase-9 expression. Overall, this study highlighted the functional and essential role of Bcl-xL in osteosarcoma cells and confirmed the therapeutic potential to target Bcl-xL in osteosarcomas.