Combination Therapeutic Effect of Psoralen and Paclitaxel-Loaded Superparamagnetic Iron Oxide Nanoparticles for the Treatment of Triple-Negative Breast Cancer

Abstract

Purpose

To overcome dose-limiting toxicity and paclitaxel (PTX) resistance in triple-negative breast cancer (TNBC) therapy, as well as the asynchronous in vivo biodistribution of PTX and psoralen (PSO) combination regimens, we developed a highly biocompatible, magnetically targeted nano-co-delivery system.

Methods

Paclitaxel/psoralen-loaded magnetic composite nanoparticles (P/P-TCNs) were fabricated via nanoprecipitation. The nanocarriers comprise a poly(lactic-co-glycolic acid) (PLGA) core co-encapsulating both therapeutic agents and superparamagnetic iron oxide nanoparticles (SPIONs), shielded by a stabilizing outer phospholipid layer.

Results

Physiochemically, P/P-TCNs exhibited uniform nanoscale dimensions, a negative surface charge, robust superparamagnetism, and synchronized drug release profiles. In vitro, P/P-TCNs significantly enhanced intracellular drug accumulation in 4T1 murine breast cancer cells, inducing profound G2/M cell cycle arrest and augmented apoptotic response. In vivo live imaging of tumor-bearing mice revealed that under an external magnetic field, P/P-TCNs selectively and efficiently accumulated at the tumor site via tail vein injection. This magnetically guided co-delivery achieved a potent tumor inhibition rate of >85%, significantly outperforming both the free drug combination and non-targeted nanoparticle controls. Furthermore, the nanoplatform demonstrated excellent hemocompatibility (<5% hemolysis) and negligible systemic toxicity in major organs.

Conclusion

P/P-TCNs facilitate the synchronized, safe, and highly efficient co-delivery of PTX and PSO. By maximizing nanocarrier‑facilitated combinatorial anti-TNBC efficacy while minimizing off-target toxicity, this biocompatible magnetic nanoplatform represents a promising translational strategy for the clinical management of refractory TNBC.

Graphical Abstract

Introduction

TNBC is a subtype of breast cancer with a high tendency to metastasize and a high recurrence rate. Its aggressiveness and inherent resistance to conventional therapies lead to a higher mortality rate than other breast cancer subtypes. TNBC exhibits high intertumor heterogeneity and a strong tendency for early distant metastasis and recurrence.^1,2 Furthermore, significant intratumoral heterogeneity in TNBC is a key factor contributing to treatment resistance and disease progression, resulting in poor efficacy of conventional therapies. These clinical and biological characteristics together constitute a significant challenge in the treatment of TNBC, making it one of the breast cancer subtypes with the worst prognosis.³

TNBC lacks specific molecular targets, so traditional chemotherapy remains its main treatment. However, its poor tumor control and unavoidable off-target effects significantly limit its clinical application. Paclitaxel (PTX) is a commonly used chemotherapy drug for TNBC, which enhances anti-tumor efficacy by inhibiting cell mitosis. However, its systemic toxicity, acquired resistance, and poor water solubility hinder its widespread clinical application. Therefore, there is an urgent need to develop strategies to improve its therapeutic index.⁴ Combination therapy can exert enhanced combinatorial anti-tumor effects by targeting distinct tumor pathways, thereby reducing the required dosage of natural phytochemicals and mitigating toxic side effects.⁵ Increasing evidence suggests that combining natural furanocoumarin PSO with PTX can produce enhanced antitumor outcomes by compensating for the shortcomings of each single agent.⁶ Paclitaxel (PTX) has attracted much attention due to its multifaceted anticancer properties. These properties include regulating tumor-related signaling pathways, inducing cell cycle arrest, and promoting apoptosis. Although the combination therapy of PTX and PSO has these advantages, their pharmacokinetic characteristics are significantly different, and their tumor targeting is insufficient.^7–10 These issues ultimately lead to inadequate drug accumulation at the tumor site and compromised combinatorial therapeutic outcomes.

Nanoparticle-based co-delivery systems have become a viable approach toaddressing these issues. Through the enhanced permeability and retention effect (EPR), these systems can enhance pharmacokinetic behavior, encapsulate multiple medications simultaneously, and increase tumor accumulation.^11,12 Among them, polymer-lipid hybrid nanoparticles combine the advantages of polymeric nanoparticles and liposomal biomimicry, allowing sustained drug release and providing a stable and flexible delivery vehicle.^13,14 Notably, superparamagnetic iron oxide nanoparticles (SPIONs) possess unique advantages over conventional nanocarriers: they allow magnetically targeted drug delivery; facilitate local drug release and strengthen antitumor activity via magnetothermal effects; possess tunable sizes and easily modifiable surfaces for versatile drug loading; and also display favorable biocompatibility and biodegradability.^15–17 SPIONs serve as multifunctional agents for the integrated diagnosis and treatment of breast cancer, owing to their excellent superparamagnetic properties, good biocompatibility, and biodegradability. Through conjugation with chemotherapeutic drugs, biomolecules, and antibodies, these nanoparticles can be functionalized to enable precise, targeted delivery to triple-negative breast cancer (TNBC) sites under the guidance of an external magnetic field. This targeted delivery strategy significantly enhances local drug accumulation and anti-tumor efficacy while effectively reducing systemic toxicity to normal tissues.¹⁸

We previously developed a phospholipid-coated SPION/PLGA nanoplatform (P/P-TCNs) via an oil-in-water emulsion, which demonstrated enhanced combinatorial therapeutic efficacy in delivering psoralen (PSO) and paclitaxel (PTX) against human MDA-MB-231 triple-negative breast cancer (TNBC) cells.¹⁹ However, the efficacy of this formulation across diverse TNBC subtypes, as well as its long-term structural stability, remained unexplored. Building upon our established synthetic protocol, the current study investigates the broader applicability of P/P-TCNs. This core-shell system comprising an aqueous lipid shell (soybean phospholipids/DSPE-PEG2000) and an organic polymeric core (PLGA, drugs, and SPIONs) leverages both passive accumulation via the enhanced permeability and retention (EPR) effect and active magnetic targeting for synchronized drug release. To validate its broad-spectrum potential, we evaluated the anti-tumor activity of P/P-TCNs in murine 4T1 TNBC cells. Furthermore, we incorporated X-ray powder diffraction (XRD) analysis and 7-day colloidal stability monitoring to systematically confirm the nanocarrier’s structural integrity. These comprehensive characterizations underscore the robust potential of this magnetic co-delivery system for combined TNBC chemotherapy.

Materials and Methods

Materials

PSO and PTX were acquired from Chengdu Refensi (Chengdu, China). Superparamagnetic iron oxide nanoparticles (SPIONs) were obtained from Sichuan Vicky Co., Ltd. (Chengdu, China). Sodium oleate and soybean lecithin were supplied by Macklin (Shanghai, China). Poloxamer F-127 (PF-127) was sourced from Shanghai Yuanye (Shanghai, China). Saline solution (0.9% NaCl) was provided by Qidu Pharmaceutical Co., Ltd. (China). Cell Cycle Assay Kit and Annexin V-FITC/PI Apoptosis Kit were procured from Elabscience (Wuhan, China). RPMI Medium 1640 basic was supplied by Thermo Fisher Scientific (Suzhou, China), and Penicillin-Streptomycin (10,000 U/mL) was purchased from Life Technologies Limited (USA).

Experimental Animals and Cell Lines

Female BALB/c nude mice (18–20 g) were acquired from the Guangdong Laboratory Animals Monitoring Institute and used for all in vivo studies. The Jinan University Animal Ethics Committee approved all experimental protocols (No. IACUC-20251017-04). Animal experiments were conducted in accordance with the guiding principles for animal experimentation set forth by the Medical Ethics Committee of the Chinese national health authorities and the ARRIVE guidelines. Upon completion of the experiments, the mice were deeply anesthetized with isoflurane and subsequently humanely euthanized via cervical dislocation. All euthanasia procedures were performed in accordance with the AVMA Guidelines for the Euthanasia of Animals. Guangzhou Cellcook Biotech Co., Ltd. supplied the murine TNBC cell line 4T1. At 37°C and 5% CO₂, cells were raised in RPMI-1640 media mixed with 10% FBS.

Nanoparticle Preparation

Preparation of Paclitaxel and Psoralen Co-Loaded Targeted Composite Nanoparticles (P/P-TCNs)

The magnetic composite nanoparticles (P/P-TCNs) were fabricated using the identical oil-in-water emulsion protocol reported in our prior publication.¹⁹ Initially, 1 g of superparamagnetic iron oxide nanoparticles (SPIONs) was dispersed in 25 mL of a 2% (w/v) sodium oleate (OA) aqueous solution and whisked for one hour at 60 °C. Following washing and drying, OA-modified magnetic nanoparticles (SPIONs@OA) were obtained. The organic phase was then created by dissolving 4.5 mg of PSO, 4.5 mg of PTX, 110 mg of PLGA, and 90 mg of SPIONs@OA in 1.6 mL of chloroform. To form the outer lipid monolayer, 83 mg of soybean lecithin and 17 mg of 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy (polyethylene glycol) −2000] (DSPE-PEG2000) were dissolved in ethanol. This lipid mixture was then added to ultrapure water preheated to 60 °C under stirring to form the aqueous phase. Finally, the organic phase containing the drug and SPIONs@OA was slow-dripped into the aqueous phase maintained at 60°C under constant mixing for one hour, ultimately yielding P/P-TCNs.

Preparation of Paclitaxel and Psoralen Co-Loaded Polymer-Lipid Nanoparticles (P/P-PLNs)

P/P-PLNs were made using a process akin to that outlined for P/P-TCNs, with the key modification that no SPIONs@OA was added to the PLGA-containing organic phase. The aqueous phase was prepared identically. The two phases were then mixed and stirred under the same conditions to obtain P/P-PLNs.

Preparation of Paclitaxel and Psoralen Co-Loaded Iron Oxide-Polymer Nanoparticles (P/P-IOPNs)

First, the identical procedure as outlined in Section 2.3.1 was used to prepare SPIONs@OA. Next, 200 mL of an aqueous solution containing 2% PF-127 was added. For ten hours, the mixture was agitated in an inert environment. Magnetic nanoparticles were then collected using a magnet, washed three times with deoxygenated water, and redispersed in fresh deoxygenated water. A chloroform solution containing 4.5 mg of PSO and 4.5 mg of PTX was introduced into this aqueous dispersion. The mixture was agitated at 30 °C for 8 hours to facilitate solvent evaporation and drug loading, resulting in the formation of P/P-IOPNs. For comparative studies, the structurally analogous P/P-PLNs (lacking the magnetic component) and P/P-IOPNs (lacking the outer lipid-PLGA hybrid shell) served as control formulations.

To facilitate in vitro and in vivo distribution studies, fluorescently labeled versions of all three nanoparticle types were prepared. Coumarin 6 (C6) or indocyanine green (ICG) was loaded into the nanoparticles during their synthesis by incorporating the fluorophore into the respective organic phases following the preparation protocols described above.

Nanoparticle Characterisation

The zeta potential and particle size of P/P-PLNs, P/P-IOPNs, and P/P-TCNs were assessed using a laser diffraction analyzer (Mastersizer 3000, Malvern, UK). The morphology of the nanoparticles was assessed by transmission electron microscopy (TECNAI G2 Spirit, FEI, USA). Compared with our previous work (Du et al, 2026), two supplementary characterization assays were incorporated in the present study: X-ray powder diffraction (XRD) was performed to identify the crystalline phase of encapsulated SPIONs; additionally, we continuously monitored hydrodynamic size and PDI over 7 days to evaluate the long-term colloidal stability of P/P-TCNs under simulated physiological conditions.¹⁹ X-ray powder diffraction (XRD) patterns of the physical mixture (PSO+PTX) and P/P-TCNs were recorded on a diffractometer (D8 Advance, Bruker AXS GmbH, Germany) with a scanning range of 5° to 90° (2θ). The colloidal stability of P/P-TCNs was evaluated by monitoring the changes in their average hydrodynamic diameter and polydispersity index (PDI) over a period of seven days. The magnetic properties of SPIONs, P/P-IOPNs, and P/P-TCNs were characterized using a vibrating sample magnetometer (VSM) at 300 K. Hysteresis loops were obtained by applying a magnetic field ranging from –30 to +30 kOe, and the corresponding saturation magnetization (Ms) values were determined.

Cellular Uptake Study

Following the preparation of C6-labeled nanoparticles (C6-PLNs, C6-IOPNs, and C6-TCNs), an initial solution of C6 was obtained in dimethyl sulfoxide (DMSO) and further diluted with phosphate-buffered saline (PBS) to produce C6, C6-PLNs, C6-IOPNs, and C6-TCNs solutions at a final C6 concentration of 200 ng/mL. The C6 concentration in each formulation was determined using a microplate reader (MAGO4, Erba, Italy). For confocal imaging, 4T1 cells were seeded at 2×10⁵ cells per dish. Cells were incubated with C6, C6-PLNs, C6-IOPNs, or C6-TCNs for 2 or 4 hours, fixed with 4% paraformaldehyde, and stained with 4′,6-diamino-2-phenylindole (DAPI) for 10 minutes. Cell uptake was observed using a confocal microscope (LSM800, Zeiss, Germany). Quantitative analysis of 4T1 cells was performed by flow cytometry: cells were incubated with a C6-containing formulation, digested with trypsin, collected, and washed. Cell fluorescence intensity at 488 nm was then quantitatively analyzed using a flow cytometer (BD FACSCanto, Becton Dickinson, China).

Cell Viability Assay

4T1 cells were seeded in 96-well plates and cultured overnight at 37°C. Subsequently, the cells were exposed to different concentrations of drug-loaded nanoparticles or free drug for 24 hours. Afterward, CCK-8 reagent was added to each well, and incubation continued for 4 hours. Optical density (OD) values were measured at 450 nm using a microplate reader (MAGO4, Erba, Italy).

Cell Cycle Analysis

4T1 cells were seeded in 6-well plates at a density of 1×10⁵ cells per well and allowed to grow for 24 hours. The cells were then treated with different drug formulations overnight. After treatment, the cells were harvested, washed with PBS, and fixed in ice-cold 70% ethanol at 4°C for at least 2 hours. Following a second wash and RNase A digestion, the fixed cells were incubated at 37°C for 30 minutes. Propidium iodide (PI) staining solution was added, and the cells were incubated on ice in the dark for another 30 minutes. Cell cycle distribution was analyzed via flow cytometry by detecting red fluorescence at an excitation wavelength of 488 nm.

Apoptosis Assay

4T1 cells (1 × 10⁵ cells per well) were treated with various drug formulations for 12 hours. After centrifuging the cells for five minutes at 300 × g, they were resuspended in 500 µL of Annexin V Binding Buffer after being rinsed with PBS. The cell suspension was stained for 20 minutes at room temperature in the dark using 5 µL of Annexin V-FITC and PI (50 µg/mL). A flow cytometer was used to measure cell apoptosis right away.

Hemolysis Assay

Blood was collected from healthy BALB/c nude mice, and red blood cells (RBCs) were isolated by centrifugation. The RBCs were diluted and incubated with blank nanoparticles (PLNs, IOPNs, TCNs), drug-loaded nanoparticles (P/P-PLNs, P/P-IOPNs), or various concentrations of P/P-TCNs (75, 150, 300, 600, and 1200 μg/mL). Normal saline (NS) and Triton X-100 (1%, v/v) served as the positive and negative controls, respectively. The samples were centrifuged after incubation, and hemolysis was assessed by using a microplate reader (MAGO4, Erba, Italy) to measure the absorbance of the supernatant at 540 nm.

In vivo Antitumor Efficacy in a 4T1 Tumour Model

To establish a triple-negative breast cancer model, 4×10⁶ 4T1 cells were subcutaneously injected into the right flank of each female BALB/c nude mouse. Once the tumor volume reached approximately 100 mm³, the mice were randomly assigned to 12 treatment groups. The groups were as follows: (1) Normal saline (NS); (2) Free PSO (3 mg/kg); (3) Free PTX (3 mg/kg); (4) PSO + PTX (3 mg/kg each); (5) P/P-PLNs (equivalent to 3 mg/kg PSO); (6) P/P-IOPNs (equivalent to 3 mg/kg PSO); (7) P/P-TCNs-Low (1.5 mg/kg PSO); (8) P/P-TCNs-Medium (3 mg/kg PSO); (9) P/P-TCNs-High (6 mg/kg PSO). Intravenous tail injections were administered every three days, followed by application of a magnetic field. Following regular monitoring, mice were deeply anesthetized with isoflurane and then humanely euthanized by cervical dislocation. Tumours were collected for weighing, and tumour inhibition rates were calculated. Tumours and organs underwent fixation, embedding, sectioning, and haematoxylin-eosin (H&E) staining for histological assessment.

In vivo Biodistribution

ICG-labeled PLNs, IOPNs, and TCNs were given intravenously to nude mice with 4T1 breast tumor models to examine the nanoparticles’ in vivo distribution. Tumor tissue was obtained for ex vivo imaging after injection, and fluorescence signal imaging was carried out at different intervals utilizing the PerkinElmer in vivo imaging system (IVIS Lumina Series III, USA, PerkinElmer).

Statistical Analysis

The mean ± standard deviation (SD) is used to display any quantitative data. GraphPad Prism software (version 8.0) was used for statistical analysis. One-way analysis of variance (one-way ANOVA) with Tukey’s multiple comparisons test was used to examine comparisons between several groups. Statistical significance was defined as a P-value of less than 0.05.

Results

Preparation and Characterisation of Nanoparticles

Figure 1A illustrates the preparation of SPIONs@OA from SPIONs and OA. The magnetically targeted composite nanoparticles co-loading Paclitaxel and Psoralen (P/P-TCNs) were fabricated using a nanoprecipitation method.²⁰ In this process, the obtained SPIONs@OA, PLGA, PSO, and PTX were dissolved in chloroform and subsequently introduced into a continuously stirred aqueous solution of phospholipids. As shown in Figure 1C and D, the prepared P/P-TCNs exhibited a particle size of 119.80 ± 2.59 nm, a PDI of 0.17 ± 0.04, and a zeta potential of −36.30 ± 0.84 mV. To evaluate the impact of magnetic properties and the phospholipid layer on subsequent anti-breast cancer efficacy, control nanoparticles were prepared (Figure 1B). Using the same method but omitting SPIONs@OA, P/P-PLNs were synthesized, with a particle size of 143.60 ± 0.15 nm, a PDI of 0.15 ± 0.02, and a zeta potential of −12.50 ± 0.60 mV (Figure 1C and D). Similarly, phospholipid-free P/P-IOPNs were prepared by encapsulating PSO and PTX with PF-127 and SPIONs@OA, yielding particles of 104.70 ± 1.57 nm in size, a PDI of 0.19 ± 0.02, and a zeta potential of −20.00 ± 0.46 mV (Figure 1C and D). All three nanoparticle types had a size below 150 nm and a PDI of less than 0.2—properties that facilitate enhanced permeation and accumulation, a key factor for inducing apoptosis in triple-negative breast cancer cells. The negative zeta potentials (below −10 mV) further support colloidal stability. Transmission electron microscopy (TEM) images showed that all three nanoparticle formulations displayed a uniform spherical morphology, with sizes consistent with those obtained from dynamic light scattering (DLS) measurements (Figure 2A).

Figure 1.

Preparation, particle size, and zeta potential of P/P-TCNs. (A) Schematic illustration of the preparation process for P/P-TCNs. (B) Physical appearance of P/P-PLNs, P/P-IOPNs, and P/P-TCNs. (C and D) Particle size and zeta potential of P/P-TCNs, P/P-PLNs, and P/P-IOPNs.

Figure 2.

Characterization of the nanoparticles. (A) Transmission electron microscopy (TEM) images of P/P-PLNs, P/P-IOPNs, and P/P-TCNs. Scale bar: 100 nm. (B) X-ray powder diffraction (XRD) patterns of PSO+PTX, P/P-PLNs, P/P-IOPNs, and P/P-TCNs. (C) Magnetization hysteresis loops of P/P-IOPNs and P/P-TCNs. (D and E) Stability of P/P-TCNs in PBS at different time points (n = 3).

XRD was used to evaluate the crystalline states of the physical mixture (PSO+PTX), P/P-PLNs, P/P-IOPNs, and P/P-TCNs (Figure 2B). Characteristic diffraction peaks of PTX at 9.6°, 15.5°, 16.5°, 19.0°, and 19.9° were absent in the patterns of P/P-PLNs, P/P-IOPNs, and P/P-TCNs. This confirms that both PSO and PTX existed in an amorphous state within the nanoparticles, demonstrating successful encapsulation and the protective role of the nano-formulations in maintaining the drugs’ non-crystalline form. The magnetic hysteresis loops of P/P-IOPNs and P/P-TCNs confirmed their superparamagnetic behavior (Figure 2C). At 300 K, the saturation magnetization (Ms) values were 10.05 emu/g and 4.89 emu/g, with corresponding remnant magnetization (Mr) values of 1.45 emu/g and 0.56 emu/g for P/P-IOPNs and P/P-TCNs, respectively. Furthermore, the hydrodynamic diameter and PDI of P/P-TCNs remained stable below 130 nm and 0.2, respectively, over a 7-day period, indicating good storage stability (Figure 2D and E).

P/P-TCNs Inhibit the Growth of 4T1 Cells

Cellular internalization efficiency is a critical determinant of nanomedicine efficacy, especially for TNBC cells with aggressive biological behavior that often evade conventional drug delivery. Fluorescence imaging and flow cytometry results (Figure 3A–C) showed that C6-labeled TCNs exhibited the highest cellular uptake among all tested formulations. After 4 hours of incubation, their mean fluorescence intensity (3784 ± 51.51) was nearly 13 times higher than that of free C6 (282 ± 2.64). This significant enhancement, superior to C6-PLNs (910 ± 24.68) and C6-IOPNs (1623 ± 29.26), highlights the unique advantages of TCN carriers in promoting intracellular drug delivery. The phospholipid components of TCNs may promote membrane fusion or endocytosis through interaction with the cell membrane, while their nanoscale size can achieve efficient cell penetration by enhancing permeability and retention effects.²¹ This enhanced cellular uptake directly addresses the key limitations of low bioavailability and insufficient intracellular accumulation of free PTX and PSO, laying the foundation for their enhanced cytotoxicity and carrier‑mediated combinatorial therapeutic effects.

Figure 3.

Evaluation of cellular uptake. (A) Uptake of C6-PLNs, C6-IOPNs, and C6-TCNs by 4T1 cells after 2 h and 4 h of incubation (scale bar: 20 μm). (B and C) Flow cytometric analysis of 4T1 cells after 4 h of incubation. Quantitative analysis of the fluorescence intensity in 4T1 cells treated with C6, C6-PLNs, C6-IOPNs, and C6-TCNs (n=3). Data are presented as mean ± SD. Statistical analysis was performed using one-way ANOVA with Tukey’s post-hoc test. Compared to the C6-TCNs group, ****P < 0.0001.

The IC50 values quantitatively demonstrated the enhanced cytotoxicity of P/P-TCNs (Figure 4A and B). The IC50 value of P/P-TCNs was 0.28 μg/mL, superior to the free drug combination (2.47 μg/mL), P/P-PLNs (0.62 μg/mL), and P/P-IOPNs (0.43 μg/mL), indicating that the TCN carrier significantly improved the antiproliferative effect of the PTX and PSO combination therapy. Notably, the blank nanocarriers (PLNs, IOPNs, TCNs) did not exhibit significant cytotoxicity (Figure 4C), confirming that the antiproliferative effect was entirely attributable to the loaded drug, rather than carrier-related toxicity. This finding is consistent with previously reported good biocompatibility of TCNs, further reinforcing their safety profile for in vivo application. For the TNBC subtype, which is highly resistant to conventional chemotherapy, the significant reduction in IC50 is clinically significant. It is possible to reduce the dosage of PTX and PSO, thereby maintaining or enhancing antitumor efficacy while mitigating systemic toxicity.

Figure 4.

Cytotoxicity and cell cycle analysis. (A) 4T1 cell viability after a 24-hour incubation period with various PTX, PSO+PTX, P/P-PLNs, P/P-IOPNs, and P/P-TCNs concentrations; (B) 4T1 cell viability after a 24-hour incubation period with various PSO concentrations; and (C) 4T1 cell viability after a 24-hour incubation period with various PLNs, IOPNs, and TCNs concentrations. (D and E) Cell cycle arrest was measured by flow cytometry in 4T1 cells treated with various drug formulations. Data are presented as mean ± SD (n=3). Statistical analysis was performed using one-way ANOVA with Tukey’s post-hoc test. P-TCNs/P-TCNs group, ****P < 0.0001, ***P < 0.001, *P < 0.05, ns P > 0.05.

Cell cycle distribution analysis revealed the mechanistic basis of the augmented combinatorial cytotoxicity of P/P-TCNs (Figure 4D and E). Consistent with known pharmacological modes of action, free PTX induces G2/M phase arrest by stabilizing microtubules and blocking mitosis, while free PSO causes S phase arrest by interfering with DNA synthesis.^22,23 All three nanoformulations and combinations of free drugs reproduced the PTX-dominated G2/M phase arrest trend, indicating that the nanocarriers did not alter the inherent cell cycle regulatory activity of the drugs. Instead, they facilitated the synchronized co‑delivery of PTX and PSO within the same tumor cell, allowing their complementary cell‑cycle‑targeting mechanisms to produce augmented antitumor activity. Importantly, P/P-TCNs exhibited the most prominent augmented combinatorial cell cycle regulatory effect, which is closely related to their superior cellular uptake efficiency. By simultaneously blocking G2/M and S phases, P/P-TCNs effectively covered the heterogeneous cell cycle states of TNBC cells, preventing cell escape that could potentially evade single-drug therapy. This augmented combinatorial cell‑cycle arrest is particularly important for TNBC, as its high intratumoral heterogeneity often leads to resistance when targeting a single cell cycle checkpoint. Furthermore, the minimal interference with the cell cycle by the blank nanocarrier further confirms the biocompatibility of TCNs.

Induction of apoptosis is a key endpoint in antitumor therapy. Flow cytometry results showed that P/P-TCNs effectively promoted apoptosis in 4T1 cells (Figure 5). The apoptosis rate of P/P-TCNs was as high as 42.9%, superior to the NS group (7.0%), the free drug combination group (24.2%), the P/P-PLNs group (36.2%), and the P/P-IOPNs group (41.6%). The significant increase in the apoptosis rate was directly attributed to enhanced cellular uptake and augmented combinatorial cell‑cycle regulation achieved by P/P‑TCNs. TCNs enabled synchronized intracellular co‑localized exposure to both agents, producing a stronger apoptotic response compared with single‑drug or non‑targeted co‑delivery systems. For TNBC characterized by impaired apoptosis pathways, the effective induction of apoptosis by P/P-TCNs is a key advantage, as it not only inhibits tumor cell proliferation but also eliminates existing cancer cells, thereby reducing the risk of recurrence and metastasis.

Figure 5.

Analysis of cell apoptosis. (A) Flow cytometric analysis of apoptosis in 4T1 cells exposed to different treatments. (B) Bar graph showing the percentage of apoptotic 4T1 cells. Data are presented as mean ± SD (n=3). Statistical analysis was performed using one-way ANOVA with Tukey’s post-hoc test. P-TCNs/P-TCNs group, ****P < 0.0001, *P < 0.05.

In vivo Biodistribution Study of P/P-TCNs

In vivo imaging techniques were used to analyze the tissue biodistribution of ICG-PLNs, ICG-IOPNs, and ICG-TCNs (Figure 6). ICG-TCNs accumulated at the tumor site within 8 hours after intravenous injection (Figure 6A). Compared to ICG-IOPNs, ICG-TCNs exhibited earlier targeting. Furthermore, the level of fluorescence at the tumor region was noticeably stronger for ICG-TCNs than for free ICG and ICG-PLNs. This enhanced tumor accumulation is primarily attributed to the combined effects of magnetic targeting and the EPR effect. Ex vivo imaging of excised tissues and organs further confirmed that the level of fluorescence in tumors was markedly higher than that in other major organs (Figure 6B).

Figure 6.

In vivo biodistribution. (A) Fluorescence images of BALB/c nude mice from different treatment groups at 2, 4, 8, 12, and 24 hours post-injection. Color bar indicates radiant efficiency (units: (B) Fluorescence distribution in harvested tumors and organs from BALB/c nude mice.

Therapeutic Efficacy of Systemically Administered P/P-TCNs in Breast Cancer

The treatment plan for BALB/c nude mice with 4T1 tumors is illustrated in Figure 7A, with intravenous drug administration commencing on day 6 after tumor inoculation. Periodic measurements revealed that tumors in the normal saline (NS) group grew the fastest, whereas tumor growth was most effectively suppressed in the high-dose P/P-TCNs (P/P-TCNs (H)) treatment group (Figure 7D and E, Supplementary Figure S1). At the study endpoint, the tumor weight in the P/P-TCNs (M) group was considerably lower than that in the PSO+PTX, P/P-PLNs, and P/P-IOPNs groups, measuring only 40.8%, 56.8%, and 62.1% of their respective weights (Figure 7C). Correspondingly, the tumor inhibition rate for PSO+PTX group (66.57 ± 3.90%) was noticeably lower than that for P/P-TCNs (M) (87.85 ± 1.02%) (P < 0.0001). The tumor inhibition rate increased progressively with the escalation of the P/P-TCNs dose (Figure 7F), indicating a potent, dose-dependent antitumor effect.

Figure 7.

In vivo antitumor efficacy and hemolysis assay of P/P-TCNs. (A) Schematic illustration of the experimental design for 4T1 tumor-bearing nude mice. (B) Changes in 4T1 tumor-bearing nude mice’s body weight over the course of treatment (n=6). (C) Tumor weights at the experimental endpoint. (D) Growth curves of tumors. (E) Images of each group’s removed tumors. (F) Tumor inhibition rates for different treatment groups. Data are presented as mean ± SD. Statistical analysis was performed using one-way ANOVA with Tukey’s post-hoc test. P-TCNs/P-TCNs (M) group: ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05. (G) Visual assessment of hemolysis after different treatments. (H) Quantified hemolysis rates induced by different nanoparticles (n=3). Data are presented as mean ± SD. Statistical analysis was performed using one-way ANOVA with Tukey’s post-hoc test. P-TCNs/P-TCNs 300 μg/mL group: ****P < 0.0001, **P < 0.01, ns P > 0.05.

Biosafety Evaluation of Systemically Administered P/P-TCNs

The biocompatibility of the nanoparticles (P/P-PLNs, P/P-IOPNs, and P/P-TCNs at various concentrations) was assessed by evaluating their hemocompatibility using an absorbance-based hemolysis assay. No significant red coloration was observed in the supernatants after co-incubation of nanoparticles with erythrocytes and centrifugation (Figure 7G). The highest tested concentration of P/P-TCNs (1200 μg/mL) did not induce significant hemolysis, with a hemolysis rate of only 4.61% ± 0.17% (Figure 7H). Both the drug-loaded nanocarrier and its blank control exhibited excellent blood compatibility, validating its suitability for intravenous administration. Furthermore, mice treated with P/P-TCNs maintained stable body weight and exhibited normal growth patterns throughout the treatment process (Figure 7B). No significant pathological damage was detected in the major organs of BALB/c nude mice in all treatment groups (Figure 8), further confirming its biocompatibility. These results consistently demonstrate the good in vivo biocompatibility of P/P-TCNs.

Figure 8.

Histological characteristics of the heart, liver, spleen, lungs, kidneys, and tumors from BALB/c nude mice following intravenous administration of different formulations. (Scale bar: 100 μm).

Discussion

TNBC is a highly aggressive and inherently drug-resistant clinical challenge. Although combining PSO and PTX can produce enhanced anti-tumor potential through complementary mechanisms such as modulating oncogenic signaling pathways and inducing cell cycle arrest, their clinical application is severely limited due to differences in their pharmacokinetic characteristics and insufficient tumor targeting. Furthermore, paclitaxel, as a classic cytotoxic drug, has been known to have dose-limiting toxicities such as myelosuppression, peripheral neurotoxicity, and allergic reactions.^24–26 Psoralen’s in vitro cytotoxicity exhibits a significant concentration-dependent effect.²⁷ To overcome these bottlenecks, we designed magnetic targeting composite nanoparticles loaded with PTX and PSO (P/P-TCNs). This design combines the advantages of a lipid-polymer hybrid matrix and SPIONs, thereby improving drug delivery and therapeutic efficacy while reducing toxic side effects.

Coating SPIONs with oleic acid introduces negatively charged groups, resulting in a magnetic localization effect that increases the accumulation of P/P-TCNs near the cell surface.¹⁵ An anionic phospholipid layer composed of DSPE-PEG2000 and soybean phospholipids also increases the distance between magnetic cores and reduces dipole coupling, further decreasing the overall magnetic response. However, under acidic conditions, PLGA and phospholipid groups undergo protonation, accelerating drug release.²⁸ Furthermore, the phospholipid shell enhances membrane affinity and biocompatibility and reduces clearance by the reticuloendothelial system (RES), thus prolonging blood circulation time.²⁸ The enhanced pharmacodynamics of P/P-TCNs is a combined effect of particle size, the magnetic targeting of SPIONs, and phospholipid shell-mediated membrane interactions. PTX stabilizes microtubules to induce G2/M arrest, while PSO triggers apoptosis via oncogenic signaling modulation. Co-delivery by P/P-TCNs enables both agents to enter the same tumor cell; PTX-mediated G2/M arrest sensitizes trapped tumor cells to PSO-induced apoptosis. Unlike free drug mixtures with mismatched pharmacokinetics, P/P-TCNs achieve synchronized intracellular drug exposure to fully unlock this complementary therapeutic effect.

Regarding biocompatibility, the metabolic fate of iron oxide cores is well-defined; internalized SPIONs undergo lysosomal degradation into labile iron pools, joining normal physiological iron homeostasis, or are cleared through hepatobiliary and renal routes.^15,29 The clinical viability of iron oxide systems is strongly reinforced by Ferumoxytol, an FDA-approved iron oxide nanoparticle whose established human safety profile bridges laboratory nanoplatforms with clinical practice. Although our study is currently bounded by the homogenous 4T1 model and the physical reach of external magnetic fields in deep lesions, the co-delivery of PTX and PSO provides compelling translational value. By equalizing the disparate kinetics of both agents, this platform addresses dose-limiting chemotoxicity while maximizing cooperative antitumor efficacy.

Although in vitro and in vivo experiments using the 4T1 cell line and xenograft models have yielded encouraging results, TNBC exhibits significant inter- and intra-tumor heterogeneity in a clinical setting.^1,30,31 Future research should validate the efficacy of P/P-TCNs in patient-derived xenograft models or genetically engineered mouse models, as these models better mimic the complexity of human TNBC. In addition, further investigation into the expression of iron uptake receptors and intratumoral iron levels would help characterize the iron metabolism perturbation triggered by SPION degradation. While apoptosis and G2/M arrest represent the primary therapeutic mechanism observed herein, subsequent studies could also evaluate whether non-apoptotic cell death pathways such as ferroptosis are involved and quantify their contribution to the overall cytotoxicity. Current research primarily focuses on the combinatorial therapeutic effects of PTX and PSO, but combining other therapeutic modalities (eg, immunomodulators or photothermal therapy) with P/P-TCNs may elicit a more comprehensive antitumor response. Magnetic targeting strategies rely on the application of external magnetic fields, which may limit their application in deep or metastatic tumors. Exploring the possibility of coupling targeting ligands (eg, integrin αvβ3 antagonists) to the surface of nanoparticles could enhance active targeting and improve precision by incorporating magnetic guidance.^32–35 Although short-term studies have observed good biocompatibility, the long-term toxicity and biodegradation kinetics of P/P-TCNs require further investigation to ensure their safety for clinical application. The pharmacokinetic and pharmacodynamic characteristics of P/P-TCNs need to be systematically evaluated in large animal models, which will help optimize dosing regimens and promote clinical translation.

In summary, the lipid-polymer magnetic hybrid P/P-TCN nanoplatform successfully overcomes the inherent defects of free PTX/PSO combined medication via magnetic active targeting, and EPR passive enrichment. Supported by independent 4T1 tumor model verification, this carrier exhibits robust anti-TNBC proliferative activity and favorable biocompatibility. Further follow-up research targeting the above limitations will continuously optimize this co-delivery system, laying a solid experimental foundation for developing safer and more effective combinatorial therapeutic regimens for TNBC patients.

Conclusions

This study successfully prepared and characterized magnetically targeted composite nanoparticles (P/P-TCNs) loaded with PTX and PSO. The P/P-TCNs system achieved the simultaneous intracellular co-delivery of PTX and PSO. PTX induced cell‑cycle arrest at the G2/M phase within the same tumor cell together with PSO, and this complementary cell‑cycle‑targeting action contributes to the augmented combinatorial antitumor effect achieved by P/P‑TCNs nanocarrier‑mediated co‑delivery. Compared with free drugs and traditional nanocarriers, P/P-TCNs effectively regulated cell cycle progression, significantly enhanced cytotoxicity and apoptosis induction in 4T1 cells, and improved antiproliferative activity. P/P-TCNs not only effectively inhibited tumor growth in mouse models but also exhibited good biocompatibility. These results provide valuable insights for advancing combination therapy strategies for TNBC and the development of novel nanomedicine platforms.

Abbreviations

TNBC, triple‑negative breast cancer; PTX, paclitaxel; PSO, psoralen; P/P‑TCNs, paclitaxel/psoralen‑loaded magnetic composite nanoparticles; PLGA, phospholipids and polylactic-co-glycolic acid copolymer; SPIONs, superparamagnetic iron oxide nanoparticles; EPR, enhanced permeability and retention effect; OA, sodium oleate; SPIONs@OA, OA-modified magnetic nanoparticles; DSPE-PEG2000, 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy (polyethylene glycol) −2000]; P/P-PLNs, Paclitaxel and Psoralen Co-loaded Polymer-Lipid Nanoparticles; P/P-IOPNs, Paclitaxel and Psoralen Co-loaded Iron Oxide-Polymer Nanoparticles; PF-127, Poloxamer F-127; C6, Coumarin 6; ICG, indocyanine green; XRD, X-ray powder diffraction; PSO+PTX, patterns of the physical mixture; PDI, polydispersity index; VSM, vibrating sample magnetometer; DMSO, dimethyl sulfoxide; PBS, phosphate-buffered saline; DAPI, 4′,6-diamino-2-phenylindole; PI, Propidium iodide; NS, Normal saline; H&E, haematoxylin-eosin; SD, standard deviation TEM, Transmission electron microscopy; DLS, dynamic light scattering; Ms, saturation magnetization; Mr, remnant magnetization.

Author Contributions

Manling Du: concept, design, data analysis, manuscript preparation, manuscript editing. Fengjie Liu: design, data analysis, manuscript preparation, manuscript editing. Panpan Wang: design, data analysis. Mulazim Hussain Asim: design. Qian Liu: design, investigation. Qianwen Li: literature search, investigation. Mengjie Xie: literature search, investigation. Kewei Zhang: literature search, investigation. Zhenjiang Yang: manuscript review, supervision. Yu Cai: manuscript review, supervision, funding acquisition. All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors report no conflicts of interest in this work.