Human breast cancer cell lines MDA-MB-231 and CAL-51, and the human prostate cancer cell line LNcap were donated from the Public Laboratory of Tianjin Medical University Cancer Institute and Hospital (Tianjin, China). The human breast cancer cell line MDA-MB-453 was purchased from the BeNa Culture Collection; Beijing Beina Chunglian Institute of Biotechnology. All of the cell lines were identified and characterised by mycoplasma assays and DNA profiling (short tandem repeat) Cells were incubated at 37°C in an incubator containing 95% air and 5% CO₂. CAL-51 and LNcap cell lines were maintained in RPMI 1640 medium (cat. no. C11875500BT; Gibco; Thermo Fisher Scientific, Inc.); MDA-MB-231 cells were maintained in Dulbecco's modified Eagle medium (cat. no. C11995500BT; Gibco; Thermo Fisher Scientific, Inc.); and MDA-MB-453 cells were maintained in L15 medium (cat. no. SH30525.01; Hyclone; Cytiva). The aforementioned media were supplemented with 10% foetal bovine serum (FBS; cat. no. 10270-106; Gibco; Thermo Fisher Scientific, Inc.) and 1% penicillin streptomycin (cat. no. 15140122; Gibco; Thermo Fisher Scientific, Inc.). Enz was purchased from Beijing Solarbio Science & Technology Co., Ltd. (cat. no. 915087-33-1) and Chid was kindly provided by Chipscreen Biosciences Ltd. Enz and Chid were stored at a concentration of 10 mM in DMSO, the maximum concentration of DMSO was 2%. Various concentrations of AMG510 (KRAS inhibitor; BeiGene) were used to treat the CAL-51 cell line.

The MTS assay (cat. no. G3582; Promega Corporation) was used to assess the inhibitory effect of Enz and Chid, both individually and in combination. MDA-MB-231, CAL-51, MDA-MB-453 and LNcap cells in the logarithmic growth phase were collected and inoculated into 96-well plates (5×10³ cells/well, six biological replicates) with 200 μl medium. Cells were pretreated with Enz (3 μM for LNcap, 10 μM for MDA-MB-453, 20 μM for CAL-51 and 20 μM for MDA-MB-231 cells) or Chid (0.125 μM for LNcap, 1 μM for MDA-MB-453, 2 μM for CAL-51 and 10 μM for MDA-MB-231 cells) for 48 h at 37°C. Subsequently, 20 μl MTS reagent was added to each well and cultured for 2 h. Absorbance was measured at 490 nm using a microplate reader (Bio-Rad Laboratories, Inc.). The half-maximal inhibitory concentration (IC50) was calculated based on cell viability rate. Statistical analyses were performed using GraphPad Prism 8.0 (Dotmatics).

LNcap, MDA-MB-231, MDA-MB-453 and CAL-51 cells were treated with Enz and Chid for 48 h at the corresponding concentrations based on IC50. The combination index (CI) was calculated using CompuSyn software (version 1.0; ComboSyn, Inc.) to evaluate the combined effect on cells. CI >1.0 indicates an antagonistic effect, CI <1.0 indicates a synergistic effect, and CI=1.0 indicates an additive effect.

Cell proliferation was analysed using a colony formation assay. A total of 800-1,000 MDA-MB-231 and CAL-51 cells were inoculated in 6-well plates were cultured overnight, then treated with Enz (20 μM for CAL-51 and 20 μM for MDA-MB-231 cells) or Chid (2 μM for CAL-51 and 10 μM for MDA-MB-231 cells) alone or in combination for 48 h at 37°C. The plates were cultured for 10-14 days and the medium was replaced every 3 days. When the colonies became visible to the naked eye, the culture was terminated. The colonies were then fixed with 4% paraformaldehyde for 30 min and stained with 0.5% crystal violet for 30 min at room temperature. Subsequently, a light microscope (Olympus Corporation) was used for observation and imaging. Adobe Photoshop software (version 22.5.0; Adobe Systems, Inc.) was used to count the colonies.

Migration was assessed using a Transwell assay (pore size, 7 μm; Corning, Inc.). The MDA-MB-231 and CAL-51 cells were pretreated with Enz (20 μM for CAL-51 and 20 μM for MDA-MB-231 cells) or Chid (2 μM for CAL-51 and 10 μM for MDA-MB-231 cells) alone or in combination for 48 h at 37°C, and 1×10⁵ pretreated cells (in 200 μl FBS-free medium) were subsequently inoculated in the upper chamber of the Transwell system, and 500 μl medium supplemented with 20% FBS was added to the lower chamber. The upper chamber was fixed and stained with a Three-Step Stain Set according to the manufacturer's instruction at room temperature (Thermo Fisher Scientific, Inc.) once a single adherent cell was detected. The membrane was then cut, placed on a slide and sealed with paraffin. Images of the migrated cells were captured under an optical microscope at ×200 magnification and the number of cells was counted using Adobe Photoshop.

Cell cycle profiles were analysed using flow cytometry. The MDA-MB-231, CAL-51 and MDA-MB-453 cells pretreated with Enz (10 μM for MDA-MB-453, 20 μM for CAL-51 and 20 μM for MDA-MB-231 cells) or Chid (1 μM for MDA-MB-453, 2 μM for CAL-51 and 10 μM for MDA-MB-231 cells) at 37°C, individually or in combination, were collected and fixed with 95% pre-cooled ethanol at -20°C overnight, followed by washing with PBS and staining with 3 μM propidium iodide and 10 μg/ml RNase in the dark at 37°C for 30 min. Cell cycle distribution was identified and quantified using a flow cytometer (FACSAria III; BD Biosciences). The cell cycle distribution was analysed using ModFit LT software (v3.3; BD Biosciences).

CAL-51 cells were lysed in RIPA buffer supplemented with 1% phenylmethylsulphonyl fluoride (cat. no. R0020; Beijing Solarbio Science & Technology Co., Ltd.) to extract total proteins. The lysates were centrifuged at 8,000 x g for 10 min at 4°C to collect the supernatants, and protein concentrations were measured using a Pierce™ BCA Protein Assay Kit (cat. no. 23227; Thermo Fisher Scientific, Inc.). The protein lysates were mixed with 5X SDS-PAGE sample loading buffer (cat. no. P0015; Beyotime Institute of Biotechnology) and boiled at 95°C for 5 min. Subsequently, 25 μg proteins were separated by SDS-PAGE on 8-12% gels and were transferred to PVDF membranes (MilliporeSigma). The membranes were blocked with 5% skim milk at room temperature for 2 h and were then incubated with the primary antibodies at 4°C overnight, according to the manufacturer's instructions. Subsequently, the membranes were incubated for 60 min at room temperature with a secondary antibody (HRP-anti-rabbit, cat. no. 15015; HRP-anti-mouse, cat. no. 15014; both 1:5,000; both from ProteinTech Group, Inc.). The protein bands were visualized using ECL luminescence reagent (cat. no. abs920; Absin) and the Western blotting Imaging System (Amersham Imager 600 RGB; Cytiva). The grey value of the target protein band was analysed using ImageJ software (V1.8.0; National Institutes of Health). The primary antibodies used for western blotting were: AR (1:1,000; cat. no. 5153S), cyclin E1 (1:1,000; cat. no. 4129) (both from Cell Signaling Technology, Inc.), p16^INK4 (1:1,000; cat. no. ab189034; Abcam), cyclin D1 (1:1,000; cat. no. sc-450; Santa Cruz Biotechnology, Inc.), mTOR (1:1,000; cat. no. 2983S), phosphorylated (p)-mTOR^Ser2448 (1:1,000; cat. no. 5536S), p70S6K (1:1,000; cat. no. 34475S), p-p70S6K^{Thr421/Ser424} (1:1,000; cat. no. 9204S) (all from Cell Signaling Technology, Inc.), insulin receptor substrate 4 (IRS4; 1:500; cat. no. sc-373778; Santa Cruz Biotechnology, Inc.), GAPDH (1:5,000; cat. no. 6004-1-Ig) and β-actin (1:5,000; cat. no. 81115-1-RR) (both from ProteinTech Group, Inc.).

All in vivo experiments were performed according to the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals (22). The animal protocol was approved by the Animal Ethics Committee of Tianjin Medical University Cancer Institute and Hospital (approval no. HSTF-AE-2023023). CAL-51 cells (4×10⁶ cells/mouse) were collected and resuspended in 100 μl pre-cooled PBS supplemented with 5 μl Matrigel^® matrix (cat. no. 354248; Corning, Inc.). Cells were injected into the mammary gland fat pad of female NSG mice (age, 4-5 weeks; weight, 18-20 g; n=24; Shanghai Biomodel Organism Science & Technology Development Co., Ltd./). Mice were maintained at a temperature of 18-22°C and humidity of 50-60% under a 12-h light/dark cycle. Each cage contained six mice, and standard rodent diet and water were freely available. Tumour volumes were measured with a Vernier calliper every 3 days and calculated using the following formula: Length x width²/2. Mice were randomly allocated to four groups (n=6 mice/group) when the average tumour volume reached ~100 mm³. The four groups were treated with the control solvent once every 2 days (CMC-Na; cat. no. S6703; Selleck Chemicals), Enz, Chid, or their combination (Enz + Chid) at approximately the same time each day. Enz was dissolved in 0.5% CMC-Na for intragastric administration at 25 mg/kg body weight (BW) every 2 days. Chid was dissolved in 0.5% CMC-Na for intragastric administration and dosed at 12.5 mg/kg BW five times a week, as described in the drug instructions. Mice were sacrificed, and tumours were collected after 2 weeks of treatment. According to the Guidelines for Euthanasia of Rodents Using Carbon Dioxide (NIH 2020) (23), without pre-charging the chamber, the animals were placed in the chamber and 100% CO₂ was introduced at a fill rate of 50% chamber volume per minute.

Total RNA was extracted from CAL-51 cells treated with 20 μM Enz, 2 μM Chid or Enz + Chid for 48 h at 37°C using TRIzol^® reagent (Thermo Fisher Scientific, Inc.). RNA was quantified using the RNA Nano 6000 Assay Kit on the Bioanalyzer 2100 system (Agilent Technologies, Inc.). Each sample (1 μg RNA) was used for cDNA library construction. The effective concentration of the library was accurately quantified by RT-PCR, and the loading concentration of the library was >1.5 nM. The library was prepared using the NEBNext^® Ultra RNA Library Prep Kit for Illumina^® (New England BioLabs) and sequenced on Illumina Novaseq 6000 platform (Illumina, Inc.) according to the manufacturer's instruction and 150 bp paired-end reads were generated.

The Kyoto Encyclopedia of Genes and Genomes (KEGG) database was used to identify enriched signalling pathways from the differentially expressed genes. KEGG pathway enrichment analysis was performed using the R packages clusterProfiler (24), enrichplot, org.Hs.eg.db and ggplot2 (v4.1.2). All packages were downloaded from https://bioconductor.org/, with P<0.05 used as the screening criteria.

Gene Set Enrichment Analysis (GSEA) was used to explore significantly dysregulated pathways between the groups. Gene expression files and phenotype label files were generated and loaded into GSEA software (v4.1.0; Broad Institute, Cambridge, MA, USA) (25). The enrichment analyses were focused on the MsigDB HALLMARKER gene set. The permutation test was carried out 1,000 times, and the criteria were normalized enrichment score absolute value >1.5, nominal P<0.05.

Fresh tumour tissues obtained from mice were fixed with 10% neutral buffered formalin immediately for at least 12 h at room temperature. Fixed tissues were embedded in paraffin blocks and cut into 4-μm sections for H&E and IHC staining.

For H&E staining, tissue sections were stained with modified Lillie-Mayer Haematoxylin Solution (cat. no. G4070; Beijing Solarbio Science & Technology Co., Ltd.) at room temperature, and counterstained with Eosin Staining Solution (cat. no. C0109; Beyotime Institute of Biotechnology) according to the manufacturer's instructions.

For IHC staining, the slides were de-paraffinized in dimethylbenzene and hydrated in a series of ethanol solutions. Antigen retrieval was performed in a microwave oven with citric acid buffer (pH 6.0) for 15 min and cooled to room temperature. The endogenous peroxidase activity was quenched using 3% H₂O₂ at room temperature for 15 min. The tissues were blocked with 5% BSA in PBS 1 h at 37°C (cat. no. A8020; Beijing Solarbio Science & Technology Co., Ltd.) and incubated with primary antibodies overnight at 4°C. Subsequently, Biotinylated Goat anti-Mouse and anti-Rabbit secondary antibodies were obtained from OriGene Technologies, Inc. (cat. nos. TA373082L and TA373083L; 1:500) for 1 h at room temperature according to the manufacturer's instructions. Slides were visualised using a DAB horseradish peroxidase colour development kit (OriGene Technologies, Inc.). The primary antibodies used were as follows: AR (1:500; cat. no. 5153S), mTOR (1:500; cat. no. 2983S) (both from Cell Signaling Technology, Inc.), IRS4 (1:50; cat. no. sc-373778; Santa Cruz Biotechnology, Inc.) and Ki-67 (1:200,cat. no. ZM-0166; OriGene Technologies, Inc.). Finally, chromogenic detection was carried out (×200 magnification) under an optical microscope and the expression of various indicators was observed.

Total RNA of CAL-51 cells was extracted using TRIzol reagent and 1,000 ng was used to generate cDNA using the PrimeScript RT Reagent Kit (cat. no. RR047A; Takara Bio, Inc.) according to the manufacturer's instructions. qPCR was performed using TB Green Premix Ex Taq (cat. no. RR420A; Takara Bio, Inc.) on a CFX96 Touch Real-Time PCR Detection System (Bio-Rad Laboratories, Inc.) in triplicate, according to the manufacturer's instructions. The thermocycling conditions were as follows: 95°C for 2 min, followed by 40 cycles at 95°C for 10 sec, 60°C for 30 sec and 72°C for 30 sec, and a final step at 72°C for 5 min. All primers were synthesised by Tsingke Biological Technology. The primer sequences are listed in Table I. The results were analysed using 2^−ΔΔCq (26) and were normalised to β-actin expression levels.

Quantification and statistical analyses were performed using GraphPad Prism 8.0 (Dotmatics) or R software (v4.0.5). For in vitro analyses, the experiments were repeated three times. Data are presented as the mean ± SD. The statistical significance of differences between groups were analysed using one-way analysis of variance followed by Bonferroni's post hoc test. P<0.05 was considered to indicate a statistically significant difference.

To evaluate the inhibitory effect of the AR inhibitor Enz and the HDACi Chid, the expression levels of AR were detected in TNBC cell lines by western blotting and comparisons were made between the four cell lines. The prostate cancer cell line LNcap, which was used as a positive control, expressed the highest level of AR, whereas the expression levels of AR in MDA-MB-453, CAL-51 and MDA-MB-231 cells were gradually lower (Fig. 1A). Androgen deprivation therapy is a standard therapy regimen for patients with prostate cancer, and Enz, a second-generation AR antagonist, can substantially increase the survival of patients with metastatic or non-metastatic, castration-resistant prostate cancer (27,28). The AR-positive TNBC cell lines may also respond to Enz, as described previously (11). TNBC cell lines were treated with Enz in a series of concentration gradients to determine the IC50 of Enz for each cell line. The IC50 value of Enz was negatively associated with AR expression. The IC50 of Enz for MDA-MB-453 cells was the lowest (16.66 μM), followed by CAL-51 cells (24.26 μM), whereas MDA-MB-231 cells had the highest IC50 value (26.51 μM) (Fig. 1B and C). The present study then evaluated the IC50 of Chid; the results revealed that the IC50 of Chid was lowest in MDA-MB-453 cells (1.44 μM), followed by CAL-51 (7.75 μM) and MDA-MB-231 cells (18.31 μM) (Fig. 1B and D). In addition, the present study verified that Chid could effectively reduce the protein expression levels of AR (Fig. S1).

To assess the synergistic effect of Enz and Chid treatment on TNBC cell lines, the CI was calculated. The CI ranged between 0.1 and 0.9, suggesting that Enz and Chid had potential synergistic effects at certain concentrations in AR-positive TNBC cell lines, including MDA-MB-453, CAL-51 and MDA-MB-231 (Fig. 1E).

AR-positive TNBC cells were treated with Enz monotherapy, Chid monotherapy or combination therapy at appropriate concentrations according to the IC50 values for the corresponding cells. Morphological changes resulted from treatment with Enz (10 μM for MDA-MB-453, 20 μM for CAL-51 and 20 μM for MDA-MB-231) and Chid (1 μM for MDA-MB-453, 2 μM for CAL-51 and 10 μM for MDA-MB-231), including cell shrinkage, cell size reduction, rounder cell morphology, nuclear chromatin condensation and fragmentation, cytoplasmic vacuolar changes and reduced synapses (Fig. 2A). The most marked changes in morphology were observed in response to the combination therapy, followed by Enz or Chid monotherapy in MDA-MB-231, CAL-51 and MDA-MB-453 cells. The colony formation assay showed that the combination therapy with Enz and Chid for 48 h significantly inhibited colony formation compared with that in the control and monotherapy groups (P<0.05; Fig. 2B). In addition, Enz + Chid inhibited cell migration more efficiently than either drug alone in AR-positive TNBC cell lines (P<0.05; Fig. 2C).

Tumour cell proliferation is partially caused by cell cycle dysregulation (29). Based on the aforementioned inhibitory effects of Enz and Chid combination therapy on proliferation, flow cytometry was used to evaluate the effect of Enz and Chid on the cell cycle. Cells were treated with Enz or Chid, alone or in combination at a certain concentration, for 48 h. The results showed that Enz or Chid monotherapy slightly increased the proportion of cells in G₂/M phase, whereas the elevation was more pronounced in the combination group compared with that in the control and monotherapy groups (Fig. 3A). This resulted in a significant decrease in the proportion of cells in the G₁/S phase.

Multiple molecules, including cyclin-cyclin-dependent kinase (CDK) complexes, have been identified as regulators of cell proliferation (30). The cell cycle is precisely regulated by multiple CDKs at the corresponding cell cycle checkpoint (31). Western blotting was used to detect Enz- and Chid-induced changes in cell cycle progression. Compared with in the control and monotherapy groups, decreased cell proliferation was associated with downregulation of cyclin D1 and cyclin E1 upon the combination treatment of Enz and Chid, whereas the individual drugs had a limited effect on AR-positive TNBC cell lines (Fig. 3B). In addition, an inhibitor of CDK4/6, p16^INK4a, was markedly upregulated in the combination treatment group (Fig. 3B) (32).

The present study then explored the potential molecular mechanisms underlying the effects of the combination therapy. Transcriptome sequencing-based analysis was performed to examine the transcriptomes of CAL-51 cells treated with vehicle, Enz, Chid, or a combination of the two for 48 h. KEGG functional enrichment analysis showed that the 'PI3K-Akt' signalling pathway was significantly enriched (P=0.00053, count=66) (Fig. 4A). Western blot analysis confirmed the inhibition of the expression of key proteins in the PI3K/Akt signalling pathway in response to the combination treatment (Fig. 4B). GSEA revealed that several important tumour suppressor pathways were upregulated only in response to combined drug treatment (NES >1.7, NOM P<0.05, FDR q<0.05), whereas the change was not significant with Enz alone (NES <1.5, NOM P>0.05, FDR q>0.05) (Figs. 4C, D and S2).

RAS has been shown to efficiently activate the 'PI3K-Akt' signalling pathway (33,34). In CAL-51 cells, genes that were downregulated by KRAS activation were upregulated only upon combined treatment with Enz and Chid (NES=1.93, P<0.0001), whereas the change was not significant with Enz alone (NES=1.27, P=0.094; Fig. 4D). These genes may be involved in the regulation of the downstream PI3K/Akt/mTOR signalling pathway in response to combination treatment. A list of genes was screened, including CHST2, TEX15, CALCB, IRS4, ADRA2C, CAMK1D, KLHDC8A, RYR2, WNT16, GDNF and BRDT, to further elucidate the association between KRAS activation and the PI3K/Akt/mTOR signalling pathway. It was revealed that the expression levels of these genes were also upregulated in the combination group compared with the other groups (Table SI). Subsequent RT-qPCR verification showed that only IRS4 was significantly upregulated by the combination treatment of Enz and Chid compared with in the control and monotherapy groups (Fig. 4E). The protein expression levels of IRS4 were then detected in CAL-51 cells treated with Enz and/or Chid, and it was revealed that IRS4 was markedly upregulated following combined treatment with Enz and Chid compared with in the control and monotherapy groups (Fig. 4F). Marked overexpression of IRS1 has been shown to reduce the growth of lung cancer, and IRS1 deficiency can drive a proinflammatory phenotype in lung adenocarcinoma with KRAS mutations (35,36). In the present study, the KRAS inhibitor AMG510 was used to treat CAL-51 for 1, 6, 12, 24 and 48 h at 37°C. The results showed that IRS4 expression increased over time. In addition, IRS4 was upregulated as a result of increased AMG510 concentration for 48 h (Fig. 4G).

The present study investigated the antitumour efficacy of combination treatment with Enz and Chid in a xenograft mouse model of CAL-51. Mice were treated with Enz or Chid, alone or in combination, after 30 days of CAL-51 xenograft model construction. The frequency of administration in each group is shown in Fig. 5A. The treatment lasted for 14 days due to the heavy tumour burden of the control group (the maximum tumour volume detected was 500 mm³). The tumour growth curves for each group are presented in Fig. 5B. It was observed that the combination of Enz and Chid significantly slowed tumour growth compared with that in the control and monotherapy groups, displaying excellent antitumour efficacy higher than that of either monotherapy (Fig. 5C). Notably, no obvious adverse drug reactions, such as vomiting and diarrhoea, were observed, indicating that the combination regimen of Enz and Chid was tolerable to these mice.

As confirmed by histological analysis, the decreased proliferation and increased area of necrosis may account for the observed response to the combination therapy (Fig. 5D). IHC staining showed decreased expression of Ki-67 and mTOR in the combination group compared with that in the control and monotherapy groups (Fig. 5D), which is consistent with the results of the in vitro experiment. These findings indicated that Chid, in combination with Enz, may sufficiently block the activation of the PI3K/Akt/mTOR signalling pathway to inhibit tumour growth and upregulate the expression of IRS4 (Fig. 6).