The 3D structure of the SBEM protein was predicted using AlphaFold software 2 (https://alphafold.ebi.ac.uk). PrankWeb (https://prankweb.cz/), a machine learning-based tool employing a Random Forest algorithm trained on known binding sites, was used to predict potential active sites (25). Through multi-sequence alignment analysis using BioXM 2.7.1, MUCL1 (UniProt ID: Q96DR8) and MUCL3 (UniProt ID: Q3MIW9) were examined to identify conserved regions. Subsequently, the 2D structure of the SBEM protein was analyzed using PyMOL 2.5 (https://pymol.sourceforge.net). Briefly, based on the protein structure confidence, 2D structure stability, active site prediction and sequence alignment, Met1 was selected as the lattice center and a docking box was established. The specific box information was as follows: center_x=6.439; center_y=−0.223; center_z=−12.452; size_x=73.15; size_y=73.15; and size_z=73.15. The SDF format of PTX was obtained from the PubChem database (https://pubchem.ncbi.nlm.nih.gov) and converted to the PDB format using Open Babel (http://openbabel.org/index.html). The protein structure was dehydrated, hydrogenated and its charge calculated before being converted to the PDBQT format using AutoDockTools 1.5.7 (https://autodocksuite.scripps.edu/adt/). The ligand was also hydrogenated, its torsional degrees of freedom were determined and it was converted to the PDBQT format. The docking box coordinates were defined and molecular docking was performed using AutoDock Vina (https://vina.scripps.edu/). Visualization and 3D analysis diagrams were generated using PyMOL 2.1.0.

Human breast cancer cell lines (SUM190PT, BT474, AU565, SKBR3 and MCF7) and the normal mammary epithelial cell line MCF10A were purchased from Qingqi (Shanghai) Biotechnology Development Co., Ltd. All cell lines were authenticated by short tandem repeat profiling to prevent cross-contamination. Cells were thawed from liquid nitrogen, resuspended and cultured in Dulbecco's Modified Eagle Medium (DMEM) (Beyotime Biotechnology; cat. no. C0891) supplemented with 10% fetal bovine serum (Beyotime Biotechnology; cat. no. C0235) and 1% penicillin-streptomycin. The cells were placed in a cell incubator (Beyotime Biotechnology; cat. no. E2392) at 37°C with 5% CO₂.

Exponentially growing SUM190PT and SKBR3 cells were treated with 2 ml trypsin digestion solution (Wuhan Servicebio Technology Co., Ltd.; cat. no. G4011) and incubated at 37°C in a 5% CO₂ atmosphere (Suzhou Jiemei Electronics Co., Ltd.; cat. no. CI-191C) for 3 min. Following incubation, 3 ml high-glucose DMEM (Wuhan Servicebio Technology Co., Ltd.; cat. no. G4524) was added and the cells were centrifuged at 1,200 × g for 3 min at 25°C. The supernatant was discarded and the pellet was resuspended in 3 ml DMEM. The cells were then seeded into 6-well plates at a density of 1×10⁵ cells/well and incubated overnight. The following day, small interfering RNA (siRNA) and overexpression vectors were mixed with Lipofectamine™ 3000 transfection reagent (Thermo Fisher Scientific, Inc.; cat. no. L3000001) and incubated at room temperature for 30 min. The transfection mixture was then added to the wells according to the experimental requirements and incubated at 37°C for 5 h. siSBEM was synthesized by GenScript Biotech Corporation, with a synthesis amount of 2.5 nmol diluted to 100 µmol/l. The overexpression vector backbone employed in this study was pcDNA3.1(+) (Thermo Fisher Scientific, Inc.; cat. no. V79020). For cell transfection, 4 µg of the pcDNA3.1(+) vector, at a concentration of 1 µg/µl, was introduced into the cells. The siRNA interference and overexpression primer sequences are shown in Tables SI and SII, respectively.

PTX-resistant breast cancer cells were developed based on previously described methods (26). Details of the drug-induction protocol are shown in Table SIII. To confirm the stability of the resistance phenotype, the PTX (MedChemExpress; cat. no. HY-B0015) concentration was gradually increased until a final concentration of 500 nmol/l was reached. The cells were then maintained at this concentration and subcultured for at least 10 generations. IC₅₀ values were measured at the 5th, 10th and 20th generations post-withdrawal of PTX. When the IC₅₀ values showed minimal variation and were significantly higher than those of the parental cell lines, the resistance model was deemed successfully established. The parental lines are denoted as SUM190PT and SKBR3, while their drug resistant counterparts are referred to as SUM190PT/PTX and SKBR3/PTX. Resistance success was confirmed by comparing cell viability and resistance coefficients before and after PTX exposure. Resistant cell lines were continuously cultured in 500 nmol/l PTX to maintain drug resistance.

To assess whether SBEM knockdown could reverse drug resistance in SUM190PT/PTX cells, the following groups were established: SUM190PT (control), SUM190PT + PTX, SUM190PT/PTX + PTX + si-negative control (NC) and SUM190PT/PTX + PTX + siSBEM. To evaluate whether SBEM overexpression induces PTX resistance, SKBR3 cells, which have relatively low SBEM expression, were divided into: SKBR3 (control), SKBR3/PTX + PTX, SKBR3 + PTX + Vector and SKBR3 + PTX + oeSBEM. AMPK activator 13 (MedChemExpress; cat. no. HY-155363), an AMPK signaling pathway activator (27), was used to investigate the role of SBEM in regulating AMPK signaling. For this, SUM190PT/PTX cells were divided into four groups: Control, AMPK activator 13, siSBEM and AMPK activator 13 + siSBEM. Both SUM190PT and SKBR3 cells were treated with PTX at a concentration of 500 nmol/l for 48 h. Drug-resistant cell lines SUM190PT/PTX and SKBR3/PTX were maintained in medium containing 500 nmol/l PTX. Cells transfected with siRNA or overexpression plasmids were incubated at 37°C for 48 h. In all groups involving AMPK activator 13, the concentration was set at 5 µmol/l with an incubation time at 37°C for 24 h (27).

Following transfection, breast cancer cells and drug-resistant cell lines in the exponential growth phase were harvested, digested and resuspended. Then, 3,000 cells per well were seeded into 96-well plates and incubated overnight. PTX was applied at varying concentrations (125, 250, 500, 1,000, 2,000, 4,000 and 8,000 nmol/l) and the cells were incubated for 48 h. Subsequently, 10 µl CCK-8 solution (Wuhan Servicebio Technology Co., Ltd.; cat. no. G4103) was added to each well and incubated for an additional 3 h. Optical density at 450 nm was measured using a microplate reader (Thermo Fisher Scientific, Inc.; cat. no. A51119500C). IC₅₀ values and PTX resistance coefficients were calculated using GraphPad Prism version 9.5.0 (Dotmatics).

Cell suspensions were collected and centrifuged at 1,200 × g for 3 min at 25°C. The supernatant was discarded and 1 ml RNA extraction reagent (Wuhan Servicebio Technology Co., Ltd.; cat. no. G3013) was added, followed by a 5-min incubation at room temperature. Next, 100 µl chloroform (Wuhan Servicebio Technology Co., Ltd.; cat. no. G3014) was added and the mixture was shaken vigorously for 15 sec, then allowed to stand for 5 min at room temperature. The sample was centrifuged at 12,000 × g for 15 min at 4°C, and the aqueous phase was transferred to a new centrifuge tube. Then, 500 µl isopropanol (Shanghai Macklin Biochemical Co., Ltd.; cat. no. 67-63-0) was added and incubated for 10 min at room temperature. After centrifugation at 12,000 × g for 15 min at 4°C, the supernatant was discarded. The RNA pellet was washed with 1 ml 75% ethanol (Shanghai Macklin Biochemical Co., Ltd.; cat. no. E809061), air-dried and resuspended in 50 µl sterile water (Wuhan Servicebio Technology Co., Ltd.; cat. no. G4700). The RNA concentration was quantified for use in subsequent experiments.

RT-qPCR was conducted according to the instructions of the SweScript One-Step RT-PCR Kit (Wuhan Servicebio Technology Co., Ltd.; cat. no. G3335). The PCR cycling program was as follows: 50°C for 30 min, 98°C for 2 min, followed by 39 cycles of 98°C for 15 sec, 55°C for 20 sec and 72°C for 20 sec. A final extension step was carried out at 72°C for 5 min. The primer sequence (5′-3′) information is as follows: SBEM forward, TCTGCCCAGAATCCGACAAC, and reverse GGGCTTCATCATCAGCAGGA; GAPDH forward, CTGGGCTACACTGAGCACC, and reverse AAGTGGTCGTTGAGGGCAATG. GAPDH was used as the internal reference gene and the 2^−∆∆Cq was used as the relative expression level of the target gene mRNA (28).

Breast cancer cell suspensions were collected and lysed using RIPA lysis buffer (Wuhan Servicebio Technology Co., Ltd.; cat. no. G2002). Lysates were incubated on ice for 30 min and then centrifuged at 12,000 × g for 15 min at 25°C. The supernatant was collected and proteins were denatured in a 95°C water bath for 15 min. Protein concentrations were determined using a BCA protein quantification kit (Wuhan Servicebio Technology Co., Ltd.; cat. no. G2026). Cell lysis and protein-loading buffers (Beyotime Institute of Biotechnology; cat. no. P0015) were added to normalize the protein concentrations across samples. Then, 10 µl (30 µg) of each protein solution was loaded into the wells of the gel (10%). Electrophoresis was performed at 80 V for 30 min, followed by 120 V for 60 min. After electrophoresis, the proteins were transferred to PVDF membranes at 260 mA for 2 h. PVDF membranes were blocked with 5% skim milk (Wuhan Servicebio Technology Co., Ltd.; cat. no. GC310001) at room temperature for 2 h, then washed three times with TBST (0.1% Tween 20) (Wuhan Servicebio Technology Co., Ltd.; cat. no. G0004) for 5 min each. Primary antibody solutions were added and incubated overnight at 4°C. After incubation, secondary antibody solutions were applied and incubated at room temperature for 1 h. Details of the antibodies, including manufacturer cat. no. and dilution ratios, are provided in Table SIV. After washing with TBST, chemiluminescent detection was performed using a luminescent substrate (Wuhan Servicebio Technology Co., Ltd.; cat. no. G2161) and images were captured using a gel imaging system (Mona Biotechnology Co., Ltd.; cat. no. GD50202). Images were analyzed with ImageJ 1.5.2a software (National Institutes of Health) for subsequent statistical analyses.

Exponentially growing breast cancer cells were treated as aforementioned in the Cell experimental grouping subsection, then centrifuged and resuspended as previously described. According to the instructions of the flow cytometry apoptosis detection kit (Wuhan Servicebio Technology Co., Ltd.; cat. no. G1511), 5 µl FITC solution was added to each cell suspension, followed by incubation in the dark at room temperature for 30 min. Subsequently, PI solution was added and incubated in the dark for an additional 5 min. Cells were analyzed using a Attune™ NxT flow cytometer (Thermo Fisher Scientific, Inc.; cat. no. A24858) and Attune™ NxT software (version 6; Thermo Fisher Scientific, Inc.) within 2 h of staining.

Exponentially growing breast cancer cells were treated as aforementioned in the Cell experimental grouping subsection, then centrifuged and resuspended as previously described. Cell lysis buffer was added according to the Co-IP kit protocol (Beyotime Institute of Biotechnology; cat. no. P2179M) and the samples were incubated at room temperature for 30 min. The lysates were then centrifuged at 1,200 × g for 3 min at 25°C and the supernatant was collected. Protein A/G magnetic beads pre-conjugated with the primary antibody were added (20 µl of magnetic bead suspension to every 500 µl of protein sample) and the mixtures were incubated overnight at room temperature. After incubation, the beads were washed with washing buffer and the immunocomplexes bound to the beads were eluted using the elution buffer. The eluted target proteins were analyzed via western blotting. Antibody details are provided in Table SIV.

Data were analyzed using SPSS version 22.0 (IBM Corp.). All experiments were independently repeated at least three times. Results are expressed as mean ± standard deviation. Comparisons between two groups were performed using unpaired t-tests. Differences among multiple groups were assessed using one-way analysis of variance followed by Tukey's post-hoc test. P<0.05 was considered to indicate a statistically significant difference.

The 2D structural formula of PTX is shown in Fig. 1A. The SBEM mRNA expression levels in various breast cancer cell and normal breast epithelial cell lines were assessed using RT-qPCR. Compared with MCF10A cells, SBEM mRNA expression was highest in SUM190PT cells and lowest in SKBR3 cells (Fig. 1B). Since SBEM was highly expressed in the SUM190PT cell line, a knockdown cell line using this cell line was initially planned. However, as the present study focuses on drug-resistant strains, a SBEM-knockdown cell line using the SUM190PT/PTX drug-resistant strain was ultimately constructed. Additionally, SKBR3 cells, which have low SBEM expression, were selected to create a SBEM-overexpressing cell line. Knockdown and overexpression experiments in cell lines representing the two extremes of SBEM expression were conducted to more clearly observe the impact of SBEM modulation on cellular behaviors, such as viability and PTX resistance. This maximized the differences in experimental outcomes and enhanced the persuasiveness of the findings.

The establishment of PTX-resistant cell lines was validated using the CCK-8 assay. After 48 h of PTX treatment, the SUM190PT and SKBR3 cell viabilities were significantly inhibited, with IC₅₀ values of 657.63±52.84 nmol/l and 732.92±137.03 nmol/l, respectively (Fig. 1C). By contrast, the PTX-resistant SUM190PT/PTX and SKBR3/PTX cells exhibited respective IC₅₀ values of 6073.67±384.43 nmol/l and 6901.00±374.46 nmol/l, with corresponding resistance coefficients of 8.95±1.33 and 9.04±1.46 (Fig. 1D). These data confirmed the successful generation of the PTX-resistant cell lines.

To further explore the association and potential mechanism of SBEM in PTX resistance, cell lines with upregulated or downregulated SBEM expression were established (Fig. 2A and B). Among the siRNAs tested, siSBEM-1 most effectively inhibited both SBEM mRNA and protein expression and was therefore selected for subsequent experiments. The effects of SBEM knockdown and overexpression on the proliferation of SUM190PT/PTX and SKBR3 cells were assessed using the CCK-8 assay. The results demonstrated that SBEM knockdown restored PTX sensitivity in SUM190PT/PTX cells, whereas SBEM overexpression reduced PTX sensitivity (Fig. 2C and D). These findings suggest that SBEM expression may be associated with PTX sensitivity in breast cancer cells.

While it was demonstrated that SBEM may be associated with PTX sensitivity, its precise regulatory mechanism remains unclear. The following analyses were conducted to screen the rational domains of the SBEM protein. The prediction results showed that the protein ipTM=-pTM=0.23, with a relatively low overall confidence level (Fig. S1A). In addition, amino acids 1–20 were predicted to form the signal peptide region (Fig. S1B). Next, PrankWeb was used, which is based on machine learning; it learns the characteristics of a large number of known binding sites through a Random Forest algorithm to predict the protein structure (28). The results are shown in Fig. S1C and the red area represents the active site. The score of this active site (residue site: 1) is >1 (score: 1.45) and the average AlphaFold score is >70 (score: 83.21), both of which indicate a high level of confidence in this site (27). Additionally, it was noted that SBEM is also known as MUCL1. MUCL has two subtypes: MUCL1 and MUCL3. To detect the conserved sequences, a sequence alignment between MUCL1 (ID: Q96DR8) and MUCL3 (ID: Q3MIW9) was performed (Fig. S1D). The alignment results were consistent with the prediction of the active site, both of which are located in the N-terminal region. The molecular docking results showed that PTX can bind tightly to SBEM (Fig. 3A). The strongest calculated binding free energy was-6.4 kcal/mol. In addition, each docking free energy was <-5.0 kcal/mol, demonstrating a good binding between the two. PTX was predicted to not only bind to the Pro84 amino acid of SBEM through a hydrogen bond but also form hydrophobic interactions with Met1, Ala5, Val6 and Pro84. Most notably, the distance of each interaction was <4 Å, indicating a relatively strong bond energy (Table SV).

Additionally, to verify the interaction between SBEM and DUSP16, a Co-IP assay was performed. DUSP16 was detected in the SBEM IP, supporting a physical interaction between the two proteins. Furthermore, SBEM knockdown led to a reduction in the phosphorylated (p-)DUSP16 levels in the input group. However, no significant change in total DUSP16 protein expression was observed compared with the siNC group, suggesting that SBEM may regulate DUSP16 phosphorylation rather than its expression (Fig. 3B).

Following SBEM knockdown, the apoptosis rate of SUM190PT/PTX cells (in culture medium containing 500 nmol/l PTX) significantly increased. The apoptosis rate of the SBEM-knockdown group increased approximately three-fold, rising from 16.53% in the control group (siNC + PTX) to 44.25% in the knockdown group (siSBEM + PTX) (Fig. 4A). By contrast, SBEM overexpression in SKBR3 cells (in culture medium containing 500 nmol/l PTX) significantly decreased apoptosis (Fig. 4B). To further investigate this observation, the expression of apoptosis-related proteins was analyzed.

In SUM190PT/PTX cells treated with PTX, compared with the siNC + PTX group, the siSBEM + PTX group had increased expression levels of Bax and cleaved-Caspase-3/Caspase-3, while the Bcl2 expression was decreased (Fig. 4C). Conversely, in SKBR3 cells treated with PTX, compared with the Vector + PTX group, overexpression of SBEM in the oeSBEM + PTX group decreased the expression of Bax and cleaved-Caspase-3/Caspase-3 and increased the expression of Bcl2 (Fig. 4D).

To elucidate the mechanism by which SBEM affects PTX sensitivity, the expression levels of DUSP16 and MAPK-related proteins were analyzed. In SUM190PT/PTX cells, compared with the siNC + PTX group, knockdown of SBEM in the siSBEM + PTX group decreased the ratios of p-DUSP16/DUSP16 and p-MAPK/MAPK (Fig. 5A). By contrast, in SKBR3 cells treated with PTX, compared with the Vector + PTX group, upregulation of SBEM in the oeSBEM + PTX group increased these ratios (Fig. 5B). These findings suggest that SBEM enhances DUSP16 phosphorylation and activates the MAPK signaling pathway, thereby potentially contributing to PTX resistance.

To further investigate the regulatory effect of SBEM on the MAPK pathway, SUM190PT/PTX cells were treated with the MAPK pathway activator, AMPK activator 13. The results showed that in SUM190PT/PTX cells treated with PTX, compared with the siSBEM group, the level of apoptosis in the siSBEM + AMPK activator 13 group decreased significantly while the cell viability increased significantly. The results indicated that AMPK activator 13 counteracted the inhibitory effects of SBEM on the viability and induction of apoptosis in SUM190PT/PTX cells (Fig. 6A and B). Furthermore, compared with the siSBEM group, the addition of AMPK activator 13 (siSBEM + AMPK activator 13 group) significantly reduced the levels of Bax and cleaved-Caspase-3 and increased the expression of Bcl2 (Fig. 6C). The effect of AMPK activator 13 on apoptosis-related proteins reversed the impact of siSBEM.

In addition, in SUM190PT/PTX cells treated with PTX, compared with the siSBEM group, the p-DUSP16/DUSP16 level decreased and the p-MAPK/MAPK level increased in the siSBEM + AMPK activator 13 group (Fig. 7). These findings suggest that SBEM may contribute to drug resistance in breast cancer cells by activating the MAPK pathway.