Mucus and tumor penetrating paclitaxel micelles for potent local therapy of cervical cancer

Abstract

The therapeutic efficacy for cervical cancer treatment is limited by insufficient drug accumulation and penetration due to physiological barriers such as mucin-rich environments after systemic administration. Thus, developing a local drug delivery system is essential to overcome these hindrances. Active transcytosis of cancer nanomedicines holds great promise for enhancing tumor extravasation, infiltration, and antitumor activity. Herein, polyzwitterionic OPDEA-PCL was developed to encapsulate paclitaxel (PTX) into micelles, serving as an intravaginal therapy for orthotopic cervical cancer. The OPDEA-PCL/PTX micelles efficiently penetrated mucus and exhibited strong resistance to mucin fouling, thereby facilitating rapid transcytosis into tumors. Furthermore, OPDEA-PCL/PTX micelles colocalized with the mitochondria of tumor cells, reversing PTX resistance. In the orthotopic cervical tumor model, the inhibition rate of OPDEA-PCL/PTX micelles was 90.0%, more than 2-fold higher than that of free PTX. In the subcutaneous cervical cancer model which is resistant to PTX. Intravenous administration of OPDEA-PCL/PTX micelles significantly overcame PTX resistance, achieving a tumor inhibition rate of 95.2%, and extending the median survival time by more than 2-fold compared to free PTX and PEG-PCL/PTX treated groups. In summary, this approach holds great promise as a potent localized nanomedicine for cervical cancer treatment with minimal side effects.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-026-04528-6.

Introduction

Cervical cancer persists as a leading gynecological malignancy worldwide, accounting for approximately 300,000 annual fatalities, with a 1–2% annual increase in incidence over recent decades [1–4]. Patients diagnosed with cervical cancer face significant challenges due to the difficulty of achieving targeted drug delivery through conventional systemic administration, resulting in suboptimal therapeutic outcomes [5–7]. On the other hand, multidrug resistance (MDR) severely compromises the effectiveness of current chemotherapeutic regimens [8–10].

Paclitaxel (PTX) remains a cornerstone agent in first-line cervical carcinoma management [11, 12]; however, its systemic administration is associated with significant toxicities. Furthermore, acquired tumor resistance induced by chemotherapy frequently diminishes the therapeutic index, and the non-selective cytotoxicity of PTX disproportionately affects mitotically active healthy tissues [13–15].

The unique anatomical accessibility of cervical lesions through transvaginal routes offers distinct pharmacokinetic advantages for drug delivery, circumventing hepatic first-pass metabolism to achieve higher bioavailability [16, 17]. Furthermore, the extensive surface area of the vaginal mucosa, combined with its inherent permeability, synergistically enhances drug permeation, enabling localized drug accumulation while minimizing systemic exposure [18]. The anatomical proximity of cervical malignancies to the vaginal canal enables transvaginal drug delivery as a promising therapeutic strategy for high drug concentration at the tumor site as well as a convenient non-invasive form of self-medication [19]. Nevertheless, transmucosal drug penetration offers tremendous challenges to intravaginally formulations [20]. Thereafter, insufficient intratumoral drug accumulation and penetration, together with drug resistance, also compromise the therapeutic outcome of cervical cancer.

Enormous progress has been made in improving mucosal penetration and enhance the therapeutic efficacy for cervical cancer [21]. The cervicovaginal mucus functions as a homeostatic defense through continuous turnover and pH-dependent structural reorganization (pH ranges from 4.0 to 5.5) [22]. Its viscoelastic hydrogel matrix contains various mucins (predominantly MUC1, MUC4, and MUC5), protecting the underlying epithelial cells by lubricating the surface and trapping foreign particles [23, 24]. In our previous studies, we developed an aqueous-soluble polyzwitterionic platform OPDEA, which is nearly neutral at physiological pH and non-fouling toward proteins, but can weakly bind to the cell membrane to induce adsorption-mediated, rapid cellular uptake and transcytosis across tumor cells [25, 26].

In contrast to conventional PEGylated nanocarriers, which often suffer from hydrogen-bond-mediated mucoadhesion and limited penetration across viscous cervical mucus [27–29], the OPDEA-PCL system exhibits a distinct antifouling mechanism and pH-responsive charge switching, which enhances electrostatic interaction with cell membranes and promotes transcytosis-mediated tumor penetration [30]. This mechanism bypasses ligand-receptor and adenosine triphosphate (ATP)-dependent transporter limitations, relying on mitochondrial energy to sustain efflux pump function [31–33]. This dual dependency positions mitochondrial targeting as a mechanism-driven therapeutic strategy to circumvent MDR by concurrently depleting ATP reservoirs and amplifying genomic instability in chemoresistant malignancies [34, 35].

The persistent disease burden and anatomical characteristics of cervical lesions necessitate engineered drug delivery systems capable of overcoming the dense mucin barrier in endocervical tissue [36]. This escalating disease burden underscores the urgent need for developing mucopenetrative drug carriers like PEG-PCL, whose low mucin-binding capacity may overcome biological barriers in the cervical tumor microenvironment. Systemic nanomedicine delivery typically follows a five-stage CAPIR cascade (Circulation, Accumulation, Penetration, Internalization, Release), where sequential optimization at each stage is mandatory for achieving the maximal therapeutic index [26]. Clinically observed translational gaps reveal a significantly attenuated permeability and retention manifestation in human malignancies compared to the animal tumor models [37], such as Abraxane (albumin-bound paclitaxel) and Lipusu (paclitaxel liposome), demonstrating important value in systemic chemotherapy by improving formulation tolerability and reducing solvent-related toxicity [38, 39]; however, they are not designed to overcome mucosal barriers or achieve active tumor penetration. The translational potential of local therapy for cervical cancer is supported by a Phase I clinical trial demonstrating complete remission in all 12 patients treated with gadolinium-based nanoparticles, with no dose-limiting toxicity (NCT03308604) [40]. Additionally, a clinical study on the vaginal gel Canesten^® showed comparable efficacy to the commercial product with minimal adverse effects (NCT06392529) [41]. These clinical precedents support the regulatory feasibility of vaginal nanoparticle delivery systems and underscore the translational relevance of OPDEA-PCL/PTX platform. Following extravasation, large-sized nanomedicines are almost motionless in the densely packed tumor microenvironment, restricting them in the periphery of tumor blood vessels rather than infiltrating tumors and thus inaccessible to the distal but highly malignant cells [42].

In this study, a tertiary amine and N-oxide based amphiphilic block copolymer, poly(2-(N-oxide-N, N-diethylamino) ethyl methacrylate)-b-poly(ε-caprolactone) copolymer (OPDEA-PCL) was used to encapsulate PTX into micelles (Fig. 1). The inherent nature of OPDEA renders OPDEA-PCL/PTX non-fouling toward proteins and capable of deep penetration. Notably, OPDEA-PCL is neutral under physiological conditions, which is different from most reported cationic mitochondrial targeting materials, endowing its improved biocompatibility. The phospholipid affinity allows OPDEA to bind weakly to cells, thereby triggering transcytosis-mediated active extravasation, tumor penetration, and efficient cellular internalization—ultimately achieving significant therapeutic outcomes. The in vivo experiments confirmed the hypothesis that OPDEA-PCL/PTX micelles induced robust antitumor efficiency in the U14 orthotopic cervical tumor and Hela/T subcutaneous cervical tumors that resistant to PTX and significantly prolonged the median survival.

Fig. 1.

Schematic illustration of amine-oxide-based paclitaxel delivery system (OPDEA-PCL/PTX micelles) for orthotopic cervical cancer treatment. (A) The chemical structure of OPDEA-PCL and its self-assembly with paclitaxel into micelles. (B) OPDEA-PCL/PTX micelles were intravaginally administered into orthotopic U14 tumor-bearing mice. (I) The protein-nonfouling feature of OPDEA-PCL enabled the micelles to penetrate through the mucus; (II) OPDEA-PCL/PTX micelles triggered rapid transcytosis to penetrate the tumor and target mitochondria; (III) in the tumor interstitium, adsorption on the cancer cell initiated rapid internalization of the OPDEA-PCL/PTX micelles, which are subsequently passed on to cells in deeper layers, leading to deep infiltration into the tumor inducing cell death

Results and discussion

Preparation, characterization and cytotoxicity assessment of OPDEA-PCL/PTX micelles

The amphiphilic block copolymer OPDEA-PCL was synthesized according to previous report (Scheme S1 and Fig. 1). Paclitaxel is modified with succinic anhydride, followed by conjugation with Cy5-NH₂, yielding Cy5-labeled paclitaxel (Scheme S2). The key characterization spectra, including ¹H-NMR (Fig. S1), ¹³C NMR (Fig. S2), GPC (Fig. S3), and FTIR (Fig. S4), are presented. The OPDEA-PCL/PTX and PEG-PCL/PTX micelles were prepared by nanoprecipitation method [25, 43]. The mass ratios of OPDEA-PCL to PTX were varied, resulting in a series of OPDEA-PCL/PTX Micelles (Fig. 2A-B). As shown in Fig. 2B-D and Fig. S5-6, the particle sizes ranged from 66 to 90 nm under pH 7.4 conditions, with nearly neutral surface and narrow size distribution. When the pH decreased from neutral to mildly acidic, the zeta potential of OPDEA-PCL/PTX micelles shifted from near-neutral to mildly positive. Notably, DLS measurements confirmed that the micelle size remained unchanged at pH 4.4 and pH 7.4 (Fig. S7), indicating good colloidal stability. With the pHs shifting from neutral to mildly acidic, the zeta potentials changed from neutral to positive. The optimized diameter of OPDEA-PCL/PTX micelles were around 72.9 nm, and the zeta potentials were nearly neutral (Fig. 2E; Fig. S8). The micelles were spherical, as observed by transmission electron microscopy (Fig. 2E inserted). For comparison, the PEG-PCL/PTX micelles had an average size of 69.4 nm and a zeta potential of nearly neutral (Fig. S8). After diluting in PBS and DMEM with 10% serum, the micelles maintained their integrities (Fig. 2F). The drug release kinetics of micelles were further investigated in neutral and acidic environment (pH = 7.4 or 4.4) to mimic the physiological condition of blood circulation and the acidic vaginal environment, respectively. Less than 30% of the encapsulated PTX was released from OPDEA-PCL/PTX within 24 h, whereas 40–60% was released from PEG-PCL/PTX (Fig. 2G), suggesting that OPDEA-PCL/PTX micelles have higher stabilities in physiological conditions (pH 7.4 and pH 4.4). These results suggested that OPDEA-PCL/PTX micelles possessed good stability, a prerequisite for drug delivery.

Fig. 2.

The preparation and characterization of OPDEA-PCL/PTX micelles. (A) Preparation of OPDEA-PCL/PTX using precipitation method. (B) The encapsulation efficacies (EE) and drug loading (DL) of OPDEA-PCL/PTX micelles. (C) Size distribution of OPDEA-PCL/PTX micelles with different mass ratios. (D) The zeta potential of OPDEA-PCL/PTX micelles with different mass ratios measured by DLS under pH 4.4 or 7.4. (E) The optimal size distribution of OPDEA-PCL/PTX micelles measured by DLS and the representative transmission electron microscopy images of OPDEA-PCL/PTX micelles. Scale bar : 100 μm. (F) Sizes of the micelles after timed incubation in different media. (G) The release profiles of OPDEA-PCL/PTX and PEG-PCL/PTX micelles at pH 4.4 and pH 7.4. (H) Cytotoxicities of free PTX, PEG-PCL/PTX and OPDEA-PCL/PTX micelles on U14 cells and Hela/T cells detected by CCK-8 assay (48 h treatment). (I) Apoptosis analysis of Hela/T cells treated with free PTX, PEG-PCL/PTX, and OPDEA-PCL/PTX by flow cytometry assay. PTX-equivalent concentration : 0.05 µg/mL; incubation time: 48 h

The cytotoxicities of OPDEA-PCL/PTX micelles were higher compared to free PTX in U14 and Hela/T cells (Fig. 2H, Table S1), which may be attributed to the sustained release of PTX from micelles. The IC₅₀ of OPDEA-PCL/PTX micelles for U14 cells was around 0.565 ng mL^− 1. In comparison, the IC₅₀ values of PEG-PCL/PTX micelles and free PTX were 2.780 ng mL^− 1 and 3.382 ng mL^− 1, respectively, indicating significantly enhanced cytotoxicity of OPDEA-PCL/PTX micelles. Furthermore, in Hela/T cells, OPDEA-PCL/PTX micelles showed superior cytotoxicity with the IC₅₀ of 0.141 µg mL^− 1, compared to 0.328 µg mL^− 1 and 0.801 µg mL^− 1 of PEG-PCL/PTX micelles and free PTX after 48 h incubation, with similar trends observed at 24 h and 72 h (Fig. 2H; Fig. S9). In normal 3T3 cells, OPDEA-PCL/PTX showed reduced toxicity (IC₅₀ = 0.070 µg/mL) versus free PTX (0.031 µg/mL); in HUVEC cells, all formulations showed comparable IC₅₀ (0.09–0.10 µg/mL). In contrast, OPDEA-PCL/PTX enhanced cytotoxicity against Hela/T and cells (Fig. S10). Besides, the OPDEA-PCL carriers had minimal effects on cell proliferation in vitro (Fig. S11). Apoptosis induction was also investigated (Fig. 2I). After 48 h incubation, apoptotic rates were 30%, 42%, and 59% for free PTX, PEG-PCL/PTX, or OPDEA-PCL/PTX respectively, demonstrating that OPDEA-PCL/PTX induced stronger apoptosis in PTX-resistant Hela/T cells, consistent with the enhanced cytotoxicity.

Cellular uptake and subcellular distribution

Zwitterionic polymers are generally nonfouling and can shed protein adsorption, thus suffering from unfavorable cellular uptake [44]. Interestingly, OPDEA-PCL, albeit zwitterionic, can be easily internalized by cells [43, 45]. As shown by confocal microscopy, OPDEA-PCL/^Cy5PTX micelle-treated cells had significantly higher Cy5 intensities than free ^Cy5PTX and PEG-PCL/^Cy5PTX micelle treated group after 15 min and 60 min incubation (Fig. 3A-B). Flow cytometry analysis demonstrated that 33.5% of U14 cells that treated with OPDEA-PCL/^Cy5PTX micelles were Cy5 positive after 15 min incubation, whereas only 1.2% and 5% were positive for PEG-PCL/^Cy5PTX micelles and free ^Cy5PTX (Fig. 3C). After 60 min incubation, the corresponding values increased to 97%, 84%, and 23%, respectively (Fig. 3D; Fig. S12). Subsequently, we investigated the subcellular distribution of the micelles. As shown in Fig. 3E-F, the Cy5 signals of the OPDEA-PCL/^Cy5PTX had weaker correlations with the LysoTracker Green signals (pearson correlation coefficient = 0.33) while overlapped with the MitoTracker Green signals remarkably (pearson correlation coefficient = 0.81). These results suggested that the intracellular OPDEA-PCL/^Cy5PTX micelles highly colocalized with mitochondria [34].

Fig. 3.

Cellular internalization of OPDEA-PCL/^Cy5PTX micelles. (A) Confocal images of U14 cells cultured with OPDEA-PCL/^Cy5PTX and PEG-PCL/^Cy5PTX micelles at Cy5-eq. dose of 0.1 µg mL^− 1 for 15 and 60 min. Cell nuclei were stained with Hoechst 33342 shown in blue, Cy5-labeled micelles were shown in red. Scale bar : 20 μm. (B) Confocal images of Hela/T cells cultured with OPDEA-PCL/^Cy5PTX and PEG-PCL/^Cy5PTX micelles at Cy5-eq. dose of 0.1 µg mL^− 1 for 15 and 60 min. Cell nuclei were stained with Hoechst 33,342 shown in blue, Cy5-labeled micelles were shown in red. Scale bar : 20 μm. (C) The cellular uptake of free ^Cy5PTX, PEG-PCL/^Cy5PTX micelles, OPDEA-PCL/^Cy5PTX micelles in U14 cells measured by flow cytometry after 15 and 60 min incubation. Cy5-eq. dose: 0.1 µg mL^− 1. (D) The endocytosis rates of ^Cy5PTX micelles in U14 cells. The cellular uptake of free ^Cy5PTX, PEG-PCL/^Cy5PTX micelles and OPDEA-PCL/^Cy5PTX micelles in U14 cells were analyzed by flow cytometry after 15 min and 60 min incubation, and quantificated by FlowJo software. Cy5-eq. dose: 0.1 µg mL^− 1, (*, P < 0.05, **, P < 0.01, ***, P < 0.001). (E) Lysosomal/Mitochondrial localization analyses of U14 cells incubated with Cy5-labeled micelles (Cy5-Eq. 0.1 µg/mL) for 1 h measured by Laser Scanning Confocal Microscope. Lysosomes were stained with lysotracker Red shown in green; mitochondrion were stained with MitoTracker Green shown in green; the cell nuclei were stained with Hoechst 33,342 shown in blue. Scale bars : 20 μm. (F) Linescan analysis and the pearson’s correlation coefficient calculations performed by ImageJ software along the randomly selected lines in panel (n = 3)

Blood clearance, biodistribution, and in vivo penetration

The intravaginal blood clearance of OPDEA-PCL/PTX micelles was subsequently evaluated in Sprague Dawley (SD) rats using ultra-performance liquid chromatography-triple quadrupole mass spectrometry (UPLC-QQQ/MS) (Figure S13-S16). After intravaginal administration of OPDEA-PCL/PTX micelle, the PTX concentration in the blood increased over time within the initial four hours and then decreased. On the contrary, intravaginally administration of free PTX gave much lower plasma concentrations at all time points (Fig. 4A). For kinetic parameters comparison, the OPDEA-PCL/PTX micelles showed higher AUC_(0−t), T_1/2, C_max, MRT_(0−t), and T_max values compared to free PTX group (Fig. 4B). At 12 h post-treatment, the mice were sacrificed, and major organs and tumor tissues were excised for ex vivo imaging. The fluorescence intensities of tumors of OPDEA-PCL/^Cy5PTX treated tumors were higher than those of free PTX group (Fig. 4C). The PTX contents in the U14-Luci tumors were determined by UPLC at 12 h post intravaginal administration. As shown in Fig. 4D, the PTX content in the tumor of OPDEA-PCL/PTX micelles group was about 6.2-fold of those administered with free PTX at 12 h post administration. These results demonstrated that the OPDEA-PCL/^Cy5PTX micelles could accumulate and retend more efficiently in tumors than free PTX. We then evaluated the in vivo tumor penetration ability of OPDEA-PCL/PTX micelles. Mice bearing orthotopic U14-Luci tumors were locally administered with OPDEA-PCL/PTX micelles, and the tumor regions were imaged (Fig. 4E). We then investigated the tumor penetration in orthotopic U14-Luci tumor-bearing mice. As shown in Fig. 4F, the OPDEA-PCL/^Cy5PTX group exhibited stronger and deeper fluorescence within tumor tissue than free ^Cy5PTX and PEG-PCL/^Cy5PTX groups. At 1 h post injection, the fluorescence intensities of OPDEA-PCL/^Cy5PTX were about 3.6- or 4.3-fold of those administered with free PTX or PEG-PCL/PTX micelles, respectively. At 6 h post injection, the OPDEA-PCL/^Cy5PTX group showed stronger fluorescence within the tumor tissue, whereas PTX fluorescence in the free ^Cy5PTX and PEG-PCL/^Cy5PTX groups still remained at the tumor periphery; the relative fluorescence intensity of the OPDEA-PCL/^Cy5PTX group was 3.1-fold or 4.9-fold of the free ^Cy5PTX and PEG-PCL/^Cy5PTX groups. At 12 h post injection, large areas of PTX fluorescence were detected deep within the tumor tissue in the OPDEA-PCL/^Cy5PTX group (Fig. 4G). This demonstrated that OPDEA-PCL/^Cy5PTX micelles have superior tumor penetration capabilities than free ^Cy5PTX and PEG-PCL/^Cy5PTX groups.

Fig. 4.

Biodistribution, pharmacokinetics and in vivo tumour penetration after vaginal administration of OPDEA-PCL/PTX micelles. (A) The plasma PTX concentration analyzed by UPLC-QQQ/MS method. (B) The corresponding pharmacokinetic parameters calculated after single administration of free PTX and OPDEA-PCL/PTX micelles, (SD rats, PTX-eq. dose, 6 mg kg^− 1, n = 3). (C) In vivo imaging of tumors and major organs (heart, liver, spleen, lung, kidney, and intestine) excised at 12 h post administration. (D) The PTX concentrations analyzed by UPLC in the U14 orthotopic cervical cancer model at 12 h post single vaginally injection of free PTX, OPDEA-PCL/PTX micelles. PTX-eq. dose, 1 mg kg^− 1, n = 3. (E) Immunofluorescence section analysis of tumor penetration of OPDEA-PCL/^Cy5PTX micelles at 1 h, 6 h, 12 h post single vaginally administration with free ^Cy5PTX, PEG-PCL/^Cy5PTX micelles, or OPDEA-PCL/^Cy5PTX micelles in the orthotopic cervical cancer model, PTX-eq. dose, 1 mg kg^− 1. Blue: DAPI; Red: Cy5; Scale bar represent 2 mm and 200 μm, respectively. (F) Linear scan analysis and corresponding quantitation of depths and intensities of free ^Cy5PTX, PEG-PCL/^Cy5PTX micelles, or OPDEA-PCL/^Cy5PTX micelles into tumors at 1 h, 6 h, and 12 h post injection using ImageJ software. (G) The relative fluorescence penetration depths ranged from 0 to 1500 μm (n = 3). The arrows represented the penetrating direction of free ^Cy5PTX, PEG-PCL/^Cy5PTX, or OPDEA-PCL/^Cy5PTX micelles

Molecular docking, mucus penetration, and in vitro penetration

The major obstacles limiting the bioavailability vaginal administration are mucus penetration and drug retention. Cervicovaginal mucus effectively blocks foreign substances and rapidly clears them through natural mucus clearance processes [16]. MUC1, MUC4, and MUC5 are representative mucin proteins in the cervicovaginal environment, and are key contributors to the mucus barrier that limits drug penetration [16, 18]. Overcoming these barriers requires nano drug efficiently penetrate mucus and rapidly enter tumor cells. To further elucidate the molecular interactions underlying mucus penetration, molecular docking simulations were performed. As shown in Fig. 5A and Fig. S17, OPDEA monomer binds to the active pockets of MUC1, MUC4, and MUC5 proteins, with binding energies of -3.0 kcal/mol, -3.2 kcal/mol and − 3.1 kcal/mol, respectively. To further validate the interaction between OPDEA-PCL/PTX micelles and mucin, isothermal titration calorimetry (ITC) analysis was performed. The equilibrium association constant (Ka) was determined to be 4.13 × 10³ M^− 1 (Fig. 5B), indicating relatively weak and reversible binding. The interaction was also validated by SPR assay, which showed a Kd value of 1.12 mM (Fig. 5C), indicating that OPDEA-PCL exhibited relatively weak binding capabilities to MUC1, MUC4, and MUC5 proteins, with higher Kd values demonstrating lower binding affinities. This suggested that the weak mucin-binding property of OPDEA-PCL could facilitate penetration through the mucus layer.

Fig. 5.

In vitro transcellular transport study of OPDEA-PCL/PTX micelles. (A) Molecular docking analysis on the interactions between N-oxide-trimethylamine moiety of OPDEA-PCL with MUC1, MUC4, and MUC5 proteins. Binding affinities were calculated by Autodock vina 1.1.2. (B) The binding constant and parameters calculated from the isothermal titration microcalorimetry (ITC) curves of 200 µM mucin with 200 µM mM of OPDEA-PCL/PTX micelles. (C) Binding affinity between mucin and OPDEA-PCL/PTX micelles were determined by surface plasmon resonance (SPR) imaging. (D) Schematic diagram showing the transport of Cy5-labeled micelles across the U14 cell monolayer on a transwell membrane. The Cy5-labeled micelles at a Cy5-eq. dose of 1 µg mL^− 1 were added in the apical side and the micelle concentrations at timed intervals in the basolateral compartment were measured by Cy5 fluorescence. CLSM images of Cy5 fluorescence signal (red) in U14 cells were taken after incubation with free ^Cy5PTX, PEG-PCL/^Cy5PTX and OPDEA-PCL/^Cy5PTX in another chamber for 4 h. The nuclei were stained with Hoechst 33,342 and shown in blue. Scale: 20 μm. ImageJ software was used to quantify the relative intensity. (E) Schematic diagram and transmucus transport of Cy5-labeled micelles across a simulated vaginal fluid mucus layer (500 μm thick) deposited on a transwell membrane. Cy5-labeled micelles solution (100 µL, Cy5-eq. dose, 1 µg mL^− 1) was in the apical compartment immersed in 600 µL PBS (pH = 7.4) in the basolateral compartment. The Cy5 fluorescence signal (red) in second batch of U14 cells were taken after incubation with free ^Cy5PTX, PEG-PCL/^Cy5PTX micelles or OPDEA-PCL/^Cy5PTX micelles in lower chamber for 4 h, and nuclei were stained with Hoechst 33,342 (blue). Scale bar: 20 μm. The relative intensities were quantitated by ImageJ software. (F) Intracellular transport of Cy5-labeled micelles between U14 cells. The cells (first batch) were cultured with micelles at a Cy5-eq. dose of 0.1 µg mL^− 1 for 6 h, washed with PBS, and imaged (i); the cells were then cultured in fresh medium for 12 h, and the medium was harvested to incubate with the second batch of cells for 6 h, followed by washing and imaging (ii); the same procedures were implemented for another two rounds (iii and iv). Scale bar: 20 μm. (G) The endocytic rates of OPDEA-PCL/^Cy5PTX into U14 cells that pretreated with endocytotic inhibitors for 0.5 h

Transepithelial transport of the micelles crossing the monolayer of U14 cells was then investigated on a transwell membrane (Fig. 5D). Free ^Cy5PTX, PEG-PCL/^Cy5PTX, or OPDEA-PCL/^Cy5PTX micelles were then applied to the apical side, and the micelle concentrations in the basolateral compartment were measured to evaluate the transcellular transport capacity. As shown in Fig. 5D, free ^Cy5PTX and PEG-PCL/^Cy5PTX hardly crossed the monolayer, while the OPDEA-PCL/^Cy5PTX micelles traveled rapidly to the basolateral side, about 3-fold of free ^Cy5PTX and PEG-PCL/^Cy5PTX. The medium from the basolateral compartment was then collected and incubated with pre-seeded U14 cells for 4 h. Confocal microscopy revealed that the relative fluorescence intensity in the OPDEA-PCL/^Cy5PTX group was significantly higher than that of the PEG-PCL/^Cy5PTX and free ^Cy5PTX groups. Figure 5E showed that both micelles penetrated through the mucus layer into the basolateral side, but the OPDEA-PCL/^Cy5PTX micelles moved much faster. The calculated permeability was 4.14 µg h⁻¹cm^− 2 for OPDEA-PCL/^Cy5PTX micelles, nearly 10-fold of PEG-PCL/^Cy5PTX micelles (0.56 µg h⁻¹cm^− 2) and free ^Cy5PTX (0.48 µg h⁻¹cm^− 2). This is likely due to the reduced interaction of OPDEA-PCL micelles with mucins, allowing them to permeate the vaginal mucus compared to PEG-PCL micelles more effectively, consistent with the reduced interaction of OPDEA-PCL micelles with mucins in the previous molecular docking data, allowing them to permeate the vaginal mucus more effectively compared to PEG-PCL micelles. PEG-PCL/PTX micelles and free PTX showed limited effectiveness in mucus penetration. Deeper tumor infiltration is critical in enhancing the therapeutic efficacy of nanomedicines [46–49]. Afterward, the medium was collected for incubation with another batch of cells. Cy5 signals could be observed in all the batches of cells in the OPDEA-PCL/^Cy5PTX group. In contrast, the Cy5 fluorescence could only be hardly observed in the first batch of cells of the free ^Cy5PTX and PEG-PCL/^Cy5PTX groups (Fig. 5F; Fig. S18).

We then investigated the cellular uptake pathways of OPDEA-PCL micelles. As shown in Fig. 5G, the endocytotic rate of OPDEA-PCL micelles was significantly reduced to 30.4% at 4 ℃, suggesting the process was energy-dependent. The micelles were primarily taken up through the macropinocytosis and actin-dependent pathways, as evidenced by the decreased endocytosis rates after wortmannin (83.03%) and cytochalasin D (79.7%) treatments. Considering the reduced mucin interaction and enhanced tumor permeability of OPDEA-PCL/PTX micelles, the in vivo fate of these micelles following intravaginal administration was expected to penetrate through the mucus layer across the vaginal mucosa, be internalized by cervical cancer cells, and subsequent transport between cancer cells, thereby achieving a highly effective therapeutic outcome.

The above results highlighted the novelty of OPDEA-PCL/PTX micelles compared with existing zwitterionic or PEG-based systems. Conventionally, PEG was served as a static hydrophilic shielding layer to reduce mucin adhesion and improve local retention. However, PEGylation did not provide dynamic modulation of surface charge or tumor microenvironment responsiveness [27, 50]. In contrast, OPDEA-PCL was a polyzwitterionic copolymer containing tertiary amine oxide functionalities that enable pH-adaptive electrostatic regulation. Under neutral vaginal mucus conditions, OPDEA-PCL micelles exhibited minimal nonspecific interactions, facilitating mucus penetration. Upon exposure to the mildly acidic tumor microenvironment, partial protonation enhanced electrostatic interaction with tumor cell membranes, promoting active transcytosis and deeper tumor infiltration. This reversible charged modulation distinguished OPDEA-PCL from conventional PEGylated or permanently zwitterionic systems.

In vivo antitumor activity of U14-luci orthotopic mouse tumor model

The U14-Luci orthotopic mouse tumor model was established to assess the anti-tumor efficiency of OPDEA-PCL/PTX micelles. The tumor-bearing mice were intravaginally treated with PBS, free PTX, PEG-PCL/PTX micelles, and OPDEA-PCL/PTX micelles (PTX equivalent dose was 3 mg kg^− 1) every 2 days for a total of 4 treatments. As shown in Fig. 6A-B, for the free PTX-treated group, the tumor growth slightly slowed down, about 44.9% of the PBS group on Day 14. It is worth noting that the OPDEA-PCL/PTX group had the smallest tumors among all the groups (only 13.5% of control group). At the end of the experiment, tumors were excised and weighed to calculate tumor inhibitory rates (TIR), and OPDEA-PCL/PTX micelles showed the highest inhibitory capacities compared to other groups. The tumor tissues were also photographed and imaged (Fig. 6C). All the treatments inhibited tumor development compared with the control group, with TIR of 42.2% for free PTX and 90.0% for OPDEA-PCL/PTX micelles, as shown by the lower tumorous luciferase intensities and slower tumor growths (Fig. 6D; Fig. S19). The pictures and the luminescence images of tumors presented the excellent antitumor efficacy of OPDEA-PCL/PTX micelles. A large number of apoptotic cells were observed in the tumor tissue sections of the OPDEA-PCL/PTX micelles group, showing severe nuclear shrinkage and reduced cell densities (Fig. 6E). TUNEL assay and PCNA further verified that the OPDEA-PCL/PTX micelles group effectively promoted tumor cell apoptosis and inhibited tumor cell proliferation. Furthermore OPDEA-PCL/PTX micelles caused substantially less mucosal damage, likely due to their sustained release and reduced local drug concentrations (Fig. S20).

Fig. 6.

Anti-tumor effect of OPDEA-PCL/PTX micelles against the orthotopic U14-Luci cervical cancer model. (A) Tumor inoculation and treatment scheme. (B) Bioluminescence imaging of mice bearing orthotopic U14-Luci tumors after different treatments using IVIS (n = 5). (C) Images of the reproductive system dissected from the tumor-bearing mice at the end of the tumor suppression experiment (n = 5). (D) Statistics analyses of tumor volume and tumor inhibition rates (n = 5). (E) HE histology, PCNA and TUNEL staining of tumors after PBS, OPDEA-PCL, free PTX, and OPDEA-PCL/PTX treatments. (F) Survivals of intravaginal administration of PBS, OPDEA-PCL, free PTX and OPDEA-PCL/PTX on the U14-Luci orthotopic mouse model (n = 5 in each group, marked with median survival days, log-rank test analysis between the two groups, *P < 0.05, **P < 0.01, ***P < 0.001)

The survival of the mice bearing U14-Luci orthotopic tumor was further studied. Free PTX prolonged mouse survival, while all the mice died on Day 25. However, OPDEA-PCL/PTX micelles significantly extended the median survival time by nearly 2-fold than that of the PBS group (Fig. 6F). The above results demonstrated that OPDEA-PCL/PTX micelles generated potent antitumor inhibitory effects with no severe side effects. The above-mentioned experiments verified that OPDEA-PCL/PTX micelles effectively possessed excellent mucus and tumor penetrating properties, likely contributing to the enhanced inhibition of tumor growths by OPDEA-PCL/PTX micelles in the aggressive U14 tumor model.

Biosafety evaluation

In clinical settings, the application of PTX can lead to a variety of side effects, significantly affecting patient prognosis [14] such as liver dysfunction, impaired neutropenia and anaphylaxis [6]. H&E staining showed no apparent pathological damage in the main organs, including the heart, lung, liver, spleen, and kidney (Fig. 7A). No remarkable difference of body weights were observed between the groups during the treatments (Fig. 7B). All the liver and kidney function markers (Fig. 7C) in the OPDEA-PCL/PTX group had no significant difference from those in the PBS group, and neither did the routine blood or inflammatory indicators (Fig. 7D). These findings suggested that OPDEA-PCL/PTX micelles had higher biological safety, making them suitable for clinical applications.

Fig. 7.

Biosafety evaluation of OPDEA-PCL/PTX micelles. (A) Representative histological features of the visceral organs in the U14-Luci orthotopic cervical cancer model after different treatments. (B) Body weight curves of mice after treatments. (C) Serum biochemical detection. (D) Blood routine analysis (n = 5)

In vivo antitumor activity of hela/T subcutaneous tumors

The antitumor activity of OPDEA-PCL/PTX micelles was also evaluated on Hela/T (resistant to PTX) subcutaneous tumor model. As shown in Fig. 8A, the i.v.-injected free PTX, OPDEA-PCL/PTX micelles, or PEG-PCL/PTX micelles at the same dose were served as positive controls. Compared with the PBS group, all PTX treatments inhibited tumor growth. The TIRs of free PTX and PEG-PCL/PTX micelles were 65.7% and 72.1%, respectively (Fig. 8B-D; Fig. S21). No remarkable difference of body weights were observed between the groups (Fig. S22). While OPDEA-PCL/PTX micelles suppressed tumor growths with TIRs as high as 95.2%. Histological examination of the tumor tissue sections showed abundant tumor apoptosis, including extensive vacuolization, severe nucleus shrinkage, and decreased cellularity in the OPDEA-PCL/PTX groups. In comparison, fewer apoptotic events were observed in the free PTX and PEG-PCL/PTX groups (Fig. 8E). The survivals of Hela/T tumor-bearing mice of the OPDEA-PCL/PTX group were also dramatically extended, and the median survivals of PBS, free PTX, and PEG-PCL/PTX groups were 18, 29, and 31 days, respectively (Fig. 8F). Importantly, no significant change in body weight occurred in the mice receiving OPDEA-PCL/PTX micelles treatments. These results demonstrated that compared to traditional PEGylated micelles, OPDEA-PCL/PTX micelles, with their efficient tumor penetration ability and remarkable mitochondrial targeting feature, could reverse cancer cells’ MDR and thus achieve high therapeutic outcomes. Recent approaches of enhancing tumor penetration and efficacy include optimizing nanoparticle size [51], introducing cationic or cell-penetrating peptides to enhance cellular uptake [52], or incorporating targeting ligands for active tumor accumulation [53]. In parallel, MDR-targeting strategies often involve increasing intracellular drug accumulation or modulating resistance-associated pathways, such as downregulating P-gp-mediated drug efflux [54]. Interestingly, the neutral-charged OPDEA segment forms a charge-balanced hydration layer that reduces mucin interactions, enabling efficient mucus permeation, transcellular transport and enhanced cellular uptake.

Fig. 8.

Anticancer activity of OPDEA-PCL/PTX micelles against Hela/T subcutaneous tumors. (A) Tumor inoculation and treatment scheme. (B) Tumor growth curves of each group during treatments. (C) Tumor images at the end of tumor suppression experiment. (D) Averaged tumor weights of each group and the calculated tumor inhibition rates. (E) H&E and TUNEL staining of tumors after treatments. (F) Survivals of intravaginal administration of PBS, OPDEA-PCL, free PTX and OPDEA-PCL/PTX on Hela/T subcutaneous tumor model (n = 5 in each group, marked with median survival days, log-rank test analysis between the two groups, *, P < 0.05, **, P < 0.01, ***, P < 0.001)

Conclusion

In conclusion, intravaginally administrated OPDEA-PCL/PTX demonstrated strong efficacies in mucus penetration and endocytosis on cervical cancer models. Furthermore, vaginal administration of OPDEA-PCL/PTX micelles offered good bioavailability and safety. Mechanism study showed that these micelles achieved substantial tumor infiltration via transcytosis-mediated active tumor penetration, effectively overcoming PTX resistance in vivo and significantly suppressed cervical cancer growths. This strategy established a transformative paradigm for advancing cervical cancer treatment by improving curative efficacy and addressing drug resistance.

Materials and methods

Cell lines and animals

The U14 cells stably transfected with lucifease was sourced from the Henan Province Industrial Microbiology Culture Collection Engineering Technology Research Center, and the Hela/T cells that resistant to PTX were obtained from Nanjing Wanmuchun Biotechnology Co., Ltd. U14 cells were cultured in high-glucose DMEM, supplemented with 10% fetal bovine serum, 1% penicillin-streptomycin solution, and 1% L-glutamine. Similarly, Hela/T cells were cultured in RPMI-1640 medium containing 10% fetal bovine serum, 1% penicillin-streptomycin solution, and 1% L-glutamine. Culture conditions were maintained at 37 °C, with 5% CO₂ concentration and 95% relative humidity in an incubator.

All animal experiments were conducted following the approved protocol by the Animal Ethics Committee of the Zhejiang Provincial Center for Experimental Animals (Ethics Number : ZJCLA-IACUC-20020139), according to the guidelines for the care and use of laboratory animals. Female Balb/c nude mice, aged 6–8 weeks and weighing 18–20 g, were acquired from the Zhejiang Provincial Center for Experimental Animals. SD rats were used for pharmacokinetic studies.

Synthesis of OPDEA-PCL

ε-Caprolactone was synthesized into PCL through ring-opening polymerization catalyzed by diphenyl phosphate, using 5-(boc-amino)-1-pentanol as the initiator. The obtained PCL was then reacted with 2-bromoisobutyryl bromide to yield PCL-Br. PDEA was synthesized through ATRP polymerization. OPDEA-PCL was synthesized according to previous studies [25]. Subsequently, the resulting PDEA was oxidized with 3-chloroperoxybenzoic acid to obtain the polymer OPDEA-PCL. Dissolve an appropriate amount of OPDEA-PCL in chloroform and perform nuclear magnetic resonance analysis. The chemical structure of OPDEA-PCL was characterized by NMR profile.

Preparation and characterization of PTX-loaded micelles

The OPDEA-PCL/PTX micelles were prepared via the precipitation method, OPDEA-PCL and PTX at the weight ratio of 5:1 were dissolved in 1 mL of tetrahydrofuran. The mixture was then added dropwise to deionized water under continuous stirring. Then the solution was transferred into a dialysis bag (MWCO: 3.5 kDa) and dialyzed against ultrapure water for 24 h resulting in purified OPDEA-PCL/PTX micelles. Micelles were also prepared at various mass ratios of OPDEA-PCL to PTX (10:1, 2.5:1, and 1:1). PEG-PCL/PTX was prepared similarly. Subsequently, the solutions were concentrated via ultrafiltration and stored at 4 °C. The sizes and zeta potentials of PTX-loaded micelles were characterized using dynamic light scattering (Zetasizer Nano ZS90, Malvern Instruments, Malvern, UK).

Detection of PTX

The content of PTX was determined using the ultra-high performance liquid chromatography-triple quadrupole tandem mass spectrometry (UPLC-QQQ/MS) technique. Chromatographic Conditions: The analysis employed an Agilent ZORBAX RRHD Eclipse XDB-C18 column (2.1 × 100 mm) with a guard column of Agilent UHPLC Grd, Ecl. XDB-C18, 2.1 mm, 1.8 μm, 3 pk. The column temperature was maintained at 45 °C. The mobile phases consisted of ultrapure water (Phase A) and methanol containing 0.1% formic acid (Phase B). The flow rate was set at 0.3 mL/min with an injection volume of 5 µL. The elution gradient was set as follows: 50% A and 50% B at 0 min, linearly changed to 12% A and 88% B at 4 min, maintained until 6 min, then returned to 50% A and 50% B at 8 min and held until 10 min.

Mass Spectrometry Conditions: The triple quadrupole tandem mass spectrometer was activated and vacuumed for 24 h prior to use. A tuning solution was employed for calibration, and the parameters were adjusted to optimize performance.

Determination of PTX encapsulation efficiency (EE) and drug loading (DL)

The PTX content was quantified using UPLC-QQQ/MS for both the standard solutions and micellar samples. Micelles were diluted to a suitable concentration, vortexed for 10 min, and then centrifuged at 14,000 rpm for 10 min. The supernatant was collected, diluted, and analyzed to determine the PTX concentration. The standard curve of PTX was plotted using the determined concentrations and corresponding peak areas. The drug loading (DL) and encapsulation efficiency (EE) of the OPDEA-PCL/PTX micelles were calculated as below:

W_E: weight of drug encapsulated in micelles; W_PTX: weight of PTX added; W_M: weight of blank micelles.

In vitro stability of micelles

Each group of micelle solutions were diluted in 4 mL of PBS (10 mM, pH 7.4) or DMEM medium containing 10% fetal bovine serum (FBS). The samples were stored at room temperature and continuously monitored for one week. At regular intervals, 50 µL of each micelle solution were placed in a cuvette for size measurement using Dynamic Light Scattering (DLS) at 25 °C. The particle sizes of each group were recorded and plotted as a function of time.

In vitro drug release study

1 mL of PEG-PCL/PTX or OPDEA-PCL/PTX micelles (PTX concentration: 0.5 mg/mL) was placed into a dialysis bag (MWCO 3500 Da). The dialysis bag was sealed and immersed in a centrifuge tube containing 10 mL of PBS (pH 4.4 or 7.4) supplemented with 0.5% Tween-80 to maintain sink conditions, then dialyzed on an orbital shaker at 37 °C, 70 rpm. At predetermined time intervals, 200 µL of the external dialysate were discarded and immediately replenished with an equal volume of fresh solution containing 0.5% Tween-80. The PTX content in the dialysate were determined using UPLC-QQQ/MS. For statistical analysis, three parallel samples were prepared and analyzed at each time point and calculated.

Cy5 labeling of PTX

PTX (50 mg) were dissolved in 1 mL of pyridine. Succinic anhydride and DMAP were added, stirred until fully dissolved. The reaction were proceeded overnight at room temperature. The mixture were concentrated under reduced pressure using a rotary evaporator and subsequently re-dissolved in 15 mL of dichloromethane. The organic layer were washed three times with 10% citric acid solution and twice with saturated saline. Then the mixture were dried over sodium sulfate, filtered, and concentrated under reduced pressure and re-dissolved in 2 mL of DMF, Cy5-NH2, PyBOP, and DIPEA were then added, stirred until fully dissolved, and allowed for reaction for 0.5 h at room temperature. A mixture of dichloromethane: methanol at 20:1 ratio were served as the eluent for column chromatography to purify the product. After purification, the product was dried under vacuum to obtain ^Cy5PTX.

Pharmacokinetics study

SD rats (200 g, 10–12 weeks old) were fasted for 12 h prior to the experiment. The rats were randomly divided into two groups for vaginal administration (n = 3): (1) free PTX, (2) OPDEA-PCL/PTX, with both groups receiving a dosing of 6 mg kg^− 1 PTX-equivalent. Blood samples were collected at various time intervals using a micro-hematocrit capillary tube from the orbital venous plexus. Samples were centrifuged at 3,500 rpm at 4 °C for 10 min, and the supernatant was collected. For chromatographic analysis, a series of standard solutions were prepared by mixing the PTX standard solution with blank plasma at concentrations. The plasma samples were collected at various time points and mixed with methanol, vortexed for 10 min, and then centrifuged at 15,000 rpm for 30 min. The supernatant was stored overnight at -80 °C and re-centrifuged at 15,000 rpm for 30 min to collect the final supernatant. PTX concentrations were quantified using UPLC-QQQ/MS method. PTX contents in rat plasma at different time points was calculated to plot the plasma clearance curve of PTX in SD rats. Pharmacokinetic parameters were analyzed using the Drug And Statistics software.

Endocytosis of OPDEA-PCL/^Cy5PTX micelles

A total of 5 × 10⁵ U14 cells were seeded in 1 mL of medium containing 10% serum in confocal culture dishes and incubated at 37 °C in a CO₂ incubator for 24 h to facilitate cell adherence. The original medium was replaced with 1 mL of fresh medium, and Hoechst 33,342 (2 µM) was added for 30-minute incubation. Subsequently, the medium was removed and the cells were washed three times with PBS. Free ^Cy5PTX, PEG-PCL/^Cy5PTX micelles, or OPDEA-PCL/^Cy5PTX micelles (Cy5-eq at 0.1 µg mL^− 1) were then added and the cells were incubated for both 15, 60 and 240 min. The medium was removed and the cells were washed three times with PBS. Images were performed using a laser confocal microscope.

U14 cells were seeded at 1.5 × 10⁵ cells per well and incubated at 37 °C in a 5% CO₂ incubator for 24 h. Cells were then treated with free ^Cy5PTX, PEG-PCL/^Cy5PTX micelles, or OPDEA-PCL/^Cy5PTX micelles (Cy5 eq. of 0.1 µg mL^− 1) for 15 min and 1 h. The medium was removed and washed with precooled PBS, followed by trypsin digestion. The resultant cell suspensions were collected and centrifuged for 5 min at 1,000 rpm to remove the supernatant. Finally, cells were resuspended in 400 µL of PBS, transferred to flow cytometry tubes, vortexed, and analyzed by BD Canto flow cytometer. Data were analyzed using FlowJo software.

Intracellular transport of Cy5-labeled micelles between U14 cells

U14 cells were seeded at a density of 5 × 10⁵ cells per mL into glass-bottom culture dishes and incubated for 12 h. Initially, the cells were incubated with either free ^Cy5PTX, PEG-PCL/^Cy5PTX, or OPDEA-PCL/^Cy5PTX micelles (Cy5-eq. of 1 µg mL^− 1) for 6 h. Subsequently, the cells were washed three times with PBS and stained with Hoechst 33,342 (2 µM) for 30 min. After staining, the culture medium was removed and the cells were washed three times with PBS. The cells were then maintained in 1 mL of fresh culture medium for another 12 h, and the medium was then collected to incubate with the second batch of cells for 6 h and subjected to confocal microscopy imaging. The described procedure was then repeated. Hoechst 33,342 was excited at 405 nm and emitted light within 430–470 nm, whereas Cy5 was excited at 640 nm and emitted within 650–750 nm.

Subcellular distribution of micelles

U14 cells cells at a density of 5 × 10⁵ cells were seeded into glass-bottom culture dishes and incubated at 37 °C in a CO₂ incubator for 24 h. The medium was replaced with 1 mL of fresh culture medium. Hoechst 33,342 (2 µM), Lysotracker Red (0.2 µM), and MitoTracker Green (0.2 µM) were then added sequentially and incubated for 30 min. The medium was then discarded, and the cells were washed three times with PBS. Fresh culture medium containing free ^Cy5PTX, PEG-PCL/^Cy5PTX micelles, and OPDEA-PCL/^Cy5PTX micelles (Cy5 eq. of 0.1 µg mL^− 1) was then added, and the cells were incubated for 1 h. Subsequently, the medium was removed, and the cells were washed three times with PBS, imaged and photographed under a confocal microscope.

Endocytotic pathway of OPDEA-PCL/^Cy5PTX micelles

U14 cells were seeded in 24-well cell culture plate at 5 × 10⁵ cells per well. Cellular uptake inhibitors including niflumic acid (7.5 µM), wortmannin (5.0 µM), chlorpromazine (30 µM), and cytochalasin D (5.0 µM) were added and incubated for 30 min. Subsequently, OPDEA-PCL/^Cy5PTX micelles (Cy5-equivalent 1 µg mL^− 1) were added and incubated for an additional 30 min. Cells were also maintained at 4 °C with all reagents pre-cooled. After incubation with micelles, the culture medium was discarded, and the cells were washed twice with PBS, digested with trypsin, and resuspended in PBS. Flow cytometry was then employed to determine the percentage of Cy5-positive cells and the fluorescence intensity for each group were measured and quantified.

Molecular docking

The protein sequence in FASTA format were downloaded and submited to the supercomputing center for AlphaFold3 modeling. The Ramachandran plot analysis method was used to determine the conformations of amino acids within the protein. The φ and ψ dihedral angles of the backbone configuration can be read from the Ramachandran plot. The 2D structure of the compound was drawn by ChemDraw 3D. The lowest-energy conformation using the MM2 module was performed and saved. The protein structure from the PDB database were downloaded and visualized with PyMOL. The structures were preprocessed using Mgtools 1.5.6, including the removal of water molecules, the addition of hydrogens, the calculation of charges, and the merging of nonpolar hydrogens. The processed ligand and receptor were saved separately as PDBQT files. Docking was then performed with Autodock Vina 1.1.2, and the conformation with the highest score was selected. The docking results were subsequently visualized using PyMOL.

Mucus penetration studies

U14 cells were seeded in the upper chamber of Transwell plates at a density of 5 × 10⁵ cells/cm², and culture medium was added to both chambers. Once the monolayer was formed, the culture medium in the upper chamber was replaced with fresh medium containing either free ^Cy5PTX, PEG-PCL/^Cy5PTX, or OPDEA-PCL/^Cy5PTX micelles (Cy5-equivalent of 1 µg mL^− 1). At predetermined intervals, 200 µL of solution was discarded from the lower chamber and replenished with the same volume of fresh culture medium. The concentration of Cy5 micelles in these samples was measured using a spectrophotometer set to an excitation wavelength of 635 nm and an emission wavelength of 675 nm. Additionally, 5 × 10⁵ U14 cells was seeded into glass-bottom culture dishes and incubated for 12 h. After incubation, the medium was removed, and the cells were washed three times with PBS. The medium from the lower chamber was then transferred and incubated for 4 h. The medium was removed, and the cells were washed three times with PBS, and imaged under confocal microscope. Relative fluorescence intensity was quantified using ImageJ software.

Simulated Vaginal Fluid (SVF) was used to create a 500 μm thick mucus layer in the upper chamber of Transwell. Subsequently, 100 µL of free ^Cy5PTX, PEG-PCL/^Cy5PTX, or OPDEA-PCL/^Cy5PTX micelles (Cy5-eq of 1 µg mL^− 1) were added to the upper chamber, and 600 µL of PBS was added to the lower chamber. The assembly was then incubated at 37 °C. Samples (200 µL) were collected from the lower chamber at 1 h, 2 h, and 4 h intervals, with each sample volume replenished with an equivalent volume of PBS. The concentration of Cy5 micelles in these samples was determined using a spectrophotometer with excitation at 635 nm and emission at 675 nm. Additionally, 1 mL of culture medium containing 5 × 10⁵ U14 cells was seeded into a glass-bottom culture dish and incubated for 12 h. Following incubation, the medium was removed, and the cells were washed three times with PBS. The solution from the lower chamber was transferred to the culture dish and incubated for 4 h. After incubation, the medium was removed, and the cells were washed three times with PBS before imaging with a confocal microscope.

For tumor penetration study, the orthotopic cervical cancer model was established. Ten days after inoculation, the mice were received an intravaginal administration of 20 µL of either free ^Cy5PTX, PEG-PCL/^Cy5PTX, or OPDEA-PCL/^Cy5PTX micelles. At 1, 6, and 12 h post-administration, the mice were euthanized. The tumors were subsequently excised, embedded in OCT compound, flash-frozen in liquid nitrogen, and preserved at -20 °C for further processing into frozen tumor tissue sections. The cryostat was set to an environmental temperature of -20 °C and a blade temperature of -25 °C. Tumor sections were cut along the sagittal plane and the largest cross-section of the tumor, producing 5 μm thick slices. These slices were then brought to room temperature, washed three times with PBS using a gel-loading pipette tip, and mounted with DAPI anti-fade mounting medium. The prepared slides were stored at 4 °C in dark. Images were acquired using a slide scanner. Line scan analyses and relative fluorescence intensities were quantified using ImageJ software.

Isothermal titration calorimetry (ITC).

For the ITC measurement, concentrated Mucin protein was diluted into the ITC buffer (20 mM HEPES, pH 7.4, 150 mM NaCl). Compounds concentrations were estimated based on their mass. All the measurements were performed at 25 °C utilizing an PEAQ-ITC (MicroCal, Inc.) microcalorimeter. The procedure is as following: the protein was loaded in the cell chamber, and compounds with a concentration of 2 mM in the syringe was injected into the cell for 20 successive injections with a spacing of 120 s between injections. The binding constant and parameters calculated from the ITC curves of 200 µM mucin with 200 µM mM of OPDEA-PCL/PTX micelles.

Surface plasmon resonance (SPR) analysis.

SPR measurements were performed using a CM5 sensor chip. The chip surface was activated by injecting a freshly prepared mixture of 400 mM EDC and 100 mM NHS at a flow rate of 10 µL/min for 420 s. Mucin was diluted to 20 µg/mL in immobilization buffer and injected into the sample flow channel (Fc2) at 10 µL/min, resulting in an immobilization level of approximately 12,600 RU. The reference channel (Fc1) was left without ligand immobilization. Residual active esters were blocked by injecting 1 M ethanolamine hydrochloride at 10 µL/min for 420 s. OPDEA-PCL/PTX micelles were diluted in running buffer to eight concentrations ranging from 7 to 500 µM. The analyte solutions were injected over both Fc1 and Fc2 channels at a flow rate of 20 µL/min, with an association phase of 100 s followed by a dissociation phase of 180 s. All binding and dissociation processes were carried out in running buffer. The analytes were tested in ascending concentration order using a multi-cycle kinetic analysis method. After each binding cycle, the sensor surface was regenerated prior to the next injection.

Establishment of U14-Luci orthotopic cervical cancer model

Female Balb/c nude mice (6–8 weeks, 18–20 g) were acquired from the Zhejiang Provincial Center for Experimental Animals. U14 cells were cultured in dishes to 80% confluence. Then the culture medium was removed, and the cells were washed thrice with PBS. 25 µL of cell suspension containing 5 × 10⁵ cells was injected submucosally into the vagina of each mouse. Seven days post-injection, tumor growth was observed at the upper part of the perineal vaginal opening of the mice. Upon dissection, tumor extension was observed from the cervix to the external membrane of the vagina.

Biodistribution of PTX micelles

After orthotopic cervical cancer model establishment, the mice were intravaginal injected OPDEAPCL/PTX micelles, PEG-PCL/PTX micelles, or free PTX at 1 mg kg^− 1 PTX-eq. dose. At timed intervals, mice were sacrificed and the main organs (heart, liver, spleen, lungs, kidneys) and tumors were excised and weighted. The PTX concentrations in the organs and tumors were measured by UPLC as mentioned above.

Cytotoxicity assay

U14 cells were plated in 96-well plate at a density of 5,000 cells per well. After 24 h, the original medium was replaced with medium containing various concentrations of free PTX, OPDEA-PCL, PEG-PCL/PTX, and OPDEA-PCL/PTX. After 48 h incubation, the medium was removed from each well, and the cells were replenished with 100 µL of fresh medium containing 10 µL of CCK-8 reagent. The plates were then incubated for an additional 2 h incubation at 37 °C.

In vivo antitumor studies

After establishment of the U14-Luci orthotopic cervical cancer model, on Day 7 post tumor inoculation, each mouse were received an intraperitoneal injection of 200 µL of luciferin substrate (D-Luciferin 150 mg kg^− 1). Mice were randomly assigned to groups using a computer-generated method, with comparable bioluminescence intensity across groups.

Tumor progression was monitored using an IVIS imaging system. The mice were grouped by bioluminescence signal intensity and divided into 4 groups (1) PBS, (2) OPDEA-PCL, (3) free PTX, and (4) OPDEA-PCL/PTX. Each treatment group was administered vaginally at 20 µL containing PTX equivalent dosage of 3 mg kg^− 1, repeated every two days for a total of four doses. Mouse body weights were recorded, and tumor fluorescence intensities were monitored using IVIS system. On Day 15 post-inoculation, the mice were euthanasia. The uterus, tumor tissue, and major organs (heart, liver, spleen, lungs, kidneys) were collected. The weights of the tumors and uterus were measured, and tumor dimensions (length and width) were determined with calipers.

Histological assay

Hematoxylin and eosin (H&E) staining was utilized to assess the integrity of tumor tissues Proliferating Cell Nuclear Antigen (PCNA) was conducted to evaluate cellular proliferation within the tumor tissues. TUNEL assay was facilitated to direct visualize tumor apoptosis.

Toxicity evaluation

At the end of the experiment, blood was collected from tumor-bearing mice and then analyzed using a hematology analyzer. White blood cells, neutrophils, lymphocytes, monocytes, eosinophils, platelets, and hemoglobin were measured. The plasma was analyzed using an LW C400 automatic biochemistry analyzer. The parameters included AST (Aspartate Aminotransferase), ALT (Alanine Aminotransferase), CR (Creatinine), BUN (Blood Urea Nitrogen), LDH (Lactate Dehydrogenase), CK (Creatine Kinase), and TBIL (Total Bilirubin). Meanwhile, major organs, including hearts, livers, spleens, lungs, and kidneys, were excised, fixed with 4% paraformaldehyde and sectioned for H&E staining.

Survival study

The in situ U14-Luci cervical cancer model was established as described above in the tumor inhibition study, and the mice were monitored until they exhibited signs of significant weight loss (exceeding 10% of normal weight), loss of mobility, persistent lack of appetite, lethargy, tightly eyes closed, coma, or other conditions indicative of a decline, in accordance with the Body Condition Scores outlined in the Rodent Tumor Monitoring Guidelines. Upon reaching these criteria, the mice were euthanized, and this event were marked the experimental endpoint for each mouse.

Anticancer activity of OPDEA-PCL/PTX micelles against Hela/T subcutaneous tumors

Hela/T cells were subcutaneously inoculated into the lumbar abdominal region of each Balb/c nude mouse (female, aged 6–8 weeks). When the tumor volume reached around 50 mm³, the tumor-bearing mice were organized into four groups: PBS, free PTX, PEG-PCL/PTX, and OPDEA-PCL/PTX (n = 5). Each group was received a 200 µL of intravenous injection via the tail vein, containing PTX equivalent dose of 12 mg kg^− 1, at every two days for a total of five doses. After treatments, mice were euthanized according to animal ethics regulations, and the subcutaneous tumor tissues were excised, weighed, and photographed.

Survival study of OPDEA-PCL/PTX micelles against Hela/T subcutaneous tumors

Hela/T subcutaneous tumor model was established as above. Once the tumors reached around 50 mm³, the tumor-bearing mice were grouped and received an intravenous injection of PBS, free PTX, PEG-PCL/PTX micelles, or OPDEA-PCL/PTX micelles (PTX equivalent dose of 12 mg kg^− 1) at every 2 days for a total of five doses. The body weights of the mice were monitored daily. After five treatments, the mice were maintained without treatment and monitored until they exhibited signs such as weight loss exceeding 10% of normal, reduced mobility, feeding difficulties, lethargy, eyes tightly closed, coma, or other conditions indicative of a decline as per the Animal Condition Scores. At this point, the mice were euthanized and marked as the experimental endpoint for each mouse.

Statistical analysis

Data are presented as mean ± standard deviation. The data were analyzed using GraphPad Prism 9.0 and SPSS 26 software. Statistical analyses were performed with one-way ANOVA and t-tests for experimental data. Survival experiments employed the Log-rank statistical method. Statistical significance was defined as *P < 0.05, **P < 0.01, ***P < 0.001. Graphs were created using GraphPad Prism 9.0 and Origin 2018 software.

Abbreviation

MDR

Multidrug resistance

PTX

Paclitaxel

TIR

Tumor inhibitory rates

Author contributions

C.Y: data collection and analysis, writing, and revision of the manuscript. X.Y, Q.X and W.J provided guidance and assistance during performance of the experiments. Q.N and S.Y: data curation and manuscript review. Z.Z: conception and design, funding acquisition, and supervision. All authors contributed to this study and approved the submitted version.

Funding

This work was supported by China Postdoctoral Science Foundation (2025M771976); Hangzhou Agriculture and Social Development Research Project (202004A12); the National Key Research and Development Program of China (2021YFA1201200); the National Natural Science Foundation (52273153); Hangzhou Biomedicine and Health Industry Development Support Science and Technology Project (2021WJCY179); Special Project of Traditional Chinese Medicine Modernization of Zhejiang Province (2021ZX013); Hangzhou Joint Fund of Zhejiang Provincial Natural Science Foundation of China under Grant (LHZQN25H300002); the Department of Science and Technology of State Administration of Traditional Chinese MedicineZhejiang Province Joint Construction Project (GZY-ZJ-KJ-24010) and Zhejiang Province Traditional Chinese Medicine Science and Technology Plan ( 2026ZL0006).

Data availability

The data that support the findings of this study are available in the supplementary material of this article.

Declarations

Competing interests

The authors declare no competing interests.