Skip to main content
NCBI home page
APL Bioengineering logoLink to APL Bioengineering
. 2025 Aug 8;9(3):036110. doi: 10.1063/5.0266780

Single-cell RNA and bulk sequencing analysis reveals that formononetin inhibits GTSF1 to exert anti-osteosarcoma effects

Tianyu Xie 1, Jiming Liang 2, Chengsen Lin 3, Wenyu Feng 4, Mingxiu Yang 2, Shanhang Li 5, Liang Xiong 1, Kai Luo 2, Feicui Li 2, Shengping Tang 2, Shangyu Liu 2, Qian Huang 1, Shijie Liao 1, Jianhong Liu 1, Yangjie Cai 1, Fuxing Tang 6, Haijun Tang 2, Qingjun Wei 4, Yunhua Lin 1,a), Yun Liu 2,a)
PMCID: PMC12335348  PMID: 40786706

Abstract

As the most common primary malignant bone tumor, osteosarcoma (OS) is characterized by drug resistance and poor prognosis, highlighting the urgent need for promising therapeutic agents. Formononetin (FMN), a natural product derived from Spatholobi Caulis, has been reported to possess anti-tumor properties. However, its role in OS has not yet been elucidated. In the present study, we established an OS patient-derived xenograft model to investigate the effects of FMN and the underlying mechanisms of its effects on OS. When FMN treatment was completed, bulk transcriptome sequencing was conducted, and the analyses were combined with OS single-cell RNA sequencing (scRNA-seq) data. Results indicated that GTSF1 was up-regulated in OS but down-regulated after FMN intervention, which may regulate the apoptosis of OS cells. Furthermore, the qRT–PCR and IHC results demonstrated that GTSF1 expression was significantly up-regulated in OS cells, whereas FMN expression was down-regulated both in vitro and in vivo. Moreover, in vitro experiments revealed that FMN effectively promoted apoptosis and suppressed the proliferation, migration, and invasion of OS cells. Therefore, this study demonstrated that FMN exerts anti-OS effects by down-regulating GTSF1 expression, thus effectively promoting the apoptosis and inhibiting the proliferation of OS cells, making FMN a promising anti-OS drug.

INTRODUCTION

As the most common primary malignant bone tumor, osteosarcoma (OS) predominantly occurs in children and adolescents, with an incidence rate of 3–4 per million, showing an increasing trend.1–3 Currently, the standard therapeutic schedule for OS includes surgical resection and adjuvant chemotherapy, with agents such as cisplatin, doxorubicin, ifosfamide, and high-dose methotrexate.4,5 However, the toxic side effects and resistance to chemotherapy limit patient benefits. For example, high-dose chemotherapeutic drugs are often associated with severe toxicity, potentially resulting in renal damage or hepatotoxicity, whereas resistance to cisplatin commonly develops, leading to suboptimal outcomes.6 Despite these interventions, the 5-year survival rate remains unsatisfactory in cases of tumor progression or distant metastasis.7,8 Given these challenges, there is an urgent need to discover novel anti-OS drugs that can improve therapeutic outcomes.

Traditional Chinese medicine, which is increasingly recognized for its multitarget actions and minimal side effects in cancer therapy, has drawn considerable attention.9,10 Formononetin (FMN), an isoflavone natural component extracted from Spatholobi Caulis, is chemically labeled as 7-hydroxy-3-(4-methoxyphenyl)-4H-chromen-4-one with C16H12O4 as its molecular formula.11–13 FMN exhibits various biological activities, including antioxidant, anti-tumor, antiviral, and neuroprotective effects.14–16 Studies have revealed its significant anti-tumor effects on various tumors, such as breast, cervical, head and neck, colon, and ovarian cancers.17–20 However, research on FMN in OS is relatively scarce, and existing studies have predominantly been conducted via in vitro cellular experiments.21–23 Only one study has explored its anti-OS effects in vivo on the basis of cell line-derived xenograft (CDX) models,24 which do not accurately reflect the complex tumor microenvironment because of significant differences from actual tumor tissue from OS patients. In contrast, patient-derived tumor xenograft (PDX) models offer a more accurate simulation of the complex tumor microenvironment of OS,25,26 thus providing more reliable validation of therapeutic effects. Despite this, the application of PDX models in OS research is limited owing to the rarity of OS samples and the demands of modeling techniques.27 To date, only a few studies have utilized this model to investigate OS.28,29 Therefore, this study is dedicated to exploring the potential of FMN as a therapeutic agent for OS and investigating its action mechanism in PDX models.

In the present study, we initially established an OS PDX model and employed FMN intervention. Subsequently, bulk and single-cell RNA sequencing (scRNA-seq) data were used to identify potential FMN targets. Multiple in vitro and in vivo experiments were performed further to explore the anti-OS effects and underlying mechanisms of FMN.

RESULTS

FMN inhibited OS growth in vivo

Here, OS tissues originating from patients were first collected (supplementary material, Fig. 1). To explore the impact of FMN on OS in vivo, PDX models were subsequently established with BALB/c-nu nude mice, which were then subjected to oral gavage intervention for 4 weeks according to the experimental groups [Fig. 1(A)]. 0.1% dimethyl sulfoxide (DMSO), which has been shown to have no significant effect on tumor growth, was used as the solvent for FMN in the above experiment [supplementary material, Figs. 2(a)–2(d)]. The results indicated that FMN significantly inhibited tumor growth [Figs. 1(B)–1(D)], and importantly, apparent liver toxicity was not observed [Fig. 1(E)]. This study confirms, for the first time, the anti-tumor efficacy of FMN in an OS PDX model, suggesting that FMN is a promising lead compound for the development of new anti-OS drugs.

FIG. 1.

FIG. 1.

Open in a new tab

FMN inhibited OS growth in vivo. (A) Flow chart illustrating the construction of the OS PDX model and the intervention process. (B)–(D) OS of BALB/c-nu nude mouse PDX models post-intervention, along with statistical analysis of tumor volume and weight. (E) HE staining of liver tissues from PDX model.

Bulk RNA-seq revealed the prospective targets of FMN

To further explore the possible targets of FMN, an in-depth analysis of bulk sequencing data from PDX models of the control and FMN groups was conducted. Differential analysis revealed the presence of differentially expressed genes (DEGs) between the two groups, including 572 mRNAs (275 down-regulated and 297 up-regulated) and 8009 lncRNAs (4262 down-regulated and 3747 up-regulated) [Figs. 2(A) and 2(B)]. Additionally, only one down-regulated circRNA was detected [Fig. 2(C)]. The heatmaps showed the top 150 genes with the most significant up- and down-regulated gene expression change trends (ranked by logFC), displayed in order of mRNAs, lncRNAs, and circRNAs between the control group and the FMN group [Figs. 2(D)–2(F)].

FIG. 2.

FIG. 2.

Open in a new tab

DEGs between the control and FMN groups based on bulk RNA-seq data. (A)–(C) Volcano diagram of DEGs, displayed in order of mRNAs, lncRNAs, and circRNAs. (D)–(F) Heatmaps of the top 150 up- and down-regulated genes with the most significant gene expression change trends (ranked by logFC), displayed in order of mRNAs, lncRNAs, and circRNAs. Top 5 genes with the most significant up-/down-regulated expression trend and GTSF1 were labeled.

ScRNA-seq and bulk RNA-seq identified GTSF1 as the target of FMN

As previously defined, quality control measures were applied to the OS scRNA-seq data. A total of 63 007 cells were extracted from 10 samples. Subsequent dimensionality reduction via PCA grouped these cells into 11 distinct clusters [Figs. 3(A)–3(C), and 3(E)]. Here, 11 cell types were identified on the basis of canonical marker genes, including macrophages (C1QA, C1QB, FCGR2A), monocytes (FCN1, MNDA, S100A8, S100A9), NK/T cells (NKG7, CD3E, CD3G, CD3D), neutrophils (MPO, DEFA3, DEFA4), fibroblasts (COL1A1, LUM, DCN, FN1), osteoblasts (IBSP, RUNX2, ALPL), B cells (CD79A, CD79B, MS4A1), plasma cells (MZB1, IGHG1), osteoclasts (ACP5, CTSK, MMP9), erythroid progenitors (HBA1, HBA2, HBB, HBD), and endothelial cells (EGFL7, PLVAP, VWF) [Fig. 3(D)]. In this study, osteoblasts derived from OS samples were initially presumed to be OS cells. Through inferCNV analysis, significant chromosomal copy number variations were observed in OS-derived osteoblasts compared with normal bone sample osteoblasts, confirming their malignant nature [Fig. 3(F)]. Furthermore, the “FindAllMarkers” function was utilized to identify DEGs between OB and OS, resulting in the selection of 175 up-DEGs presented in a volcano plot [Fig. 3(G)]. Moreover, 275 down-DEGs were identified from the OS PDX models between the FMN and control groups via DESeq2. After these results were intersected, a unique candidate target gene, GTSF1, was identified [Fig. 4(A)]. Single-cell analysis indicated that, compared with that in OB cells, GTSF1 expression was significantly elevated in OS cells, as illustrated in Figs. 4(B)–4(D).

FIG. 3.

FIG. 3.

Open in a new tab

ScRNA-seq analyses of the data from normal bone samples and OS samples. (A)–(C) Single-cell transcriptomic maps of 11 cell types identified from normal bone samples and OS samples. (D) Bubble plot showing the expression of marker genes used to define the aforementioned cell types. (E) Composition ratios of each cell type in the normal and OS groups. (F) InferCNV analysis identifying malignant osteoblasts. (G): Volcano plot displaying genes that were differentially expressed between the normal and OS groups.

FIG. 4.

FIG. 4.

Open in a new tab

ScRNA-seq and bulk RNA-seq identified GTSF1 as the target of FMN. (A) Venn diagram showing the unique intersection of up-regulated DEGs identified from scRNA-seq data in OS and down-regulated DEGs from bulk RNA-seq data in the FMN-treated group, pinpointing GTSF1 as the promising target of FMN. (B) and (D) t-SNE plot and violin plot showing the expression of GTSF1 in OS and normal samples on the basis of scRNA-seq data. (C) Volcano plot displaying the differential expression of GTSF1 between the FMN and control groups on the basis of the scRNA-seq data. (E) Expression of GTSF1 in osteoblasts and OS cell lines. (F)–(G) GTSF1 expression under FMN treatment at 0, 80, and 160 μM in U2OS and SAOS2 cells. (H)–(I) IHC analysis of GTSF1 expression in the control and FMN groups via statistical analysis.

FMN inhibited the expression of GTSF1

Before the regulatory effects of FMN on GTSF1 expression were assessed, preliminary screening was conducted across OS cell lines. Initially, the mRNA expression levels of GTSF1 were detected in five OS cell lines and the hFOB1.19 cell line. The results indicated that GTSF1 expression was significantly higher in U2OS and SAOS2 cells than in the other three OS cell lines, with hFOB1.19 cells used as the reference [Fig. 4(E)]. Consequently, these two cell lines were selected for further experiments. Interventions with FMN in the above two cell lines and subsequent qRT–PCR analyses revealed that FMN significantly down-regulated the expression of GTSF1 in vitro, with the extent of inhibition increasing with increasing concentrations of FMN [Figs. 4(F) and 4(G)]. IHC results also confirmed that FMN effectively reduced GTSF1 expression in vivo [Figs. 4(H) and 4(I)]. However, how FMN inhibits OS progression remains unclear.

FMN promoted the apoptosis and inhibited the proliferation of OS cells

To investigate the changes in signaling pathways in OS following treatment with FMN, GSVA was first performed on the scRNA-seq data. The results revealed that, compared with the cells from the normal bone sample, the cells from the OS sample presented a notably lower enrichment score for apoptosis signaling pathways, suggesting a state of apoptosis resistance in these cells [Fig. 5(A)]. Additionally, several cell cycle-associated pathways, including MYC_TARGETS_V1, E2F_TARGETS, and G2M_CHECKPOINT, presented significantly increased enrichment scores in OS cells, indicating a potential state of excessive proliferation [Fig. 5(A)]. However, these trends were notably reversed in OS cells treated with FMN. Specifically, the enrichment scores of apoptosis pathways in the FMN group were significantly greater than those in the control group, whereas the scores for proliferation-related pathways were lower [Fig. 5(B)]. On the basis of these findings, we hypothesize that FMN may exert anti-OS effects through promoting apoptosis and negatively regulating the cell cycle to inhibit the proliferation of OS cells. To further validate the inhibitory impact of FMN, flow cytometry and EdU assays were carried out to detect OS cell apoptosis and proliferation following FMN treatment. Here, our results demonstrated that FMN significantly promoted OS cell apoptosis, with the effects intensifying at higher FMN concentrations [Figs. 5(C)–5(F)]. Additionally, FMN inhibited OS cell proliferation in a concentration-dependent manner [Figs. 6(A)–6(C)]. Furthermore, the results of the transwell assay revealed that FMN also significantly inhibited the migration and invasion of OS cells [Figs. 6(D)–6(I)]. However, the role of GTSF1 in OS and whether FMN exerts its effects by regulating GTSF1 expression still need further validation.

FIG. 5.

FIG. 5.

Open in a new tab

FMN promoted the apoptosis of OS cells. (A)–(B) GSVA results indicating the potential pathways involved in the differences between osteosarcoma and normal OB cells on the basis of the scRNA-seq data and between the control and FMN groups on the basis of the bulk RNA-seq data. (C)–(F) Flow cytometric analysis of apoptosis induced by FMN at 0, 80, and 160 μM in U2OS and SAOS2 cells via statistical analysis.

FIG. 6.

FIG. 6.

Open in a new tab

FMN inhibited the proliferation, migration, and invasion of OS cells. (A)–(C) Results of the EdU assay and statistical analysis of U2OS and SAOS2 cells treated with FMN at 0, 80, and 160 μM. (D)–(F) Migration abilities of U2OS and SAOS2 cells under FMN treatment at 0, 80, and 160 μM, as determined by statistical analysis. (G)–(I) Invasion abilities of U2OS and SAOS2 cells under FMN treatment at 0, 80, and 160 μM, as determined by statistical analysis.

FMN inhibited OS progression by regulating GTSF1 expression

Before the specific roles of FMN and GTSF1 in OS were elucidated, the efficacy of GTSF1 overexpression plasmids (oe-GTSF1) was validated through qRT–PCR, confirming their effectiveness [Figs. 7(A) and 7(B)]. A rescue experiment was set up, which revealed that oe-GTSF1 inhibited but FMN promoted the apoptosis of OS cells. Thus, the anti-apoptosis effects of oe-GTSF1 were significantly reduced when intervened with FMN [Figs. 7(C)–7(F)]. Similarly, the results of the EdU assays aligned with these findings [Figs. 7(G)–7(I)], demonstrating that FMN could effectively promote apoptosis and suppress proliferation in OS cells by down-regulating GTSF1 expression, thereby exerting its anti-tumor effects.

FIG. 7.

FIG. 7.

Open in a new tab

FMN inhibited OS progression by regulating the GTSF1 expression. (A)–(B) Expression of GTSF1 in U2OS and SAOS2 cells after oe-GTSF1 intervention. (C)–(F) Flow cytometric apoptosis analysis of U2OS and SAOS2 cells under various FMN treatment conditions. (G)–(I) EdU assay in U2OS and SASO2 cells under various treatment conditions with FMN.

DISCUSSION

OS is the most common primary malignant tumor of the bone, characterized by rapid progression, metastatic tendency, and poor prognosis.30–32 Chemotherapy still remains the cornerstone of OS treatment, but commonly used chemotherapy drugs such as cisplatin are limited by serious side effects and the possibility of developing drug resistance.33–35 In recent years, active compounds derived from natural herbs have attracted significant interest owing to their multi-target and low-toxicity characteristics.36 The anti-tumor efficacy of FMN in OS remains superficial, and the specific mechanisms of action are still not well elucidated. In our study, we utilized the PDX model for the first time to explore the anti-OS effects of FMN. Through sequencing analysis and experiments, we confirmed that FMN can effectively promote apoptosis and inhibit the proliferation of OS cells through down-regulating the GTSF1 expression, thereby exerting its anti-tumor effects.

FMN extracted from Spatholobi Caulis has been reported to inhibit the progression of various cancers.17,37–41 However, there are currently only a few studies based on in vitro environments, and CDX models have initially revealed the anti-tumor activity of FMN in OS.21,22 In view of this, we established OS PDX model and intervened with FMN, which resulted in a significant reduction in tumor volume and weight. Thus, to explore the possible mechanism of FMN, we conducted bulk RNA-seq of OS samples from the PDX model and integrated these data with scRNA-seq data, identifying GTSF1 as an essential target of FMN in OS. GTSF1, located on chromosome 12q13.13, is involved in DNA methylation and the activation of retrotransposons in germ cells.42 Reports on GTSF1 in human tumors are rare, with only a few studies reporting its role in hepatocellular carcinoma and myeloid leukemia.42,43 To the best of our knowledge, no research has yet revealed the role of GTSF1 in OS. Therefore, we performed GSVA analysis and found that the roles of FMN and GTSF1 may be related to the apoptotic and proliferative phenotypes of OS. Furthermore, we conducted a series of experiments to validate our hypothesis. As anticipated, in vitro experiments confirmed that FMN promoted the apoptosis and inhibited the proliferation of U2OS and SAOS2 cells. However, whether the effects of FMN on OS are related to the regulation of GTSF1 is still unclear. Hence, through experimentation, FMN inhibited the expression of GTSF1 in a concentration-dependent manner in OS cells. Moreover, oe-GTSF1 inhibited apoptosis and promoted proliferation in OS cells, whereas FMN treatment resulted in the opposite effects. When oe-GTSF1 was combined with FMN intervention, it effectively countered the pro-oncogenic effects of oe-GTSF. Additionally, FMN effectively suppressed the invasion and migration of OS cells. As a result, FMN has shown great potential for the development of anti-tumor drugs aimed at treating OS.

While the current study has elucidated the anti-OS effects of FMN and its possible mechanisms through a series of in vitro and in vivo experiments, the conclusions drawn require validation with larger sample sizes and more extensive experimental investigations.

CONCLUSION

Overall, our findings indicated that FMN could induce apoptosis in OS cells by inhibiting the expression of GTSF1 and suppressing the proliferation, invasion, and migration of OS cells. These findings position FMN as a highly promising candidate for anti-OS drug development, offering new directions for future therapeutic research in OS.

METHODS

Acquisition and processing of scRNA-seq data

This study incorporated two publicly available scRNA-seq datasets from the GEO database, including GSE162454 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE162454) and GSE169396 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE169396). GSE162454, extracted from our preliminary study, includes six OS samples not exposed to radiotherapy or chemotherapy, while GSE169396 consists of four samples from normal bone samples.44,45 All analyses were performed via the “R” and “Seurat” packages. The “merge” function of “Seurat” was utilized to integrate the scRNA-seq data from both the GSE162454 and GSE169396 datasets, and the “Harmony” function was used to eliminate batch effects. Data quality control was performed on the following criteria: (1) nFeature_RNA > 300; (2) nFeature_RNA < 4500; and (3) percent.mt < 15. Data normalization was subsequently conducted via the “NormalizeData” function, followed by principal component analysis (PCA) via the default settings of the “RunPCA” function to obtain the first 15 principal components. The “FindVariableFeatures” function identified the top 2000 variable genes for further analysis (parameters: selection. method = “vst”, nfeatures = 2000), and the data were scaled via the “ScaleData” function. The cell clusters were identified via the “FindNeighbors” and “FindClusters” functions with a resolution of 0.05. Differentially expressed genes (DEGs) for each cluster were identified via the “FindAllMarkers” function (parameters: min.pct = 0.25, logfc. threshold = 0.25), and cell types were annotated on the basis of previously reported marker genes.46–49 Additionally, copy number variation (CNV) analysis was performed to infer the presence of malignant osteoblasts through the “inferCNV” R package, with the following settings: cutoff = 0.1, denoise = TRUE, noise_logistic = TRUE, hclust_method = “ward. D2,” with all other parameters set to default values. Monocytes, macrophages, neutrophils, NK/T cells, plasma cells, and B cells were selected as reference cells for this analysis.

Construction and administration of the PDX model

Four-week-old female NCG mice and BALB/c-nu nude mice were purchased from Gempharmatech Co., Ltd. OS tissue was obtained from patients at the First Affiliated Hospital of Guangxi Medical University. After excision, the OS tissue was cut into granules approximately 3 mm in diameter and 30 mm3 in volume, which were subsequently transplanted subcutaneously into the dorsal region between the right iliac crests of NCG mice and passaged. Before the tumor volume of the third-generation NCG mice reached 1500 mm3, the tumors of the NCG mice were excised, cut into granules on similar sides, and transplanted subcutaneously into BALB/c-nu nude mice. Ten nude mice were randomly divided into control and FMN groups 1-week post-transplantation. After 1 week of modeling, drug intervention was performed under the following conditions: The FMN group received oral administration at a dose of 50 mg/kg every other day, while the control group received an equal volume of physiological saline. After 4 weeks, the tumors were excised and measured for volume and weight, after which sequencing and further experiments were performed to explore the action mechanism of FMN. Additionally, the liver was isolated, fixed in formalin, and subjected to pathological examination. This study was conducted following the guidelines of the Helsinki Declaration and approved by the Ethics Committee of the First Affiliated Hospital of Guangxi Medical University.

Bulk transcriptome sequencing

Total RNA was extracted from samples via TRIzol reagent according to the manufacturer's instructions. The RNA purity and concentration were assessed through the NanoDrop 2000 spectrophotometer (Thermo Scientific, USA), and RNA integrity was evaluated with an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). The transcriptome library was constructed via the VAHTS Universal V6 RNA-seq Library Prep Kit. Sequencing was performed on the Illumina NovaSeq 6000 platform, generating 150 bp paired-end raw reads in fastq format, which were processed via fastp software to remove low-quality reads. Alignments to the reference genome were completed via HISAT2 software, and gene expression levels (FPKM) were calculated. Read counts for each gene were obtained via HTSeq-count. The DESeq2 package was used for differential expression analysis, and genes with p-value < 0.05 and absolute values of log fold change (logFC)>1 were defined as DEGs.

Identification of FMN candidate target genes

The “FindAllMarkers” function of “Seurat” was utilized to preliminarily identify DEGs between normal osteoblasts and malignant osteoblasts, with parameters set at min.pct = 0.25 and logfc. threshold = 0.25. This analysis facilitated the initial screening for gene expression variations indicative of malignancy. Subsequently, genes that were up-regulated (up-regulated DEGs) under malignant conditions were subsequently further filtered according to stringent criteria: p-value < 0.05 and logFC>1. Similarly, the DESeq2 package was employed to identify the down-regulated DEGs between the control and FMN groups with selection criteria of p-value < 0.05 and logFC<−1. Ultimately, the intersection of up-regulated DEGs and down-regulated DEGs was considered as the candidate target genes for FMN in OS.

Gene set variation analysis

Gene set variation analysis (GSVA) was utilized to investigate the differences in pathway enrichment between normal bone samples and OS samples, as well as between the control group and the FMN group. The gene set “h.all.v7.4.symbols.gmt” was used as the reference set for pathway analysis. The results were visualized as heatmaps, which provide a clear and concise representation of the enriched pathways across different conditions.

Cell culture and reagent

The OS cell lines 143B, HOS, MG63, SAOS2, U2OS, and hFOB1.19 were all acquired from the Cell Bank of the Chinese Academy of Sciences. The cell lines 143B, HOS, MG63, and hFOB1.19 were cultured with DMEM (Gibco, USA), while SAOS2 and U2OS cell lines were maintained in McCoy's 5A medium (Gibco, USA). All the aforementioned cell lines, except hFOB1.19, were incubated at 37 °C in a 5% CO2 incubator. In contrast, hFOB1.19 cells were cultured at a lower temperature of 33.5 °C in a 5% CO2 incubator. Adherent cells were dissociated using 0.25% trypsin (Solarbio, Beijing). The media for the 143B, HOS, MG63, U2OS, and hFOB1.19 cell lines were supplemented with 1% penicillin/streptomycin (Solarbio, Beijing) and 10% fetal bovine serum (FBS; Gibco, USA). The SAOS2 medium additionally contained 15% FBS (Gibco, USA). FMN was dissolved in dimethyl sulfoxide (DMSO) and diluted with PBS to form a working solution for subsequent experiments.

Quantitative real-time polymerase chain reaction

First, the 143B, HOS, MG63, SAOS2, U2OS, and hFOB1.19 cell lines were cultured in 6-well plates as described above, and the cells were collected after 24 h. Total RNA was subsequently extracted via an RNA extraction kit (Beyotime Biotech, Shanghai), and cDNA was synthesized via the PrimeScriptTM RT reagent Kit (Takara bio, Beijing) according to the manufacturer's instructions. Quantitative real-time polymerase chain reaction (qRT–PCR) was performed via the SYBR Green PCR Kit (Sigma-Aldrich, Germany) to detect the mRNA expression of GTSF1 in different cell lines, with GTSF1 (Sangon Biotech, Shanghai) as the target gene and GAPDH (Sangon Biotech, Shanghai) as the internal control gene. Quantitative analyses were conducted on the 7500 Real-Time PCR System (Thermo Applied Biosystems, USA). Gene expression levels were calculated via the 2−ΔΔCt method. The primer sequences of GTSF1 are listed in supplementary material, Table 1. Furthermore, OS cell lines, which exhibit notably higher GTSF1 expression than the hFOB1.19 cell line, undergo intervention with FMN at various concentrations (80 and 160 μM). The expression of GTSF1 will then be assayed via the aforementioned methodology to elucidate the impact of FMN on GTSF1 expression.

HE staining and IHC

HE staining and IHC were conducted according to methods described in previous publications.50 Specifically, paraffin blocks of OS tissue were subjected to sectioning and deparaffinization, followed by rehydration with a gradient of ethanol. For H&E staining, the nuclei were stained with hematoxylin, differentiated with 1% hydrochloric acid in alcohol, and counterstained with eosin. Additionally, for immunohistochemical analysis, antigen retrieval was carried out via citrate buffer, endogenous peroxidase activity was quenched with 3% hydrogen peroxide, and the sections were blocked with goat serum for 30 min before incubation with the primary antibody (GTSF1; Proteintech, China) overnight. Afterward, staining was performed via an immunohistochemistry kit (Elabscience, China), the samples were counterstained with hematoxylin, and images were acquired via an orthogonal fluorescence microscope. The positive areas were then quantified via ImageJ software.

Cell transfection

U2OS and SAOS2 cells were seeded in 6-well plates at a density of 200 000 cells per well, establishing a control group (serving as the negative control for overexpression) and an oe-GTSF1 group (for the overexpression of GTSF1). Once more than 50% of the U2OS and SAOS2 cells had adhered to the plate surface, plasmid transfection was carried out in accordance with the manufacturer's instructions for the transfection reagent (Invitrogen, USA). Following 48 h of transfection, qRT–PCR was employed to evaluate GTSF1 expression in the altered cell lines. After the overexpression of these genes was verified, subsequent experiments were conducted via the above transfection protocol. Supplementary material, Table 2 provides the positive clone sequence of the plasmid utilized for overexpressing GTSF1.

Flow cytometry assay

U2OS and SAOS2 cells were seeded into 6-well plates at the density of 200 000 cells per well, respectively. To assess the impact of FMN on OS cell apoptosis, we established three groups: the control, 80 μM, and 160 μM groups. The latter two groups were administered 80 and 160 μM FMN, respectively. Furthermore, to investigate the influence of FMN-modulated GTSF1 expression on OS cell apoptosis, we designed four experimental groups: the control, oe-GTSF1, FMN, and oe-GTSF1+FMN groups. Among these groups, the FMN group received 80 μM FMN, the oe-GTSF1 group was transfected with the GTSF1 overexpression plasmid, and the oe-GTSF1+FMN group underwent both GTSF1 overexpression plasmid transfection and treatment with 80 μM FMN. The cells were harvested for apoptosis analysis via an Annexin V-APC/7-AAD Apoptosis Kit (MULTI SCIENCES, Zhejiang). The stained cells were subsequently analyzed via a Cytomics FC500 flow cytometer (Beckman Coulter, USA).

EdU assay

U2OS and SAOS2 cells were seeded into 24-well plates, each well containing a cell slide, at the density of 50 000 cells per well. The grouping settings and processing followed the same protocol as described in the subsection on the flow cytometry assay. After intervention according to experimental grouping, EdU working solution was added to the cells, which were incubated, fixed, permeabilized, and then treated with Click Additive and DAPI according to the protocol of the EdU kit (Beyotime Biotech, Shanghai). Observations and imaging were subsequently conducted via an inverted phase contrast microscope (Olympus BX53, Japan). The cell counts for each group were quantified via ImageJ software (Version Fiji, ImageJ Software, USA). Statistical significance was set at P < 0.05.

Cell migration and invasion assays

U2OS and SAOS2 cells were seeded in a 24-well plate, with an 8 μm pore size transwell chamber (Jet Biofiltration, Guangzhou) per well, with 50 000 cells per chamber. The cell suspensions were added to the chambers according to the experimental groups. For the migration assay, no Matrigel was applied to the upper chamber; whereas for the invasion assay, a layer of Matrigel (Corning, USA) was precoated on the upper chamber to simulate the extracellular matrix barrier. McCoy's 5A medium supplemented with 10% FBS was used as the chemoattractant in the lower chamber. After a 48-h intervention, the cells were fixed with 4% paraformaldehyde and stained with crystal violet. The cells inside the chamber were removed, and then the migrated cells were inspected and photographed through an inverted phase contrast microscope. The cell counts for each group were quantified via ImageJ software.

Statistical analysis

Statistical analysis was performed via R software and GraphPad Prism 8.0. Independent samples t-tests were used to compare GTSF1 expression between the two groups, and rank sum tests were used for comparisons between multiple groups. P < 0.05 was considered statistically significant unless otherwise stated. Significance levels were defined in this study as follows: * indicates p < 0.05, ** indicates p < 0.01, *** indicates p < 0.001, and **** indicates p < 0.0001.

SUPPLEMENTARY MATERIAL

See the supplementary material for more information on the sequence of the primer of GTSF1 (supplementary material, Table 1); positive clone sequence of GTSF1 overexpression (supplementary material, Table 2); HE staining of OS tissue (supplementary material, Fig. 1); effect of the solvent DMSO on OS growth in vivo; (A-B) in vivo tumor growth and gross tumor morphology as well as (C-D) measurement of tumor volume and weight (supplementary material, Fig. 2).

ACKNOWLEDGMENTS

This work was supported by the National Natural Science Foundation of China (No. 82260814, Yun Liu), the Guangxi Natural Science Foundation (Grant No. 2025GXNSFBA069554, Tianyu Xie), the Guangxi Natural Science Foundation Youth Science Fund Project (Grant No. 2025GXNSFBA069555, Chengsen Lin), the Youth Science Foundation of the Guangxi Medical University (No. GXMUYSF202533, Tianyu Xie), the “Medical Excellence Award” Funded by the Creative Research Development Grant from the First Affiliated Hospital of Guangxi Medical University (Yun Liu), and the First-class discipline innovation-driven talent program of Guangxi Medical University (Yun Liu).

Contributor Information

Yunhua Lin, Email: mailto:mlinyunhua@126.com.

Yun Liu, Email: mailto:liuyun200450250@sina.com.

AUTHOR DECLARATIONS

Conflict of Interest

The authors have no conflicts to disclose.

Ethics Approval

Ethics approval for experiments reported in the submitted manuscript on animal or human subjects was granted. This study was conducted in accordance with the guidelines of the Helsinki Declaration and approved by the Ethics Committee of the First Affiliated Hospital of Guangxi Medical University.

Author Contributions

Tianyu Xie and Jiming Liang contributed equally to this work.

Tianyu Xie: Data curation (equal); Formal analysis (equal); Investigation (equal); Methodology (equal); Visualization (equal); Writing – original draft (equal). Jiming Liang: Data curation (equal); Formal analysis (equal); Methodology (equal); Validation (equal); Visualization (equal); Writing – original draft (equal). Chengsen Lin: Formal analysis (equal); Methodology (equal); Visualization (equal); Writing – review & editing (equal). Wenyu Feng: Data curation (equal); Formal analysis (equal); Methodology (equal); Visualization (equal); Writing – review & editing (equal). Mingxiu Yang: Formal analysis (equal); Methodology (equal); Visualization (equal); Writing – review & editing (equal). Shanhang Li: Data curation (equal); Formal analysis (equal); Software (equal); Visualization (equal); Writing – review & editing (equal). Liang Xiong: Data curation (equal); Formal analysis (equal); Methodology (equal); Visualization (equal); Writing – review & editing (equal). Kai Luo: Data curation (equal); Formal analysis (equal); Visualization (equal); Writing – review & editing (equal). Feicui Li: Conceptualization (equal); Formal analysis (equal); Methodology (equal); Visualization (equal); Writing – review & editing (equal). Shengping Tang: Formal analysis (equal); Validation (equal); Visualization (equal); Writing – review & editing (equal). Shangyu Liu: Data curation (equal); Formal analysis (equal); Visualization (equal); Writing – review & editing (equal). Qian Huang: Formal analysis (equal); Methodology (equal); Visualization (equal); Writing – review & editing (equal). Shijie Liao: Formal analysis (equal); Methodology (equal); Visualization (equal); Writing – review & editing (equal). Jianhong Liu: Formal analysis (equal); Methodology (equal); Visualization (equal); Writing – review & editing (equal). Yangjie Cai: Data curation (equal); Methodology (equal); Software (equal); Visualization (equal); Writing – review & editing (equal). Fuxing Tang: Formal analysis (equal); Visualization (equal); Writing – review & editing (equal). Haijun Tang: Formal analysis (equal); Methodology (equal); Visualization (equal); Writing – review & editing (equal). Qingjun Wei: Data curation (equal); Formal analysis (equal); Investigation (equal); Validation (equal); Writing – review & editing (equal). Yunhua Lin: Conceptualization (equal); Project administration (equal); Supervision (equal); Writing – review & editing (equal). Yun Liu: Conceptualization (equal); Funding acquisition (equal); Project administration (equal); Supervision (equal); Writing – review & editing (equal).

DATA AVAILABILITY

The data that support the findings of this study are openly available in GSA at https://ngdc.cncb.ac.cn/gsa, Ref. 51; The data that support the findings of this study are openly available in NCBI at https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE162454, Ref. 52; and The data that support the findings of this study are openly available in NCBI at https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE169396, Ref. 53.

References

  • 1.Yuanzhuang Z., Yeqiu X., Yuxin B., Yinzhou L., Guanzhen Q., Ming H., Jie L., Jian X., Bin C., and Yong W. J. E., “N6-methyladenosine (m6A) modification in osteosarcoma: Expression, function and interaction with noncoding RNAs - an updated review,” Epigenetics 18(1), 2260213 (2023). 10.1080/15592294.2023.2260213 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Hongde Y., Bolun Z., Lin Q., Jian H., Mingyang G., Jiaze L., and Qingtao M. J. P., “AP003352.1/miR-141-3p axis enhances the proliferation of osteosarcoma by LPAR3,” PeerJ 11, e15937 (2023). 10.7717/peerj.15937 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Lei Y., Yan S., Hai-Long W., Qi-Yuan D., Jing G., Li-Sheng H., Wen-Hao W., Zhi-Ping G., and Bin-Sheng Y. J. F. O., “p53 inhibits CTR1-mediated cisplatin absorption by suppressing SP1 nuclear translocation in osteosarcoma,” Front. Oncol. 12, 1047194 (2023). 10.3389/fonc.2022.1047194 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Shaoyan S., Qian W., and Xiaolong D., “Comprehensive bioinformatics analysis reveals the oncogenic role of FoxM1 and its impact on prognosis, immune microenvironment, and drug sensitivity in osteosarcoma,” J. Appl. Genet. 64(4), 779–796 (2023). 10.1007/s13353-023-00785-5 [DOI] [PubMed] [Google Scholar]
  • 5.Le Z., Longzhou L., Wen-Juan N., Fuhua X., and Xiao-Min L. J. I. J. O., “Circular RNAs in osteosarcoma: An update of recent studies (review),” Int. J. Oncol. 63(5), 123 (2023). 10.3892/ijo.2023.5571 [DOI] [PubMed] [Google Scholar]
  • 6.Rathore R., Caldwell K. E., Schutt C., Brashears C. B., Prudner B. C., Ehrhardt W. R., Leung C. H., Lin H., Daw N. C., Beird H. C., Giles A., Wang W.-L., Lazar A. J., Chrisinger J. S. A., Livingston J. A., and Van Tine B. A., “Metabolic compensation activates pro-survival mTORC1 signaling upon 3-phosphoglycerate dehydrogenase inhibition in osteosarcoma,” Cell Rep. 34(4), 108678 (2021). 10.1016/j.celrep.2020.108678 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Feifei P., Haoyu G., Deyao S., Fengxia C., Yizhong P., Xin H., Jianxiang L., Zhicai Z., and Zengwu S. J. G. D., “The generation and use of animal models of osteosarcoma in cancer research,” Genes Dis. 11(2), 664–674 (2023). 10.1016/j.gendis.2022.12.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Park J. A. and Cheung N. K. V., “Promise and challenges of T cell immunotherapy for osteosarcoma,” Int. J. Mol. Sci. 24(15), 12520 (2023). 10.3390/ijms241512520 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Qu J., Qiu B., Zhang Y., Hu Y., Wang Z., Guan Z., Qin Y., Sui T., Wu F., Li B., Han W., and Peng X., “The tumor-enriched small molecule gambogic amide suppresses glioma by targeting WDR1-dependent cytoskeleton remodeling,” Signal Transduction Targeted Ther. 8(1), 424 (2023). 10.1038/s41392-023-01666-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Wang R., Gao C., Yu M., Song J., Feng Z., Wang R., Pan H., Liu H., Li W., and Fan X., “Mechanistic prediction and validation of Brevilin A therapeutic effects in lung cancer,” BMC Complementary Med. Ther. 24(1), 214 (2024). 10.1186/s12906-024-04516-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Zhang B., Hao Z., Zhou W., Zhang S., Sun M., Li H., Hou N., Jing C., and Zhao M., “Formononetin protects against ox-LDL-induced endothelial dysfunction by activating PPAR-γ signaling based on network pharmacology and experimental validation,” Bioengineered 12(1), 4887–4898 (2021). 10.1080/21655979.2021.1959493 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Mayuko S., Risa K., Hiromi G., Atsushi K., Makio S., and Ko F. J. N., “Formononetin attenuates H(2)O(2)-induced cell death through decreasing ROS level by PI3K/Akt-Nrf2-activated antioxidant gene expression and suppressing MAPK-regulated apoptosis in neuronal SH-SY5Y cells,” Neurotoxicology 85, 186 (2021). 10.1016/j.neuro.2021.05.014 [DOI] [PubMed] [Google Scholar]
  • 13.Park Y., Choo S. P., Jung G. S., Kim S., Lee M. J., Im W., Park H., Lee I., Lee J. H., Cho S., and Choi Y. S., “Formononetin inhibits progression of endometriosis via regulation of p27, pSTAT3, and progesterone receptor: In vitro and in vivo studies,” Nutrients 15(13), 3001 (2023). 10.3390/nu15133001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Zi-Wen Z., Xue-Yan Z., Shu-Ying L., Si-En L., Yu-Han Z., Kunmei J., and Jia-Jie C. J. M., “Formononetin inhibits mast cell degranulation to ameliorate compound 48/80-induced pseudoallergic reactions,” Molecules 28(13), 5271 (2023). 10.3390/molecules28135271 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Vijay Kumar B. and Rituraj P. J. C. P., “A comparative study on inclusion complex formation between formononetin and β-cyclodextrin derivatives through multiscale classical and umbrella sampling simulations,” Carbohydr. Polym. 310, 120729 (2023). 10.1016/j.carbpol.2023.120729 [DOI] [PubMed] [Google Scholar]
  • 16.Selepe M. A., “Isoflavone derivatives as potential anticancer agents: synthesis and bioactivity studies,” ChemMedChem 19(23), e202400420 (2024). 10.1002/cmdc.202400420 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Almatroodi S. A., Almatroudi A., Khan A. A., and Rahmani A. H., “Potential therapeutic targets of formononetin, a type of methoxylated isoflavone, and its role in cancer therapy through the modulation of signal transduction pathways,” Int. J. Mol. Sci. 24(11), 9719 (2023). 10.3390/ijms24119719 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Xu Lou I., Yu X., and Chen Q., “Exploratory review on the effect of Astragalus mongholicus on signaling pathways,” Front. Pharmacol. 15, 1510307 (2024). 10.3389/fphar.2024.1510307 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Tay K.-C., Tan L. T.-H., Chan C. K., Hong S. L., Chan K.-G., Yap W. H., Pusparajah P., Lee L.-H., and Goh B.-H., “Formononetin: A review of its anticancer potentials and mechanisms,” Front. Pharmacol. 10, 820 (2019). 10.3389/fphar.2019.00820 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Wang J. Y., Jiang M. W., Li M. Y., Zhang Z. H., Xing Y., Ri M., Jin C. H., Xu G. H., Piao L. X., Jin H. L., Ma J., Jin Y., Zuo H. X., and Jin X., “Formononetin represses cervical tumorigenesis by interfering with the activation of PD-L1 through MYC and STAT3 downregulation,” J. Nutr. Biochem. 100, 108899 (2022). 10.1016/j.jnutbio.2021.108899 [DOI] [PubMed] [Google Scholar]
  • 21.Liu Y., He J., Chen X., Li J., Shen M., Yu W., Yang Y., and Xiao Z. J. C. P., “The proapoptotic effect of formononetin in human osteosarcoma cells: Involvement of inactivation of ERK and Akt pathways,” Cell. Physiol. Biochem. 34(3), 637–645 (2014). 10.1159/000363029 [DOI] [PubMed] [Google Scholar]
  • 22.Lizhi C., Yue Z., Zheng W., Shuang L. T. L., Yanfang W., Yang Y., Hongmei L., and Wenhua H. J. A., “Anti-cancer targets and molecular mechanisms of formononetin in treating osteosarcoma based on network pharmacology,” Aging (Albany NY) 15(20), 11489 (2023). 10.18632/aging.205139 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Kang L., Hongyuan S., Minghui L., Jiarong C., Qingshui Y., and Pingyue L. J. T. C. R., “Formononetin inhibits osteosarcoma cell proliferation and promotes apoptosis by regulating the miR-214-3p/phosphatase and tensin homolog pathway,” Transl. Cancer Res. 9(8), 4914 (2020). 10.21037/tcr-20-2296 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Wei H., Xianpei W., Jiandong T., Niansu X., Guoping Z., Li Z., and Luanhai O. J. J. C. P., “In vitro and in vivo studies of antiosteosarcoma activities of formononetin,” J. Cell. Physiol. 234(10), 17305–17313 (2019). 10.1002/jcp.28349 [DOI] [PubMed] [Google Scholar]
  • 25.Li Z., Zheng W., Wang H., Cheng Y., Fang Y., Wu F., Sun G., Sun G., Lv C., and Hui B., “Application of animal models in cancer research: Recent progress and future prospects,” Cancer Manage. Res. 13, 2455–2475 (2021). 10.2147/CMAR.S302565 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Chen J., Jin Y., Li S., Qiao C., Peng X., Li Y., Gu Y., Wang W., You Y., Yin J., Shan Y., Wang Y.-X., Qin M., Li H., Cai Y., Dong Y., Peng S., and Pan L., “Patient-derived xenografts are a reliable preclinical model for the personalized treatment of epithelial ovarian cancer,” Front. Oncol. 11, 744256 (2021). 10.3389/fonc.2021.744256 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.da Costa M. E. M., Droit R., Khneisser P., Gomez-Brouchet A., Adam-de-Beaumais T., Nolla M., Signolles N., Torrejon J., Lombard B., Loew D., and Ayrault O., “Longitudinal characterization of primary osteosarcoma and derived subcutaneous and orthotopic relapsed patient-derived xenograft models,” Front. Oncol. 13, 1166063 (2023). 10.3389/fonc.2023.1166063 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Qian Z., Zhiqiang Z., Guodi L., Dehua L., Zhangjie G., Linsong Z., Yingjiao P., Xingbing C., Lu W., Guoping L., Xiaoli T., and Ziming Z. J. B. C., “B7-H3 targeted CAR-T cells show highly efficient anti-tumor function against osteosarcoma both in vitro and in vivo,” BMC Cancer 22(1), 1124 (2022). 10.1186/s12885-022-10229-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Dafu C., Chengyue L., Weifeng L., Meiyu S., Meizhou S., Jianxun G., Jingjing C., Jing-Jun N., Peng L., Yuwen L., Bingran Y., Renxian W., Shun D., and Fu-Jian X. J. B. M., “Reduction-responsive nucleic acid nanocarrier-mediated miR-22 inhibition of PI3K/AKT pathway for the treatment of patient-derived tumor xenograft osteosarcoma,” Bioact. Mater. 28, 376–385 (2023). 10.1016/j.bioactmat.2023.05.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Jian-Hua W., Chen G., Feng-Jin G., Xiao Z., Ming-Sheng Z., Hong Q., Teng-Fei C., Yang L., Liang Q., and Hui-Hua X. J. I. J. M. M., “Knockdown of STIP1 inhibits the invasion of CD133-positive cancer stem-like cells of the osteosarcoma MG63 cell line via the PI3K/Akt and ERK1/2 pathways,” Int. J. Mol. Med. 46(6), 2251–2259 (2020). 10.3892/ijmm.2020.4764 [DOI] [PubMed] [Google Scholar]
  • 31.Mao P., Feng Z., Liu Y., Zhang K., Zhao G., Lei Z., Di T., and Zhang H., “The role of ubiquitination in osteosarcoma development and therapies,” Biomolecules 14(7), 791 (2024). 10.3390/biom14070791 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Wang Q., Wu Y., Lin M., Wang G., Liu J., Xie M., Zheng B., Shen C., and Shen J., “BMI1 promotes osteosarcoma proliferation and metastasis by repressing the transcription of SIK1,” Cancer Cell Int. 22(1), 136 (2022). 10.1186/s12935-022-02552-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Gao X., Zhang C., Wang Y., Zhang P., Zhang J., and Hong T., “Berberine and cisplatin exhibit synergistic anticancer effects on osteosarcoma MG-63 cells by inhibiting the MAPK pathway,” Molecules 26(6), 1666 (2021). 10.3390/molecules26061666 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Banche Niclot A. G. S., Marini E., Ferrero I., Barbero F., Rosso E., Fenoglio I., Barge A., Pessina A., Coccè V., Paino F., Mareschi K., and Fagioli F., “A new paclitaxel formulation based on secretome isolated from mesenchymal stem cells shows a significant cytotoxic effect on osteosarcoma cell lines,” Pharmaceutics 15(9), 2340 (2023). 10.3390/pharmaceutics15092340 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Li M., Wang M., Huang J., Tang S., Yang J., Xu Z., Xu G., Chen X., Liu J., and Yang C., “High-performance pyrite nano-catalyst driven photothermal/chemodynamic synergistic therapy for Osteosarcoma,” J. Nanobiotechnol. 22(1), 141 (2024). 10.1186/s12951-024-02419-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Zhou H., Zhang M., Cao H., Du X., Zhang X., Wang J., and Bi X., “Research progress on the synergistic anti-tumor effect of natural anti-tumor components of chinese herbal medicine combined with chemotherapy drugs,” Pharmaceuticals (Basel) 16(12), 1734 (2023). 10.3390/ph16121734 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Shan L., Linlian Z., Yufeng H., and Ting S. J. A. B. P., “Formononetin enhances the chemosensitivity of triple negative breast cancer via BTB domain and CNC homolog 1-mediated mitophagy pathways,” Acta Biochim. Pol. 70(3), 533–539 (2023). 10.18388/abp.2020_6466 [DOI] [PubMed] [Google Scholar]
  • 38.Ji-Su O., Tae-Hyeon K., Jong-Hyun P., Hyangi L., In-A C., Jae-Seek Y., Gyeong-Je L., Yo-Seob S., Do Kyung K., Chun Sung K., Sun-Kyoung Y., Heung-Joong K., Su-Gwan K., and Jae-Sung K. J. O. R., “Formononetin induces apoptotic cell death through the suppression of mitogen-activated protein kinase and nuclear factor-κB phosphorylation in FaDu human head and neck squamous cell carcinoma cells,” Oncol. Rep. 43(2), 700–710 (2019). 10.3892/or.2019.7432 [DOI] [PubMed] [Google Scholar]
  • 39.Yang Y., Zhao Y., Ai X., Cheng B., and Lu S., “Formononetin suppresses the proliferation of human non-small cell lung cancer through induction of cell cycle arrest and apoptosis,” Int. J. Clin. Exp. Pathol. 7(12), 8453 (2015). [PMC free article] [PubMed] [Google Scholar]
  • 40.Xing Z., Lingyun B., Yu Y., and Jian C. J. N. C., “Formononetin induces apoptosis in PC-3 prostate cancer cells through enhancing the Bax/Bcl-2 ratios and regulating the p38/Akt pathway,” Nutr. Cancer 66(4), 656–661 (2014). 10.1080/01635581.2014.894098 [DOI] [PubMed] [Google Scholar]
  • 41.Liu X. J., Li Y. Q., Chen Q. Y., Xiao S. J., and Zeng S. E., “Up-regulating of RASD1 and apoptosis of DU-145 human prostate cancer cells induced by formononetin in vitro,” Asian Pac. J. Cancer Prev. 15(6), 2835 (2014). 10.7314/APJCP.2014.15.6.2835 [DOI] [PubMed] [Google Scholar]
  • 42.De-Yong G., Yun L., Xiao-Li L., Ying-Ying W., and Liang-Ming L. J. O. L., “GTSF1 gene may serve as a novel potential diagnostic biomarker for liver cancer,” Oncol. Lett. 15(3), 2835–2839 (2017). 10.3892/ol.2017.7695 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Chunmei Z., Guanchen B., Weijie Z., Dongfang B., and Gaofeng B. J. M. S. M., “Identification of miRNA-mRNA network associated with acute myeloid leukemia survival,” Med. Sci. Monit. 23, 4705 (2017). 10.12659/msm.903989 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Xiang Q., Ying L., Hui S., Zun W., Yun G., Junxiao Y., Xiaohua L., Huixi Z., Yu C., Cui Z., Wanqiang L., Liang C., Yihe H., Boyang L., Wendi S., Xuezhen Z., Li-Jun T., Hong-Mei X., and Hong-Wen D. J. A., “Single-cell RNA sequencing of human femoral head in vivo,” Aging (Albany NY) 13(11), 15595 (2021). 10.18632/aging.203124 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Yun L., Wenyu F., Yan D., Mengying B., Zhenchao Y., Mingwei H., Zhaojie Q., Shijie L., Juliang H., Qian H., Zhenyuan Y., Yanyu Z., Binqian G. R. H., Rirong Y., Yonghua J., Jinling L., Zengming X., Xinli Z., Chengsen L., Jiake X., Yu Y. J. M., Qingjun W., and Zengnan M. J. F. O., “Single-cell transcriptomics reveals the complexity of the tumor microenvironment of treatment-naive osteosarcoma,” Front. Oncol. 11, 709210 (2021). 10.3389/fonc.2021.709210 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Guo R., Kong J., Tang P., Wang S., Sang L., Liu L., Guo R., Yan K., Qi M., Bian Z., Song Y., Jiang Z., and Li Y., “Unbiased single-cell sequencing of hematopoietic and immune cells from aplastic anemia reveals the contributors of hematopoiesis failure and dysfunctional immune regulation,” Adv. Sci. (Weinh) 11(10), e2304539 (2024). 10.1002/advs.202304539 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Liu Y., He M., Tang H., Xie T., Lin Y., Liu S., Liang J., Li F., Luo K., Yang M., Teng H., Luo X., He J., Liao S., Huang Q., Feng W., Zhan X., and Wei Q., “Single-cell and spatial transcriptomics reveal metastasis mechanism and microenvironment remodeling of lymph node in osteosarcoma,” BMC Med. 22(1), 200 (2024). 10.1186/s12916-024-03319-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Wang Z., Tao P., Fan P., Wang J., Rong T., Hou Y., Zhou Y., Lu W., Hong L., Ma L., Zhang Y., and Tong H., “Insight of a lipid metabolism prognostic model to identify immune landscape and potential target for retroperitoneal liposarcoma,” Front. Immunol. 14, 1209396 (2023). 10.3389/fimmu.2023.1209396 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Zhu J., Yuan J., Arya S., Du Z., Liu X., and Jia J., “Exploring the immune microenvironment of osteosarcoma through T cell exhaustion-associated gene expression: A study on prognosis prediction,” Front. Immunol. 14, 1265098 (2023). 10.3389/fimmu.2023.1265098 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Chen Y., Liu Y., Jiang K., Wen Z., Cao X., and Wu S., “Linear ubiquitination of LKB1 activates AMPK pathway to inhibit NLRP3 inflammasome response and reduce chondrocyte pyroptosis in osteoarthritis,” J. Orthop. Transl. 39, 1 (2023). 10.1016/j.jot.2022.11.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.GSA (2025). “Genome sequence archive,” GSA. https://ngdc.cncb.ac.cn/gsa.
  • 52.NCBI (2021). “Single-cell RNA sequencing of human osteosarcoma,” NCBI. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE162454.
  • 53.NCBI (2021). “Single-cell RNA sequencing of human femoral head in vivo,” NCBI. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE169396. [DOI] [PMC free article] [PubMed]

Associated Data

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

Data Availability Statement

The data that support the findings of this study are openly available in GSA at https://ngdc.cncb.ac.cn/gsa, Ref. 51; The data that support the findings of this study are openly available in NCBI at https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE162454, Ref. 52; and The data that support the findings of this study are openly available in NCBI at https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE169396, Ref. 53.

Articles from APL Bioengineering are provided here courtesy of American Institute of Physics

RESOURCES