Main materials

Cs (1105508, Sigma-Aldrich), EDTA·2Na (324503, Sigma-Aldrich), acetone (650501, Sigma-Aldrich), glutaraldehyde (8.20603, Sigma-Aldrich), dialysis membrane (MWCO 14 kDa, FDM514m Beyotime), Cur (2 mg, 08511, Sigma-Aldrich), FBS (10099, Gibco), RPMI 1640 medium (11875101, Gibco), DMEM/F-12 medium (21041025, Gibco), PMA (16561–29-8, Sigma-Aldrich), IL-4 (20 ng/mL, 200–04, PeproTech), IL-13 (20 ng/mL, 200–13, PeproTech), 0.5% Triton X-100 (P0096, Beyotime), anti-CD86 rabbit monoclonal antibody (ab239075, 1:100, Abcam, UK), anti-CD163 mouse monoclonal antibody (ab156769, 1:100, Abcam, UK), Alexa Fluor 488/594-conjugated secondary antibodies (ab150077/ab150116, 1:200, Abcam, UK), DAPI (10 μg/mL, D3571, Thermo Fisher, USA), polybrene (TR-1003, Merck), puromycin (540222, Sigma-Aldrich), RIPA lysis buffer (P0013B, Beyotime, China), BCA protein assay kit (P0011, Beyotime, China), HRP-conjugated goat anti-rabbit IgG (ab6721, 1:2000, Abcam), EdU (C0071S, Beyotime), 0.4% crystal violet (C0121, Beyotime), anti-PD-L1 antibody (BE0101, Bio X Cell), isotype control antibody (BE0090, Bio X Cell), anti-Ki67 rabbit antibody (ab16667, 1:200, Abcam), anti-PCNA rabbit antibody (ab92552, 1:500, Abcam), RBC Lysis Solution (158904, Qiagen), anti-CD80-PE (553769, BD Biosciences), anti-CD86-APC (561964, BD Biosciences), anti-CD206-FITC (141703, BioLegend), anti-CD163-APC (155305, BioLegend).

Preparation and characterization of TAMs M2pep-Cs-Cur NPsSynthesis of M2pep-Cs NPs

The synthesis procedure of Cs NPs entailed the gradual introduction of 3 mL acetone (650501, Sigma-Aldrich) into 2 mL Cs solution (5 mg/mL, 1105508, Sigma-Aldrich) with the incorporation of 1.25 mg EDTA·2Na (324503, Sigma-Aldrich) over a period of 5 min under magnetic stirring at 25 °C and 300 rpm, resulting in a milky-white non-crosslinked NP suspension. Subsequently, 20 μL of glutaraldehyde (25% solution in water, 8.20603, Sigma-Aldrich) was added for crosslinking at pH 6.5 for 4 h. The NPs were then purified by dialysis (MWCO 14 kDa, FDM514m, Beyotime) in distilled water for 48 h. The M2pep peptide (sequence: YEQDPWGVKWWY, 10.27517/d.cnki.gzkju.2021.000532), labeled with Cy5, was synthesized and conjugated to the Cs NPs. To achieve this, 0.5 mg of M2pep (0.27 mmol/L) was included in 1.0 mL of 1.0 mg/mL Cs NPs (< 0.07 mmol/L) and stirred at 200 rpm at room temperature (RT) for 16 h. The NPs, post-centrifugation at a speed of 10,000 rounds per minute for 15 min, were rinsed and rehydrated in deionized water, with the repetitive nature of this rinsing process being threefold. The conjugation efficiency was determined using UV–visible spectroscopy at 280 nm, with an efficiency of approximately 85% (Pang et al. 2019).

Loading of Cur and efficiency calculation

The loading of Cur was carried out using the following steps. 200 μL of a Cur solution (4 mg) (08511, Sigma Aldrich) was added to separate 1 mL suspensions of M2pep-Cs NPs (10 mg, prepared in a Cs solution dissolved in 1% acetic acid and modified to pH 5.0 with 0.1 M NaOH). Sodium triphosphate (TPP, 1 mg/mL, pH 5) was introduced, and the solution was gently mixed in a dimly lit environment at RT for a duration of 12 h. Subsequently, the mixture experienced centrifugal forces at a rate of 10,000 rounds per minute for 10 min to eliminate any Cur that had not been taken up, following which the residue was dispersed in 2 ml of deionized water to generate Cur-loaded NPs, named as M2pep-Cs-Cur NPs. Assessment of the free Cur concentration in the supernatant was executed at 425 nm with the aid of a UV–Vis spectrophotometer (UV-1900, Shimadzu). By applying specific mathematical expressions, the drug loading efficiency (DLE) and drug loading capacity (DLC) were quantified: DLC (%) = (weight of loading)/(weight of NPs + Mass of NPs) × 100%; DLE (%) = (weight of loading)/(Total drug added) × 100%. Measurement of free Cur concentration in the supernatant was carried out via UV–Vis spectroscopy at 425 nm. Under equivalent experimental setups, the calibration chart for Cur was constructed, with interference from degradation products and environmental factors evaluated and accounted for. The trials were triplicated for each data point. Results are expressed as mean ± standard deviation (SD), and statistical significance was assessed through paired or unpaired t-tests for normally distributed data, with P < 0.05 indicating statistical significance (Zhou et al. 2014).

Particle size and charge analysis

Through the utilization of dynamic light scattering (DLS) with a Nano-ZS90 apparatus (Malvern, UK) featuring a 633 nm He–Ne laser, the dimensions, polydispersity index (PDI), and zeta potential on the surface of Cs NPs and M2pep-Cs NPs were assessed. The evaluations were completed under constant conditions at 25 degrees Celsius, repeated thrice (Zhou et al. 2014, Khan et al. 2016).

Stability of NPs

Studying the physiological endurance of NPs, they were dissolved in water, saline, PBS, and RPMI 1640, and then cultured for a day. Furthermore, suspension of M2pep-Cs NPs occurred in solutions with different pH levels (4.5, 5.5, 6.5, 7.5, and 8.5). Post a 24-h incubation duration, the dimensions, dispersion, and radiance intensity of NPs were analyzed utilizing DLS. Direct observation indicated the morphology of the NPs using a Hitachi H7650 transmission electron microscope (Hitachi, Japan). Measurements were performed in triplicate (Zhou et al. 2014).

Transmission electron microscopy (TEM) analysis

TEM (Hitachi H7650, Japan) was employed to analyze the morphology of the NPs. Specimens of M2pep-Cs NPs were placed onto copper grids coated with carbon, dried in the air, and subjected to negative staining with a 1% w/v uranyl acetate solution for a duration of 15 min. Once dried, the samples underwent examination utilizing an accelerating voltage of 120 kV. Captured TEM pictures were scrutinized and interpreted with the Soft Imaging Viewer software for additional investigation. All experiments were conducted in triplicate (Zhou et al. 2014, Khan et al. 2016).

Cur release and cytotoxicity in Vitro

The Cur release curves of M2pep-Cs NPs were assessed in three distinct environments: pH 7.4, pH 6.5, and pH 5.(adjusted from PBS buffer at pH 7.4 using HCl). Put simply, 1 mL of Cur NP(10 µg/mL)-loaded solution with an identified concentration was moved into a dialysis sack (MWCO 3.5 kDa), ensuring the NPs (molecular weight > 14 kDa) remained inside while Cur (molecular weight ~ 0.37 kDa) dispersed and soaked in 5 mL of PBS. Subsequently, all extracts were placed in a 50 ml rotor tube and swirled leisurely in obscurity at 37 degrees Celsius moving at a velocity of 100 rpm. At set time intervals, 5 mL of the buffer solution was extracted and substituted with fresh 5 mL of PBS. The Cur release quantity was detected at 425 nm through UV spectrophotometric analysis. Each measurement was executed thrice. Results are reported as mean ± SD. Analytical computations involved assessments of normality and uniformity of variance, proceeded by unpaired t-tests to compare groups.

Cytotoxicity of Cur

THP-1 cellular entities were introduced into a 96-section container and managed at a temperature of 37 degrees Celsius under 5% CO₂ for a period of 24 h. Following removal of the original medium, 200 μL of new medium with unbound Cur and Cur-loaded NPs (at concentrations of 1.5625 to 50 μg/mL) or M2pep-Cs NPs (at concentrations of 31.25 to 1000 μg/mL, carrying around 5% drug content) was added. Cs NPs without Cur were used as controls. After 48 h of incubation, removal of the medium was conducted, and 200 μL of a fresh medium containing MTT (0.5 mg/mL) was introduced into each well, followed by an incubation in darkness lasting for 4 h. Subsequently, the suspended liquid was gathered, and 150 μL of the solvent DMSO was poured into every well to disintegrate the formazan crystals. After vibrating in the dark for 5 min, each well's light absorption was recorded at 570 nm employing a microplate photometer (E8051, Promega). Computation of cell viability was executed based on the provided formula: Cell viability = (Sample OD570 / Control OD570) × 100% (Wang et al. 2021a, b, c). The investigation was carried out in triplicate, and results from diverse time points were evaluated using two-way analysis of variance (ANOVA).

Cellular uptake

THP-1 cells in 12-well plates were treated with 5 μM Cur, M2pep-Cs NPs (100 μM), and M2pep-Cs-Cur NPs (100 μM, with a drug loading capacity of approximately 5%) for 4 h. After treatment, the cells received three PBS washes and were then set using cold paraformaldehyde for 15 min. Following cell washing, DAPI was applied for a 10-min staining duration. Finally, observations of fluorescence were made utilizing a confocal microscope (Leica, CH) and analyzed using the Image J software. Additionally, cells treated with the drugs were collected, fixed with cold paraformaldehyde, stained with DAPI, and analyzed using flow cytometry (FC500 Beckman, USA) (Zhang et al. 2022a, b). The investigation was duplicated on three occasions.

Construction of the "Curcumin-targets-disease" network

A total of 966 and 63 target genes for Cur were retrieved from the Comparative Toxicogenomics Database (CTD) (https://ctdbase.org/) and the SwissTargetPrediction database (STP) (http://swisstargetprediction.ch/), respectively. By employing the "Venn diagram" software, intersecting genes were distinguished through the Venn diagram visualization. Subsequently, the GeneCards database (https://www.genecards.org/) yielded the leading 100 genetic markers correlated with MΦ. Identifying six common genes between Cur targets and genes associated with MΦ was facilitated by a Venn diagram. Application of the "clusterProfiler" software facilitated the implementation of enrichment analyses for Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG). The STRING online platform was utilized to conduct an analysis of protein–protein interactions (PPI). The assembly of the network "Drug-Targets-Macrophages" was accomplished with the aid of Cytoscape 3.7.2 tool, and network topology parameters were calculated using NetworkAnalyzer for network analysis (Wang et al. 2022a, b, c).

Molecular docking simulation

The Protein Data Bank (https://www.rcsb.org) provided crystal structures for six selected proteins. PubChem data provided the configuration of Cur, which was then transformed into a three-dimensional setup using Chem3D Ultra 14.0 tool, and subsequently underwent energy reduction using the MM2 technique. The utilization of PyMOL software facilitated the preparation of target protein receptors by dehydrating and eliminating organic molecules. The target protein receptor molecules were hydrogenated, charges were calculated, and the conversion of both the compound and target protein receptors to "pdbqt" files was conducted by AutoDockTools 1.5.6. Accurate case focal points and grid parameters were fixed for the docking assessments. Finally, the molecular docking assessment utilized Vina 1.1.2 for calculating the energy values associated with docking (Rosa 2021). Molecular dynamics (MD) simulations of the six protein–ligand complexes were conducted using Amber20 for 100 ns. Root-mean-square deviation (RMSD) trajectories were analyzed, and MMGB/SA modules were used to calculate binding free energies based on stable MD trajectories. Protein systems were parameterized using the AMBER FF14SB force field and solvated in TIP3P water boxes extended by 10 Å. Appropriate counterions were added to neutralize the system.

In vitro cell cultures

Throughout this research, human monocyte MΦ THP-1 (TIB-202, ATCC) were nurtured in RPMI 1640 medium enriched with 10% FBS, while HEK-293 T cells (CRL-3216, ATCC) were maintained in DMEM/F12 medium also supplemented with 10% FBS. Human TNBC cells MDA-MB-231 (CRM-HTB-26, ATCC) and BT-549 (HTB-122, ATCC), as well as mouse TNBC cells 4T1 (CRL-2539, ATCC), were all maintained in a CO2-regulated incubator (model 51032874, ThermoFisher, USA) with an unchanging 37 °C temperature. Gibco (USA) was the source for the FBS (Catalog number 10099), RPMI 1640 medium (Catalog number 11875101), and DMEM/F-12 medium (Catalog number 21041025).

MΦ activation was initiated by treating THP-1 cells with 320 nM phorbol myristate acetate (PMA) (Sigma-Aldrich, 16561–29-8) for a period of 24 h. The differentiation of M2 MΦ was initiated by treating THP-1 cells with PMA (320 nm) for 6 h, followed by stimulation with IL-4 (20 ng/mL) (PeproTech, 200–04) and IL-13 (20 ng/mL) (PeproTech, 200–13) for a period of 18 h. Subsequently, examination of cell morphology was carried out, and the profiling of cell surface markers was executed (Zhang et al. 2021).

For the co-culture of MΦ and TNBC cells, cancerous cells (1 × 105) were placed in the top compartment of Falcon* Cell Culture Inserts (Corning, Corning, NY), whereas THP-1 monocytes (1 × 106) that had undergone prior treatment were concurrently cultured in the lower compartment. The assessment of invasion was performed by seeding pre-treated THP-1 cells in the lower section of Cell Invasion Inserts (Corning) with RPMI medium containing 10% FBS or a control solution, followed by the analysis of MDA-MB-231 cell invasion (2 × 105) in the upper section after 24 h (Weng et al. 2019).

As per experimental requirements, the cell groups were categorized as follows (NP treatment time and concentration were determined based on cell uptake and toxicity experiments): MΦ group (MΦ), M2 group (M2 MΦ), Control group (M2 cells treated with PBS), M2pep-Cs-Cur NPs group (M2 cells treated with 5 μM M2pep-Cs-Cur NPs for 24 h), oe-NC group (M2 cells as the control group for COX-2 overexpression), oe-COX-2 group (M2 cells with COX-2 overexpression), NPs + oe-NC group (oe-NC cells treated with 5 μM M2pep-Cs-Cur NPs for 24 h), NPs + oe-COX-2 group (COX-2 cells treated with 5 μM M2pep-Cs-Cur NPs for 24 h).

Immunofluorescence experiment

The cells underwent cultivation in glass-bottomed 35 mm cell culture plates (706001, NEST). Immobilization of cellular units was accomplished at RT using 4% paraformaldehyde for 15 min, then subjected to dual PBS washes and permeabilized with 0.5% Triton X-100 (P0096, Beyotime) for a duration of 10 min. The cells then experienced an overnight incubation at 4 °C with primary antibodies: rabbit anti-CD86 (ab239075, 1:100, Abcam, UK) and mouse anti-CD163 (ab156769, 1:100, Abcam, UK). Upon completion of three PBS washes, the sections were exposed to secondary antibodies tagged with Alexa Fluor 488/594 (ab150077/ab150116, 1:200, Abcam, UK) for one hour. Upon completion of three more PBS washes, the specimens underwent staining with DAPI (10 μg/mL, D3571, Thermo Fisher, USA) for a time span of 10 min at RT. Laser scanning confocal microscopy (Leica, CH) was utilized for the visualization process. ImageJ tool was applied in the analysis of fluorescence (Chen et al. 2022a, b, c). Experimental Groups for M2 MΦ Induction: Control group: PBS-treated M2 MΦ; M2pep-Cs-Cur NPs group: M2 MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h; NPs + oe-NC group: oe-NC MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h; NPs + oe-COX-2 group: oe-COX-2 MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h. The study was replicated in triplicate, and the data derived from the two categories were evaluated through an independent samples t-test.

Cell transfection and cell grouping

Implementing control plasmid oe-NC, target plasmid oe-COX-2 (constructed in pCDH vector), auxiliary plasmid pMD2.G (12259, Addgene), and psPAX2 (12260, Addgene), the lentivirus packaging kit (V48820, Invitrogen, USA) was employed for the transfection of HEK-293 T cells. Post 48 h of transfection completion, the supernatant underwent collection and concentration employing a lentivirus enrichment reagent (631231, Takara), and the concentrated lentivirions were preserved at −80 °C. Upon reaching the logarithmic growth period, THP-1 cells were subjected to trypsin for cell disintegration, and 1 × 105 cells were distributed in each well within a 6-well plate. After incubating the cells for 24 h, and with cell density reaching around 75%, an appropriate dose of lentiviral vectors (initial experiments displayed diverse MOIs for cellular transduction, ultimately validating MOI = 10 at a concentration of roughly 5 × 106 TU/mL) along with 5 μg/mL of polybrene (Merck, TR-1003) were supplemented to the growth medium for transfection. Cell choosing occurred 2 days post lentiviral transduction, with puromycin (540222, Sigma-Aldrich) at a concentration of 10 μg/mL, and the cells with stable transfection were maintained for at least one week (Zhou et al. 2020). Plasmid construction was performed by Shanghai Hanheng Biotechnology Co., Ltd.

RT-qPCR

Trizol reagent (Cat. No. 16096020, Thermo Fisher Scientific, USA) was utilized for the extraction of total RNA from tissues and cells. Utilizing the NanoDrop One/OneC microvolume nucleic acid and protein analyzer by Thermo Scientific, the RNA concentration and quality were inspected, achieving an A260/A280 ratio of 2.0 and a concentration exceeding 5 μg/μL. By employing the first-strand cDNA synthesis kit (D7168L, Beyotime, Shanghai), the RNA was transcribed in reverse and converted into cDNA. Utilizing the RT-qPCR kit (Q511-02, Vazyme Biotech, Nanjing, China), the RT-qPCR investigations were executed according to the manufacturer's instructions. 20 μL of the reaction mixture comprised 2 μL of cDNA template, 0.2 μL of both forward and reverse primers, and 10 μL of RT-qPCR Mix, mixed with RNase-free water. Implementing the Bio-Rad CFX96 real-time PCR system, the PCR amplification process took place under the prescribed settings: initial denaturation at 95 ℃ for 30 s, followed by 40 cycles of denaturation at 95 ℃ for 10 s, annealing at 60 ℃ for 30 s, extension at 72 ℃ for 30 s, and a melting curve analysis from 65 ℃ to 95 ℃. Shengong Biotech (Shanghai, China) furnished and formulated the primer sequences, which can be found in Table S1. The 2−ΔΔCt method was employed to analyze the relative expression of the target gene in the experimental group as opposed to the control group, wherein GAPDH mRNA acted as the reference gene. The formula used was ΔΔCt = ΔCt test – ΔCt control, where ΔCt = Ct target – Ct reference (Zhang et al. 2019). Experimental Groups: Control group: PBS-treated M2 MΦ; M2pep-Cs-Cur NPs group: M2 MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h; oe-NC group: M2 MΦ with COX-2 overexpression control plasmid; oe-COX-2 group: M2 MΦ with COX-2 overexpression plasmid; NPs + oe-NC group: oe-NC MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h; NPs + oe-COX-2 group: oe-COX-2 MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h. Repetition occurred thrice during the experiment iterations. Two groups were subjected to an independent samples t-test, and multiple groups underwent a one-way ANOVA.

Western blot

Lysis of the total proteins derived from tissues and cells was conducted in RIPA lysis buffer (P0013B, Beyotime, Shanghai) supplemented with 1% PMSF. Quantification of total protein content in every individual sample was conducted through the employment of a BCA assay kit (P0011, Beyotime, Shanghai). With the aim of facilitating electrophoresis separation, SDS gels ranging from 8 to 12% were meticulously prepared to align with the desired sizes of the target protein bands, and loading of protein samples into each lane was uniformly executed with the aid of a microloader pipette. Transferring the proteins from the gel to a PVDF membrane (1,620,177, BIO-RAD, USA) was followed by blocking them with 5% skim milk for one hour at RT. An overnight incubation at 4 °C was conducted following the addition of primary antibodies (Refer to Table S2). Following this, the membrane underwent three washes using 1 × TBST at RT, lasting for 5 min per wash. Post the initial incubation process, the membrane was exposed to an HRP-conjugated goat anti-rabbit IgG secondary antibody (ab6721, 1:2000). This exposure lasted for 1 h at RT and the antibodies utilized were obtained from respected suppliers such as Abcam and Cell Signaling Technology. Following three rounds of membrane washing in 1 × TBST buffer for 5 min each at RT, ECL substrate (1705062, Bio-Rad, USA) was introduced, enabling the visualization of protein bands via an Image Quant LAS 4000C gel imaging system (GE Healthcare, USA). The internal referencing of GAPDH facilitated the measurement of total protein levels within cells. Grayscale value analysis was utilized with Image J software to determine the relative expression levels of proteins, where the relationship between the target band's grayscale value and the internal reference band's grayscale value indicated the protein expression levels (Wu et al. 2022). Experimental Groups; Control group: PBS-treated M2 MΦ; M2pep-Cs-Cur NPs group: M2 MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h; oe-NC group: M2 MΦ with COX-2 overexpression control plasmid; oe-COX-2 group: M2 MΦ with COX-2 overexpression plasmid; NPs + oe-NC group: oe-NC MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h; NPs + oe-COX-2 group: oe-COX-2 MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h. The repetition was done threefold to validate the results. An independent t-test was executed comparing two groups, with a one-way ANOVA conducted to analyze multiple groups.

Cell growth evaluation utilizing CCK-8 assay

Post cell dissociation and rehydration, the cells were portioned into each well of a 96-well plate at a density of 1 × 103 cells, where they underwent overnight culturing. Cell viability was examined at 0 h, 24 h, 48 h, and 72 h post-culturing based on the instructions provided by the manufacturer of the CCK-8 kit (C0041, Beyotime, Shanghai, China) using the CCK-8 method. The CCK-8 detection solution (10 μL) was introduced cyclically at specified time intervals, undergoing a 1-h incubation in a CO2 incubator, and then the quantification of optical density at 450 nm was performed with the assistance of a microplate reader (E8051, Promega) (Zhang et al. 2023a, b, c, d). Experimental Groups for Induced M2 MΦ: Control group: PBS-treated M2 MΦ; M2pep-Cs-Cur NPs group: M2 MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h; NPs + oe-NC group: oe-NC MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h; NPs + oe-COX-2 group: oe-COX-2 MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h. Each experiment was conducted in triplicate for statistical reliability. Examination of data involved a two-way ANOVA at disparate time instances.

Detection of cell proliferation rate using EdU labeling

Allocation of three replicates per group was done for cell seeding in 24-well plates. The culture medium received an addition of EdU (C0071S, Beyotime) at a concentration of 10 µmol/L, following which the cells were placed in a CO₂ incubator for 2 h. Upon removal of the culture medium, the cells were immobilized by exposure to a PBS solution containing 4% paraformaldehyde for 15 min at normal RT, followed by permeabilization with 0.5% Triton-100 (HFH10, Invitrogen™, USA) in PBS for 20 min also at RT. Post this step, each well received 100 µL of EdU staining solution and was stored in a light-free environment at regular RT for 30 min. A 5-min incubation period was allotted for the application of DAPI stain (Shanghai Biyun Tian Biotechnology Co., Ltd, Catalog C1002) to specifically target and label the cellular nuclei. Following mounting, a fluorescence microscope (FV-1000/ES, Olympus, Japan) was utilized to observe 6–10 random fields of view, and the quantity of cells showing positivity in each field was documented. Here is how the EdU labeling rate was computed: EdU labeling rate (%) = (number of positive cells) / (number of positive cells + number of negative cells) × 100% (Shi et al. 2022). Experimental Groups for Induced M2 MΦ: Control group: PBS-treated M2 MΦ; M2pep-Cs-Cur NPs group: M2 MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h; NPs + oe-NC group: oe-NC MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h; NPs + oe-COX-2 group: oe-COX-2 MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h. The trial was reiterated thrice to ensure consistency. The independent t-test method was applied to compare two sets, whereas for multiple sets, a one-way ANOVA analysis was utilized.

Clonogenic assay

In the clonogenic test, the cells designated for examination were dissociated to form individual cell suspensions, with 1000 cells inoculated into every 6 cm petri dish. Scheduled alterations of the complete growth medium were made in each culture dish every three days to ensure optimal growth conditions. The cells were subjected to PBS washing, 4% paraformaldehyde fixation, and a 15-min staining process with 0.4% crystal violet (C0121, Beyotime) post a 10-day incubation period. Manual counting was employed for colonies comprising more than 10 cells, and the average count was obtained from replicate wells for statistical analysis (Zhao et al. 2020). Experimental Groups: Control group: PBS-treated M2 MΦ; M2pep-Cs-Cur NPs group: M2 MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h; NPs + oe-NC group: oe-NC MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h; NPs + oe-COX-2 group: oe-COX-2 MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h. Every trial was duplicated three times. Two groups were subjected to an independent t-test, and in the case of multiple groups, an ANOVA test was utilized.

Transwell invasion assay

The evaluation of in vitro cell invasion was accomplished through the utilization of Transwell chambers (3422, 8 μm pore size, Corning, USA) within a 24-well plate in the performance of the Transwell invasion assay. A culture medium enriched with FBS was introduced into the lower segment of the Transwell unit, which featured a polycarbonate membrane with an 8 μm pore size pre-coated with Matrigel, amounting to 600 mL. Subsequent to incubation at 37 °C for 1 h, the system reached equilibrium. Cells, post enzymatic breakdown, were restored in DMEM medium lacking FBS, and were introduced into the upper compartment with a concentration of 2 × 104 cells/mL, thereafter they were subjected to incubation at 37 °C alongside 5% CO2 for a duration of 24 h. Upon completion of the experiment within the Transwell chambers, removal of these chambers was carried out, followed by a double rinse with PBS, and later fixed using 5% glutaraldehyde at a temperature of 4 °C. Subsequent to this process, a 0.1% crystal violet staining persisted for 5 min duration, accompanied by the double washing with PBS, and the removal of surface cells was executed by the application of a cotton ball. To facilitate observation, the chambers were inverted and viewed under a fluorescence microscope (Nikon TE2000, China). 5 random fields were photographed, and the mean number of cells that infiltrated the chamber was subsequently computed (Li et al. 2022a, b, c, d). Experimental Groups: Control group: PBS-treated M2 MΦ; M2pep-Cs-Cur NPs group: M2 MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h; NPs + oe-NC group: oe-NC MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h; NPs + oe-COX-2 group: oe-COX-2 MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h. Three repetitions were performed for every individual experiment. The utilization of an independent samples t-test allowed for the comparison of two separate groups, while a one-way ANOVA was utilized to analyze data from multiple groups.

Wound healing assay

Cells from each group in good growth condition were prepared into single-cell suspensions using the previously described method. A six-well plate received a 2 mL cell suspension (4 × 105 cells/mL) after the cell counting process. Cultivation in a CO₂ incubator at 37 °C continued until the cells reached a growth density of 90%−100%. The cell monolayer was subjected to a perpendicular mark using a 200 μL pipette tip, resembling a "scratch". A serum-free medium was poured into the plate after it underwent two PBS washes to eliminate the suspended cells. After a subsequent 24-h incubation, images were captured, and the scratch distance was measured using Image J software. The calculation method for the relative migration speed is as stated below: relative migration rate = (migration distance of experimental group cells/migration distance of control group cells) × 100% (Jiang 2011). Experimental Groups: Control group: PBS-treated M2 MΦ; M2pep-Cs-Cur NPs group: M2 MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h; NPs + oe-NC group: oe-NC MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h; NPs + oe-COX-2 group: oe-COX-2 MΦ treated with 5 μM M2pep-Cs-Cur NPs for 24 h. The experiment was iterated threefold to validate the findings thoroughly. Two groups were assessed with an independent t-test, whereas multiple groups were analyzed using one-way ANOVA.

Establishment of the TNBC xenograft mouse model

Procured from Beijing Vital River Laboratory Animal Technology Co., Ltd., China, twenty-four female C57BL/6 J mice, aged 4–5 weeks, were accommodated in a controlled environment at 26–28 °C with a humidity level of 50–65%. Strict adherence to ethical standards was maintained in all animal experiments, as approved by the Institutional Animal Ethics Committee.

During the orthotopic transplantation process, a mixture was prepared containing 4T1 cells, 50% FBS/PBS, and Matrigel (1:1, v/v), to be introduced into the mammary fat pads of female nude mice aged 6 weeks, with each fat pad receiving 2 × 105 cells for the establishment of the 4T1 model. When the tumor reached a size of 10 mm, the animals were randomized into treatment groups. Based on previous studies (Larasati et al. 2018) and in vitro experimental results, dosages of M2pep-Cs-Cur NPs at 10 mg/kg were delivered via tail vein injection every two days, with the control group being treated with an equal volume of PBS. The intraperitoneal injection regimen received by both the Anti-PD-L1 (Bio X Cell, #BE0101) at 12.5 mg/kg and the control group with their specific isotype antibody (Bio X cell, #BE0090) was set at once every 3 days. After the final treatment, mouse serum samples were gathered, and the levels of CK, LDH, ALT, AST, CREA, and urea were gauged employing a fully automated biochemical analyzer (BK-400, Shandong Beike Biotechnology Co., Ltd.). Mice were euthanized by CO₂ inhalation 4 weeks after orthotopic injection, and tissue samples from mammary tumors and essential organs (heart, liver, spleen, lungs, kidneys) were obtained. Preservation methods included fixation in formalin with paraffin embedding or rapid freezing in liquid nitrogen and storage at −80 °C (Lee et al. 2022; Pérez-Núñez et al. 2022).

According to the experimental requirements, the animal grouping is as follows: Control group (PBS-treated TNBC mice), M2pep-Cs-Cur NPs treatment group (TNBC mice treated with M2pep-Cs-Cur NPs), M2pep-Cs-Cur NPs and anti-PD-L1 combination treatment group (TNBC mice treated with M2pep-Cs-Cur NPs and anti-PD-L1), with 6 mice in each group. For RNA-seq analysis: Control group and M2pep-Cs-Cur NPs treatment group, with 3 mice in each group.

Immunohistochemistry

Paraffin embedding was performed for the fixation of tumor tissues in formalin, followed by sectioning. Staining of the sections was conducted through the application of hematoxylin and eosin (H&E), involving a 2-min incubation in hematoxylin and a subsequent 1-min exposure to eosin. Subsequent to dehydration using 70%, 80%, 90%, and 95% ethanol, the sections were treated with xylene twice, with each treatment lasting for 5 min. They were then encapsulated in neutral resin for examination under a BX63 light microscope from Olympus, Japan. The immunohistochemical staining was performed following standard procedures: rabbit anti-Ki67 (ab16667, 1:200, Abcam) and rabbit anti-PCNA (ab92552, 1:500, Abcam) served as the primary antibodies. These antibodies were incubated overnight, followed by a 30-min incubation with goat anti-rabbit IgG (ab6721, 1:1000, Abcam) the succeeding day. The 37 °C incubated sections were treated with SABC (Strept Avidin–Biotin Complex, Vector Labs, USA) for 30 min, then subjected to staining with DAB chromogen solution (P0203, Beyotime Biotechnology, Shanghai) for 6 min before a quick 30-s hematoxylin counterstain. The specimens underwent a dehydration procedure involving a sequence of immersions in solutions of 70%, 80%, 90%, and 95% ethanol, followed by two rounds of exposure to xylene lasting 5 min each. They were subsequently encased in neutral resin for observation using a light microscope (BX63, Olympus, Japan). Measurement of the staining findings was accomplished via the use of Image J software to gauge both the average gray scale value (intensity of staining) and the quantity of cells exhibiting positive outcomes. Statistical examination was conducted to differentiate variances among distinct sets, and the graphical representation of findings facilitated the assessment of dye impacts on cellular or tissue slices (Zhang et al. 2023a, b, c, d). Employed an independent samples t-test to compare two groups, and conducted one-way ANOVA for analyses involving multiple groups.

Flow cytometry

Tumors obtained were subjected to the following analyses using flow cytometry. Following the manufacturer's instructions, the tumors underwent mincing and enzymatic digestion with the assistance of the MAcs Miltenyi Mouse Tumor Dissociation Kit (130–096–730, Miltenyi Biotec). The separated tumor cells were washed with RPMI Medium 1640, and utilizing the RBC Lysis solution (158904, Qiagen) enabled the lysis of red blood cells. Enumeration and suspension of cells were conducted in PBS solution supplemented with 0.5% BSA and 2 mM EDTA. M1-like and M2-like MΦ were labeled with Anti-CD80-PE (16-10A1, 553769, BD Biosciences), Anti-CD86-APC (GL1, 561964, BD Biosciences), Anti-CD206-FITC (C068C2, 141703, Biolegend), and Anti-CD163-APC (S15049I, 155305, Biolegend) individually (Mehta et al. 2020; Zhang et al. 2021). Two groups were subjected to an independent t-test comparison, and multiple groups underwent a one-way ANOVA analysis for statistical as