Skip to main content
NCBI home page
Cell Proliferation logoLink to Cell Proliferation
. 2018 Aug 7;51(6):e12508. doi: 10.1111/cpr.12508

RNA‐binding protein PUM2 suppresses osteosarcoma progression via partly and competitively binding to STARD13 3′UTR with miRNAs

Ruixi Hu 1,†, Xiaodong Zhu 1,†, Chao Chen 1, Ruijun Xu 1, Yifan Li 1, Wei Xu 1,✉
PMCID: PMC6528862  PMID: 30084199

Abstract

Objectives

This work aims to reveal the roles and related mechanisms of RNA binding protein PUM2 in osteosarcoma progression.

Materials and methods

Transcriptome analysis based on RNA sequencing data, real‐time quantitative PCR (RT‐qPCR), and western blot analysis were used to detect the expression of RBPs and miRNAs in OS and normal adjacent tissues, and the correlation between them in OS tissues. RT‐qPCR, western blot, cell viability, transwell migration, tumour spheres formation and in vivo tumour formation assays were used to examine the effects of RBP PUM2 on OS progression. Additionally, RNA immunoprecipitation (RIP) assay combined with RNA sequencing was performed to determine the binding site of RBP PUM2 on STARD13 3′UTR. Luciferase reporter and RIP assays were used to confirm the binding of miRNAs or PUM2 on STARD13 3′UTR.

Results

PUM2 and STARD13 expression was significantly decreased in OS tissues, and positively correlated. Overexpression of PUM2 or STARD13 3′UTR inhibited OS cells proliferation, migration, and stemness. Mechanistically, PUM2 competitively bound to STARD13 3′UTR with miR‐590‐3p and miR‐9. The inhibition of PUM2 on OS cells progression was attenuated by STARD13 knockdown or related miRNAs overexpression.

Conclusion

PUM2 suppresses OS progression via partly and competitively binding to STARD13 3′UTR with miRNAs.

1. INTRODUCTION

At present, OS treatment is still based on the combination of surgery and chemotherapy, and the research on gene therapy and neoadjuvant therapy has also made some progress, which is expected to improve the survival rate of patients.1 However, a major and unsolved problem in clinical practice is that the prognosis of patients who cannot be resected or relapsed is extremely poor.2 Thus, it is still an urgent need to find novel targets facilitating new methods for OS treatment or improving prognosis.

RNA‐binding proteins (RBPs) could regulate transcripts expression via directly binding to them and enhancing their mRNA stability, and thus promote transcripts expression.3 Amount of evidence have indicated that RBPs play important roles in cancer progression3 and early testis germ cell development.4 Yu et al5 showed that RBPs CUGBP1 and HuR could regulate E‐cadherin translation by altering recruitment of E‐cadherin mRNA to processing bodies. And Poria et al. indicated that HuR could sequester miR‐21 to prevent translation repression of proinflammatory tumour suppressor gene programmed cell death 4.6 Additionally, our previous study demonstrated that HuR could increase OS cells migration, invasion, and stemness through activating YAP and decreases susceptibility to chemotherapeutic agents.7 PUM2 (Pumilio2), as a RBP, has been shown to play critical roles in developing mammalian neural stem cells,8 epilepsy 9 and human germ cells progression.10 However, the roles and targets of PUM2 in cancer progression are still unclear. Here, we found that PUM2 expression was significantly decreased in OS tissues, and assumed that PUM2 held inhibitory effects in OS progression.

miRNAs could regulate transcripts expression via directly binding to mRNAs 3′UTR and thus inhibit their expression, or promote their degradation.11 Importantly, miRNAs play critical roles in cancer progression, like in tumour angiogenesis,12 metastasis,13 apoptosis,14 and drug resistance.15 Previous studies have shown that RBPs could competitively bind to mRNAs 3′UTR via miRNAs, for example, transformer 2beta and miR‐204 could regulate apoptosis through competitively binding to 3′UTR of BCL2 mRNA.16 The RBP ARPP21 could control dendritic branching by functionally opposing the miRNA it hosts.17 These studies demonstrate that RBPs and miRNAs could co‐regulate transcripts expression collectively.

STARD13 (StAR‐related lipid transfer domain 13) has been confirmed to hold suppressive roles in various tumour progression, like Li et al and Hanna et al have shown that STARD13 inhibits breast cancer metastasis.18, 19 Guo et al20 have indicated that STARD13 3′UTR could promote breast cancer apoptosis and Zhang et al21 have demonstrated that STARD13 promotes hepatocellular carcinoma apoptosis by acting as a ceRNA for Fas. However, the roles of STARD13 3′UTR and related mechanisms by which STARD13 is regulated in OS progression are still unclear.

In the present study, we showed that RBP PUM2 and STARD13 expression was significantly decreased in OS tissues compared with that in normal adjacent tissues, and exhibited positive correlation in OS tissues. Furthermore, we identified that PUM2 directly and competitively bound to STARD13 3′UTR with miR‐590‐3p and miR‐9. Additionally, PUM2 and STARD13 3′UTR inhibited OS cells migration, proliferation, and stemness, and PUM2 exerted its effects in a STARD13 3′UTR‐ and miRNAs‐dependent manner. Importantly, overexpression of STARD13 3′UTR could inactivate RhoA/Rock pathway. Thus, our results indicate that RBP PUM2 could suppress OS progression via directly and competitively binding to STARD13 3′UTR with related miRNAs, and thus inhibiting RhoA/Rock pathway.

2. MATERIALS AND METHODS

2.1. Clinical samples and cells culture

Thirty‐nine OS and normal adjacent paraffin‐embedded tissue samples, four fresh OS, and normal adjacent tissues were randomly selected from the TongRen Hospital from October 2014 to June 2017. Written informed consent from all patients and approval of the Hospital Ethic Review Committees were obtained. Isogenic OS cell lines MG63, U2OS, Saos2, 143B were purchased from the Chinese Academy of Sciences Cell Bank and cultured in Dulbecco's minimum essential medium (DMEM) (Gibco, Grand Island, NY, USA) supplemented with 10% foetal bovine serum (Gibco), 80 U/mL penicillin, and 0.08 mg/mL streptomycin at 37°C under humidified atmosphere with 5% CO2.

2.2. Real‐time quantitative PCR

For RNA extraction in cells, OS cells were washed with ice‐cold PBS and followed by extracted using TRIeasy™ Total RNA Extraction Reagent TRIeasyTM (Yeasen, Shanghai, China) following the manufacturer's protocols. For RNA extraction in paraffin‐embedded tissue samples, paraffin‐embedded tissues were processed and total RNA was extracted by total RNA extraction kit for paraffin embedded tissues (Cat # DP439; TianGen Biotech, Beijing, China) following the manufacturer's recommendation. Then total RNA was reverse transcribed into cDNA using Hifair™ III 1st Strand cDNA Synthesis SuperMix (Yeasen) according to the standard protocols. Afterwards, mRNAs expression was measured with Hieff™ qPCR SYBR® Green Master Mix (No Rox) (Yeasen) and performed on an ABI Prism 7500 Detection System (Applied Biosystems, Inc., Foster City, CA, USA). Then, the expression of each transcript was normalized to GAPDH and calculated using the 2−△△ct method.

2.3. RNA immunoprecipitation

The detailed procedure was denoted in our previous work.7 Briefly, OS cells were lysed 25 mmol/L Tris‐HCl buffer (pH 7.5) and 100 U/mL RNase inhibitor (Sigma, St. Louis, MO, USA), and then incubated with protein‐A Sepharose beads precoated with 3 μg anti‐PUM2 antibody, or anti‐Ago2 antibody or control rabbit IgG for 1.5 h at 4°C. The RNA‐protein complexes were pulled‐down by protein‐A/G agarose beads and RNA extracted with Trizol, followed by detecting mRNAs and miRNAs expression with real‐time quantitative PCR (RT‐qPCR).

2.4. Transcripts sequencing and data analysis

We thanked Novengene Co. Ltd (Beijing, China) for conducting this experiment and analysing data. Total RNA extracted from three OS and normal adjacent tissues, and RNA immunoprecipitation (RIP) were used for RNA‐seq. RNA‐seq libraries were acquired using Illumina's TruSeq Stranded Total RNA Library Prep Kit with Ribo‐Zero Gold according to the manufacturer's instructions, and sequenced by HiSeq 2000 sequencer. We used Cuff‐diff to estimate fragments per kilobase of transcript per million (FPKM) values for known transcripts, and analyse differentially expressed transcripts. P < 0.05 was considered as significant. Heatmap of gene expression was generated based on log2 (FPKM) using HemI 1.0.3.7 (http://hemi.biocuckoo.org/down.php). For the RIP‐seq analysis, the detailed procedure was referred to previous work.22

2.5. Lentivirus package and stable expression cell lines construction

The detailed procedure was mentioned in our previous work.7 PUM2 overexpression (Lenti‐PUM2), PUM2 knockdown (Lenti‐PUM2‐shRNA), STARD13 3′UTR overexpression (Lenti‐S‐UTR), RhoA overexpression (Lenti‐RhoA), and control virus plasmids were packaged. Afterwards, virus with 2 μg/mL polybrene was added into OS cells for 72 hours and infected cells were selected with puromycin (Sigma; 2 μg/mL) for 2 weeks, followed by RT‐qPCR and western blot verification.

2.6. Western blot

Cells or tissues were lysed with RIPA lysis buffer (10 mmol/L Tris‐HCl pH 8.0, 1 mmol/L EDTA, 1% Triton X‐100, 0.1% SDS, 140 mmol/L NaCl and protease inhibitor cocktail). Protein concentration was measured by Enhanced BCA Protein Assay Kit (Beyotime, Beijing, China). Equal amounts of protein (30 μg) were loaded and separated by SDS‐PAGE, and transferred onto PVDF membranes. The membranes were blocked by 5% nonfat milk at room temperature for 1.5 hours and incubated with primary antibodies. HRP‐labelled goat anti‐mouse IgG(H+L) (Beyotime) and HRP‐labelled goat anti‐rabbit IgG(H+L) (Beyotime) were used as secondary antibodies. Membrane‐bound immune complexes were detected using an enhanced chemiluminescence kit (Tanon, China) followed by exposure with Tanon 5200.

2.7. Cell proliferation assay

Cell Counting Kit‐8 (CCK8; Beyotime) was used to evaluate cell proliferation. Cells with different treatment were seeded in 96‐well plates at 4000 cells/well. After 24, 48, and 72 hours, cell viability was determined by CCK8 analysis following the manufacturer's recommendation. Each assay was performed at least three times.

2.8. Cell migration assay

The detailed procedure was reported in the previous study.23

2.9. Cell spheroid formation assay

OS cells with different treatment were cultured in ultra‐low attachment 24‐well plates (Corning, Union City, CA, USA) at 500 cells/well with MammoCult™ Human Medium Kit (Cat#05620; Stemcell Technologies, Vancouver, BC, Canada). After 10‐day culture, mammospheres number and size were evaluated using a microscope fitted with a ruler.

2.10. ALDH1 activity assay

ALDH1 activity was assayed by ALDH Activity Assay Kit (Colorimetric) (Cat # KA3742; Abnova, San Diego, CA, USA) according to the manufacturer's protocols.

2.11. Transfection

Cells were seeded at plates. When cells were grown to 80% density, 2.5 μg of plasmids or 50 nmol/L miRNAs mimics, or mimics NC, or inhibitors or NC was transfected or cotransfected into cells using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) following the manufacturer's instructions.

2.12. Luciferase reporter assay

Luciferase reporter analysis was performed to examine the binding of PUM2 and miRNAs (miR‐590‐3p, miR‐9, miR‐653, and miR‐182) on STARD13 3′UTR. The sequence of STARD13 promoter, 5′UTR, coding area or 3′UTR was cloned into PMIR‐Reporter plasmid, respectively, named as Luc‐S‐P, Luc‐S‐5′UTR, Luc‐S‐CDS, and Luc‐S‐3′UTR. Then these above plasmids were cotransfected with Lenti‐PUM2 infection in OS cells, or with miRNA mimics or inhibitors, or NC into OS cells with β‐gal (Ambion, Grand Island, NY, USA) plasmid using Lipofectamine 2000 for 72 hours. After that, the luciferase activity of Luc‐S‐UTR was measured using a Luciferase Reporter Assay Kit (cat. no. K801‐200; BioVision, Inc., Milpitas, CA, USA). β‐gal activity was determined using a β‐Galactosidase Enzyme Assay System with Reporter Lysis Buffer (cat. no. E2000; Promega Corporation, Madison, WI, USA) following the manufacturer's protocols, which was used as a normalization control for luciferase activity.

2.13. RhoA GTPase assay

RhoA GTPase activity was analysed by RhoA G‐LISA (Cat # BK132; Cytoskeleton, Denver, CO, USA) following the manufacturer's protocol. Briefly, cells were lysed and protein concentration was measured. Lysis buffer was added to equalize the cell extracts to give identical protein concentrations in each sample. Samples were handled following the technical guide. Finally, absorbance at 490 nm was measured using a microplate spectrophotometer to reflect the RhoA GTPase activity shown as relative value.

2.14. F‐actin visualization

F‐actin was stained with rhodamine phalloidin (Cat BK005; Cytoskeleton) for 30 minutes at room temperature following the manufacturer's protocol. The nuclei was stained by DAPI (C1002; Beyotime) for 10 minutes and observed with the confocal microscopy. Images were analysed using OLYMPUS FLUOVIEW ver.3.0 Viewer software.

2.15. In vivo tumorigenic assay

Male athymic BALB/c nude mice (4‐6 week) were purchased from Model Animal Research Center of Nanjing University, housed and fed in standard pathogen‐free conditions. For tumour‐limiting dilution assays, tumour cells were mixed 1:1 with matrigel matrix (BD Biosciences, Franklin Lakes, NJ, USA) and subcutaneously implanted in the mice. On day 8, all mice were killed and tumour tissues were collected and weighed. All animal studies were approved by the Shanghai JiaoTong University Animal Care and Use Committee.

2.16. Statistical analysis

All data were obtained from at least three independent experiments, and presented as the mean ± standard deviation. Data sets with only two groups were analysed using a Student's t‐test. Differences between multiple groups were analysed using one‐way analysis of variance with the Tukey‐Kramer post hoc test. P < 0.05 was considered to indicate a statistically significant difference.

3. RESULTS

3.1. RBP PUM2 expression is significantly decreased in OS tissues and positively correlated with the metastasis free survival of OS patients

Firstly, total RNA from OS and normal adjacent tissues were subjected to RNA sequencing, and we found that total 73 genes were upregulated in OS tissues and 80 genes were downregulated. Then we focused on RBPs which were listed in starBase v2.0 (http://starbase.sysu.edu.cn/rbpMrna.php), and significantly decreased in OS tissues. As shown in Figure 1A, several RBPs expression was significantly decreased in OS tissues. RT‐qPCR analysis was further performed to confirm the change of these RBPs expression, and we found that the change of RBP PUM2 expression was the most significant (Figure 1B). Additionally, western blot assay confirmed that PUM2 expression was remarkably decreased in OS tissues (Figure 1C). Importantly, KM‐Plotter assay from online TCGA data set (http://hgserver1.amc.nl/cgi-bin/r2/main.cgi) which included 127 OS tissues showed that PUM2 expression was positively correlated with the metastasis‐free survival of OS patients (Figure 1D). Thus, PUM2 was chosen for the following research.

Figure 1.

Figure 1

Open in a new tab

PUM2 expression is significantly decreased in OS tissues and positively correlated with the metastasis free survival of OS patients. (A) RNA‐seq was performed to detect the expression of RBPs in OS and normal adjacent tissues. (B) RBPs mRNA expression level was examined in OS and normal adjacent tissues via RT‐qPCR analysis. (C) PUM2 protein expression was detected in OS and normal adjacent tissues via western blot analysis. (D) KM‐Plotter assay was performed to analyse the correlation between PUM2 expression and the overall survival of OS patients. Data were presented as the mean ± SD; *P < 0.05, **P < 0.01 vs OS tissues

3.2. Overexpression of PUM2 inhibits OS cells proliferation and migration

We further detected PUM2 expression in various OS cell lines, and found that PUM2 expression was significantly decreased in MG63 and Saos2 cells compared with other cell lines (Figure 2A). Then we constructed MG63 and Saos2 cells with PUM2 stable overexpression with lentivirus infection. RT‐qPCR and western blot analysis confirmed the infection efficiency (Figure 2B,C). Furthermore, we examined whether PUM2 could affect OS cells proliferation. As shown in Figure 2D,E, PUM2 overexpression significantly decreased OS cells viability. Additionally, the effect of PUM2 on OS cells migration was evaluated, and consistent results that PUM2 overexpression suppressed OS cells migration were obtained (Figure 2F,G). Since epithelial‐mesenchymal transition (EMT) could contribute to cells migration, the effects of PUM2 on EMT process were further determined. As expected, PUM2 overexpression suppressed EMT process in OS cells characterized as the decrease of mesenchymal marker (Vimentin) expression, and increase of epithelial marker (E‐cadherin) expression (Figure 2H,I).

Figure 2.

Figure 2

Open in a new tab

Overexpression of PUM2 inhibits OS cells proliferation and migration. (A) PUM2 expression was detected in OS cell lines (MG63, U2OS, Saos2, and 143B). (B, C) The infected efficiency of Lenti‐PUM2 was confirmed in MG63 and Saos2 cells via RT‐qPCR (B) and western blot (C) analysis. (D, E) Cell viability of MG63 and Saos2 cells with PUM2 overexpression or not was determined by CCK8 assay. (F, G) The ability of cell migration was examined in cell depicted in (D) and quantified (G). (H, I) The expression of EMT makers (E‐cadherin and Vimentin) was evaluated in cells depicted in (D) via RT‐qPCR (H) and western blot (I) analysis. Data were presented as the mean ± SD; *P < 0.05 vs control

3.3. Overexpression of PUM2 inhibits OS cells stemness

As tumour cells stemness could lead to tumours recurrence and chemoresistance,24, 25 we further investigated whether PUM2 could suppress OS cells stemness. Firstly, we found that overexpression of PUM2 inhibited the expression tumour stemness markers (Nanog and ALDH1) (Figure 3A‐C). Meanwhile, cell spheroid formation was remarkably attenuated by PUM2 overexpression in OS cells characterized as the decrease of spheres size and number (Figure 3D,E). Additionally, ALDH1 activity was also downregulated in OS cells with PUM2 overexpression via ALDH Activity Assay Kit analysis (Figure 3F). Thus, our results suggest that PUM2 could inhibit OS cells progression.

Figure 3.

Figure 3

Open in a new tab

Overexpression of PUM2 inhibits OS cells stemness. (A, B) ALDH1 (A) and Nanog (B) mRNA level was detected in OS cells with PUM2 overexpression or not. (C) ALDH1 and Nanog protein level was evaluated in OS cells depicted in (A). (D, E) The mammospheres size (D) and numbers (E) were examined in cells depicted in (A). (F) ALDH1 activity was measured in cells depicted in (A). Data were presented as the mean ± SD; *P < 0.05 vs control

3.4. Overexpression of PUM2 inhibits OS cells progression in vivo

We continue exploring whether PUM2 regulates tumour‐initiating potential of OS cells in vivo. We compared the capacity of PUM2‐overexpressed cells to seed tumours at limiting dilutions. Although all cell lines could form tumours at the density of 1 × 106 SaoS2 cells, PUM2‐overexpressed cells showed a decrease of tumour size and weight (Figure 4B,C). Notably, the tumour‐seeding ability of PUM2‐overexpressed cells was significantly decreased at the density of 1 × 106 MG63 cells, 1 × 105 and 1 × 104 MG63 and SaoS2 cells respectively (Figure 4A,B). Taken together, these results indicate that PUM2 could inhibit the tumour initiation ability of OS cells.

Figure 4.

Figure 4

Open in a new tab

Overexpression of PUM2 inhibits OS cells progression in vivo. (A, B) Images of tumours were harvested when serially diluted MG63 and Saos2 with PUM2 overexpression or not were planted. (C) Weight of tumours was harvested when 1 × 106 Saos2 cells with PUM2 overexpression or not were planted. Data were presented as the mean ± SD; *P < 0.05 vs control

3.5. Overexpression of PUM2 increases STARD13 expression via directly binding to STARD13 3′UTR

Since PUM2 belongs to RBPs which have their targets, we tried to search the potential targets of PUM2 in OS cells. Bioinformatics methods (http://starbase.sysu.edu.cn/index.php) predicted the potential targets of PUM2, among which STARD13 attracted our attention based on the result that STARD13 expression was positively correlated with the metastasis free survival of OS patients (Figure 5A). RT‐qPCR and western blot analysis confirmed that STARD13 expression was indeed increased in OS cells with PUM2 overexpression (Figure 5B,C). In addition, when PUM2 was knocked down and the de novo synthesis was blocked with actinomycin D, the decay rate of STARD13 was faster in PUM2 knockdown cells than in control cells (t 1/2 = 5.3 ± 0.3 h and t 1/2 = 5.8 ± 0.4 h vs t 1/2 = 3.2 ± 0.3 h and t 1/2 = 2.2 ± 0.3 h, respectively) (Figure 5D,E). In particular, STARD13 promoter activity was unaffected in PUM2 knockdown cells (Figure 5F), indicating that PUM2 could not affect STARD13 promoter activity. Importantly, further luciferase reporter assays with Luc‐S‐5′UTR, Luc‐S‐CDS, or Luc‐S‐3′UTR transfection, respectively, as well as Lenti‐PUM2 infection, indicated that PUM2 overexpression just increased the activity of Luc‐S‐3′UTR, but unaffected the activity of Luc‐S‐5′UTR or Luc‐S‐CDS (Figure 5G). Furthermore, RIP assays with RT‐qPCR primers expanding STARD13 3′UTR showed that PUM2 could interact with STRAD13 3′UTR directly in OS cells (Figure 5H). These results indicated that PUM2 could directly bind to STRAD13 3′UTR, and thus enhance STARD13 mRNA stability and expression.

Figure 5.

Figure 5

Open in a new tab

Overexpression of PUM2 increases STARD13 expression via directly binding to STARD13 3′UTR. (A) KM‐Plotter assay was performed to analyse the correlation between STARD13 expression and the overall survival of OS patients. (B, C) STARD13 mRNA (B) and protein (C) expression was determined in OS cells with PUM2 overexpression or not. (D, E) OS cells with PUM2 overexpression or not were treated with actinomycin D (2.5 μg/mL) for the indicated times. STARD13 mRNA level was measured by RT‐qPCR and the percentage of mRNA that remained was plotted. (F) The promoter activity of STARD13 was detected in cell depicted in (A). (G) The luciferase activity of vector, Lus‐S‐5′UTR, Luc‐S‐CDS, and Lus‐S‐3′UTR was examined in cells depicted in (A). RT‐qPCR was used to measure the abundance of STARD13 mRNA that was present in the PUM2‐IP materials after the RIP assay. Data were presented as the mean ± SD; *P < 0.05 vs control

3.6. PUM2 exerts its inhibitory effects dependent on STARD13 3′UTR

We further investigated the detailed binding sites of PUM2 and whether PUM2 exerted its inhibitory effects via STARD13 3′UTR. Indeed, overexpression of STARD13 3′UTR significantly decreased OS cells progression characterized as the decrease of cell proliferation, migration, and stemness (Figure S1). Then we performed genome‐wide RIP‐seq analysis to identify the PUM2‐bound mRNA regions with antibody against PUM2. Peak and motif analysis revealed that the 5′‐UAA‐3′ motif was discovered to be highly enriched (P = 1e−8) (Figure 6A). Furthermore, PUM2 knockdown was then performed with Lenti‐PUM2‐shRNA infection in U2OS cells which has been shown to hold relative high expression level of PUM2. As shown in Figure 6B,C, Lenti‐PUM2‐shRNA infection significantly downregulated PUM2 and STARD13 expression. Then STARD13 3′UTR was overexpressed in OS cells with PUM2 knockdown via Lenti‐S‐UTR infection. Overexpression of STARD13 3′UTR attenuated the promotion of PUM2 knockdown on OS cells proliferation (Figure 6D,E), migration (Figure 6F,G), and stemness (Figure 6H,I). Thus, these results indicate that PUM2 suppresses OS cells progression at least partly through STARD13 3′UTR.

Figure 6.

Figure 6

Open in a new tab

PUM2 exerts its inhibitory effects dependent on STARD13 3′UTR. (A) RNA extracted from RIP was subjected to RNA‐seq and the binding site enriched was shown. (B, C) PUM2 mRNA and protein level was detected in OS cells with PUM2 knockdown or not. (D, E) Cell viability was examined in OS cells with PUM2 knockdown plus STARD13 3′UTR overexpression or not. (F, G) Cell migration ability was measured in cells depicted in (D). (H, I) Cell spheres size and numbers were evaluated in cells depicted in (D). Data were presented as the mean ± SD; *P < 0.05 vs control

3.7. Overexpression of PUM2 competitively bound to STARD13 3′UTR with miR‐590‐3p and miR‐9

Since miRNAs could bind to mRNA 3′UTR, miRNA.org (www.miRNA.org) predicted that various miRNAs could bind to STARD13 3′UTR, and we searched the miRNAs that could bind to 5′‐UAA‐3′ motif and found that miR‐590‐3p, miR‐9, miR‐653, and miR‐182 could bind to STARD13 3′UTR containing 5′‐UAA‐3′ motif (Figure 7A). Notably, miR‐9 had been proved to bind to STARD13 3′UTR in breast cancer in previous study.18 Luciferase reporter analysis indicated that miR‐590‐3p and miR‐9 mimics or inhibitors significantly decreased or increased the activity of Luc‐S‐3′UTR, respectively, but miR‐653 and miR‐182 could not (Figure 7B), indicating that STARD13 was a potential target of miR‐590‐3p and miR‐9 but not miR‐653 and miR‐182. To confirm the direct interaction between miR‐590‐3p, miR‐9 and STARD13 at endogenous levels, we performed RIP analysis to pull down endogenous miRNAs associated with Ago2 in Luc‐S‐3′UTR‐overexpressed MG63 cells. The precipitated miRNAs were subjected to RT‐qPCR analysis and results showed that miR‐590‐3p and miR‐9 were enriched in RNAs retrieved from Luc‐S‐3′UTR‐overexpressed cells (Figure 7C). Furthermore, RT‐qPCR and western blot assay indicated that STARD13 expression was inhibited or promoted by miR‐590‐3p or miR‐9 mimics, or inhibitors, respectively (Figure 7D,E), supporting that STRAD13 was the bona fide target of miR‐9 and miR‐590‐3p in OS cells. Next, we assumed that PUM2 could competitively bind to STARD13 3′UTR with miR‐590‐3p and miR‐9. As expected, the promotion of PUM2 overexpression on STARD13 expression was attenuated by miR‐590‐3p or miR‐9 overexpression (Figure 7F,G). Importantly, the inhibition of PUM2 on OS cells progression was rescued by miR‐590‐3p or miR‐9 overexpression too (Figure S2). Therefore, our results suggest that PUM2 could suppress OS cells progression via competitively binding to STARD13 3′UTR with miR‐590‐3p and miR‐9.

Figure 7.

Figure 7

Open in a new tab

Overexpression of PUM2 competitively bound to STARD13 3′UTR with miR‐590‐3p and miR‐9. (A) The diagram of STARD13 3′UTR with miRNA‐binding sites and 5′‐UAA‐3′ motif. (B) The luciferase activity of Luc‐S‐3′UTR was detected in MG63 cells with NC, miR‐590‐3p, miR‐9, miR‐653, or miR‐182 overexpression. (C) miR‐590‐3p and miR‐9 expression was measured in the Ago2‐IP materials after the RIP assay. (D, E) STARD13 mRNA and protein level was examined in MG63 cells with miR‐590‐3p or miR‐9 mimics or inhibitor transfection. (F, G) STARD13 mRNA and protein level was evaluated in MG63 cells with PUM2 overexpression plus miR‐590‐3p or miR‐9 overexpression or not. Data were presented as the mean ± SD; *P < 0.05 vs control

3.8. Overexpression of PUM2 inhibits OS cells progression via suppressing RhoA/Rock pathway

Finally, we tried to explore the downstream signalling of PUM2/STARD13 3′UTR axis. Previous studies have identified that STARD13 could block RhoA‐ROCK signalling axis by acting as a RhoGTPase‐activating proteins (GAP)26 and thus disorganizing F‐actin structure and RhoA‐ROCK signalling could contribute to tumour stemnes.27 Firstly, we detected whether STARD13 3′UTR could regulate RhoA/Rock signalling in OS cells. Indeed, G‐LISA RhoA activation assay showed that STARD13 3′UTR‐overexpressed cells displayed a lower basal level of RhoA activity compared with that in control cells (Figure S3A), and immunofluorescence staining of filamentous actin (F‐actin) with rhodamine‐labelled phalloidin revealed that stress fibre formation was markedly inhibited in cells with STARD13 3′UTR overexpression (Figure S3B). We further determined the effects of STARD13 3′UTR overexpression on RhoA downstream effector MLC phosphorylation level; pMLCS19 level was significantly decreased in STARD13 3′UTR‐overexpressed OS cells. Notably, STARD13 3′UTR unaffected RhoA expression in OS cells (Figure S3C). Importantly, overexpression of PUM2 exerted similar effects with STARD13 3′UTR on RhoA/Rock signalling (Figure 8A‐C). Next, we tried to prove that PUM2 could exert inhibitory effects through RhoA/Rock signalling, the downstream effector of STARD13 3′UTR. RhoA was overexpressed in OS cells with PUM2 overexpression by Lenti‐RhoA infection. RT‐qPCR and western blot confirmed the infection efficiency (Figure 8D,E). Consistently, RhoA overexpression attenuated the inhibition of PUM2 overexpression on OS cells proliferation (Figure 8F,G), migration (Figure 8H,I), and stemness (Figure 8J,K). Taken together, these results demonstrate that overexpression of PUM2 could inhibit OS cells progression via suppressing RhoA/Rock pathway.

Figure 8.

Figure 8

Open in a new tab

Overexpression of PUM2 inhibits OS cells progression via suppressing RhoA/Rock pathway. (A) RhoA activity was examined in OS cells with PUM2 overexpression or not. (B) The F‐actin formation was measured in cells depicted in (A). (C) The expression of p‐MLC, MLC, and RhoA was detected in cells depicted in (A). (D, E) RhoA mRNA and protein level was measured in OS cells with PUM2 overexpression plus RhoA overexpression or not. (F, G) Cell viability was evaluated in cells depicted in (D). (H, I) Cell migration was detected in cells depicted in (D). (J, K) Cell spheres size and numbers were examined in cells depicted in (D). Data were presented as the mean ± SD; *P < 0.05 vs control

4. DISCUSSION

OS is a primary malignancy that is most common in adolescents; although the survival of OS patients is significantly upregulated with adjuvant chemotherapy and surgery, metastasis and chemoresistance represent a clinical challenge for OS.28 Hence, it is important to elucidate the mechanisms and find novel targets for facilitating OS treatment.

In the present study, we identified that RBP PUM2 expression was significantly decreased in OS tissues, and inhibited OS cells proliferation, migration, and stemness. To the best of our knowledge, this is the first study showing the roles of PUM2 in OS progression. Further RIP‐seq and RNA sequencing were used to identify the potential targets of PUM2, and STARD13 was identified as a direct target of PUM2. Importantly, the 5′‐UAA‐3′ motif in STARD13 3′UTR was determined as the binding sites of PUM2. mRNA stabilization is regulated by RBPs and miRNAs; miRNAs associate with the RNA‐induced silencing complex (RISC) and recognize target mRNAs containing 3′UTRs with partially complementary sequence.15 RBPs influence mRNA stability and translation by interacting mainly with the 3′UTR.6 Here, we found that the binding sites of PUM2 were overlapped with the binding sites of miR‐590‐3p and miR‐9, and PUM2 regulated STARD13 expression via competitively binding with these two miRNAs. Notably, STARD13 has been identified as the target of miR‐9 in previous reports in breast cancer.18, 29 This suggests that miR‐9/STARD13 might be a common regulatory axis in tumours. Several studies have shown that RBPs posttranscriptionally regulate target mRNAs via the joint influence of miRNAs. For example, RBP HuR could recruit let‐7/RISC to repress c‐Myc expression.30 RNPC1 could inhibit nonsmall lung cancer progression via regulating miR‐181a/CASC2 axis and thus promote CASC2 expression31 and HuR facilitates stemness of lung cancer cells via competitively binding to CDK3 with miR‐873 and miR‐125a‐3p and thus increasing CDK3 expression.32 Thus, when miRNA‐binding sites overlap with present near binding sites of RBPs, RBPs could either antagonize or cooperate with miRNAs to regulate target genes expression. Here, we confirmed that PUM2 could antagonize miR‐590‐3p and miR‐9 functions on STARD13 expression.

By informatics methods analysis, PUM2 was predicted to have other targets, indicating that PUM2 likely has wide‐ranging roles in gene expression, and the regulatory mechanism for STARD12 expression may be more complex than simply competitive interaction between PUM2 and miR‐590‐3p and miR‐9. However, the competitive interaction between PUM2 and miR‐590‐3p, and miR‐9 may provide new insight into the regulation of STARD13 expression in OS cells. Since STARD13 has been identified as a tumour suppressor in other cancers,18, 19, 20, 21, 26 we are wondering whether PUM2/STARD13 3′UTR regulatory axis is a common phenomenon in other tumours; this could be explored in the future studies.

In sum, we showed that PUM2/STARD13 3′UTR regulatory axis could suppress OS cells progression through RhoA/Rock signalling. Despite more details need to be elucidated, we proposed that drugs promoting PUM2/STARD13 3′UTR regulatory axis might target OS and could be used as a combinative treatment with chemotherapy in OS therapy.

CONFLICTS OF INTEREST

The authors declare no conflict of interest.

Supporting information

 

Click here for additional data file. (3.4MB, tif)

 

Click here for additional data file. (2.7MB, tif)

 

Click here for additional data file. (2.5MB, tif)

ACKNOWLEDGEMENTS

This work is supported by National Natural Science Foundation of China (81501897) and Shanghai Municipal Commission of Health and Family Planning of science and Research Fund (20154Y0070).

Hu R, Zhu X, Chen C, Xu R, Li Y, Xu W. RNA‐binding protein PUM2 suppresses osteosarcoma progression via partly and competitively binding to STARD13 3′UTR with miRNAs. Cell Prolif. 2018;51:e12508 10.1111/cpr.12508

REFERENCES

  • 1. Martin‐Broto J, Redondo A, Valverde C, et al. Gemcitabine plus sirolimus for relapsed and progressing osteosarcoma patients after standard chemotherapy: a multicenter, single‐arm phase II trial of Spanish Group for Research on Sarcoma (GEIS). Ann Oncols. 2017;28(12):2994‐2999. [DOI] [PubMed] [Google Scholar]
  • 2. Schuetze SM. Incremental improvement in osteosarcoma chemotherapy? Ann Oncol. 2017;28(12):2911‐2913. [DOI] [PubMed] [Google Scholar]
  • 3. Nahas GR, Murthy RG, Patel SA, Ganta T, Greco SJ, Rameshwar P. The RNA‐binding protein Musashi 1 stabilizes the oncotachykinin 1 mRNA in breast cancer cells to promote cell growth. FASEB J. 2016;30(1):149‐159. [DOI] [PubMed] [Google Scholar]
  • 4. Sutherland JM, Sobinoff AP, Fraser BA, et al. RNA binding protein Musashi‐1 directly targets Msi2 and Erh during early testis germ cell development and interacts with IPO5 upon translocation to the nucleus. FASEB J. 2015;29(7):2759‐2768. [DOI] [PubMed] [Google Scholar]
  • 5. Yu TX, Gu BL, Yan JK, et al. CUGBP1 and HuR regulate E‐cadherin translation by altering recruitment of E‐cadherin mRNA to processing bodies and modulate epithelial barrier function. Am J Physiol Cell Physiol. 2016;310(1):C54‐C65. [DOI] [PubMed] [Google Scholar]
  • 6. Poria DK, Guha A, Nandi I, Ray PS. RNA‐binding protein HuR sequesters microRNA‐21 to prevent translation repression of proinflammatory tumor suppressor gene programmed cell death 4. Oncogene. 2016;35(13):1703‐1715. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Xu W, Chen C, Xu R, et al. Knockdown of HuR represses osteosarcoma cells migration, invasion and stemness through inhibition of YAP activation and increases susceptibility to chemotherapeutic agents. Biomed Pharmacother. 2018;102:587‐593. [DOI] [PubMed] [Google Scholar]
  • 8. Zahr SK, Yang G, Kazan H, et al. A translational repression complex in developing mammalian neural stem cells that regulates neuronal specification. Neuron. 2018;97(3):520‐537 e526. [DOI] [PubMed] [Google Scholar]
  • 9. Follwaczny P, Schieweck R, Riedemann T, et al. Pumilio2‐deficient mice show a predisposition for epilepsy. Dis Model Mech. 2017;10(11):1333‐1342. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Ginter‐Matuszewska B, Kusz K, Spik A, et al. NANOS1 and PUMILIO2 bind microRNA biogenesis factor GEMIN3, within chromatoid body in human germ cells. Histochem Cell Biol. 2011;136(3):279‐287. [DOI] [PubMed] [Google Scholar]
  • 11. Naidu S, Magee P, Garofalo M. MiRNA‐based therapeutic intervention of cancer. J Hematol Oncol. 2015;8:68. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Zheng L, Li X, Gu Y, Lv X, Xi T. The 3′UTR of the pseudogene CYP4Z2P promotes tumor angiogenesis in breast cancer by acting as a ceRNA for CYP4Z1. Breast Cancer Res Treat. 2015;150(1):105‐118. [DOI] [PubMed] [Google Scholar]
  • 13. Zhou Q, Zheng X, Chen L, et al. Smad2/3/4 pathway contributes to TGF‐beta‐induced MiRNA‐181b expression to promote gastric cancer metastasis by targeting Timp3. Cell Physiol Biochem. 2016;39(2):453‐466. [DOI] [PubMed] [Google Scholar]
  • 14. Adams CM, Hiebert SW, Eischen CM. Myc induces miRNA‐mediated apoptosis in response to HDAC inhibition in hematologic malignancies. Can Res. 2016;76(3):736‐748. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Zheng L, Meng X, Li X, et al. miR‐125a‐3p inhibits ERalpha transactivation and overrides tamoxifen resistance by targeting CDK3 in estrogen receptor‐positive breast cancer. FASEB J. 2017;32:588‐600. [DOI] [PubMed] [Google Scholar]
  • 16. Kuwano Y, Nishida K, Kajita K, et al. Transformer 2beta and miR‐204 regulate apoptosis through competitive binding to 3’ UTR of BCL2 mRNA. Cell Death Differ. 2015;22(5):815‐825. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Rehfeld F, Maticzka D, Grosser S, et al. The RNA‐binding protein ARPP21 controls dendritic branching by functionally opposing the miRNA it hosts. Nat Commun. 2018;9(1):1235. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Li X, Zheng L, Zhang F, et al. STARD13‐correlated ceRNA network inhibits EMT and metastasis of breast cancer. Oncotarget. 2016;7(17):23197‐23211. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Hanna S, Khalil B, Nasrallah A, et al. StarD13 is a tumor suppressor in breast cancer that regulates cell motility and invasion. Int J Oncol. 2014;44(5):1499‐1511. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20. Guo X, Xiang C, Zhang Z, Zhang F, Xi T, Zheng L. Displacement of Bax by BMF mediates STARD13 3'UTR‐induced breast cancer cells apoptosis in an miRNA‐dependent manner. Mol Pharm. 2018;15(1):63‐71. [DOI] [PubMed] [Google Scholar]
  • 21. Zhang H, Wang F, Hu Y. STARD13 promotes hepatocellular carcinoma apoptosis by acting as a ceRNA for Fas. Biotech Lett. 2017;39(2):207‐217. [DOI] [PubMed] [Google Scholar]
  • 22. Benegiamo G, Mure LS, Erikson G, et al. The RNA‐binding protein NONO coordinates hepatic adaptation to feeding. Cell Metab. 2018;27(2):404‐418 e407. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Xu CZ, Jiang C, Wu Q, Liu L, Yan X, Shi R. A feed‐forward regulatory loop between HuR and the long noncoding RNA HOTAIR promotes head and neck squamous cell carcinoma progression and metastasis. Cell Physiol Biochem. 2016;40(5):1039‐1051. [DOI] [PubMed] [Google Scholar]
  • 24. Fernandez HR, Gadre SM, Tan M, et al. The mitochondrial citrate carrier, SLC25A1, drives stemness and therapy resistance in non‐small cell lung cancer. Cell Death Differ. 2018;25:1239‐1258. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Liu R, Shi P, Nie Z, et al. Mifepristone suppresses basal triple‐negative breast cancer stem cells by down‐regulating KLF5 expression. Theranostics. 2016;6(4):533‐544. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Hu J, Li X, Guo X, et al. The CCR2 3'UTR functions as a competing endogenous RNA to inhibit breast cancer metastasis. J Cell Sci. 2017;130(19):3399‐3413. [DOI] [PubMed] [Google Scholar]
  • 27. Yoon C, Cho SJ, Aksoy BA, et al. Chemotherapy resistance in diffuse‐type gastric adenocarcinoma is mediated by RhoA activation in cancer stem‐like cells. Clin Cancer Res. 2016;22(4):971‐983. [DOI] [PMC free article] [PubMed] [Google Scholar] [Retracted]
  • 28. Lian Z, Han J, Huang L, et al. A005, a novel inhibitor of phosphatidylinositol 3‐kinase/mammalian target of rapamycin, prevents osteosarcoma‐induced osteolysis. Carcinogenesis. 2018; https://doi.org.10.1093/carcin/bgy036. [Epub ahead of print]. [DOI] [PubMed] [Google Scholar]
  • 29. D'Ippolito E, Plantamura I, Bongiovanni L, et al. miR‐9 and miR‐200 regulate PDGFRbeta‐mediated endothelial differentiation of tumor cells in triple‐negative breast cancer. Can Res. 2016;76(18):5562‐5572. [DOI] [PubMed] [Google Scholar]
  • 30. Kim HH, Kuwano Y, Srikantan S, Lee EK, Martindale JL, Gorospe M. HuR recruits let‐7/RISC to repress c‐Myc expression. Genes Dev. 2009;23(15):1743‐1748. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Yang L, Zhang Y, Ling C, Heng W. RNPC1 inhibits non‐small cell lung cancer progression via regulating miR‐181a/CASC2 axis. Biotech Lett. 2018;40(3):543‐550. [DOI] [PubMed] [Google Scholar]
  • 32. Zhang Y, Yang L, Ling C, Heng W. HuR facilitates cancer stemness of lung cancer cells via regulating miR‐873/CDK3 and miR‐125a‐3p/CDK3 axis. Biotech Lett. 2018;40(4):623‐631. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

 

Click here for additional data file. (3.4MB, tif)

 

Click here for additional data file. (2.7MB, tif)

 

Click here for additional data file. (2.5MB, tif)

Articles from Cell Proliferation are provided here courtesy of Wiley

RESOURCES