Therapeutic Potential of Salvia miltiorrhiza in Glioblastoma: Evidence from in vitro and in vivo Studies Authors: Cheng-Chi Lee, Woon-Man Kung, Yuan-Chieh Yeh, Yu-Li Chen, Chuan-Hsin Chang, Chih-Cheng Chien, Kuo-Chen Wei, Chi-Cheng Chuang, Peng-Wei Hsu, Tsong-Long Hwang, Yin-Cheng Huang Published: 2026 ABSTRACT AbstractBackgroundGlioblastoma (GB) is a highly aggressive and invasive brain tumor characterized by a poor prognosis. The Danshen (Salvia miltiorrhiza Bunge) crude extract (DCE) demonstrated effective anticancer effects against glioblastoma. This study investigated the in vitro and in vivo efficacy of DCE on GB and explored its potential mechanisms of action.Materials and MethodsCell viability and migration were assessed using 3-(4, 5-dimethyl-2-thiazolyl)-2, 5-diphenyl-2H-tetrazolium bromide and wound healing assays, respectively. The cell cycle and apoptosis were evaluated using flow cytometry. The in vivo effect of DCE was determined using a U87-MG GB xenograft model in BALB/c-nude mice. The gene expression profile of DCE-treated tumor tissue was assessed using RNA sequencing, with validation of key genes by using quantitative reverse transcription polymerase chain reaction (qRT-PCR).ResultsDCE treatment significantly reduced cell viability in both the U87-MG and DBTRG-05MG GB cell lines. Additionally, DCE notably induced G0/G1 cell cycle arrest and promoted apoptosis in both cell lines. Furthermore, DCE effectively inhibited cell migration in both GB cell lines. The in vivo results revealed that DCE significantly suppressed tumor growth in the GB xenograft mouse model. qRT-PCR analysis indicated that the anticancer activity of DCE is associated with the downregulation of CXCL2, CXCL3, CXCL8, and PTGS2 gene expression.ConclusionDCE effectively suppressed the progression of GB in both in vitro and in vivo models, likely through modulation of modulating pro-inflammatory signaling pathways. These findings highlight DCE as a promising candidate for therapeutic development against GB.Keywords: Danshen, radix Salvia miltiorrhiza, glioblastoma, in vivo animal study, signaling pathway Introduction Glioblastoma (GB) is a highly aggressive and invasive brain tumor with a dismal prognosis. The current standard treatment for GB includes surgical resection followed by a combination of radiotherapy and chemotherapy. Despite these interventions, the median survival time remains poor—typically ranging from 12 to 16 months over the past decade.1,2 Temozolomide (TMZ), the first-line chemotherapeutic agent, has shown limited efficacy and is often associated with significant adverse effects, highlighting the urgent need for novel therapeutic strategies.3 Traditional Chinese Medicine (TCM), with a history spanning thousands of years, is increasingly recognized as a complementary or alternative approach in cancer therapy. Its perceived superior safety profile compared to conventional treatments has garnered growing attention. An increasing body of evidence suggests that TCM interventions may enhance the physical and emotional well-being of cancer patients, as well as mitigate the adverse effects of chemotherapy and surgery.4–8 Moreover, several representative TCM-derived phytochemicals, such as curcumin, resveratrol, berberine, wogonin, quercetin, and celastrol, have been demonstrated to exhibit antitumor effects in various types of cancer.9,10 For example, curcumin (a constituent of Curcuma longa) has been shown to strongly induce apoptosis in colorectal cancer cells.11 Quercetin, which is commonly found in daily foods such as nuts and tea, as well as in the general diet, has been reported to induce cell cycle arrest and promote tumor regression.12 Resveratrol (a constituent of Polygonum cuspidatum) promotes apoptosis via NF-κB downregulation, leading to decreased Bcl-2 expression and activation of caspase-3 and caspase-8.13 Licorice has also been shown to possess antioxidant, anti-inflammatory, antiviral, and anticancer properties. Its anticancer effects are primarily mediated through cell cycle arrest and induction of apoptosis in leukemia, breast, and prostate cancer cells.14,15 Despite the growing number of studies elucidating the anticancer mechanisms of TCM-derived phytochemicals across various tumor types, reports specifically focusing on GB remain limited. Among these TCMs, Curcuma longa has been relatively well studied in the context of glioblastoma.16 However, our data indicated that Salvia miltiorrhiza Bunge (Danshen) exhibits comparable or even more potent inhibitory effects on GB cells (Supplementary Figure 1). Salvia miltiorrhiza Bunge, commonly known as Danshen in Chinese, is a widely used TCM well-documented efficacy in the treatment of cardiovascular,17 cerebrovascular17 and inflammatory18 diseases. In addition, Danshen has been reported to possess anticancer properties, including antiproliferative, pro-apoptotic, and antiangiogenic effects. Several studies have shown that Danshen can inhibit the progression of GB19 as well as other cancer types.20 Moreover, active compounds isolated from Danshen have demonstrated the ability to suppress GB proliferation and infiltration through multiple mechanisms and signaling pathways.21,22 However, most investigations into the use of Danshen against GB have been limited to in vitro experiments, with little exploration of gene expression changes. Furthermore, prior studies have primarily focused on individual purified compounds derived from Danshen—such as dihydrotanshinone, tanshinone I, cryptotanshinone, dihydrotanshinone I, tanshinone IIA, and tanshinone IIB,21–24 despite the fact that clinical applications generally involve crude extracts containing a complex mixture of bioactive constituents rather than pure compound alone.25 To address these gaps in current knowledge, the present study aimed to evaluate the therapeutic efficacy of a crude Danshen extract (DCE) against GB using both in vitro and in vivo models. In addition, this study investigated the potential underlying mechanisms of action, including pathway modulation and gene expression changes, associated with DCE treatment. To the best of our knowledge, this is the first comprehensive study to assess the effects of DCE on GB, incorporating in vitro and in vivo assays, pathway analysis, and transcriptomic profiling. Materials and Methods Chemicals and Reagents Tanshinone IIA was purchased from Sunhank Technology (Tainan, Taiwan). 3-(4, 5-dimethyl-2-thiazolyl)-2, 5-diphenyl-2H-tetrazolium bromide (MTT) assay kits and other necessary chemicals were obtained from Sigma-Aldrich (St. Louis, MO, USA). Sample Preparation and Extraction Dried roots of Danshen were purchased from https://www.natural-grass-co.com/, a distributor of herbal remedies in Taipei, Taiwan (Figure 1A). Genomic analysis was conducted by the Bioproduction Engineering Technology Department, Biomedical Technology and Device Research Laboratories, ITR (Hsinchu, Taiwan). The plant material used in this study was identified as Danshen based on 99.69% similarity of its internal transcribed spacer sequence to the reference sequence in the National Center for Biotechnology Information genome database (Figure 1B). Five grams of ground Danshen root material were loaded into a 66-mL extraction cell of a Dionex ASE 350 auto-extractor (Thermo Fisher Scientific, USA). Plant material was then extracted using 95% ethanol at 50°C for 15 min. Solvent was removed from the collected filtrate under vacuum to yield the ethanolic extract of Danshen. The yield was 4.2%. Identification of Salvia miltiorrhiza and its bioactive compound (A and B) Internal transcribed spacer sequence of purchased S. miltiorrhiza exhibited 99.69% similarity to the reference sequence (ID: KR082761.1) in the National Center for Biotechnology Information nucleotide database (https://www.ncbi.nlm.NationalInstitutesofHealth.gov/nucleotide), confirming the identity of the plant material. (C) Ethanolic extract of S. miltiorrhiza was prepared using an auto-extraction device and chromatographically analyzed using an RP-PDA-HPLC system. A peak corresponding to tanshinone IIA was identified by comparison with a reference standard. High-Performance Liquid Chromatography Analysis High-performance liquid chromatography (HPLC) was performed using a Nexera-I LC-2040C-3D HPLC system (Shimadzu, Kyoto, Japan) equipped with a UniverSilHS C18 HPLC column (250 × 4.6 mm; Fortis Technologies, Cheshire, UK). The ethanolic extract (10 mg/mL) of Danshen and tanshinone IIA reference standard (0.5 mg/mL) were dissolved in methanol and filtered through a 0.22-μm filter before analysis. A methanol gradient elution was employed as follows: 80%, 0−10 min; 80%–100%, 10–20 min; 100%, 20–25 min. The flow rate was set at 1.2 mL/min, and the detection wavelength and injection volume were 270 nm and 20 μL, respectively. Treatment with DCE Dimethyl sulfoxide (DMSO, Sigma-Aldrich, St. Louis, MO, USA) was used as the solvent to prepare stock solutions of the indicated drugs at a concentration of 0.01% Stock solutions were stored at 4°C. For cell culture experiments, a series of dilutions were prepared from the stock solutions to achieve the desired final concentrations of DCE (0, 10, 20, 40, 50, 60, 80, and 100 μg/mL). U-87 MG and DBTRG-05MG cell lines were obtained from the American Type Culture Collection (Manassas, VA, USA). DCE solutions were added to cultured U87-MG and DBTRG-05MG cells upon attachment to evaluate the effects of DCE treatment. 3-(4, 5-Dimethyl-2-Thiazolyl)-2, 5-Diphenyl-2H-Tetrazolium Bromide Assay Cell viability was assessed using an MTT assay. Briefly, U87-MG and DBTRG-05MG cells were seeded in 96-well plates at a density of 1×104 cells/100 μL cells in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% fetal bovine serum in 96-well culture plates. Incubated cells in a humidified chamber with 5% CO2 at 37 °C. The following groups were included: background group (blank), control group, and the experimental group in which increasing concentrations of DCE (0, 10, 20, 40, 50, 60, 80, and 100 μg/mL) were used for treatment. After incubation for 24, 48, and 72 h, MTT (5 mg/mL; Sigma-Aldrich) was added to each well and incubated until formazan formation and then replaced the medium with fresh medium to 10% of the culture medium volume. Incubated the plates in incubator for 4 hours at 37°C. Removed the MTT containing medium and added DMSO. The plates were shaken for 10–15 minutes. The absorbance was measured at 570 nm. The plates were then inverted and dried, after which 0.1 mL of DMSO was added to dissolve the formazan. The absorbance or optical density of the formazan solution was measured at a wavelength of 595 nm by using a multiplate reader. The rate of cell growth inhibition was calculated using the following formula: Cell Apoptosis Assay Cell apoptosis was assessed using an Annexin V-fluorescein isothiocyanate (FITC)/propidium iodide (PI) kit (Bd Pharmingen; Bd Biosciences, CA, USA), and the rate of apoptosis was analyzed using a flow cytometer (FAcScalibur; BD Biosciences; Becton, Dickinson and Company). U87-MG and DBTRG-05MG were seeded in 6-well plates at a density of 1×106 cells/well and treated with various concentrations of DCE (0, 0.625, 1.25, and 2.5 μg/mL) for 24 h. The cells were then resuspended in a 5-mL culture tube containing 1 × binding buffer (provided with the Annexin V-FITC/PI kit) at a density of 1×106 cells/mL and stained with Annexin V-FITC and counterstained using PI. After incubation at room temperature for 15 min, the samples were analyzed using a flow cytometer (FACSCalibur; BD Biosciences) to quantify the apoptotic cell population. Cell Cycle Assay The cell cycle distribution of U87-MG and DBTRG-05MG cells was determined using flow cytometry and PI staining. The cells were seeded in 6-well plates at a density of 1×106 cells/well and treated with various concentrations of DCE (0, 0.625, 1.25, and 2.5 μg/mL) for 24 h. After treatment, the cells were fixed in cold 70% ethanol for 2 h and then stained with a solution containing RNase A (60 μg/mL) and PI (50 μg/mL) in phosphate-buffered saline and incubated for 30 min in the dark at room temperature. The cell cycle distribution was analyzed using flow cytometry (FAcScalibur; Bd Biosciences). A total of 10,000 events were acquired for each cell sample. The percentage of cells in the subG0/G1, G2/M, and S phases was quantified using Modfit LT version 2.0 software (Verity Software House). Three independent experiments were performed. Animal Experiments and Xenograft Model All animal procedures were performed in accordance with institutional guidelines for animal care and were approved by the Institutional Animal Care and Use Committee (IACUC). BALB/c-nude mice (Strain NO. D000521) aged 6–8 weeks were purchased from GemPharmatech for xenograft studies. Prior to surgery, all instruments and the surgical field were sterilized with 70% ethanol. The surgical site was disinfected sequentially with povidone–iodine and 70% ethanol. Mice were anesthetized with inhalational isoflurane (induction at 3–4%, maintenance at 1–2%) and secured in a stereotaxic apparatus following hair removal. Body temperature and respiratory rate were continuously monitored throughout the procedure. For tumor xenograft establishment, 5×105 U87 glioblastoma cells suspended in 5 μL were injected subcutaneously into mice. Three days after implantation, tumor-bearing mice received DCE (0.67 mg/kg) five times per week until sacrifice. The in vivo dosage of DCE was determined based on the IC50 values obtained from cell viability assays and effective concentrations identified in prior in vitro experiments. At the experimental endpoint (4 weeks post-implantation), tumors were excised and measured for volume and weight. Excised tumor tissues were weighed, fixed in 4% paraformaldehyde, embedded in paraffin, and processed for immunohistochemical (IHC) staining. Following surgery, mice were monitored for 10 minutes for vital signs, including body temperature, respiration, mobility, and mucosal color. During the first 3 postoperative days, wound healing and signs of discharge or infection were assessed at least once daily. Any postoperative complications were promptly reported to a veterinarian. Pain assessment was conducted according to the institutional pain-scoring guidelines. Analgesia was administered preoperatively or prior to experimental procedures. Ketoprofen (1 mg/kg, subcutaneous injection) was used as the primary analgesic. For moderate pain (score 5–14), ketoprofen was administered immediately and repeated as necessary; combination analgesia with xylocaine or buprenorphine was provided when indicated. Severe pain (score 15–20) warranted humane euthanasia. Humane endpoints included any of the following conditions: body weight loss of 20–25%, anorexia for 24 hours, food intake <50% of normal for 3 consecutive days, inability to eat or drink, body temperature <37°C, central nervous system depression, tremors, or limb paralysis. Mice meeting any humane endpoint criteria were euthanized by carbon dioxide inhalation, followed by tissue collection and incineration. All animal experiments were conducted in accordance with National Institutes of Health guidelines and were approved by the Institutional Animal Care and Use Committee of Chang Gung Memorial Hospital (protocol number: 2021042201). The study protocol was approved by the Institutional Ethics Committee of Chang Gung Memorial Hospital. Animals were housed under pathogen-free conditions at the Animal Core Facility of Chang Gung Memorial Hospital. RNA Isolation and Transcriptome Sequencing Total RNA was extracted from samples by using TRIzol (Invitrogen) and chloroform and then precipitated with isopropanol in accordance with the manufacturer’s instructions. RNA concentration and quality were assessed. Additionally, RNA integrity was assessed using the RNA integrity number determined using the Agilent 2100 Bioanalyzer (Agilent, CA, USA) and RNA 6000 Nano Kit in accordance with the manufacturer’s instructions. Barcoded mRNA libraries were generated using the Illumina TruSeq RNA Library Prep kit (Illumina, San Diego, CA, USA) in accordance with the manufacturer’s protocol. The libraries were sequenced on an Illumina Nova Seq 6000 platform (Illumina) in accordance with the manufacturer’s instructions. The sequencing reads were mapped to the human genome reference (GRCh38 ensembl version 84) by using a STAR RNAseq read aligner.26 Raw read counts were generated using RSEM27 and subsequently normalized by counts per million (CPM). Batch effects were corrected using the ComBat function from the sva package in R. Differentially expressed gene (DEG) analysis was performed by comparing control animals with DCE-treated U87 xenograft mice (n = 3 per group). DEGs were identified using the edgeR package in R, with thresholds set at a fold change (FC) > 1.5 and a false discovery rate (FDR) < 0.05 based on the corrected counts. Pathway enrichment analysis of DEGs was performed using Gene Ontology and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway tools.28 Statistical Analysis Data are expressed as means ± standard errors (SEs). Student’s t tests and analyses of variance were conducted using Prism, GraphPad Software (version 9.0.2, San Diego, CA, USA). A P value of <0.05 was considered statistically significant. Results Identification of Potential Chemical Compositions Through RP-PDA-HPLC Analysis To identify potential bioactive compounds, reverse-phase photodiode array high-performance liquid chromatography (RP-PDA-HPLC) was performed. Chemical constituents in the DCE were identified by comparison with a reference standard, confirming the presence of tanshinone IIA (Figure 1C). Following this initial confirmation, the DCE was subjected to preparative flash column chromatography, yielding multiple fractions. These fractions were further analyzed by liquid chromatography–mass spectrometry (LC-MS) and RP-PDA-HPLC to identify those enriched in potential active compounds. Enriched fractions were subsequently purified by additional chromatographic methods until pure compounds were obtained. The chemical structures of the isolated compounds were elucidated by comparing their ultraviolet, infrared, and nuclear magnetic resonance spectra with reference data from scientific databases, enabling definitive structural identification. DCE Treatment Significantly Inhibits the Growth of GB Cells To evaluate the anticancer activity of DCE on GB cell viability, a 24-h cytotoxicity assay was conducted using human GB cell lines U87-MG and DBTRG-05MG because IC50 concentration was significant after treatment with DCE after 24 h. As depicted in Figure 2 and Table 1, DCE significantly inhibited the growth of both cell lines (at a concentration of 10 g/mL), with IC50 values of 11.75 ± 1.02 and 91.82 ± 1.12 μg/mL, respectively. Additionally, our results showed the IC50 values of tanshinone IIA, one of the identified active compounds in DCE, were 55.50 μM and 53.13 μM for U87-MG and DBTRG-05mg, respectively. These results demonstrate the in vitro antiproliferative effect of DCE on GB cells. TMZ was used as positive control and according to the literature, the IC50 ranged from 100 to 200 μM for U87-MG and 90 μM for DBTRG-05mg.29,30 Inhibitory effect of DCE on GB cells. U87-MG cells (A) and DBTRG-05MG cells (B) were treated with various concentrations of DCE for 24 h. U87 or DBRTG-05MG cells were seeded into 96-well plates and incubated in 10% fetal bovine serum containing either different concentrations of DCE or 0.01% DMSO (solvent control). After the 24-h incubation period, MTT reagent was added, followed by incubation for an additional 1 h. Subsequently, absorbance was measured using a microplate reader. Data are presented as the mean ± SD of three independent experiments (n=3). *P < 0.05, ***P < 0.001, compared with the control group. Table 1Inhibitory Effects of DCE Against Human GBM CellsU87-MGDBTRG-05MGIC50 (μg/mL)aIC50 (μg/mL)aDCE11.75 ± 1.0291.82 ± 1.12Note: Results are presented as mean ± SE (n = 3). aConcentration necessary for 50% inhibition (IC50). Inhibitory Effects of DCE Against Human GBM Cells Note: Results are presented as mean ± SE (n = 3). aConcentration necessary for 50% inhibition (IC50). DCE Treatment Significantly Inhibits the Migration of GB Cells The effect of DCE on the migratory ability of U87-MG and DBTRG-05MG GB cells was evaluated using a wound healing assay. Cell migration was quantified by measuring the distance traversed by the monolayer in both DCE-treated and vehicle (DMSO) control groups, expressed as migration index (gap density %). As shown in Figure 3, DCE treatment significantly inhibited cell migration in U87-MG (P < 0.001) and DBTRG-05MG (P < 0.001) cells at concentrations of 20 and 80 μg/mL, respectively, following 24- and 48-hour incubations. These results demonstrate that DCE effectively suppresses the migratory capacity of GB cells in vitro. Effect of DCE on U87-MG and DBTRG-05MG cell migration. The migration ability of U87-MG and DBTRG-05MG GB cells was evaluated using a wound healing assay. Representative bright-field images of U87-MG (A) and DBTRG-05MG GB cells (C) were treated with either 20 μg/mL or 80 μg/mL DCE. (B–D) Cell migration is quantified as a percentage, indicating the remaining wound gap size 24 or 48 h after scratch, compared with the control group. Data are presented as the mean ± SE of three independent experiments (n=3). ***P < 0.001, ###P < 0.01 vs. respective DMSO condition, compared with the control group. DCE Treatment Induced G0/G1 Phase Arrest in GB Cells Dysregulation of the cell cycle and apoptotic processes is a hallmark of GB pathogenesis. Therefore, therapeutic interventions aimed at inducing cell cycle arrest and apoptosis in GB cells hold considerable promise.31 To investigate the effect of DCE on cell cycle distribution, U87-MG and DBTRG-05MG cells were treated with DCE, and then cell cycle analysis was performed. As depicted in Figure 4, compared with the control group, DCE treatment induced G0/G1 phase arrest in U87-MG (P < 0.001) and DBTRG-05MG (P < 0.001) cells, suggesting its potential to inhibit tumor cell proliferation. DCE-induced G0/G1 cell cycle arrest in U87-MG and DBTRG-05MG GB cells. Cell cycle distribution as determined using PI staining and flow cytometry and the percentages of cells in the G0/G1, S, and G2/M phases displayed as histograms of U87-MG (A and B) and DBTRG-05MG cells (C and D) after treated with or without 20 and 80 μg/mL DCE, respectively, for 24 or 48 h. Data are presented as the mean ± SE of three independent experiments (n=3). *P < 0.05; ***P < 0.001, compared with the control. DCE Treatment Promotes Early Apoptosis and Cell Death Evasion of apoptosis is a key hallmark of GB development.31 To investigate whether DCE induces apoptosis in GB cells, Annexin V-FITC/PI dual staining was performed using the TACS Apoptosis Detection Kit. Based on fluorescence signals, cells were categorized into early apoptotic (Annexin V+/PI−), late apoptotic (Annexin V+/PI+), and dead (Annexin V+/PI+) populations. As shown in Figure 5, DCE treatment (20 μg/mL) markedly increased early and late apoptotic populations in U87-MG cells after 24 h (21.9 ± 13.2%), while a modest increase in overall cell death was observed at 48 h (5.4 ± 4.9%). A similar trend was observed in DBTRG-05MG cells (Figure 6), where 80 μg/mL DCE induced apoptosis at 24 h (5.8 ± 0.3%) and significantly elevated cell death at 48 h (15.0 ± 1.0%). DCE-induced apoptosis in U87-MG cells. U87-MG cells were treated with (A) 0.1% DMSO (control) and (B and C) DCE (20 g/mL). Data are presented as mean ± SE of three independent experiments (n=3) (D and E). Marginally significant differences were observed compared with the control group. DCE-induced apoptosis and death in DBTRG-05MG cells. DBTRG-05MG cells were treated with (A) 0.1% DMSO (control) and (B and C) DCE (80 g/mL). Data are presented as the mean ± SE of three independent experiments (n=3) (D and E). ***P < 0.001), compared with the control. DCE Treatment Inhibits Tumor Growth in a U87-MG Xenograft Mouse Model The antitumor effects of DCE in vivo were evaluated in a U87-MG xenograft mouse model. Once the tumors reached a measurable size, the mice were treated with a vehicle control, DCE (0.67 mg/kg, 5 times per week for 8 days), or TMZ (12.5 mg/kg, 5 times per week for 8 days). Compared to the vehicle control, DCE suppressed tumor growth more effectively (P < 0.0001; Figure 7A–C). Notably, the antitumor ability of DCE was comparable to that of TMZ. Additionally, treatment with DCE had no effect on body weight (Figure 7B), suggesting minimal toxicity. Antitumor effect of DCE in GB xenograft model. BALB/c-nude mice bearing U87-MG xenografts were treated with DCE,“tumeric extract” or TMZ (positive control). The tumor growth curves illustrate that DCE significantly inhibited tumor growth compared with the control group (A–C). Additionally, no change in body weight was observed after DCE or TMZ treatment (B). Tumor size and body weight were also measured. Data are expressed as mean ± SE. ***P < 0.001, compared with the control group. RNA-Sequencing Transcriptome Analysis of DCE-Regulated Gene Expression in U87-MG Xenografted Tumors in BALB/c-Nude Mice To assess gene expression alterations induced by DCE, total RNA was extracted from tumor tissues of both control and DCE-treated groups and subjected to RNA sequencing. Differentially expressed genes (DEGs) were identified using thresholds of adjusted P < 0.05 and |log2(fold change)| > 0.5. Although no genes reached statistical significance for differential expression, a downward trend in the expression of inflammatory mediators—including CXCL2, CXCL3, CXCL8, and PTGS2—was observed in the DCE-treated group, suggesting potential modulation of inflammatory pathways in the U87-MG xenograft model. Kyoto Encyclopedia of Genes and Genome Pathway Enrichment Analyses and Validation of DCE-Induced Changes in Gene Expression by qRT-PCR To elucidate the molecular mechanisms underlying the antitumor activity of DCE, KEGG pathway enrichment analysis was conducted based on the identified DEGs. Gene set enrichment analysis (GSEA) indicated potential involvement of the IL-17, NF-κB, and TNF signaling pathways (data not shown). To validate these transcriptomic findings, the mRNA expression levels of representative downstream targets were quantified by real-time PCR. Consistent with the RNA-seq trends, DCE treatment significantly downregulated the expression of CXCL2, CXCL3, CXCL8, and PTGS2 in tumor tissues from the U87-MG xenograft model (Figure 8). qRT-PCR analysis of DCE-treated GB tumor tissues from xenograft mice. U87-injected BALB/c-nude mice were treated with DCE. The mRNA levels of CXCL2, CXCL3, CXCL8, and PTGS2 were determined using qRT-PCR and normalized to -actin mRNA. Bars in the figure present the means ± SEs. *P < 0.05, ***P < 0.0001, compared with control. Discussion Clinical Application of Danshen Crude Extracts Our experimental findings demonstrated that both Curcuma longa and Salvia miltiorrhiza (Danshen) crude extracts exhibited anti- GB activity, with DCE exerting more pronounced inhibitory effects (Supplementary Figure 1). Consistently, previous studies have reported the anti- GB properties of Danshen-derived compounds and proposed potential mechanisms of action.32–34 In particular, a recent systems bioinformatics analysis identified STAT3, Akt, and IκB signaling pathways as potential therapeutic targets mediating Danshen’s effects in GB.35 While these findings offer important insights, further in vivo studies are necessary to clarify the exact molecular mechanisms. Notably, in clinical practice, crude herbal extracts are more commonly prescribed than isolated pure compounds. Therefore, we selected DCE for further investigation of its anti-cancer efficacy in this study. Effect of the Extraction Method on the Ability of DCE to Suppress Tumor Growth and Migration The biological activity of a compound may be influenced by the extraction method employed. Traditional Chinese medicinal herbs are particularly valued for their ability to exert therapeutic effects with minimal cytotoxicity to normal cells. In this study, DCE demonstrated the capacity to induce apoptosis in tumor cells while preserving the viability of non-tumor cells, highlighting its potential as a safer therapeutic agent. The extraction approach used facilitated the isolation of bioactive constituents with comparable or enhanced efficacy relative to individual pure compounds. As shown in Figure 1, DCE contains at least two known active compounds, which may act synergistically to exert a stronger anti-cancer effect than each compound alone. This synergism may translate into reduced clinical side effects, improved patient compliance, and enhanced quality of life. Furthermore, in vivo experiments showed that DCE significantly reduced tumor volume, with antitumor efficacy comparable to that of TMZ. This highlights the therapeutic potential of multi-compound herbal formulations in the treatment of GB, particularly in settings where conventional therapies cause substantial adverse effects. Further mechanistic studies and clinical evaluations are warranted to confirm the safety and efficacy of DCE as a complementary or alternative treatment option. DCE Effectively Inhibits Proliferation, Migration, Promotes Apoptosis, Cell Death and Alters Cell Cycle Our results demonstrated that DCE exerted potent anticancer effects against glioma cell lines (U87-MG and DBTRG-05MG), including inhibition of proliferation, migration, and invasion, as well as induction of apoptosis. Among the major bioactive constituents identified in DCE, tanshinone IIA—detected as a prominent peak in the extract—exhibited significant antiproliferative effects, with an IC50 value of 55.50 μM for U87-MG and and 53.13 μM for DBTRG-05MG. Consistent with our findings, DCE has previously been reported to induce cell cycle arrest in the SHG-44 glioma cell line.36 Similarly, tanshinone IIA has been shown to promote G0/G1 cell cycle arrest and apoptosis through STAT3 pathway modulation.37,38 In the present study, DCE-treated cells exhibited G1 phase arrest, indicating disruption of oncogenic cell cycle progression. Analogous to other TCMs such as curcumin.6,11 DCE appears to exert its therapeutic effects by targeting key signaling pathways involved in cell cycle regulation, thereby halting cancer cell proliferation at specific checkpoints. Furthermore, DCE effectively induced early apoptosis and cell death in glioma cells.21 Given its composition as a multi-component extract—containing active compounds such as tanshinone IIA—DCE may possess broader and potentially synergistic anticancer activity compared to single agents. This broad-spectrum efficacy highlights DCE’s potential as a multifaceted therapeutic agent capable of concurrently targeting multiple hallmarks of GB. qRT-PCR Revealed Downregulation of CXCL2, CXCL3, CXCL8, and PTGS2 After DCE Treatment As shown in Figure 8, DCE treatment markedly downregulated the expression of CXCL2, CXCL3, CXCL8, and PTGS2 in tumor tissues derived from the U87-MG xenograft model. CXCL chemokines are well-documented contributors to GB progression, regulating key processes such as tumor cell proliferation, migration, and survival.39 Notably, in vivo studies have shown that prolonged temozolomide (TMZ) treatment can lead to resistance in astroglioma cells, a phenomenon accompanied by increased expression of CXCL2, CXCL3, and CXCL8.40 Among these, CXCL2 has been strongly correlated with enhanced tumorigenicity and poor prognosis in GB.41,42 In addition, CXCL1 and CXCL2 have been implicated in immune evasion by promoting myeloid cell recruitment while simultaneously impairing CD8⁺ T-cell infiltration into the tumor microenvironment, thereby accelerating disease progression.41 Consequently, targeting CXCL2 may represent a promising immunomodulatory strategy for GB therapy.41,43,44 CXCL8 has also been associated with glioma tumorigenesis, progression, and recurrence. Elevated CXCL8 expression contributes to neovascularization in GB and is linked to poor clinical outcomes.45 Ahn et al reported that necrotic tumor cell can stimulate NF-B and activator protein-1 production, leading to increased CXCL8 production.46 Additionally, IL-8 promotes the growth and metastasis of GB by stimulating CXCR1 and CXCR2.47 Therefore, targeting CXCL8 and its downstream effectors may result in favorable GB treatment outcomes through various pathways.45,48 Prostaglandin-endoperoxide synthase 2 (PTGS2), commonly overexpressed in various cancers including breast49 and liver50 malignancies, is also upregulated in gliomas. Its expression is associated with increased cell proliferation and resistance to radiotherapy, possibly through NF-κB–dependent signaling mechanisms.51–53 Our data revealed that DCE treatment significantly reduced the expression of CXCL2, CXCL3, CXCL8 and PTGS2 associated with proangiogenic signaling pathways. Taken together, the suppression of CXCL2, CXCL3, CXCL8, and PTGS2 by DCE suggests that its antitumor effects may be mediated through modulation of inflammatory and angiogenic signaling networks. Possible Downregulated Signaling Pathways Following DCE Treatment As mentioned above, RNA sequencing analysis did not identify statistically significant downregulation of specific signaling pathways following DCE treatment. However, several genes involved in pro-inflammatory signaling and tumor progression—CXCL2, CXCL3, CXCL8, and PTGS2—exhibited consistent downregulation trends. Although these changes did not reach statistical significance at the transcriptomic level, their biological relevance prompted further validation. Subsequent qRT-PCR analysis confirmed significant downregulation of these genes in tumor tissues from DCE-treated U87-MG xenograft mice compared to controls. The IL-17 signaling pathway has been implicated in the pathogenesis of several cancers.54,55 Although IL-17 can exert both tumor-promoting and tumor-suppressive effects depending on context,56,57 in gliomas it is generally associated with tumor progression through inflammation and angiogenesis.58 Moreover, inhibition of PI3K, Akt1, and NF-κB-p65 has been shown to suppress IL-17-mediated glioma cell proliferation and migration.59 The NF-κB signaling pathway has been widely implicated in the development and progression of multiple cancers, including lung,60,61 breast,62 colorectal,63 liver,64 pancreatic,65 and prostate66 cancers. In glioma specifically, NF-κB activation contributes to tumor invasion, angiogenesis, and therapy resistance.67,68 Inhibition of NF-κB signaling has been shown to suppress tumor proliferation and progression.69,70 In addition, TNF signaling pathway is known to regulate various biological processes, including cellular proliferation, inflammation, vascular invasion, and destruction of tumor vasculature.71 Inhibition of this pathway has been demonstrated to effectively suppress tumor growth and metastasis in some cancer types,60,63,72–74 including GB.75,76 In our study, DCE treatment downregulated IL-17, NF-κB and TNF signaling pathways and the inhibition of these interconnected networks may underlie DCE’s capacity to suppress tumor proliferation, migration, and angiogenesis in glioblastoma. Limitations The limitations of our study should be acknowledged. First, it remains challenging to identify and validate the specific bioactive compounds responsible for the therapeutic effects of DCE and other TCMs, owing to their complex and heterogeneous compositions. This complexity poses substantial difficulties for mechanistic investigations and standardization in scientific research. Second, the in vivo experiments were limited by a relatively small sample size and the absence of comprehensive toxicity and pharmacokinetic evaluations. Therefore, the long-term safety profile and optimal therapeutic dosage of DCE remain to be fully elucidated. In addition, the translational applicability of these findings to clinical settings requires further validation. Furthermore, experiments focusing on the potential ability of the crude extract or its active components to cross blood-brain-barrier (BBB) warrants further investigation. Recently, Salvia miltiorrhiza-derived nanovesicles and brain delivery system were demonstrated efficient BBB penetration in traumatic brain injury77 and ischemic stroke,78 indicating its potential therapeutic effect on brain tumors. Future studies involving larger animal cohorts, detailed toxicological assessments, pharmacokinetic, pharmacodynamic studies, and well-designed mechanistic investigations are warranted to better clarify the therapeutic potential and safety of TCM-derived compounds. Conclusion DCE exhibited potent antitumor activity against glioblastoma (GB) in both in vitro and in vivo models. Its efficacy was associated with the downregulation of CXCL2, CXCL3, CXCL8, and PTGS2, leading to the suppression of tumor cell proliferation and migration, as well as the induction of apoptosis. These findings highlight the therapeutic potential of DCE as a novel multi-targeted agent for GB treatment and underscore the value of traditional Chinese medicine as a source of anticancer compounds. Further investigations are warranted to elucidate the precise molecular mechanisms involved and to evaluate the clinical applicability of DCE in patients with GB. Acknowledgments We acknowledged Chang Gung Memorial Hospital, Chang Gung University and Chang Gung University of Science and Technology for the support. This manuscript was edited by Wallace Academic Editing. Funding Statement This study was funded by Chang Gung Memorial Hospital (CMRPG1N0121, CORPG2P0121–122, and NMRPG2P0051). Abbreviations DCE, Danshen crude extract; DEGs, differentially expressed genes; DMSO, dimethyl sulfoxide; FITC, fluorescein isothiocyanate; GB, glioblastoma; HPLC, high-performance liquid chromatography; KEGG, Kyoto Encyclopedia of Genes and Genomes; MTT, 3-(4, 5-dimethyl-2-thiazolyl)-2, 5-diphenyl-2H-tetrazolium bromide; NF-κB, NF-kappaB; PI, propidium iodide; PTGS2, prostaglandin-endoperoxide synthase 2; TMZ, temozolomide; qRT-PCR, quantitative reverse transcription polymerase; RP-PDA-HPLC, reverse-phase photodiode array HPLC; RSEM, RNA-Seq by Expectation Maximization; SE, standard error. Data Sharing Statement The datasets during and/or analyzed during the current study available from Yin-Cheng Huang (one of the corresponding authors) on reasonable request. Ethics Approval and Consent to Participate All animal experiments were approved by the Institutional Animal Care and Use Committee (2021042201) at Chang Gung Memorial Hospital and were performed in accordance with National Institutes of Health and institutional guidelines. Disclosure The authors have no competing interests to declare for this work. References Associated Data Data Availability Statement The datasets during and/or analyzed during the current study available from Yin-Cheng Huang (one of the corresponding authors) on reasonable request.