MiR-18b-3p promotes cell proliferation and metastasis by directly inhibiting PTEN expression in osteosarcoma

Abstract

Background

See More

Osteosarcoma (OS), the most prevalent primary malignant bone tumor in adolescents, demonstrates aggressive clinical behavior and poor prognosis. This study aimed to elucidate the functional role and molecular mechanism of miR-18b-3p in OS pathogenesis.

Methods

The miRNA microarray data from the GEO database were evaluated through GEO2R. The expression patterns of miR-18b-3p and PTEN were systematically analyzed using RT-qPCR and western blotting. A dual-luciferase reporter assay validated the direct targeting relationship between miR-18b-3p and PTEN 3′-UTR. Functional experiments included CCK-8, wound healing, and transwell invasion assays, and xenograft mouse models. Bioinformatics predictions were further supported by Gene Ontology enrichment analysis, protein–protein interaction network construction, and single-cell RNA sequencing analysis.

Results

Integrated bioinformatics interrogation identified miR-18b-3p as a prognostically characteristic gene, with its overexpression strongly correlating with reduced overall survival in OS patients (p = 0.046). Clinical specimens and cellular models revealed marked upregulation of miR-18b-3p in OS tissues and cell lines (U2OS, MG63, HOS, SAOS2) compared to normal controls. PTEN was subsequently identified as a direct downstream target through complementary binding site verification, showing inverse correlation with miR-18b-3p expression. Functional experiments demonstrated that miR-18b-3p overexpression significantly enhanced malignant phenotypes, including cellular proliferation, migration, and invasion capacity, while PTEN restoration effectively reversed these oncogenic effects. Additionally, OS cells (HOS and SAOS2) secreted miR-18b-3p, and this paracrine mechanism potentially reprogrammed the tumor microenvironment through suppressing PTEN expressed in various immune and stromal cells.

Conclusions

Our findings establish miR-18b-3p as a novel oncogenic regulator in OS pathogenesis through direct PTEN targeting, highlighting its potential as a therapeutic target.

Background

Osteosarcoma (OS), a highly aggressive primary bone malignancy originating from mesenchymal progenitors, predominantly affects adolescents and young adults [1]. Despite multimodal therapeutic advances including neoadjuvant chemotherapy and surgical resection, the global mortality burden remains substantial, with 5-year survival rates stagnating at 60–70% for localized disease [2]. Prognosis deteriorates dramatically in metastatic or recurrent cases, where survival probabilities plummet below 30% at five years [3]. These clinical realities underscore the critical need to elucidate molecular drivers of OS pathogenesis to inform novel therapeutic strategies.

MicroRNAs (miRNAs), a conserved class of endogenous noncoding RNAs (~ 21–25 nucleotides), orchestrate post-transcriptional regulation through sequence-specific binding to 3′-untranslated regions (3′-UTRs) of target mRNAs, mediating either transcript degradation or translational repression [4]. This regulatory paradigm enables miRNAs to modulate fundamental cellular processes of musculoskeletal diseases [5, 6]. Emerging oncogenomic investigations have established miRNA dysregulation as a hallmark of multiple malignancies, with relevance to OS progression and treatment resistance [7, 8]. For instance, Jiang et al. recently identified exosome-mediated transfer of miR-144-3p as a ferroptosis-inducing mechanism that constrains OS proliferation and invasion [7]. Intriguingly, miR-18b-3p (previously designated miR-18b) demonstrates paradoxical roles across tumor types, functioning as an oncogenic driver in nasopharyngeal carcinoma [9], hepatocellular carcinoma [10], and colorectal cancer [11], while exhibiting tumor-suppressive activity in melanoma and breast malignancies [12, 13]. Despite these context-dependent functionalities, the biological significance and mechanistic involvement of miR-18b-3p in OS remain entirely unexplored.

Our study systematically characterized miR-18b-3p’s role in OS pathogenesis. We identified significant miR-18b-3p upregulation in clinical OS specimens and cellular models, mechanistically elucidating its tumor-promoting effects through direct PTEN suppression, positioning this miRNA as a potential therapeutic target.

Methods

Clinical samples

Twenty-six paired OS tissue samples and matched adjacent normal tissues (≥ 5 cm from tumor margins) were prospectively collected from patients undergoing surgical resection at Yancheng Fist Hospital Affiliated to Nanjing University Medicine School. All specimens underwent histopathological confirmation by two independent board-certified pathologists. Clinicopathological parameters, including age, gender, tumor stage, and metastasis status, are summarized in Table 1. Written informed consent was obtained from all participants, and the study protocol received ethical approval from Yancheng First Hospital.

Table 1 Patients clinicopathological characteristics

Cell culture

Human osteosarcoma cell lines (U2OS, MG63, HOS, SAOS2) and the normal osteoblastic cell line hFOB1.19 were acquired from the Shanghai Institute of Biological Sciences, Chinese Academy of Sciences. Cells were maintained in Dulbecco’s Modified Eagle Medium (DMEM; Gibco) supplemented with 10% fetal bovine serum (FBS; Gibco) under standardized culture conditions (37 °C, 5% CO₂ humidified atmosphere). All cell lines underwent routine mycoplasma contamination testing.

RNA isolation and quantitative real-time PCR

Total RNA was extracted using TRIzol reagent (Invitrogen) following the manufacturer’s guidelines. Complementary DNA (cDNA) synthesis was performed with PrimeScript RT Master Mix (Takara) for mRNA analysis, while miRNA-specific cDNA was generated using the Hairpin-it™ miRNA RT-PCR Quantitation Kit (GenePharma). Quantitative PCR amplification was performed using SYBR Premix Ex Taq™ (Takara) on a QuantStudio 6 Flex system (Applied Biosystems), with GAPDH and U6 snRNA serving as endogenous controls for mRNA and miRNA normalization, respectively. Primer sequences: PTEN: Forward 5′-GCGTGCAGATAATGACAAGG-3′, Reverse 5′-GGATTTGACGGCTCCTCTAC-3′; GAPDH: Forward 5′-GAAGGTGAAGGTCGGAGTC-3′, Reverse 5′-GAAGATGGTGATGGGATTTC-3′.

Relative expression levels were calculated using the 2^−ΔΔCt method.

Gene modulation

Chemically modified miR-18b-3p inhibitors, PTEN-targeting siRNAs (siPTEN), and scrambled negative controls (NC) were synthesized by GenePharma. Transient transfection was performed using Lipofectamine 3000 (Invitrogen) at 50 nM final concentration, with transfection efficiency validated 48 h post-transfection via RT-qPCR and immunoblotting.

Western blot analysis

Cellular proteins were extracted using RIPA lysis buffer (Beyotime) containing protease inhibitors. Protein lysates (30 μg/lane) were resolved by 10% SDS-PAGE and electrotransferred to PVDF membranes (Millipore). After blocking with 5% non-fat milk, membranes were incubated overnight at 4 °C with anti-PTEN antibody (1:2000; Proteintech, 22034-1-AP) followed by HRP-conjugated secondary antibody (1:5000; Beyotime). Protein bands were visualized using an ECL detection system (Keygen), with GAPDH serving as a loading control.

CCK-8 assay

For the Cell Counting Kit-8 (CCK-8) assay, HOS and SAOS2 cells were seeded into 96-well plates at a density of 1 × 10³ cells/well and were maintained in DMEM supplemented with 10% (v/v) FBS. The absorbance was measured at 450 nm after culture for 0, 1, 2, and 3 days by using a CCK-8 kit (Keygen, China) according to the protocol.

Scratch healing assay

After transfection, HOS and SAOS2 cells were planted into 6-well plates and cultured in DMEM supplemented with 10% (v/v) FBS for 48 h. Subsequently, pipette tips (200 μl) were used to scratch three parallel lines, and cells were washed with PBS twice. After that, cells were cultured in DMEM supplemented with low FBS concentration (1%) at an atmosphere of 5% CO₂ at 37 °C. Photographs were taken at 0 and 48 h after wounding under an Olympus FSX100 microscope (Olympus, Tokyo, Japan). The relative migration index is calculated using the following equation: (0 h scratch width–48 h scratch width)/0 h scratch width.

Transwell invasion assay

The top transwell chambers (BD Biosciences, CA, USA) were coated with Matrigel mix (BD Biosciences, CA, USA). Then DMEM containing 10% (v/v) FBS was added to the bottom chamber while OS cells seeded in the top chamber were cultured in the DMEM without FBS. After 24 h, OS cells that passed through the matrigel to the underside of the filter were fixed with methanol and stained with 0.1% crystal violet. Thereafter, the stained cells were counted under an Olympus FSX100 microscope (Olympus, Tokyo, Japan). The relative invasion index is calculated using the following equation: the number of cells in the treatment or control groups/ the average number of cells in the control group.

Establishment of xenograft mouse model

Six-week-old male BALB/c nude mice were procured from Gempharmatech Biotechnology Co., Ltd. (Jiangsu, China). For tumor implantation, HOS cells were prepared as a single-cell suspension (2 × 10⁶ cells/mouse) in 100 μl of ice-cold Matrigel matrix (Corning, NY, USA) and subsequently administered via subcutaneous injection into the right flank region. Tumor growth was monitored weekly, with tumor volume calculated according to the modified ellipsoid formula: Volume (mm³) = (short axis² × long axis)/2. The tumor grafts were achieved after five weeks. All experimental procedures were conducted in strict compliance with the Institutional Animal Care and Use Committee guidelines established by Yancheng First Hospital Affiliated to Nanjing University Medical School, following the NIH Guide for the Care and Use of Laboratory Animals.

Luciferase reporter assay

Wild-type (WT) and mutant (MUT) PTEN 3′-UTR sequences were cloned into pmirGLO dual-luciferase vectors (GeneCreat). 293 T cells were co-transfected with reporter constructs (100 ng) and miR-18b-3p inhibitor/NC (50 nM) using Lipofectamine 2000. Firefly/Renilla luciferase activities were measured 48 h post-transfection using a GloMax Navigator system (Promega).

Bioinformatics analysis

OS-related miRNA expression dataset (GSE28423) and scRNA-seq dataset were retrieved from the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/) and further analyzed by using the online tool GEO2R and Tumor Immune Single-cell Hub 2 (TISCH2) database. Gene Ontology (GO) analysis was conducted by using the DAVID website (https://david.ncifcrf.gov). Protein–protein interaction (PPI) analysis was carried out by using String (https://string-db.org) and Cytoscape.

Statistical analysis

Data represent mean ± SD from triplicate experiments. Intergroup differences were assessed using two-tailed Student’s t-tests or ANOVA with Tukey’s post-hoc test. Clinical correlations were analyzed via χ² tests. Survival associations were determined by Kaplan–Meier/log-rank methods. All analyses were performed using SPSS 25.0 and GraphPad Prism 9.0, with p < 0.05 considered statistically significant.

Results

MiR-18b-3p was upregulated in OS and associated with advanced disease progression

To systematically identify dysregulated miRNAs in osteosarcoma pathogenesis, we performed differential expression analysis of the GSE28423 miRNA microarray dataset encompassing 11 OS cell lines versus normal bone tissues. Under stringent thresholds (|log2FC|> 1 and p < 0.05), 238 miRNAs demonstrated significant expression alterations, comprising 127 upregulated and 111 downregulated species (Fig. 1A). The ten most differentially expressed miRNAs included miR-18b-3p, miR-18a, miR-301a, miR-9, miR-31-5p, miR-451, miR-144, miR-142-3p, miR-223, and miR-1 (Fig. 1B). LASSO regression with identified miR-18b-3p, miR-144, miR-223, and miR-451 as core discriminative features for OS classification (Fig. 1C). Given the unreported status of miR-18b-3p in OS biology, we prioritized this candidate for functional investigation. Survival analysis of the GSE39052 cohort (n = 26) revealed significantly reduced overall survival in patients with high miR-18b-3p expression (Fig. 1D). Validation in clinical specimens through qPCR demonstrated significant higher miR-18b-3p levels in OS tissues compared to matched normal counterparts (Fig. 1E), with cellular models confirming overexpression across four OS cell lines relative to hFOB1.19 osteoblasts (Fig. 1F). Clinico-molecular correlation analysis established significant associations between elevated miR-18b-3p expression and aggressive disease features: tumor size, advanced TNM stage, and pulmonary metastasis (Table 1). Consistently, metastatic OS specimens exhibited significant higher miR-18b-3p levels than non-metastatic cases (Fig. 1G), underscoring its potential role in disease progression.

Fig. 1

miR-18b expression was significantly upregulated in OS. A Analysis of dysregulated miRNAs in OS cell lines based on GSE28423. B The top ten dysregulated miRNAs in OS cell lines compared to normal human bones. C The characteristic dysregulated miRNAs of OS identified through LASSO regression analysis. D The association between miR-18b-3p expression and OS patients’ overall survival. E The expression of miR-18b was significantly upregulated in OS tissues compared to adjacent normal tissues. F The expression of miR-18b was significantly upregulated in OS cell lines compared to hFOB1.19 cells. G The expression of miR-18b was significantly upregulated in OS patients with metastasis compared to those without metastasis. **p < 0.01, ***p < 0.001

Functional characterization of miR-18b-3p in OS oncogenesis

To investigate the regulatory role of miR-18b-3p in OS, we first knocked down miR-18b-3p expression. Genetic ablation of miR-18b-3p through inhibitor transfection achieved efficient silencing in HOS and SAOS2 cells (Fig. 2A). CCK-8 assay showed that miR-18b-3p knockdown decreased cellular viability after transfection (Fig. 2B). Scratch healing assay showed that wound closure rates decreased at 48 h compared to controls (Fig. 2C). Moreover, the transwell invasion assay showed that matrigel-transmigrated cells were reduced following miR-18b-3p inhibition (Fig. 2D). In vivo validation using subcutaneous xenografts demonstrated concordant results, with miR-18b-3p-silenced tumors exhibiting significantly smaller volumes versus controls at endpoint (Fig. 2E).

Fig. 2

miR-18b knockdown significantly inhibited the proliferation, migration, and invasion of OS cells. A miR-18b inhibitor significantly suppressed miR-18b expression in HOS and SAOS2 cells. B CCK-8 assay showed that miR-18b inhibition significantly suppressed the proliferation of OS cells. C Scratch healing assay showed that miR-18b inhibition significantly suppressed the migration of HOS and SAOS2 cells. D Transwell invasion assay showed that miR-18b inhibition significantly suppressed the invasion of OS cells. E In vivo models proved that miR-18b inhibition significantly suppressed the growth of OS cells. **p < 0.01, ***p < 0.001

Systematic bioinformatics approach identifies PTEN as a key functional target of miR-18b-3p

A multi-platform bioinformatics pipeline was employed to delineate miR-18b-3p’s molecular targets. Intersection analysis of TargetScan and miRDB predictions yielded 226 conserved target genes (Fig. 3A). Functional enrichment through DAVID revealed a significant association of these targets with biological processes including “cell migration” and “positive regulation of cell proliferation” (Fig. 3B). Protein interaction network construction using STRING identified PRKAC8 and PTEN proteins as top network hubs (Fig. 3C). Subsequent MCODE module analysis (Cytoscape; node score cutoff = 0.4) pinpointed PTEN protein as the central node within the highest-scoring cluster, exhibiting maximal connectivity to downstream effectors (Fig. 3D). Experimental validation confirmed PTEN mRNA levels were significantly reduced in OS tissues versus normal counterparts (Fig. 3E), demonstrating a strong inverse correlation with miR-18b-3p expression (Fig. 3F).

Fig. 3

Selecting potential targets of miR-18b in OS. A The potential targets of miR-18b predicted by TargetScan and miRDB. B GO analysis of putative targets of miR-18b. C PPI analysis of putative targets of miR-18b. D MCODE identified a functional module among these miR-18b targets. E PTEN expression was significantly downregulated in OS tissues. F PTEN expression was positively correlated with miR-18b expression in OS tissues. ***p < 0.001

Validating PTEN as a direct target of miR-18b-3p in OS

Consistent with clinical observations, comparative analysis revealed significantly lower PTEN mRNA levels (Fig. 4A) and reduced protein expression (Fig. 4B) in OS cell lines versus hFOB1.19 osteoblasts. Functional modulation experiments demonstrated that miR-18b-3p inhibition elevated PTEN transcript levels (Fig. 4C) and protein abundance (Fig. 4D), confirming miRNA-mediated regulation. Dual-luciferase reporter assays provided direct evidence of molecular interaction: miR-18b-3p suppression increased wild-type PTEN 3′-UTR luciferase activity, while mutant PTEN 3′-UTR constructs showed no responsiveness (Fig. 4E, F). This binding specificity conclusively establishes PTEN as a direct miR-18b-3p target through 3′-UTR complementarity in OS pathogenesis.

Fig. 4

Identifying PTEN as a direct target of miR-18b in OS. A, B The expression of PTEN was significantly downregulated at mRNA and protein levels in OS cell lines compared to hFOB1.19. C, D PTEN expression was significantly elevated at mRNA and protein levels after miR-18b inhibition. E, F miR-18b inhibitor significantly increased the luciferase activity of the wild-type (Wt) reporter rather than the Mut reporter. *p < 0.05, **p < 0.01, ***p < 0.001

PTEN knockdown significantly reverses the suppressive role of the miR-18b-3p inhibitor in OS

To establish PTEN as the critical mediator of miR-18b-3p’s oncogenic activity, we performed combinatorial genetic perturbations through co-transfection of miR-18b-3p inhibitor and PTEN-targeting siRNA (50 nM each). Immunoblot analysis confirmed that siPTEN effectively counteracted the PTEN upregulation induced by miR-18b-3p inhibition (Fig. 5A). The CCK-8 assay showed that PTEN knockdown reversed the growth-inhibitory effect of miR-18b-3p suppression (Fig. 5B, C). Consistently, scratch healing assay and transwell invasion assay showed that both cell migration and invasion abilities suppressed by miR-18b-3p depletion were re-established after PTEN knockdown in OS cells (Fig. 5D, E). These results mechanistically confirm that miR-18b-3p exerts its pro-tumorigenic effects primarily through PTEN suppression in OS pathogenesis.

Fig. 5

PTEN knockdown significantly reversed the suppressive role of miR-18b inhibitor in OS. A Co-transfecting miR-18b inhibitor and siPTEN significantly reversed the suppressive effect of miR-18b inhibitor on PTEN expression. B The CCK-8 assay showed that miR-18b knockdown impaired cell growth, whereas PTEN downregulation reversed such an effect. C, D Scratch healing assay and transwell invasion assay showed that the migration and invasion abilities of OS cells suppressed by miR-18b depletion were re-established after PTEN knockdown. *p < 0.05, **p < 0.01, ***p < 0.001

The miR-18b-3p/PTEN axis may modulate the tumor microenvironment in OS

Emerging evidence indicates tumor-derived miRNAs can orchestrate microenvironmental remodeling through paracrine regulation of stromal and immune compartments [14]. Our qPCR analysis revealed that miR-18b-3p is actively secreted by OS cells, with extracellular miR-18b-3p levels in conditioned media significantly reduced following cellular knockdown (Fig. 6A). To delineate the axis’s microenvironmental relevance, we interrogated the GSE162454 scRNA-seq dataset encompassing 46,544 cells from OS specimens. Uniform Manifold Approximation and Projection (UMAP) clustering identified seven major stromal/immune populations: conventional CD4⁺T cells, exhausted CD8⁺T cells, endothelial cells, cancer-associated fibroblasts, tumor-associated macrophages, osteoblasts, and plasma cells (Fig. 6B), whose representative markers were exhibited in Fig. 6C. PTEN exhibited pan-microenvironmental expression across all cell types, with particularly high abundance in macrophages (Fig. 6D). This ubiquitous PTEN distribution supports the hypothesis that tumor-secreted miR-18b-3p may simultaneously target malignant and stromal compartments to drive microenvironmental reprogramming.

Fig. 6

The miR-18b-3p/PTEN axis may modulate the tumor microenvironment in OS. A The change of miR-18b-3p expression in cultured medium after intercellular miR-18b-3p knockdown. B The major cell types identified by scRNA-seq analysis. C The representative markers of the identified major cell types. D The distribution of PTEN in the identified major cell types of OS tissues. *p < 0.05

Discussion

Emerging evidence underscored the critical involvement of non-coding RNA dysregulation in musculoskeletal diseases, with specific miRNAs demonstrating either tumor-suppressive or oncogenic propensities in OS pathogenesis [15, 16]. For instance, miR-486-3p and miR-371b-5p have been shown to suppress malignant phenotypes by inhibiting proliferation and metastasis in OS [17, 18], whereas miR-624-5p drives oncogenic transformation through Hippo pathway inactivation [19]. This functional dichotomy highlights the complexity of miRNA-mediated regulation in bone malignancies.

The present study elucidates the context-dependent duality of miR-18b-3p (previously miR-18b), which exhibits tissue-specific oncogenic or tumor-suppressive roles across malignancies. While it promotes tumorigenesis in nasopharyngeal carcinoma [9], hepatocellular carcinoma [10], and colorectal cancer [11], it conversely suppresses progression in melanoma and breast cancer [12, 13]. Our findings resolve its previously unexplored role in OS, demonstrating significant miR-18b-3p upregulation in clinical specimens and cellular models, with functional assays confirming its tumor-promoting effects through enhanced proliferation, migration, and invasion.

The tumor suppressor PTEN protein is a lipid phosphatase that constrains oncogenic signaling via phosphatidylinositol-3,4,5-trisphosphate (PIP3) dephosphorylation, thereby inhibiting AKT/mTOR activation [20]. While PTEN protein is frequently inactivated through genomic alterations in solid tumors [21], our study revealed miRNA-mediated post-transcriptional regulation as an alternative inactivation route in OS. Through systems biology approaches integrating target prediction algorithms and network topology analysis, we identified PTEN as the highest-confidence miR-18b-3p target within a functionally enriched module. Experimental validation confirmed direct 3′-UTR binding (luciferase activity modulation) and inverse miR-18b-3p/PTEN expression correlation (R = − 0.60), with PTEN reconstitution reversing miR-18b-3p-driven oncogenicity. Inhibition of miR-18b-3p, using synthetic small interfering RNAs (siRNAs) or locked nucleic acids (LNAs), represents a potential strategy to restore PTEN expression and suppress OS progression. Many genes and cytokines participate in the pathological process of musculoskeletal diseases and have been targeted for regulation with siRNA [22,23,24]. Translating this strategy into a viable in vivo therapy requires overcoming significant challenges in delivery. Promising approaches include nanoparticle-based carriers (e.g., lipid nanoparticles (LNPs), polymeric nanoparticles, or inorganic nanoparticles) designed to protect the oligonucleotides from degradation, enhance tumor accumulation via the enhanced permeability and retention (EPR) effect or active targeting ligands (e.g., peptides, antibodies against OS surface markers), and facilitate cellular uptake.

Notably, our work extended beyond cell-autonomous effects to microenvironmental modulation. Tumor-derived miRNAs are increasingly recognized as mediators of microenvironment reprogramming [25]. We demonstrated that OS cells actively secrete miR-18b-3p into extracellular compartments (significant reduction upon cellular knockdown), suggesting a potential paracrine signaling mechanism. Our scRNA-seq analysis of OS ecosystems revealed ubiquitous PTEN expression across stromal and immune subsets, including tumor-associated macrophages and fibroblasts. Given PTEN’s established role in immune cell function, particularly its capacity to enhance anti-tumor immunity through macrophage polarization [21, 22], our findings suggest a plausible dual mechanism whereby miR-18b-3p could simultaneously drive intrinsic tumor progression and potentially extrinsic immunosuppression via PTEN repression in stroma and immune cells. However, direct experimental validation of phenotypic alteration (e.g., macrophage polarization shift, fibroblast activation) in specific microenvironmental cell types upon exposure to OS-derived miR-18b-3p is required to fully substantiate this paracrine reprogramming hypothesis.

While our study provides compelling evidence for the miR-18b-3p/PTEN axis in OS, several limitations should be acknowledged. First, the clinical cohort size is relatively small; validation in larger, independent patient cohorts would strengthen the clinical relevance of miR-18b-3p as a prognostic biomarker. Second, our functional in vivo validation was limited to subcutaneous xenograft models assessing primary tumor growth; orthotopic or metastatic models would provide more physiologically relevant insights into the role of miR-18b-3p in OS invasion and metastasis. Third, the study utilized established cell lines; incorporating primary patient-derived OS cells or organoids could enhance the translational relevance of the findings.

Conclusions

In conclusion, our study established miR-18b-3p as a novel therapeutic vulnerability in OS through its direct targeting of PTEN and microenvironmental reprogramming effects. The miR-18b-3p/PTEN axis represents a promising therapeutic target. Future research focusing on overcoming delivery barriers for miR-18b-3p inhibitors and validating its paracrine role in the TME is warranted to translate these findings into clinical applications.