The human GBM cell lines A172, U251, U87, used in this study, were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai China). All cell lines were cultured in Dulbecco’s modified eagle medium (DMEM, Gibco, United States, #8121300), supplemented with 10% fetal bovine serum (FBS, Gibco, United States, #10099-141) and 100U/mL penicillin/streptomycin (Gibco, United States, #15140-122) and were incubated in a humidified atmosphere with 5% (v/v) CO₂ at 37°C. Bru was purchased from MCE (Medchemexpress, United States, #HY-19543) and dissolved in DMSO.

This study was approved by the Clinical Research Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University (permission: 2015-063), and written informed consents were obtained from all patients. The study protocol was carried out according to the principles of the Helsinki Declaration. Ten GBM patients (Grade IV), who had no history of preoperative radiotherapy or chemotherapy and underwent surgery between 2019 and 2021 at the First Affiliated Hospital of Wenzhou Medical University, were included in this retrospective research to be analyzed for ECM1 expression. The samples were taken from nine males and one female, aged between 48 and 77 years (mean 66.7 years). The clinical characteristics of the patients are shown in Table 1.

For the isolation of human primary GBM cells, all steps were carried out under sterile conditions in a laminar flow cell culture hood to reduce the risk of contamination of the primary culture. Firstly, the tumor specimens were washed with phosphate buffered saline (PBS, Gibco, United States, #70011069) and dissected into small fragments. Then, the fragments were enzymatically digested by using human Tumor Dissociation kit (Miltenyi Biotec, Germany, #130-095-929) in combination with the gentleMACS™ Dissociator (Miltenyi Biotec, Germany, #DXT-130-096-730). Fibroblasts containing in the cell suspension were removed by using human Anti-Fibroblast MicroBeads (Miltenyi Biotec, Germany, #130-050-601) and erythrocytes were removed by using Red Blood Cell Lysis Solution (10×) (Miltenyi Biotec, Germany, #130-094-183). Finally, a total of 1 × 10⁶ cells were seeded in T25 cell culture flasks with a medium composed of DMEM medium (Gibco, United States, #8121300), 10% FBS (Gibco, United States, #10099-141) and 100U/mL penicillin/streptomycin (Gibco, United States, #15140-122) at 5% (v/v) CO2 humidified atmosphere with 37°C for further research.

Cell Counting Kit-8 (CCK-8; Dojindo, Japan, #CK04) was used to analyze cell viability. Three GBM cell lines (A172, U251 and U87) were plated in 96-well plates with a density of 10⁴ cells/well and incubated for 24 h as per the manufacturer’s instructions. At the indicated times after treatment, 10 μL of CCK-8 solution was added to 90 μL of culture medium. The cells were subsequently incubated for 3 h at 37°C in 5% (v/v) CO₂ and the optical density was measured at 450 nm using a plate reader (TECAN, Switzerland). In colony formation assay, the cells were plated into six-well plates with a density of 500 cells/well and cultured for 2 weeks. Then, different concentrations of Bru or DMSO alone as control were added to treat the cells for 24 h. Next, cells were stained with the Crystal Violet Staining Solution (Beyotime, China, #C0121) according to the manufacturer’s instructions. Furthermore, microscopic examination was performed using an Axiovert 200 microscope (Carl Zeiss, Oberkochen, Germany), and cell colonies were counted using PhotoShop CS6. Three independent experiments were performed.

Six-week-old female BALB/c (nu/nu) athymic nude mice were purchased from Shanghai Experimental Animal Center (Shanghai, China). U87 cells (1×10⁷) were resuspended in PBS with Matrigel (1:1; BD Biosciences, United States, #356234) and then subcutaneously injected into the right back of each nude mouse. All mice had free access to food and water. When the transplanted tumors reached an average size of 50 mm³, the mice were randomly assigned to two groups. Vehicle or Bru was administered (2 mg/kg) daily by oral gavage. Tumor volumes were measured with calipers at two perpendicular diameters and calculated individually using the formula (length×width²)/2. Twenty-five days later, all mice were sacrificed, and the tumors were harvested. All procedures were approved by the Administration Committee of Experimental Animals, Laboratory Animal Center, Wenzhou Medical University (Permit Number: wydw 2020-146).

To determine the effect of Bru on cell apoptosis, a FITC Annexin V apoptosis detection kit I (BD Biosciences, United States, #559763) was used according to manufacturer’s instructions. First, the cells were treated with Bru for 0–72 h and collected. Then, a total of ≥10,000 cells were resuspended in binding buffer and incubated with 5 µL of Annexin V-FITC and 5 µL of PI for 15 min in dark. The proportions of cells in early apoptosis and late apoptosis were reported as the percentage of Annexin V⁺/PI⁻- and Annexin V⁺/PI⁺-labeled cells, respectively. The stained cells were analyzed directly by flow cytometry using a FACSCalibur with the Cell Quest program (BD Biosciences, United States) for data analysis.

Cell lysates and tumor samples were extracted with cell lysis buffer (Beyotime, China, #P0013J), and the protein concentrations in the lysates were quantified using an Enhanced BCA Protein Assay Kit (Beyotime, China, #P0010). Protein samples (30–50 µg) were separated by SDS-PAGE and then transferred to PVDF membranes (Millipore, United States, #IPVH00010). The membranes were immunoblotted with primary antibodies followed by HRP-conjugated secondary antibodies. Immunoreactive proteins were visualized using an ECL Western blot detection kit (Advansta, United States, #K12045), and images were developed using a Bio-Rad system (Bio-Rad, United States). The indicated antibodies are listed as follow: rabbit anti-cleaved-Caspase-3 antibody (molecular weight: 17 kDa, 1:1,000; Cell Signaling Technology; #9664T), rabbit anti-pro-Caspase-3 antibody (molecular weight: 35 kDa, 1:1,000; Cell Signaling Technology; #14220T), rabbit anti-cleaved-Caspase-9 antibody (molecular weight: 37 kDa, 1:1,000; Cell Signaling Technology; #52873T), mouse anti-pro-Caspase-9 antibody (molecular weight: 47 kDa, 1:1,000; Cell Signaling Technology; #9508T), rabbit anti-Bax antibody (molecular weight: 21 kDa, 1:1,000; Affinity; #AF0120), rabbit anti-Bcl-2 antibody (molecular weight: 26 kDa, 1:1,000; Affinity; #AF6139), rabbit anti-ECM1 antibody (molecular weight: 61 kDa, 1:1,000; Abcam; #ab126629), rabbit anti-MMP1 antibody (molecular weight: 54 kDa, 1:1,000; Affinity; #AF0209), rabbit anti-TIMP1 antibody (molecular weight: 24 kDa, 1:1,000; Affinity; #AF7007), rabbit anti-MMP2 antibody (molecular weight: 72 kDa, 1:1,000; Abcam; #ab181286), rabbit anti-TIMP2 antibody (molecular weight: 24 kDa, 1:1,000; Affinity; #AF0264), rabbit anti-MMP9 antibody (molecular weight: 78 kDa, 1:1,000; Abcam; #ab76003), mouse anti-beta-Actin antibody (molecular weight: 42 kDa, 1:1,000; Affinity; #T0022), rabbit anti-Tubulin beta antibody (molecular weight: 55 kDa, 1:1,000; Affinity; #AF7011). Three independent experiments were performed.

In total, 3 µg of RNA per sample was used as input material for the preparations. Sequencing libraries were generated using the NEB Next® UltraTM RNA Library Prep Kit for Illumina® (NEB, United States, #E7530L) following the manufacturer’s recommendations, and index codes were added to attribute sequences to each sample. Clustering of the index-coded samples was performed on a cBot Cluster Generation System using a TruSeq PE Cluster Kit v3-cBot-HS (Illumina, United States, #PE-401-3001). After cluster generation, the library preparations were sequenced on an Illumina platform, and 125bp/150bp paired-end reads were generated.

Total RNA was extracted from GBM cells by using TRIzol™ reagent, according to the manufacturer’s instructions. The RNA concentrations were examined by a spectrophotometer (NanoDrop 1000, Thermo). The first-strand cDNAs were synthesized using a cDNA Synthesis SuperMix kit (YEASEN, China, #11123ES10). Each cDNA (2 μL) was amplified using the Hieff qPCR SYBR Green Master Mix (final volume, 20 ml, YEASEN, China, #11203ES03) and then analysed on the Applied Biosystems 7900 Real-time PCR Detection System. Thermal cycling conditions were performed as follows: melting step (95°C for 30 s), annealing step (40 cycles of 95°C for 10 s), and elongation (60°C for 30 s). Actin was used as internal control. Melting curves were used for verifying the specificity of each PCR reaction. The cycle threshold (Ct) was the mean value of three Ct values. The relative expression level of RNA was analyzed by the 2^−ΔΔCt method.

RNA primers are listed as follows:

ECM1 (human-forward): 5′-TGA​ACC​AAA​TCT​GCC​TTC​CTA​AC-3’

Three GBM cell lines (A172, U251 and U87) were transfected with small interfering (si)RNA oligonucleotides by riboFECT^TM CP Transfection kit (RIBOBIO, China, #C10511). Briefly, siRNA was first diluted with riboFECT^TM CP Buffer and then co-incubated with riboFECT^TM CP Reagent for 15min at room temperature. The mixture was then applied to cells plated in 2 ml of medium (final concentration of siRNA, 50 nM). All siRNAs were purchased from RIBOBIO.

The sequences of siRNAs were as follows:

siECM1: 5′-GCT​TCA​ACA​TCA​ATT​ATC​T-3'

Nucleotide sequence coding for ECM1 was cloned into pLV-CMV-ECM1-EF1-ZsGreen1-T2A-Puro (Youze Bio, China, #LV-0368) packaged by lentivirus. According to the manufacturer’s instructions, the three GBM cell lines (A172, U251 and U87) were infected with lentivirus expressing ECM1 (ECM1) or empty lentivirus (Control) for 96 h and then used separately for further studies. All lentiviruses were labeled with ZsGreen1.

Migration of 3 cell lines (A172, U251 and U87) was measured by determining the ability of cells to move into cell-free spaces. Cells were treated with siRNA or lentivirus as mentioned. When cells reached confluence, sterile plastic pipette tips were used to form identical linear wounds. At 0 and 48 h, the scratches were photographed and evaluated using an inverted phase contrast microscope. Migration distances were then calculated.

Matrigel solution was prepared by diluting Matrigel with serum-free cell culture medium at a dilution of 1:8. The 24-well Transwell chambers (Corning Costar, United States, #3422) were coated with Matrigel solution for 12 h at 37°C before cell seeding. Cells were cultured in the chamber with serum-free medium containing 1% bovine serum albumin in triplicate at 10⁵ cells per well. After incubation at 37°C for 48 h, the chambers were gently wiped with a cotton swab. Cells invading the polycarbonate membrane were fixed with 4% paraformaldehyde for 15 min and then washed with PBS twice for a total of 10 min. Staining was done with 0.1% crystal violet for 10min and then washed twice with PBS. The cell numbers on the lower surface of the membrane were counted under Axiovert 200 microscope (Carl Zeiss, Oberkochen, Germany).

Formalin-fixed tumor tissues were fixed in 4% paraformaldehyde for 24 h, embedded in paraffin, and then serially sectioned (4 μm). For hematoxylin and eosin (H&E) staining, sections were stained with H&E (Beyotime, China, #P0010). For IHC, sections were stained with primary antibodies against ECM1 (Abcam, United States, #ab126629). Briefly, the tissue sections were dehydrated and subjected to peroxidase blocking. Primary antibody was added and incubated at room temperature for 30 min on the Dako AutoStainer using the Dako Cytomation EnVision + System-HRP (DAB) detection kit (Carpinteria, United States, #K406511). Slides were counterstained with hematoxylin and then observed under a microscope.

All statistical analyses were performed using the SPSS 17.0 statistical software (SPSS Inc., Chicago, IL, United States). Continuous variables were presented as mean ± SE. Statistical significance between two measurements was determined by the two tailed unpaired Student’s t-test, and among groups, it was determined by one-way analysis of variance (ANOVA) followed by Tukey’s post-hoc multiple comparison test. p values <0.05 were considered as statistically significant.

Firstly, cell viability and colony formation assay were used to clarify the cytotoxic effect of Bru on GBM cells. The results of cell viability assay showed that cell viability of three GBM cell lines (A172, U251, U87) significantly decreased in both dose- (Figure 1A) and time-dependent (Figure 1B) manner in response to Bru treatment. Similar results were also obtained from the plate clone formation assays in three GBM cell lines, which were treated with different concentrations of Bru for 24 h, determined by the formation count (Figure 1C). The formation number of A172 cells colonies was 79.8 ± 0.86 in 0 nM group, 71.2 ± 0.86 in 25 nM group, 45.4 ± 0.51 in 50 nM group, and 34.4 ± 1.29 in 75 nM group. The formation number of U251 cell colonies was 94.8 ± 0.80 in 0 nM group, 81 ± 1.18 in 50 nM group, 54 ± 0.71 in 75 nM group, and 40.4 ± 0.75 in 100 nM group, whereas the formation number of U87 cell colonies was 81.2 ± 1.43 in 0 nM group, 59.8 ± 1.07 in 25 nM group, 40 ± 0.71 in 50 nM group, and 25.6 ± 0.93 in 75 nM group. To further evaluate whether Bru-induced cell death was also effective in vivo, we investigated tumor growth in U87 xenograft models (Figure 1D). The results showed that the average tumor volume of Bru-treated group was significantly lower than that of Control group after transplanting for 25 days (3.13 ± 1.87 mm³ vs. 14.53 ± 3.87 mm³, n = 5, p = 0.0293; Figure 1E). The average tumor weight was 1,320 ± 373.36mg and 292 ± 169.54 mg in Bru-treated group and Control group, respectively (p = 0.0365, Figure 1F).

To conclude, these results suggested that Bru possessed a potent inhibitory effect on GBM cells in vivo and in vitro.

Recently, mounting evidence has proved that Bru can inhibit the tumor growth via inducing cell apoptosis (Chen et al., 2018; Wang et al., 2018). In order to clarify whether this inhibitory effect of Bru on GBM cells was also related to the induction of apoptosis, we subjected Bru-treated cells to flow cytometry and western blot analyses. The results of Annexin V/PI apoptosis analysis demonstrated that Bru treatment increased the numbers of both early and late apoptotic cells in a time-dependent manner (Figure 2A). After being treated with Bru for 72 h, the percent of early apoptotic cells increased by 0.53% in A172 cells (p < 0.001), by 1.61% in U251 cells (p < 0.001), and by 5.92% in U87 cell (p < 0.001). Meanwhile, the percent of late apoptotic cells increased by 1.41% in A172 cells (p < 0.001), by 0.82% in U251 cells (p < 0.001), and by 4.03% in U87 cells (p < 0.001), compared with those in Control group (treated with DMSO alone). To further clarify the mechanism of Bru’s inhibitory effect on GBM cells, we investigated the expression of the main mitochondrion-dependent apoptotic proteins, including Caspase-3, Caspase-9, Bcl-2, and Bax in the three GBM cell lines by Western blotting analysis. As shown in Figure 2B, after Bru treatment, the levels of pro-Caspase-3, pro-Caspase-9, and Bcl-2 decreased, while the levels of cleaved-Caspase-3, cleaved-Caspase-9, and Bax significantly increased in a time-dependent manner.

To sum up, these results revealed that Bru could exert its inhibitory effect on GBM cells by inducing cellular apoptosis.

In order to further study the underlying mechanism of Bru’s action on GBM, we performed RNA sequencing array on GBM cells treated by Bru with IC₅₀ for 48 h. A log2 fold change >1.0 and a p < 0.05 were used as cutoff criteria. The volcano plots results showed that Bru significantly changed the mRNA expression pattern in U251 (819 upregulated genes and 954 downregulated genes) and U87 (485 upregulated genes and 266 downregulated genes) cells (Figure 3A). Then, we subjected these differentially expressed genes to the Reactome pathway database (https://reactome.org), which provides molecular details of signal transduction, transport, DNA replication, metabolism, and other cellular processes as an ordered network of molecular transformations in a single, consistent data model. Reactome enrichment analysis indicated that the collagen and extracellular matrix related pathways (Italic Script with *) were mainly involved in Bru-treated GBM cells (Figure 3B). Subsequently, we used the Venn diagram analysis (VDA) to display the shared genes between U251 and U87 cells (Figure 3C, listed in Supplementary Table S1) and finally determined that ECM1, an important factor in extracellular matrix formation, changed obviously by Bru treatment. In order to verify the accuracy and repetition of RNA sequencing array results, three GBM cell lines treated with Bru for 48 h were subjected to perform qRT-PCR analysis and western blot assay. The results showed that, compared with Control group, Bru treatment significantly reduced the mRNA expression (Figure 3D) and protein level (Figure 3E) of ECM1 in these three GBM cell lines.

The results above showed that the expression of ECM1 had a close relationship with Bru treatment.

To verify the effects of ECM1 on proliferation, migration, and invasion of GBM cells, ECM1 siRNA was transfected into the three GBM cell lines to down-regulate the expression of ECM1. Western blot analysis showed that the expression of ECM1 in siECM1 group was significantly lower than that in siControl group (Figure 4A). In cell viability assay, the three GBM cell lines showed significant reduction of cell viability after silencing ECM1 (Figure 4B). 48 h after siRNA transfection, the cell viability in siECM1 group was significantly reduced by 50.13 ± 4.47% in A172 cells (p < 0.001), 63.14 ± 6.20% in U251 cells (p < 0.001), and 39.45 ± 4.36% in U87 cells (p < 0.001), compared with siControl group. In order to explore whether ECM1 knockdown could inhibit the migration and invasion, as a novel cellular response to regulate cell proliferation, we knocked down the expression of ECM1 in GBM cell lines and subjected these cells to wound healing and transwell assays. The results showed that, compared with siControl group, the wound healing migration ability of GBM cells in siECM1 group was significantly reduced by 45.51 ± 6.03% in A172 cells (p < 0.001), 21.03 ± 2.55% in U251 cells (p < 0.01), and 37.35 ± 3.24% in U87 cells (p < 0.001) at 48 h (Figure 4C). In transwell assay, ECM1 silencing could significantly decrease cellular invasion ability of A172 cells by 92.75 ± 1.41% (p < 0.001), U251 cells by 90.68 ± 1.41% (p < 0.001), and U87 cells by 75.02 ± 1.09% (p < 0.001) at 48 h (Figure 4D). Moreover, we also found that ECM1 silencing was sufficient to increase the chemosensitivity of Bru by 31.12 ± 3.74% (p < 0.001), 30.10 ± 2.59% (p < 0.001), and 94.92 ± 4.50% (p < 0.001) at 48 h in A172, U251, and U87, respectively (Figure 4E).

These findings suggested that ECM1 played an important role in promoting the growth and invasion of GBM cells, and ECM1 silencing could sensitize GBM cells to Bru treatment.

In order to further investigate the protein expression profile of ECM1 on GBM cells, the three GBM cell lines (A172, U251, U87) were transfected with lentivirus to overexpress ECM1. After transfection for 96h, most of the cells showed positive signals of ZsGreen1, indicating that the lentiviruses infected GBM cells efficiently (Figure 5A). The structures of the plasmid overexpressing ECM1 (pLV-CMV-ECM1-EF1-ZsGreen1-T2A-Puro) and the Control plasmid (pLV-CMV-MCS-EF1-ZsGreen1-T2A-Puro) were shown in Figure 5B. Western blot analysis was also used to measure the transfection efficiency, and the results showed that the expression level of ECM1 was significantly up-regulated in ECM1 group (Figure 5C). In cell viability assay, the results showed that overexpressed ECM1 could distinctively promote cell proliferation, compared with Control group (Figure 5D). Cell viability of GBM cells in ECM1 group was significantly increased by 117.28 ± 6.40% in A172 cells (p < 0.001), 92.73 ± 8.70% in U251 cells (p < 0.001), and 42.53 ± 5.74% in U87 cells (p < 0.001) at 48 h, compared with Control group. In addition, the wound healing assay showed that overexpression of ECM1 enhanced the migration ability of GBM cells by 26.14 ± 5.63% in A172 cells (p < 0.01), 24.85 ± 3.24% in U251 cells (p < 0.01), and 19.38 ± 4.80% in U87 cells (p < 0.05) at 48 h (Figure 5E). As shown in Figure 5F, ECM1 overexpression could significantly increase cellular invasion ability of A172 cells by 308.73 ± 19.17% (p < 0.001), U251 cells by 47.53 ± 1.55% (p < 0.001), and U87 cells by 459.72 ± 32.66% (p < 0.001) at 48 h. Besides, we also found that up-regulating the expression of ECM1 was sufficient to reduce chemosensitivity of Bru by 49.78 ± 9.23% (p < 0.001), 37.78 ± 7.15% (p < 0.001), and 56.21 ± 9.43% (p < 0.001) at 48 h in A172, U251, and U87, respectively (Figure 5G).

All these results above revealed that ECM1 was required for the growth of GBM cells and overexpressed ECM1 could reverse the inhibitory effect of Bru treatment on GBM cells.

In order to further explore the molecular mechanism of Bru treatment, we performed western blot to examine the expression of migration-related biomarkers, including MMP1, MMP2, MMP9, TIMP1, and TIMP2. The results showed that knockdown of ECM1 significantly reduced the expression of MMP1, MMP2 and MMP9, while significantly increasing the expression of TIMP1 and TIMP2 (Figure 6A). On the contrary, overexpression of ECM1 significantly increased the protein levels of MMP1, MMP2 and MMP9, while reducing the expression of TIMP1 and TIMP2 (Figure 6B). The expression of ECM1 and other related genes in RNA sequencing array were also shown in Figure 3A (indicated as volcano plots). In addition, as shown in Figure 6C, the expression levels of ECM1, MMP1, MMP2 and MMP9 in xenograft tumors tissues from Bru-treated group significantly decreased, compared with the Control group, while the expression levels of TIMP1 and TIMP2 significantly increased (Figure 6C). Consistently, IHC staining also showed that the expression levels of ECM1 in xenograft tumors from Bru-treated mice decreased, compared with the levels in tumors from Control group (n = 5, Figure 6D and Supplementary Figure S1A).

These results demonstrated that Bru could exert its inhibitory effect by down-regulating the expression of ECM1 in vitro and in vivo.

In order to determine whether the expression of ECM1 was an important prognostic factor for Bru treatment on human primary GBM cells, we collected human GBM tissue samples to culture in vitro, and the cells were then treated with Bru for 12–24 h and subjected to cell viability assay. The results showed that Bru treatment (50 nM) resulted in a viability decrease in six out of 10 human GBM tissue samples (Figure 7A). The HE staining results were shown in Supplementary Figure S1B, and the expression of ECM1 protein was high in the majority of these Bru-treated sensitive human primary GBM cells (Figure 7B). Interestingly, based on the clinical characteristics of GBM patients, tumor cells collected from the MGMT-negative patients were also sensitive to Bru treatment at 48h, indicating that Bru successfully inhibited the proliferation of TMZ-resistant human primary GBM cells (case No. 4, 8, 10 in Table 1).

In all, these results suggested that Bru was also effective in the treatment of human primary GBM cells.