Human MDA-MB-231 and mouse 4T1 cell lines were obtained from the Chinese Academy of Sciences (Shanghai, China). The MDA-MB-231 cells were maintained in complete Leibovitz L15 medium (Gibco, USA), while 4T1 cells were cultured in complete Dulbecco’s modified Eagle medium (Gibco, USA). Cells were incubated at 37 °C in a humidified 5% CO₂ atmosphere.

PBMCs were isolated from heparinized blood of healthy donors using Ficoll-Paque density gradient centrifugation. To induce immature dendritic cells (imDCs), the harvested PBMCs were cultured in the presence of recombinant human GM-CSF (rhGM-CSF) and interleukin-4 (rhIL-4) over a 5-day differentiation period. Mature dendritic cells (mDCs) were subsequently induced by exposing imDCs to tumor necrosis factor-alpha (TNF-α) and interferon-gamma (IFN-γ).

From this same PBMC pool, a portion was differentiated into immature DCs, while CD3⁺ T cells were simultaneously isolated via immunomagnetic separation. This approach ensured that the DCs and T cells used in any given co-culture experiment were autologous. The purified cell pellet was carefully reconstituted in 10 mL of ImmunoCult-XF T Cell Expansion Medium and cultivated under tightly controlled physiological conditions within T75 culture flasks.

Cell viabilities under gradient concentrations of nab-PTX (1 nM-100 µM) (YBH16872021, Hengrui Medicine Co., Ltd) exposure were systematically quantified employing the CCK-8 assay kit (Abcam, USA) in strict adherence to the manufacturer’s standardized protocols.

The cell apoptosis was quantified through flow cytometry employing an Annexin V-FITC/PI apoptosis detection kit (Merck Millipore, USA) after a 48-hour treatment period with incremental concentrations of nab-PTX.

DCs were pretreated with conditioned medium for 24 hours before the migration assay to assess their motility. After incubation, DCs were seeded into the upper chambers of 6-well BD Falcon^® Transwells. After 24 hours, migrated cells were counted. Migration efficiency was calculated as (migrated DCs/total DCs) ×100%, providing a quantitative measure of motility.

Western blotting was performed according to standard protocols. Briefly, total protein was extracted after treatment, separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and transferred to nitrocellulose membranes. Membranes were incubated overnight with primary antibodies at specified dilutions, including anti-cyclic GMP-AMP synthase (cGAS) (1:10,000, Abcam, USA, ab302671), anti-stimulator of interferon genes (STING) (1:10,000, Abcam, USA, ab239074), anti-phospho-STING (1:10,000, Cell Signaling Technology, USA, 50907T), anti-TANK-binding kinase 1 (TBK1) (1:10,000, Abcam, USA, ab40676), anti-phospho-TBK1 (1:10,000, Abcam, USA, ab109272), anti-interferon regulatory factor 3 (IRF3) (1:10,000, Abcam, USA, ab68481), and anti-phospho-IRF3 (1:10,000, Abcam, USA, ab76493). Protein band intensities were quantified using Image J for statistical analysis.

Tumor specimens were embedded in paraffin blocks and incubated with anti-CD11c primary antibody (diluted at 1:100 ratio; CST #97586S, Cell Signaling Technology, USA) at 4 °C for 16 hours. Specimens were then treated with HRP-conjugated secondary antibody and imaged using high-resolution digital microscopy.

Cells were immunostained with primary antibodies: anti-HMGB1 (1:100 dilution, Abcam, USA, ab18256) and anti-CD11c (1:100 dilution, Abcam, USA, ab52632). Samples were then treated with Alexa Fluor 488-conjugated secondary antibody (1:1000 dilution, Thermo Fisher Scientific, USA) followed by nuclear counterstaining with 4’,6-diamidino-2-phenylindole (DAPI).

Total RNA was extracted from cellular specimens utilizing TRIzol™ Reagent (Thermo Fisher Scientific, USA) in strict accordance with the manufacturer’s guidelines. Subsequently, first-strand cDNA synthesis was performed with the High Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific, USA). Quantitative reverse transcription polymerase chain reaction (qRT-PCR) assays were carried out on a Bio-Rad CFX platform using iTaq Universal SYBR Green (Bio-Rad, USA), with primer sequences specified in Table 1.

The analytical measurements were performed using commercially available ELISA kits, with all protocols meticulously followed according to manufacturers’ specifications. Optical density values were then precisely measured through microplate spectrophotometry at designated wavelengths.

Female BALB/c mice (6–8 weeks old, 17 ± 3g body weight) sourced from Animal Experimental Center of Guizhou Medical University, were housed in room temperature (20-26 °C) and specific pathogen-free (SPF) environment. To avoid rejection and foreign body response, BALB/c mice underwent subcutaneous implantation of mouse 4T1 cells in the right flank region. When tumors reached an average volume of ~100 mm³, the mice were randomly assigned to two experimental cohorts: a control group receiving PBS solution and a treatment group administered nab-PTX via therapeutic intraperitoneal injection on days 7, 11, and 15. After 19 days post-tumor injection, the mice were euthanized by isoflurane inhalation (Isofluorane, R510-22, RWD Life Science Co, China), and the tumors were removed and weighed.

Conditioned media from MDA-MB-231 cells, treated with 5 nM nab-PTX for 48 hours or untreated controls (NC) in Leibovitz L15 medium, were added to 6-well plates. ImDCs were cultured in these conditioned media (imDCs+nab-PTX and imDCs+NC) under standard conditions. After 48 hours, cells were harvested for further experiments.

DCs were primed with conditioned medium and co-cultured with T cells at a 1:10 ratio for 48 hours. After incubation, supernatants were collected and aliquoted into sterile microcentrifuge tubes for ELISA.

Proteomic data were collected from tumor tissues of three mice per group (nab-PTX treated and control) and analyzed by liquid chromatography-mass spectrometry (LC-MS). Differentially expressed proteins were identified through comparative profiling and subsequently subjected to Gene Ontology (GO) classification and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses. Functional annotation was performed using EggNOG-mapper (v2.0) for GO categorization, and KEGG pathway annotation was conducted via BLAST searches against the KEGG database. The disease-free survival (DFS) and overall survival (OS) were evaluated by the Gene Expression Profiling Interactive Analysis 2 (GEPIA 2) platform (http://gepia2.cancer-pku.cn/#survival). Prognostic correlations across subtypes, with a specific emphasis on TNBC, were examined using bc-GenExMiner v4.9. Protein-protein interaction networks were mapped via the STRING database (https://cn.string-db.org/), while UALCAN (https://ualcan.path.uab.edu/) facilitated comparative analyses of GZMB and CD8α expression profiles between malignant and adjacent normal tissues.

All the data were analyzed from at least three independent experiments and the quantification results were expressed as mean ± SD. Data met the assumption of the statistic tests. Statistical comparisons between two groups were conducted by using the Student’s t test and between multiple groups using one or two-way ANOVA using the Prism software (v.7.0, GraphPad, San Diego, CA), whereas image J program was used for picture analysis. Statistical significance was defined as *p < 0.05.

To determine the optimal nab-PTX treatment conditions, MDA-MB-231 cells were exposed to graded concentrations. CCK-8 assays revealed a significant, dose- and time-dependent inhibition of cell proliferation (Figure 1A, P < 0.001). Treatment with 1 nM for 24, 48, and 72 hours resulted in proliferation rates of approximately 80-90%. When the concentration was increased to 5 nM for the same durations, proliferation decreased to roughly 60-80%. Treatment with 100 nM or higher concentrations led to a more pronounced reduction in proliferation; notably, after exposure to 100 μM for 72 hours, the cells were almost completely unable to adhere, and the proliferation rate dropped to around 10%. Based on these findings, subsequent experiments were conducted using untreated controls and cells treated with 5 nM nab-PTX for 48 hours. Flow cytometry analysis confirmed that nab-PTX induced apoptosis (Figure 1B). To further evaluate its immunogenic potential, RT-qPCR was performed to quantify canonical ICD biomarkers (HMGB1, CRT, HSP70A1A, and Caspase-3), which were significantly upregulated across treatment groups (Figure 1C). Consistently, ELISA demonstrated elevated extracellular levels of HMGB1, CRT, and HSP70 in the supernatants of nab-PTX–treated cells (Figure 1D). Immunofluorescence imaging additionally revealed nuclear-to-cytoplasmic HMGB1 translocation in treated cells compared with controls (Figure 1E). Collectively, these findings demonstrate that nab-PTX triggers the release of ICD associated DAMPs.

Given the central role of DCs in T cell activation, we investigated the impact of conditioned medium on DC maturation. PBMCs from healthy donors were differentiated into imDCs and mDCs using standard protocols (Figure 2A). Confocal microscopy confirmed cell integrity and distinct morphology (Figure 2B). Conditioned medium from MDA-MB-231 cells exposed to nab-PTX for 48 hours was used in subsequent experiments. ImDCs cultured in conditioned medium showed significant upregulation of maturation markers (CD80, CD86, CD40, HLA-DR) compared with controls, as measured by qRT-PCR (Figure 2C). Transwell assays compared migratory behavior under different conditions. Spontaneous migration was similar between controls (23-33%) and treated groups (24-35%), but conditioned medium–treated DCs exhibited enhanced chemotaxis (Figure 2D). Phagocytosis assays revealed reduced FITC-dextran uptake in conditioned medium-treated DCs (Figure 2E). Cytokine profiling showed increased secretion of IFN-α, IFN-β, IL-2, IL-12, CCL5, and CXCL10 in treated groups (Figure 2F). Collectively, these findings demonstrate that conditioned medium promotes imDC maturation into mDCs, characterized by enhanced chemotaxis and elevated cytokine and chemokine secretion.

Conditioned medium-activated imDCs were co-cultured with purified T cells for 24 hours (Figure 3A). To evaluate the immunomodulatory effects on T cell proliferation, CD3⁺T cells were isolated from PBMCs using magnetic sorting, reaching 92.3% purity confirmed by flow cytometry (Figure 3B). Co-culture with stimulated DCs significantly increased CD3⁺CD8⁺T cell subsets, rising from 28.7% in controls to 32.4% in the experimental group (Figure 3C, D) (p<0.05). These results suggest that conditioned medium–activated imDCs promote T cell differentiation. To assess cytokine secretion, TNF-α, IFN-γ, and Granzyme B (GZMB) were quantified by ELISA. TNF-α secretion increased from 12.1 ± 1.3 pg/mL in controls to 30.4 ± 2.7 pg/mL in treated groups (Figure 3E). IFN-γ levels rose from 79.8 ± 6.5 pg/mL to 198.2 ± 12.1 pg/mL (Figure 3F). GZMB production increased 2.1-fold in treated groups compared with controls ((Figure 3G). Collectively, these findings show that conditioned medium–activated imDCs enhance T cell effector responses by upregulating pro-inflammatory cytokines and cytotoxic mediators.

To assess the immunomodulatory effects of nab-PTX in vivo, TNBC mouse models were generated by orthotopic implantation of 4T1 cells into BALB/c mammary fat pads, with a schematic summarizing dosing and sample collection (Figure 4A). During the 19-day treatment, longitudinal monitoring showed comparable physiological profiles, particularly body weight, between nab-PTX and saline groups (Figure 4F). Quantitative analysis showed significantly greater tumor suppression in nab-PTX-treated mice, with marked reductions in tumor volume (Figures 4B, D, E) and mass (Figure 4C) compared with controls. HE staining confirmed no toxic damage in major organs, including heart, liver, spleen, lung, and kidney (Figure 4G). Immunohistochemistry revealed robust DC infiltration in nab-PTX-treated tumors, with a 2.8-fold increase in CD11c⁺cell density compared with controls (Figure 4H). Immunofluorescence showed increased tumor-infiltrating CD8⁺T cells after nab-PTX treatment (Figure 4I). Molecular profiling revealed strong upregulation of GZMB in treated tumors, with western blot confirming a 2.1-fold protein increase versus controls (p < 0.01, Figures 4J, K). These findings demonstrate that nab-PTX enhances antitumor immunity by promoting DC activation and cytotoxic T cell-mediated tumor elimination.

To identify key prognostic factors in TNBC, the expression levels of CD8ɑ gene and its clinical implications were investigated. The data revealed that elevated CD8ɑ expression correlated with improved DFS and OS (Figure 5A, B). Statistical analysis confirmed significantly enhanced DFS and OS in the high CD8ɑ expression group compared to the low-expression cohort (Figure 5C, D). Notably, CD8ɑ downregulation was consistently observed in breast cancer tissues versus adjacent normal tissues (Figure 5H). Given GZMB’s established role as a key cytotoxic effector in CD8⁺ T cell immunity, a strong positive correlation between CD8ɑ and GZMB expression patterns (Figure 5E-G) was identified. Survival analysis further demonstrated that GZMB levels emerged as independent predictors of prolonged DFS (p = 0.035, HR = 0.67) and OS (p = 0.032, HR = 0.7) (Figure 5I-J), with this prognostic relationship proving most pronounced in TNBC subtypes (Figure 5K-L).

To assess the immunomodulatory effects of nab-PTX in vivo, proteomic profiling was performed on tumor tissues from mouse models. Differential expression analysis revealed distinct molecular signatures between groups, illustrated in the hierarchical clustering heatmap (Figure 6A). Mass spectrometry identified 7,143 proteins, of which 85 were upregulated and 58 downregulated (Figure 6B). Granzyme family members and chemokine-related proteins were among the most strongly upregulated (Figure 6C). KEGG pathway analysis showed that upregulated proteins were mainly involved in immune processes (Figure 6D). Functional annotation identified three main effects: regulation of bone marrow leukocyte differentiation, acute inflammatory responses, and T-cell cytokine production (Figure 6E). Molecular analysis highlighted chemokine activity and cytokine receptor binding as the most affected functions (Figure 6F). Subcellular localization showed that 64.3% of differentially expressed proteins were extracellular (Figure 6G), suggesting nab-PTX may enhance immune infiltration and antitumor responses via extracellular matrix remodeling.

Western blot was performed to assess key protein expression in the cGAS-STING pathway of DCs after conditioned media exposure. Quantitative analysis showed marked upregulation of p-STING, p-TBK, and p-IRF3 in conditioned media-treated DCs compared with controls (Figures 7A, B). Treatment with H-151, a selective STING inhibitor, reduced STING expression in imDCs (Figures 7C, D). Conversely, the STING agonist G10 increased STING expression in complementary experiments (Figures 7E, F). Cytokine profiling showed that G10-stimulated imDCs had increased IL-2 and IL-12 secretion, along with elevated CCL5 and CXCL10. In contrast, H-151-treated imDCs maintained baseline IL-2, CCL5, and CXCL10 but showed reduced IL-12 production (Figure 7G). The cGAS-STING pathway significantly regulated TNF-α, IFN-γ, and GZMB secretion (Figure 7H).

These findings found that conditioned media from nab-PTX-treated MDA-MB-231 cells enhanced imDC immunogenicity via cGAS-STING activation, promoting maturation and inducing IL-2, IL-12, CCL5, and CXCL10 production. This cytokine response drives CD8⁺T cell differentiation and activation, resulting in synergistic enhancement of anti-tumor immunity (Figure 8).