Osteopontin derived from hypoxia-induced M2 macrophages promotes osteosarcoma progression through modulation of EGR3/ISG15 signaling and RIG-I expression

Abstract

Background

Osteosarcoma (OS) is one of the most common malignancies arising in bone. Hypoxia and immune regulation are pivotal in tumor biology. However, their combined effects and mechanisms in OS remain understudied. This study aimed to explore the role and mechanism of hypoxic-induced M2 macrophages in promoting the progression of OS.

Methods

Differentially expressed proteins in hypoxic macrophage supernatants were detected by antibody array. Cell functional experiments, siRNA-mediated gene silencing, and overexpression transfection were used to study osteopontin (OPN) and its supernatant effect. Bioinformatics analysis was applied to investigate downstream targets and pathways, and a xenograft model was established to assess in vivo effects.

Results

Our data revealed that hypoxic M2 macrophage supernatant enhanced OS malignancy and epithelial-mesenchymal transition, activating cancer pathways. Hypoxia upregulated OPN in M2 macrophages, and OPN inhibition reduced its tumor-promoting effect. Early growth response 3 (EGR3) was differentially expressed in OS cells treated with the supernatant, and its overexpression inhibited OS cell migration, reversing tumor promotion. Interferon-stimulated gene 15 (ISG15), a key differentially expressed gene related to OPN and EGR3 overexpression, inhibited OS cell proliferation and migration. Additionally, OPN increased retinoic acid-inducible gene I (RIG-I) expression and enhanced signal transducer and activator of transcription 3, nuclear factor kappa B, and extracellular signal-regulated kinase signaling, while EGR3 and ISG15 overexpression inhibited these effects. Silencing ISG15 restored pathway activation and reversed the inhibitory effect of EGR3 on OS cell migration. Dual-Luciferase reporter gene assay confirmed that EGR3 activates ISG15 transcription. OPN treatment upregulated DNA (cytosine-5)-methyltransferase 1 (DNMT1) expression, and ChIP assays demonstrated that EGR3 overexpression enhanced DNMT1 binding to the EGR3 promoter. These findings suggest that OPN promotes OS malignancy by downregulating EGR3 and ISG15, and by enhancing RIG-I expression, as validated in a xenograft model of OS.

Conclusion

Our findings demonstrate that hypoxic-induced M2 macrophages promote OS progression through OPN-dependent mechanisms, including inhibition of EGR3 and ISG15 expression and upregulation of RIG-I.

Graphical abstract

Introduction

Osteosarcoma (OS), the most common primary malignant bone tumor, shows a unique bimodal age distribution. The primary peak occurs in children and adolescents (median age of 18), and a second, smaller peak is observed in older adults over 60 years of age [1]. OS mainly occurs in long bones like the femur, tibia, and humerus, near growth plates, and rarely in the skull, jaw, and pelvis, and is characterized by malignant osteoblastic cells producing osteoid matrix, a feature associated with aggressive behavior and high metastatic potential [2]. While 68% of localized OS patients achieve 5 + years survival, 20–30% still develop metastasis or recurrence, and the prognosis for advanced cases has remained poor despite available anticancer therapies. [3–6]. Therefore, uncovering the molecular mechanisms of OS progression and identifying reliable biomarkers is essential to advance precision oncology and immunotherapeutic strategies.

Bone tumors, notably OS, develop within a unique bone tumor microenvironment (TME), which is a specialized and dynamic condition comprising bone cells (osteoclasts, osteoblasts, and osteocytes), stromal cells (mesenchymal stem cells and fibroblasts), vascular cells (endothelial cells and pericytes), immune cells (macrophages and lymphocytes), and a mineralized extracellular matrix (ECM) [7]. Under physiological conditions, a precise interplay between these cells maintains bone homeostasis, while in primary and secondary tumors, the bone TME provides a favorable microenvironment for tumor cell growth and progression [2]. The crosstalk between OS and the bone TME encompasses an intricate network of environmental signals triggered by myriad cytokines, chemokines, and soluble growth factors [8]. Hypoxia, a hallmark of the TME, reprograms gene and protein expression, contributing to tumor progression and poor prognosis [9, 10]. For example, slowly proliferating cells in low-oxygen zones can resist cytotoxic drugs targeting fast-growing cells. In contrast, cancer stem cells may thrive there, facilitating epithelial-to-mesenchymal transition (EMT) and driving tumor progression [11].

Tumor-associated macrophages (TAMs) are a major immune cell group within the TME and can account for up to 50% of the cellular content in certain solid tumors [12]. They promote disease progression and therapy resistance by supplying nutrients and growth factors to malignant cells. TAMs exhibit significant heterogeneity across cancer patients, lesions, and tumors. This arises from their ability to adapt to the environment, adopting diverse phenotypes, metabolism, and functions, ranging from pro-inflammatory M1-like to anti-inflammatory M2-like states [13]. TAMs exhibit significant heterogeneity shaped by the complex TME, with hypoxia acting as a key driver that recruits TAMs to low-oxygen regions and promotes their polarization toward tumor-supportive phenotypes. For instance, studies have found that hypoxic niches drive TAM immunosuppression—a process exacerbated by vascular abnormalities induced by adaptive immunity in glioblastoma. Targeting these niches can disrupt TAM organization and improve tumor control [14]. Hypoxia also alters tumor-macrophage communication, favoring M2 polarization, which boosts tumor growth, angiogenesis, and immune suppression, unlike M1 macrophages, which inhibit tumors. For example, activation of hypoxia-inducible factor-1α (HIF-1α) and HIF-2α is one of the main initiators of macrophage-head and neck squamous cell carcinoma cell interactions [15]. Additionally, hypoxia regulates the communication between TAMs and tumor cells by influencing the expression and release of mediators such as extracellular vesicles, cytokines, and growth factors. These interactions are crucial for tumor survival, proliferation, and invasion [13].

Increasing evidence has shown that hypoxia or the immune microenvironment plays a vital role in the development of OS [16–18]. However, reliable gene signatures based on the combination of hypoxia and the immune status for OS prognostic prediction have not been identified. This study aims to identify gene expression and biomarkers associated with both hypoxia and immune status and provide novel insights into the molecular mechanisms underlying the progression of OS.

Materials and methods

Isolation and differentiation of bone marrow-derived macrophages (BMDMs)

Bone marrow-derived macrophages (BMDMs) were generated from BALB/c mice aged 6 to 8 weeks, which were obtained from Shanghai SLAC Laboratory Animal Co., Ltd. (License No. SCKK 2022-0004). Following euthanasia via cervical dislocation, femurs and tibias were collected under sterile conditions. After removing surrounding muscle tissues, both ends of the bones were excised, and the bone marrow was flushed out using a 1 mL syringe. The resulting suspension was passed through a 70 μm filter to eliminate tissue debris. Subsequently, red blood cell lysis buffer (2–3 mL) was applied to eliminate erythrocytes. Cells were pelleted by centrifugation at 300 × g for 5 min, and the resulting pellet was resuspended in complete Roswell Park Memorial Institute 1640 medium (RPMI-1640 medium containing 10% fetal bovine serum (FBS, s-FBS-x-015, SERANA, China) and 1% penicillin/streptomycin (P/S, 15140-122, Gibco, USA). For macrophage differentiation, cells were seeded at 1.5 × 10⁶ per dish in a medium supplemented with 10 ng/mL macrophage colony-stimulating factor (M-CSF, HY-P70553, MCE, USA). Cultures were maintained at 37℃ in an incubator with 5% CO₂ and high humidity. On the third day, half of the medium was replaced with fresh medium, followed by complete medium replenishment on day 5. By day 7, the adherent macrophage population was collected and designated mature BMDMs for subsequent assays.

Cell culture

The cell lines HOS (YDT-0265), MG-63 (YDT-0400), THP-1 (YDT-0666), K7M2 (YDT-0317), and 293T (YDT-0019) utilized in this study were procured from INDIT Bio-Technology Co., Ltd (Hangzhou, China). The HOS and 293T cell lines were cultivated in Minimum Essential Medium (MEM, C11095500BT, BDBIO, China), while Dulbecco’s Modified Eagle Medium (DMEM, 02-5062EJ, BDBIO, China) was used to maintain MG-63 and K7M2 cells. The THP-1 cell line was grown in RPMI-1640. All cultures were incubated at 37 °C in a humidified atmosphere containing 5% carbon dioxide, and all media were supplemented with 10% fetal bovine serum and 1% penicillin–streptomycin.

Cells under hypoxic conditions were cultured in a tri-gas incubator (1% O₂, 5% CO₂, 94% N₂) to simulate a low-oxygen environment. Normoxic conditions were maintained at 21% O₂ and 5% CO₂. These oxygen levels were strictly controlled and consistently applied across all experiments involving normoxia-hypoxia comparisons.

Induction of M1-type macrophages in vitro

THP-1 cells underwent differentiation into M0 macrophages following 24 h exposure to 100 ng/mL phorbol 12-myristate 13-acetate (PMA, HY-18739, MCE, USA). Following a 24 h recovery in a fresh complete medium, cells were cells were treated with 100 ng/mL lipopolysaccharide (LPS, L2630, Sigma, USA) and 20 ng/mL interferon-γ (IFN-γ, HY-P7025A, MCE, USA) for 48 h to drive M1 polarization.

Induction of M2-type macrophages in vitro

To induce M2-type polarization, THP-1 cells were initially treated with 100 ng/mL PMA for 24 h to facilitate differentiation. Subsequently, the cells were incubated for 48 h with 10 ng/mL of Interleukin-4 (IL-4, 200-04-20UG, PEPROTECH, China) and Interleukin-13 (IL-13, 200-13-10UG, PEPROTECH, China). BMDMs were polarized to the M2 phenotype by treatment with 20 ng/mL IL-4 and 20 ng/mL IL-13 for 24 h.

Macrophage-OS coculture assay

To investigate the function of hypoxic tumor-associated macrophages (TAMs) in the progression of OS, human THP-1-derived macrophages were induced and polarized into both M1 and M2 subtypes under normoxic (21% O₂) and hypoxic (1% O₂) conditions, while primary mouse BMDMs were polarized into the M2 subtype under the same conditions. The supernatant from M2 macrophages was co-cultured with OS cell lines for an additional 24 h to assess its impact on the malignant phenotypes of OS cells.

Methylthiazolyl diphenyltetrazolium bromide (MTT) assay

HOS, MG-63, and K7M2 cell lines were plated in 96-well plates at a density of 1 × 10⁴ cells per well. After 24 h of incubation, MTT (M2128, Sigma, USA) was added to each well, and the cells were incubated for an additional 3 h. Following incubation, the formazan produced was solubilized using dimethyl sulfoxide (DMSO, D2650, Sigma, USA). The absorbance at 570 nm was recorded using a microplate reader (51,119,570, Thermo Fisher, USA) to assess cell viability.

Transwell assay

HOS, MG-63, and K7M2 cells were collected and seeded in the upper Transwell chamber (3422, COSTAR, USA) at a density of 1 × 10⁴ cells per well without adding serum. A culture medium supplemented with 10% FBS was added to the lower Transwell chamber. After 24 h of culture, the cells were fixed with methanol and stained with crystal violet (C110703, Aladdin, China). Invasion assays were performed by adding Matrigel (356,234, Corning, USA) to the upper chamber half an hour in advance. Cell migration and invasion were observed and imaged using a microscope (IX73, Olympus Corp, China).

Colony formation assay

HOS, MG-63, and K7M2 cells were seeded in 96-well plates at a density of 5 × 10⁵ cells per well. After cell adhesion, the medium was carefully removed, and cells were fixed with 4% paraformaldehyde (1 mL per well) for 30 min. Cells were then stained with 1% crystal violet solution (1 mL per well) for 20 min to visualize colonies. Colonies containing more than 50 cells were considered viable and counted under an inverted microscope (ECLIPSE Ts2, Nikon, Japan). Colony images were captured and analyzed using ImageJ software (v1.8.0).

Western blotting (WB) analysis

Cell lysates were generated using RIPA buffer, and protein concentrations were determined with a bicinchoninic acid (BCA) assay kit (PC0020, Solarbio, China). WB was performed in accordance with the published description [19]. The protein samples were separated using SDS–polyacrylamide gel electrophoresis and subsequently transferred to a polyvinylidene difluoride (PVDF) membrane. The ECL kit (BL520B, Biosharp, China) was used to visualize those specifically bound to primary and secondary antibodies. The primary antibodies used in this experiment are listed in Table 1.

Table 1.

Details of antibodies

Name	Brand	Number
β-actin	CST	4970
HIF1A	Abcam	ab179483
E-cadherin	Proteintech	20,874-1-AP
N-cadherin	Abclonal	A19083
Vimentin	Abclonal	A19607
p-STAT3	CST	9145S
STAT3	Proteintech	10,253-2-AP
p-AKT	CST	4060s
AKT	Proteintech	60,203-2-lg
p-P13K	CST	4228
P13K	Proteintech	60,225-1-Ig
p-ERK	CST	p-ERK
ERK	Proteintech	16,443-1-AP
p-NFκB	CST	3033S
NFκB	CST	#8242
OPN	Abcam	ab69498
EGR3	Abcam	ab232820
RIG-I	Santa Cruz	sc-376845
PD-L1	CST	13684T
ISG15	Sigma	HPA004627
DNMT1	CST	5032
Anti-mouse IgG	CST	7076P2
Anti-rabbit IgG	CST	7074P2
HRP-conjugated anti-mouse IgG	Servicebio	G1214
HRP-conjugated anti-rabbit IgG	Servicebio	G1213

Cytokine antibody array analysis

To identify differentially expressed proteins in the supernatant of THP-1 cells differentiated into M2 macrophages and treated under hypoxic conditions for 24 h, the supernatant was collected and centrifuged to remove cellular debris, and samples were diluted as required. Cytokine and chemokine expression was assessed using the Human Cytokine Antibody Array (ab133998, Abcam, UK), a multiplexed, semi-quantitative array designed to simultaneously detect and quantify multiple cytokines and chemokines in a single sample. The antibody array membrane was initially blocked with 1 × blocking buffer for 30 min at room temperature to prevent non-specific binding. Subsequently, diluted supernatant samples were added to the array and incubated for 2 h at room temperature. After a series of buffer washes, biotin-conjugated secondary antibodies were applied, followed by a 2 h incubation at room temperature. Additional washes were performed before adding HRP-conjugated streptavidin, with a further incubation of 2 h at room temperature. Finally, the membranes were visualized using a chemiluminescence detection system to compare signal intensities at specific antigen–antibody spots, determining relative protein expression levels among the samples. Signal intensity was quantified using ImageLab software.

Real-Time quantitative polymerase chain reaction (RT-qPCR)

Cells were lysed using Trizol reagent (15,596,018, Ambion, China) to extract total RNA. RNA extraction was performed using isopropanol (327,270,010, Acros Organics, USA), followed by a 75% ethanol wash, and then dissolved in Diethyl pyrocarbonate (DEPC)-treated water (BL510B, Biosharp, China). The subsequent procedures were performed using the RevertAid First Strand cDNA Synthesis Kit (K1622, Thermo, USA) following the manufacturer's instructions. RT-qPCR amplification was conducted on a Thermo 7300 RT-qPCR system with the Genious™ 2 × SYBR Green Fast qPCR Mix (RM21204, Abclonal, China). Relative mRNA expression levels were calculated using the 2^−ΔΔCt method, with β-actin as the reference gene. RT-qPCR was performed using optimized primer pairs and commercial SYBR Green reagents under conditions recommended by the manufacturer. Only Ct values within the accepted linear range were included in the analysis. The primer sequences can be found in Table 2.

Table 2.

The sequences of RNA primers

Names	Forward(5′-3′)	Reverse(5′-3′)
β-actin	AGCAGTTGTAGCTACCCGCCCA	GGCGGGCACGTTGAAGGTCT
OPN	AAAATAGAGCTGCCTTGGGGG	GGCTTTCGTTGGACTTACTTGG
ENO4	CGATTTGGCTGTTGGGCTTG	CCTGTTTCGGCTGATTCCTCT
GRIN3B	TTCAGTATCAACTCCGCCCG	GAAGAGGATGGCGTAGCACA
WNT11	ACTGAACCAGACGCAACACT	GGTCCCTCTCTCCAGGTCAA
DLX5	ACTTTGCCCGAGTCTTCAGC	TCTTTCTCTGGCTGGTTGGT
EGR3	TGCCTGACAATCTGTACCCC	TCCCAAGTAGGTCACGGTCT
BATF2	CCCAAGGAGCAACAAAGGC	GCTGAGCAGGAGGCACAAT
ISG15	GACCTGACGGTGAAGATGCTG	TGCTGCGGCCCTTGTTAT
TRIM5	GACAAGATACCAGACATT	CTCCTTCCTCTAACCCTAT
PTPN7	AGGTCACCCTACACTTTCTGC	CTCCTTCCCGTCATAGCC

ELISA quantification of osteopontin (OPN)

To evaluate the levels of OPN in the supernatant of THP-1 cells that had been induced to the M2 phase and subjected to hypoxic treatment, we employed an ELISA kit (EK2135, MultiSciences, China). The process commenced with the addition of 50 μL of diluent to each well of the microplate, which was then followed by the introduction of 50 μL of either the standard, control, or sample. The plate was left to incubate at room temperature for a duration of 2 h. After this, the plate underwent four washes before a conjugate reagent was applied to each well, with further incubation at room temperature for 2 h. After another set of four washes, 100 μL of substrate solution was added to each well. The enzymatic reaction was halted after 30 min with the addition of the stop solution. Finally, the absorbance was determined at 450 nm using a spectrophotometer (840-297000, Thermo, USA).

Cell transfection

The stable lentiviral constructs for early growth response 3 (EGR3) or interferon-stimulated gene 15 (ISG15) overexpression (Lenti-EGR3 and Lenti-ISG15), as well as for OPN knockdown (sh-OPN), along with their corresponding negative controls, were generated using 293 T cells. The overexpression plasmids were obtained from Suzhou Genewiz Biotechnology Co., Ltd., and the shRNA plasmids were sourced from Tsingke Biotechnology Co., Ltd. (Hangzhou, China). Plasmids and transfection reagent (Lipofectamine 2000, 11,668,019, Thermo, USA) were diluted in serum-free medium, combined, and allowed to incubate at room temperature for 15 min to form complexes. These complexes were subsequently added to the cell medium, mixed gently, and incubated at 37 °C with 5% CO₂ for 24 h.

HOS and MG63 cells were cultured in a serum-free medium for 12 h before the start of the experiment. To construct HOS and MG63 cell lines overexpressing EGR3 and ISG15, plasmids overexpressing EGR3 (NM_004430) and ISG15 (NM_005101) were inserted into the pcDNA3.1 vector, which contains a CMV promoter commonly used to achieve high expression levels. For gene silencing, siRNAs targeting ISG15 and a negative control siRNA (siR-NC) were used. EGR3 or ISG15 and their corresponding empty control plasmids or siRNAs were transfected into the corresponding HOS or MG63 cells for 24 h using Lipofectamine 2000. The transfection efficiency was detected by WB. All plasmids were obtained from Youbio (Hunan, China). The sequences used in this experiment are listed in Table 3.

Table 3.

Information of sequences for shRNA and siRNA

Name	Sequences
sh-OPN-1	GATCGCCGGCCACAAGCAGTCCAGATTATACTCGAGTATAATCTGGACTGCTTGTGGTTTTTT
sh-OPN-NC	GATCGGTTCTCCGAACGTGTCACGTCTCGAGACGTGACACGTTCGGAGAACCTTTTTT
siR-ISG15	sense: CAUGUCGGUGUCAGAGCUGAA antisense: UUCAGCUCUGACACCGACAUG
siR-ISG15-NC	sense: UUCUCCGAACGAGUCACGUTT antisense: ACGUGACUCGUUCGGAGAATT

Dual-Luciferase reporter gene assay

To evaluate the impact of wild-type OPN on promoter activity in a hypoxic environment, M2 macrophages (THP-1) were co-transfected with a luciferase reporter plasmid containing the wild-type human OPN promoter region (− 1000 to + 1) cloned into the pGL3-basic vector. The cells were cultured for 24 h in a hypoxic environment with normoxic (21% O₂) or hypoxic (1% O₂) conditions, lysed with 1 × lysis buffer, and then centrifuged to remove debris. Luciferase activity was measured in a 96-well plate using the Dual-Luciferase® Reporter Assay System (E1910, Promega, USA). LAR II and Stop & Glo® Reagents were added sequentially, and relative luciferase units were calculated and normalized to the empty vector control.

For assessing EGR3 transcriptional activation of ISG15, luciferase reporter constructs containing wild-type or mutant EGR3-binding sequences within the ISG15 promoter were generated. HOS cells were simultaneously transfected with reporter plasmids and either pcDNA3.1-EGR3 or an empty pcDNA3.1 vector. Luciferase activity was assessed 24 h after transfection.

Chromatin immunoprecipitation (ChIP)

Cells were crosslinked with 1% formaldehyde, lysed, and chromatin was sheared using ultrasound. Overnight incubation of the sheared chromatin with a DNA (cytosine-5)-methyltransferase 1 (DNMT1)-specific antibody was followed by immunoprecipitation using protein G magnetic beads at 4 °C. After crosslink reversal and protein digestion with NaCl and Proteinase K at 65 °C for 2 h, DNA was purified for subsequent analysis. PCR amplification was performed using primers targeting the DNMT1-binding region within the EGR3 promoter (forward: 5′-CTTTCTAACTGCTGGGGTG-3′, reverse: 5′-CTATTTATCTGAGCCCCGG-3′).

RNA extraction and library construction

Total RNA was isolated using Trizol (15,596,018, Ambion, USA) following the manufacturer's instructions. A NanoDrop ND-1000 was used to assess RNA concentration and purity, while an Agilent Bioanalyzer and agarose gel electrophoresis confirmed the integrity (RIN > 7.0). From 1 µg of total RNA, Poly(A) RNA was extracted using Dynabeads Oligo(dT)25. The purified RNA was then fragmented at 94 °C for 5–7 min and converted to cDNA using SuperScript™ II. The second strand synthesis involved E. coli DNA polymerase I, RNase H, and dUTP. An adenine base was introduced to the cDNA for adapter ligation, and size selection was performed using AMPureXP beads. PCR amplification was performed under defined thermal conditions, resulting in an average insert size of 300 ± 50 bp. Paired-end sequencing of 150 bp was subsequently conducted on an Illumina NovaSeq™ 6000 (LC-Bio Technology, Hangzhou).

Bioinformatics analysis of RNA-seq

For gene expression profiling, two independent RNA-seq analyses were conducted. In Fig. 4, HOS cells were treated for 24 h with supernatants from THP-1-derived M2 macrophages cultured under hypoxic (1% O₂) or normoxic (21% O₂) conditions. In Fig. 6, MG-63 cells were transfected with either a plasmid for EGR3 overexpression or a control vector, followed by recombinant OPN treatment.

Fig. 4.

DEGs and their enriched pathways were identified by bioinformatics analysis. HOS cells were treated with the supernatant from hypoxic M2 macrophages under hypoxic or normoxic conditions for 24 h, followed by HTS analysis to identify DEGs. A Bar chart of DEGs counts. B Volcano plot of DEGs. (C) A heatmap was utilized to visualize the significant DEGs. D and E The biological processes and pathways in which DEGs were enriched were identified by GO D and KEGG E enrichment analysis

Fig. 6.

Identified DEGs and enriched pathways in EGR3-transfected MG-63 cells with OPN treatment. MG-63 cells were transfected with EGR3 and combined with OPN treatment, followed by HTS analysis to identify DEGs. A Bar chart of DEGs counts. B Volcano plot of DEGs (upper panel: OPN_EGR3 vs. OPN; lower panel: OPN vs. pCDH). C A heatmap was utilized to visualize the significant DEGs (left panel: OPN_EGR3 vs. OPN; right panel: OPN vs. pCDH). D and E The biological processes and pathways in which DEGs were enriched were identified by GO (D) and KEGG (E) enrichment analysis (left panel: OPN_EGR3 vs. OPN; right panel: OPN vs. pCDH)

Raw sequencing data were obtained in FASTQ format and subjected to quality control using FastQC. Adapter sequences and low-quality reads were removed using fastp (https://github.com/OpenGene/fastp) to generate clean reads. Clean reads were aligned to the human reference genome using HISAT2 (https://ccb.jhu.edu/software/hisat2), and gene-level quantification was performed using featureCounts. Raw count matrices were normalized and analyzed for differential gene expression using DESeq2. Differentially expressed genes (DEGs) were defined using a threshold of |log₂FC|≥ 1 and adjusted q-value < 0.05. DEG statistics, including the number of upregulated and downregulated genes, were visualized using ggplot2 (https://github.com/gpertea/gffcompare/). Functional enrichment analyses for Gene Ontology (GO) terms and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were conducted using the clusterProfiler R package, and the top 20 enriched pathways were presented. For the Hypoxia vs. Normoxia comparison, the top 100 DEGs (ranked by q-value) were selected for heatmap visualization using the pheatmap package, with expression values shown as Z-scores. For OPN_EGR3 vs. OPN and OPN vs. pCDH groups, heatmaps were similarly generated based on DEGs, with standardized expression displayed using Z-transformed values.

Construction of the xenograft tumor models in vivo

Xenograft tumor models were established by subcutaneously injecting MG-63 cells with stable overexpression of EGR3 or ISG15 or control cells (5 × 10⁶/200 μL) into 5-week-old female nude mice (n = 6) obtained from Slaccas (Shanghai, China). The mice were assigned to six groups: Lenti-control + PBS, Lenti-EGR3 + PBS, Lenti-ISG15 + PBS, Lenti-control + OPN, Lenti-EGR3 + OPN, and Lenti-ISG15 + OPN. After tumor formation, the treatment group mice were administered 100 μL of a 150 μg/mL OPN solution through intratumoral injection twice a week, while the control group mice received an equivalent volume of PBS. All mice were kept under pathogen-free conditions with a 12 h light/dark cycle at 22°C for approximately one week before treatment. Tumor growth was closely monitored, and once the control group tumors reached a volume of 1500 mm³, all the mice were euthanized, tumors were stripped, and their volumes and weights were measured. The experimental procedures conducted in this study adhered strictly to the Chinese Regulations for the Administration of Laboratory Animals. The animal experiments were approved by the Laboratory of Animal Experimental Ethical Inspection of Dr. Can Biotechnology (Zhejiang) Co., Ltd. (Approval No. DRK-20240914021).

Immunohistochemistry (IHC) assay

The expression of EGR3, ISG15, phosphorylated signal transducer and activator of transcription 3 (p-STAT3), phosphorylated nuclear factor kappa B (p-NFκB), and phosphorylated extracellular signal-regulated kinase (p-ERK) of the tumor tissues was determined by the IHC assay. After dewaxing and rehydration, longitudinal sections of 5 μm thickness were stained with hematoxylin for 5 min, followed by cytoplasmic staining with eosin. The sections were rinsed in distilled water for 30 s and then dehydrated using a graded alcohol series. Afterward, the sections were cleared with xylene and examined and photographed using a fluorescence microscope (BX53, Olympus). For immunohistochemistry, the sections were blocked with 3% BSA and incubated overnight with specific primary antibodies. Following this, they were treated with a secondary antibody at room temperature for 30 min. Staining was visualized using DAB (G1212, Servicebio), and counterstaining was performed with hematoxylin. Finally, three randomly selected fields were photographed under the microscope. The primary antibodies used in this experiment are listed in Table 1.

Terminal deoxynucleotidyl transferase-mediated dUTP-biotin nick end labeling (TUNEL) assay

The TUNEL assay was performed on isolated xenograft tumors to detect apoptosis. Briefly, the tumors were paraffin-embedded and sectioned at a thickness of 5 μm. The tumor sections were stained using the TUNEL FITC Apoptosis Detection Kit (A111-01, Vazyme, China) according to the manufacturer's guidelines. Following this, the nuclei were stained with DAPI solution (C1002, Beyotime, China). The stained sections were then analyzed and photographed with a fluorescence microscope (BX53, Olympus, Japan).

Statistical analysis

Statistical analyses were conducted using SPSS version 18 (SPSS, Inc.) and GraphPad software 6.0. Measurement data were expressed as the mean ± standard deviation (SD). A two-tailed unpaired Student's t-test was employed for group comparisons, while one-way ANOVA followed by Tukey's post hoc test was utilized for comparisons among multiple groups. A p-value of less than 0.05 was deemed statistically significant.

Results

Hypoxic M2 macrophages promote malignant phenotypes in OS cells

To investigate the function of hypoxic TAMs in the progression of OS, we selected MG-63, HOS, and K2M7 cell lines for further study. These cell lines are widely used in OS research, with stable genetic profiles, and have been employed in studies related to metastasis, drug response, hypoxia, and hypoxia-driven signaling [20, 21]. Human THP-1-derived macrophages were polarized into both M1 and M2 subtypes under normoxic and hypoxic conditions. Primary mouse BMDMs were also polarized into the M2 subtype under the same conditions to assess their impact on the malignant phenotypes of OS cells. WB analysis was performed to assess HIF1A expression in K7M2 cells after co-culture with supernatants from primary mouse BMDMs polarized to the M2 phenotype under normoxic or hypoxic conditions. The results showed that HIF1A expression was significantly upregulated in K7M2 cells co-cultured with supernatants from hypoxic M2 macrophages, validating the induction of hypoxia (Fig. 1A). Subsequently, we demonstrated that hypoxic M2 macrophage supernatants significantly promoted cell proliferation, migration, invasion, and colony formation in OS cells. In contrast, supernatants from M1 macrophages had no significant effect on cell migration and invasion (Fig. 1B–E). The detection of epithelial-mesenchymal transition (EMT) pathway-related proteins indicated that hypoxic M2 macrophages enhanced the activity of the EMT pathway (Fig. 1F). Furthermore, hypoxic M2 macrophages activated multiple cancer-related signaling pathways, including the STAT3, PI3K/Protein kinase B (AKT), Mitogen-activated protein kinase (MAPK)/ERK, and NFκB signaling pathways (Fig. 1G). These findings suggested that hypoxic M2 macrophages play a significant role in the promotion of malignant phenotypes in OS cells.

Fig. 1.

Hypoxic M2 macrophages promote malignant phenotypes in OS Cells. A WB assay was used to detect HIF1α expression in K7M2 cells after co-culture with supernatants from primary BMDMs polarized to the M2 phenotype under normoxic or hypoxic conditions. B MTT assay was used to examine the effect of hypoxic M2 macrophages on the proliferation of OS cells. C The effect of hypoxic M2 macrophages on the colony formation of OS cells. D Transwell assay was employed to test the migration and invasion capabilities of K7M2 cells co-cultured with M2 macrophages under normoxic and hypoxic conditions. E The migratory and invasive capabilities of HOS and MG-63 cells were assessed using Transwell assays in response to normoxic or hypoxic M0, M1, and M2 macrophages. F WB assay was used to detect the effect of hypoxic M2 macrophages on the expression of EMT pathway-related proteins in OS cells. G WB assay was used to detect the effect of hypoxic M2 macrophages on cancer-related signaling pathways, including STAT3, PI3K/AKT, ERK, and NFκB in OS cells. All data came from at least three replicate experiments. A value of p below 0.05 is considered significant, ^*p < 0.05, ^**p < 0.01 vs the control group

Hypoxia promotes the polarization of M1 to M2 macrophages and upregulates OPN expression in M2 macrophage

To investigate the effect of hypoxia on M2 macrophage secretory profile, an antibody array was utilized to identify differentially expressed proteins in the supernatant of M2 macrophages with hypoxic treatment for 24 h. The results demonstrated there were eight significantly differentially expressed proteins, including epithelial-derived neutrophil-activating peptide 78 (ENA-78/CXCL5), growth-regulated oncogene (GRO/CXCL1), monocyte chemoattractant protein-1 (MCP-1/CCL2), macrophage-derived chemokine (MDC/CCL22), regulated on activation, normal T cell expressed and secreted (RANTES/CCL5), tumor necrosis factor-alpha (TNF-α), and tissue inhibitor of metalloproteinases-1 (TIMP-1), OPN (Fig. 2A). Among these, OPN showed the most notable increase and was selected for further functional investigation based on its previously reported involvement in OS progression and immune modulation [22, 23]. To corroborate the identification outcome derived from the antibody array, we further examined the levels of OPN in the supernatant and intracellular environment of hypoxic M2 macrophages. The results revealed an upregulation of OPN in both locations (Fig. 2B,C). Consistently, the dual-luciferase reporter gene assay demonstrated that hypoxic treatment enhanced the fluorescence intensity of OPN in M2 macrophages (Fig. 2D).

Fig. 2.

Hypoxia upregulates OPN expression in M2 macrophages. A The antibody array was utilized to identify differentially expressed proteins in the supernatant of M2 macrophages after 24 h hypoxia or normoxia treatment. B ELISA kit was used to examine the levels of OPN in the supernatant of hypoxic M2 macrophages. C WB assay was used to detect the expression levels of OPN in the intracellular environment of hypoxic M2 macrophages. D Dual-luciferase reporter gene assay was used to detect the fluorescence intensity in THP-1 cells transfected with wild-type OPN combined with hypoxia or normoxia treatment. All data came from at least three replicate experiments. A value of p below 0.05 is considered significant, ^*p < 0.05, ^**p < 0.01 vs the control group

Knockdown of OPN in M2 macrophages attenuates hypoxia-induced malignant phenotypes of OS cells

To gain a comprehensive understanding of the function of OPN within M2 macrophages in the progression of OS, OPN was knocked down in THP-1 cells (Fig. 3A). The shRNA with the most effective knockdown, sh-OPN-1, was selected for utilization in the subsequent experiments. The results demonstrated that hypoxic M2 macrophages significantly facilitated cell proliferation, migration, and colony formation, whereas knockdown of OPN within M2 macrophages attenuated these effects (Fig. 3B–D). This indicated that the tumor-promoting effects of hypoxia-treated M2 macrophages were achieved through OPN-dependent mechanisms, at least in part.

Fig. 3.

Knockdown of OPN in M2 macrophages attenuates hypoxia-induced malignant phenotypes of OS cells. A WB assay was used to detect the knockdown efficiency of OPN in THP-1 cells. B MTT assay was used to examine the effect of OPN knockdown in M2 macrophages combined with hypoxic treatment on the proliferation of OS cells. C Transwell assay was used to examine the effect of OPN knockdown in M2 macrophages on the migration of OS cells after a 24 h hypoxia or normoxia treatment. D The effect of OPN knockdown in M2 macrophages on the colony formation of OS cells after a 24 h hypoxia or normoxia treatment. All data came from at least three replicate experiments. A value of p below 0.05 is considered significant, ^*p < 0.05, ^**p < 0.01 vs the control group

OPN contributes to malignant phenotypes via EGR3 downregulation in OS cells

To further investigate the mechanisms underlying the tumor-promoting effects of hypoxic M2 macrophages, HOS cells were treated with the supernatant from THP-1-derived M2 macrophages cultured under hypoxic (1% O₂) or normoxic (21% O₂) for 24 h, followed by high-throughput sequencing (HTS) analysis to identify differentially expressed genes (DEGs). The results revealed a total of 124 upregulated genes and 136 downregulated genes among the DEGs (Fig. 4A,B). A heatmap was utilized to visualize the significant DEGs (Fig. 4C). Additionally, Gene Ontology (GO) enrichment analysis indicated that the DEGs were intimately involved in positive regulation of peptidyl-serine phosphorylation, oxygen carrier activity, and cytokine-mediated signaling pathway (Fig. 4D). Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis indicated that the DEGs were predominantly enriched in pantothenate and CoA biosynthesis, TNF signaling pathway, cytokine-cytokine receptor interaction, and transcriptional misregulation in cancer (Fig. 4E).

After filtering the DEGs based on fold change and p-value, five genes were identified as worthy of further investigation: Enolase 4 (ENO4), Glutamate Receptor, Ionotropic, N-Methyl-D-Aspartate 3B (GRIN3B), Wnt Family Member 11 (WNT11), Distal-Less Homeobox 5 (DLX5), and Early Growth Response 3 (EGR3). The analysis of their expression levels revealed that EGR3 exhibited the most significant differential expression in HOS cells treated with the supernatant from hypoxic M2 macrophages, with a tendency towards downregulation (Fig. 5A). Therefore, EGR3 was selected for further investigation. The results demonstrated that overexpression of EGR3 markedly suppressed the migratory capacity of OS cells (Fig. 5B,C). Furthermore, overexpression of EGR3 resulted in a reduction in the migration of OS cells promoted by hypoxic M2 macrophages (Fig. 5D). Importantly, the tumor-promoting effects of recombinant OPN were reversed by overexpression of EGR3 (Fig. 5E). These findings suggested that OPN in hypoxic M2 macrophages promoted malignant phenotypes by inhibiting the expression of EGR3 in OS cells.

Fig. 5.

OPN promotes malignant phenotypes by inhibiting EGR3 expression in OS cells. A RT-qPCR assay was used to detect the relative mRNA expression levels of the ENO4, GRIN3B, WNT11, DLX5, and EGR3 genes in HOS cells treated with the supernatant from hypoxic M2 macrophages. B WB assay was used to detect the overexpression efficiency of EGR3 overexpression in OS cells. C Transwell assay was used to examine the effect of EGR3 overexpression on the migration of OS cells. D Transwell assay was used to examine the effect of overexpression of EGR3 on the migration of OS cells induced by hypoxic M2 macrophages. E Transwell assay was used to examine the effect of EGR3 overexpression on the migration of recombinant OPN (100 ng/mL) treated OS cells. All data came from at least three replicate experiments. A value of p below 0.05 is considered significant, ^*p < 0.05, ^**p < 0.01 vs the control group

OPN regulates EGR3, ISG15, and RIG-I expression and associated immune signaling in OS cells

As OPN was previously found to be upregulated in hypoxia-treated M2 macrophages, we further investigated whether recombinant OPN exerts a similar regulatory effect on immune-related gene expression in OS cells. MG-63 cells were transfected with either an EGR3 overexpression plasmid or a control vector, followed by treatment with OPN. Total RNA was extracted, and high-throughput sequencing was performed to identify DEGs. The results demonstrated that, in comparison between the control group and the OPN treatment group, there were 298 upregulated genes and 612 downregulated genes. In the comparison between the OPN treatment group and the EGR3-transfected plus OPN treatment group, there were 304 upregulated genes and 143 downregulated genes (Fig. 6A,B). The heatmaps illustrated the gene expression data (Fig. 6C). Subsequently, GO enrichment analysis revealed that the DEGs were predominantly enriched in various responses to viruses, immune system processes, innate immune response, protein binding, and interferon signaling pathways (Fig. 6D). KEGG enrichment analysis demonstrated that the DEGs identified from both comparisons were closely associated with multiple signaling pathways, including the NOD-like receptor signaling pathway, Cytokine-cytokine receptor interaction, Toll-like receptor signaling pathway, TNF signaling pathway, NFκB signaling pathway, and RIG-I-like receptor signaling pathway (Fig. 6E).

Among the DEGs, four genes were identified as warranting further investigation: Basic Leucine Zipper ATF-Like Transcription Factor 2 (BATF2), Interferon-Stimulated Gene 15 (ISG15), Tripartite Motif Containing 5 (TRIM5), and Protein Tyrosine Phosphatase, Non-Receptor Type 7 (PTPN7). The assessment of the mRNA expression levels demonstrated that OPN treatment significantly downregulated the expression of these proteins (BATF2, ISG15, and TRIM5) in MG-63 cells. Overexpression of EGR3 reversed the inhibitory effects of OPN on these proteins (Fig. 7A). Notably, ISG15 is known to be involved in the RIG-I-like (Retinoic acid-inducible Gene I-like) receptor signaling pathway [24]. Given the results from 6E, the DEGs were enriched in the RIG-I signaling pathway. These results suggested that OPN in M2 macrophages and EGR3 in OS cells exerted their influence on the malignant phenotype of OS cells, potentially through the regulation of the RIG-I-like receptor signaling pathway via ISG15. These results suggested that OPN suppresses ISG15 and associated immune-related genes, while EGR3 overexpression reverses these effects, potentially influencing the malignant phenotype of OS cells via the RIG-I-like receptor signaling pathway.

Fig. 7.

OPN promotes OS cell growth and alters the expression of EGR3, ISG15, and RIG-I, potentially affecting RIG-I-like receptor signaling. A RT-qPCR assay was used to detect mRNA levels of BATF2, ISG15, TRIM5, and PTPN7 in MG-63 cells with EGR3 overexpression and OPN treatment. B WB assay was used to detect the expression levels of ISG15, RIG-I, and PD-L1 in OS cells after OPN treatment. C WB assay was used to examine the effect of EGR3 overexpression on the expression levels of ISG15, RIG-I, and PD-L1 in OS cells. D WB assay evaluated the effect of ISG15 overexpression on the expression levels of RIG-I, PD-L1, STAT3, NFκB, PI3K, and ERK in OS cells. E MTT assay was used to examine the effect of overexpression of ISG15 on the proliferation of OS cells. F Transwell assay was used to examine the effect of overexpression of ISG15 on the migration of OS cells. G WB assay was used to examine the effect of overexpression of EGR3 combined with OPN treatment on the expression levels of STAT3, NFκB, PI3K, and ERK signaling pathway proteins in OS cells. H WB assay was used to examine the effect of EGR3 overexpression combined with ISG15 knockdown on the expression levels of ISG15, RIG-I, PD-L1, STAT3, NFκB, ERK, and PI3K signaling pathway proteins in OS cells. I Transwell assay was used to assess the migration capacity of HOS cells with EGR3 overexpression and ISG15 knockdown. J Dual-luciferase reporter gene assay was used to detect the fluorescence intensity in HOS cells transfected with wild-type or mutant ISG15 and EGR3 overexpression. K WB assay was used to detect DNMT1 expression in HOS cells stimulated with recombinant OPN. L ChIP assay evaluated DNMT1 binding to the EGR3 promoter. All data came from at least three replicate experiments. A value of p below 0.05 is considered significant, ^*p < 0.05, ^**p < 0.01 vs the control group

ISG15 is a ubiquitin-like protein that modifies target proteins through ISGylation [24, 25]. It has been reported that RIG-I, a key protein in the RIG-I-like receptor signaling pathway, is ISGylated by the ISG15 E3 ligase [24, 26, 27]. Additionally, RIG-I can enhance the stability of PD-L1 through post-translational modifications [28]. The results demonstrated that OPN treatment reduced the expression of ISG15 while upregulating the expression levels of RIG-I and PD-L1 (Fig. 7B). Conversely, overexpression of EGR3 exerted the opposite effect on these proteins (Fig. 7C). These findings suggested that OPN may modulate RIG-I-like receptor signaling in OS cells through coordinated changes in the expression of EGR3, ISG15, and RIG-I.

It has been reported that the RIG-I-like receptor signaling pathway plays a role in regulating tumor cell survival, proliferation, and immune evasion by activating downstream pathways such as NFκB, STAT3, and ERK/MAPK [29–31]. In our study, ISG15 overexpression in OS cells was associated with reduced RIG-I protein expression and decreased activation of its downstream signaling components, including STAT3, NFκB, PI3K, and ERK (Fig. 7D). Furthermore, overexpression of ISG15 inhibited cell proliferation and migration in OS cells (Fig. 7E,F). Importantly, OPN treatment led to reduced ISG15 levels and upregulation of RIG-I, along with enhanced phosphorylation of STAT3, NFκB, ERK, and PI3K. Conversely, the tumor-promoting effects of OPN were reversed by overexpression of EGR3 (Fig. 7G).

To further investigate the impact of EGR3 on these pathways, WB analysis showed that EGR3 overexpression significantly upregulated ISG15 expression and concurrently inhibited downstream signaling pathways, including phosphorylation of STAT3, NFκB, ERK, and PI3K. Silencing ISG15 restored the activation of these pathways (Fig. 7H). Transwell assays demonstrated that knockdown of ISG15 reversed the inhibitory effect of EGR3 overexpression on OS cell migration (F7g. 7I). Dual-Luciferase reporter gene assay confirmed that EGR3 transcriptionally activates ISG15 expression by binding to its promoter (Fig. 7J). Additionally, OPN treatment markedly increased the protein expression of DNMT1, a key maintenance DNA methyltransferase [32], in OS cells (Fig. 7K), suggesting a potential epigenetic regulatory mechanism. ChIP assays further demonstrated that EGR3 overexpression enhanced DNMT1 binding to the EGR3 promoter, indicating that DNMT1 contributes to the transcriptional repression of EGR3 through promoter methylation (Fig. 7L). Taken together, these findings suggested that OPN exerted a pro-cancer effect on OS cells by suppressing EGR3 and ISG15, thereby increasing RIG-I expression and activating downstream signaling pathways.

OPN promotes OS tumor growth in vivo accompanied by downregulation of EGR3 and ISG15

To validate the role and mechanism of OPN in vivo, the xenograft tumor models were constructed with MG-63 cells. Consistent with the results from cell experiments in vitro, OPN treatment promoted tumor growth in vivo, while the overexpression of EGR3 or ISG15 effectively reversed the pro-cancer effects of OPN (Fig. 8A and 8). The results of apoptosis assays on tumor samples demonstrated that OPN treatment inhibited tumor cell apoptosis in OS, whereas overexpression of EGR3 or ISG15 promoted apoptosis and attenuated the inhibitory effect of OPN (Fig. 8C). IHC analysis of relevant indicators in tumor samples demonstrated that OPN treatment results in a decrease in the expression of EGR3 and ISG15, while significantly increasing the levels of p-STAT3, p-NFκB, and p-ERK. Conversely, overexpression of EGR3 or ISG15 has been observed to diminish the upregulation of these markers induced by OPN (Fig. 8D). This is consistent with our cellular findings, in which OPN treatment was associated with enhanced activation of the STAT3, NFκB, and ERK/MAPK signaling pathways, accompanied by downregulation of EGR3 and ISG15 in OS cells.

Fig. 8.

OPN promoted tumor growth in vivo by downregulating EGR3 and ISG15 expression in OS. Mice were subcutaneously inoculated with MG-63 cells (5 × 10⁶/200 μL). When the tumor volume reached 100 mm³, mice were randomly divided into six groups: Lenti-control + PBS, Lenti-EGR3 + PBS, Lenti-ISG15 + PBS, Lenti-control + OPN, Lenti-EGR3 + OPN, and Lenti-ISG15 + OPN, with six mice in each group. Treatment group mice received 100 μL of 150 μg/mL OPN via intratumoral injection while control group mice received an equivalent volume of saline solution. A and B When the tumors reached a volume of 1500 mm³, tumor tissues were isolated from the mice and then the size (A), volume, and weight of the tumors (B) were recorded. C TUNEL assay was used to detect the apoptotic cells in tumor tissue (n = 3). The images are magnified 400 × and 800 ×. D IHC assay was used to detect the changes in the levels of EGR3, ISG15, p-STAT3, p-NFκB, and p-ERK protein in tumor tissues (n = 3). The images are magnified 200 × and 400 ×. The scale has been marked on the diagram

In summary, this study demonstrated that hypoxia-induced OPN in M2 macrophages contributes to OS progression by altering the expression of EGR3, ISG15, and RIG-I, thereby modulating RIG-I-like receptor signaling and activating multiple downstream tumor-related pathways.

Discussion

Recently, OS has become well-known for highly aggressive bone tumors with an elevated tendency of metastasis [1]. The emerging evidence highlights the role of hypoxia or the immune microenvironment in OS progression, reflecting the complex interplay behind tumor aggressiveness and treatment difficulties [33, 34]. Despite knowing the OS immune evasion and tumor hypoxia's poor prognosis link, current studies mainly focus on analyzing and predicting this relationship [35, 36], and their underlying mechanisms are unclear. With the development of bioinformatics and HTS technology, many aberrantly expressed oncogenes have been identified and can serve as prognostic signatures for OS [37–39]. However, these single-characteristic-based prognostic signatures fail to adequately reflect OS's disease features, and reliable gene signatures combining hypoxia and immune status for OS remain unidentified. Here, our findings have provided compelling evidence that hypoxia stimulated the expression of OPN in M2 macrophage supernatants, which in turn inhibited the expression of EGR3 and promoted the malignant phenotype of OS through the RIG-I-like receptor signaling pathway and immune evasion of PD-L1. Our findings highlight the contribution of the hypoxia-immune microenvironment to OS progression and offer new perspectives on its potential regulatory interactions.

Immune evasion plays a critical role in OS progression by enabling tumor cells to escape immune surveillance.. Studies indicate that OS cells can upregulate CD47 to inhibit macrophage phagocytosis and express PD-L1 to suppress T cell activity via PD-1 interaction [33, 40]. Additionally, TME remodeling and impaired antigen presentation further contribute to immune evasion in OS [41, 42]. Tumor hypoxia is another important factor in OS progression, promoting angiogenesis, metastasis, and resistance to therapy by activating various signaling pathways. For instance, hypoxia has been shown to upregulate vascular endothelial frowth factor (VEGF) expression via stabilization of HIF-1α, thereby facilitating angiogenesis and tumor progression in OS cells [43]. In addition, hypoxia-mediated downregulation of spindle and kinetochore-associated complex subunit 1 (SKA1) has been associated with increased chemoresistance in OS [44]. These findings underscore that immune response or tumor hypoxia is important for OS cell growth and metastasis. In this study, we were surprised to find that hypoxic-treated M2 macrophage supernatant promoted the malignant phenotype of OS cell lines, which was consistent with the first research to preliminarily establish a prognostic signature based on hypoxia and immune response in OS cells [16]. To validate hypoxia in our model, we confirmed HIF1A upregulation in OS cells treated with supernatants from hypoxia-exposed M2 macrophages. Unexpectedly, HIF1A was also detectable under normoxia. This may result from M2-derived cytokines (e.g., IL-10, TGF-β, VEGF) activating PI3K/AKT and NFκB pathways, which enhance HIF1A transcription and translation. Notably, NFκB can bind the HIF1A promoter and upregulate its transcription even in normoxia [45, 46]. In addition, oxygen consumption in dense co-cultures may create localized hypoxia that contributes to this effect [47]. More importantly, this supernatant promoted the expression of EMT-related proteins and the phosphorylation of STAT3, P13K, AKT, ERK, and NFκB in OS cells, alterations of which are vital for the malignant progression of cancers [48]. We observed that, under hypoxic conditions, macrophages differentiated into the M2 type. Antibody arrays were subsequently used to screen and validate hypoxia-induced differential proteins within macrophages. Among these, the upregulation of OPN attracted our great research interest. OPN, initially identified in OS cells [49], is a highly glycosylated and phosphorylated protein widely distributed in tissues, cells, and body fluids. It plays crucial roles in bone metabolism, immune regulation, tumor biology, cardiovascular health, neurobiology, and gastrointestinal protection [50]. Recent studies highlight the context-dependent role of OPN in OS, which may vary due to differences in OPN source, cellular context, hypoxic conditions, or tumor stage. For instance, hypoxia-induced OPN expression occurs via αvβ3 integrin–mediated GLUT1/3 upregulation, activating the FAK pathway [51]. IRF8 downregulation increases OPN levels, impairing T-cell activation and facilitating immune escape [52]. Conversely, some studies suggest that OPN downregulation inhibits mesenchymal stem cell or osteoblast maturation, potentially leading to OS [53]. In our study, we specifically focused on OPN derived from hypoxia-induced M2 macrophages, which was associated with reduced expression of EGR3 and ISG15 and increased RIG-I levels in OS cells, along with activation of downstream signaling pathways. This context-dependent regulatory pattern may help explain discrepancies observed in previous studies.

The hypoxia-immune-related gene signature has already been associated with the overall survival in multiple types of tumors, such as bladder cancer [54], hepatocellular carcinoma [55], and triple-negative breast cancer [56]. These studies have demonstrated that the hypoxia-immune-based genes possess high prognostic potential and clinical guidance values. Notably, our study has first identified and validated the significant upregulation of OPN in hypoxic M2 macrophage supernatants, and silencing OPN markedly reversed the stimulatory effect of these macrophages on the malignant phenotypes of OS, suggesting its pivotal role in mediating the pro-tumorigenic actions of hypoxic M2 macrophages for OS. Using HTS, we conducted an analysis of the DEGs and their enrichment in HOS cells exposed to supernatants from hypoxic M2 macrophages, uncovering a significant downregulation of EGR3-a transcription factor, pivotal in regulating gene expression during cellular differentiation, inflammatory responses, and cell migration [57]. EGR3 exhibits tumor-suppressive functions in various cancers. In breast cancer, EGR3 triggered apoptosis by directly binding to the promoter of the anti-apoptotic factor MCL1 gene, facilitating its transcription [58]. In prostate cancer, EGR3 regulated the expression of approximately 330 genes, with 35% involved in immune responses and inflammatory processes, and 15% intersecting with the NFκB signaling pathway, and promoted prostate cancer progression by upregulating IL6 and IL8, two interleukins highly relevant to prostate cancer [59]. Additionally, EGR3 regulated gene expression, transcriptionally activating zinc finger protein 36 (ZFP36), growth arrest and DNA-damage-inducible beta (GADD45B), and suppressor of cytokine signaling 3 (SOCS3) in prostate cancer, contributing to its antitumor effects [60]. Here, we confirmed EGR3's inhibitory effect on OS migration. Surprisingly, EGR3 overexpression reversed the promotion effect of hypoxic M2 macrophage supernatant or OPN treatment on OS migration. Another in-depth analysis of the DEGs and their enrichment pathways uncovered the important role of OPN and EGR3 in OS immune responses. This revelation stemmed from our observations that the DEGs were predominantly enriched in various viral responses, immune system processes, and signaling pathways such as Toll-like receptor, NFκB, and RIG-I-like receptor. Furthermore, we observed that ISG15, a ubiquitin-like protein involved in antiviral immune responses [61], was downregulated following OPN treatment, and its overexpression significantly inhibited OS cell proliferation and migration. EGR3 overexpression reversed the pro-tumorigenic effects of OPN in vitro and in vivo, potentially through coordinated regulation of ISG15 and RIG-I. RIG-I plays a crucial role in antiviral defense and immune regulation [62]. Notably, ISG15 is typically induced downstream of RIG-I activation. In this study, we observed an alternative relationship wherein OPN suppresses EGR3 and ISG15, leading to increased RIG-I protein levels. This may reflect reduced ISGylation-dependent degradation of RIG-I rather than classical activation [63, 64]. Given the immune regulatory role of RIG-I and the associated PD-L1 changes, our findings our findings raise the possibility that restoring EGR3 and ISG15 expression may influence RIG-I-mediated immune responses in OS, which warrants further investigation.

We acknowledge several limitations in this study. First, the ectopic overexpression of EGR3 and ISG15 using a CMV promoter may not reflect physiological levels, and supraphysiological expression could have influenced some outcomes. Inducible systems or endogenous modulation strategies (e.g., CRISPRa, hypoxia mimetics) should be considered in future studies. Second, although xenograft models were used, the tumor immune microenvironment was not assessed, limiting our understanding of the ISG15/RIG-I axis in an immune context. Immunocompetent or humanized models will be needed to evaluate immune infiltration and checkpoint activity. Third, while our data suggest a correlation between OPN stimulation, ISG15 suppression, and RIG-I upregulation, direct mechanistic links remain unconfirmed. The increase in RIG-I may reflect reduced ISGylation-dependent degradation rather than classical activation, and further validation is required. Fourth, among the upregulated cytokines, only OPN was functionally characterized; other candidates merit investigation. Lastly, the therapeutic relevance of modulating OPN and its downstream regulatory interactions involving EGR3, ISG15, and RIG-I should be further evaluated in orthotopic or patient-derived models to assess long-term efficacy and safety.

Conclusion

This study proposed innovatively that hypoxic M2 macrophages, through OPN-dependent mechanisms, played a pivotal role in accelerating the progression of OS by inhibiting the expression of EGR3 and ISG5. This finding enhanced our understanding of OS mechanisms and offered potential therapeutic targets.

Abbreviations

BCA

Bicinchoninic acid

BMDMs

Bone marrow–derived macrophages

ChIP

Chromatin immunoprecipitation

DEPC

Diethyl pyrocarbonate

DMSO

Dimethyl sulfoxide

DNMT1

DNA (cytosine-5)-methyltransferase 1

ENA-78/CXCL5

Epithelial-derived neutrophil-activating peptide 78

ECM

Extracellular matrix

EGR3

Early growth response 3

ERK

Extracellular signal-regulated kinase

GADD45B

Growth arrest and DNA-damage-inducible beta

GRO/CXCL1

Growth-regulated oncogene

HIF-1α

Hypoxia-inducible factor-1α

HIF1A

Hypoxia-inducible factor 1 alpha

IHC

Immunohistochemistry

IL-13

Interleukin-13

ISG15

Interferon-stimulated gene 15

M-CSF

Macrophage colony-stimulating factor

MCP-1/CCL2

Monocyte chemoattractant protein-1

MDC/CCL22

Macrophage-derived chemokine

MTT

Methylthiazolyl diphenyltetrazolium bromide

NFκB

Nuclear factor kappa B

OPN

Osteopontin

OS

Osteosarcoma

PMA

Phorbol 12-myristate 13-acetate

PVDF

Polyvinylidene difluoride

PKB/AKT

Protein kinase B

RANTES/CCL5

Regulated on activation, normal T cell expressed and secreted

RPMI-1640

Roswell Park Memorial Institute 1640

RT-qPCR

Real-Time quantitative polymerase chain reaction

SOCS3

Suppressor of cytokine signaling 3

STAT3

Signal transducer and activator of transcription 3

TAMs

Tumor-associated macrophages

TME

Tumor microenvironment

TIMP-1

Tissue inhibitor of metalloproteinases-1

TNF-α

Tumor necrosis factor-alpha

TUNEL

Terminal deoxynucleotidyl transferase-mediated dUTP-biotin nick end labeling

VEGF

Vascular endothelial frowth factor

WB

Western blotting

ZFP36

Zinc finger protein 36

Author contributions

CX conceptualized the study, curated the data, conducted the investigation, and took the lead in writing the original draft and subsequent revisions. WH contributed to the conceptualization, data curation, investigation, and writing of the original draft. LZ performed the formal analysis and contributed to the methodology. All authors read and approved the final manuscript.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

All data generated or analysed during this study are included in this published article.

Declarations

Ethics approval and consent to participate

All mice experiments were approved by the Laboratory of Animal Experimental Ethical Inspection of Dr. Can Biotechnology (Zhejiang) Co., Ltd. (Approval No. DRK-20240914021).

Consent for publication

Not applicable.

Competing interests

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest, and all authors report no conflicts of interest in this work.