A redox-responsive self-assembling COA-4-arm PEG prodrug nanosystem for dual drug delivery suppresses cancer metastasis and drug resistance by downregulating hsp90 expression

Abstract

Metastasis and resistance are main causes to affect the outcome of the current anticancer therapies. Heat shock protein 90 (Hsp90) as an ATP-dependent molecular chaperone takes important role in the tumor metastasis and resistance. Targeting Hsp90 and downregulating its expression show promising in inhibiting tumor metastasis and resistance. In this study, a redox-responsive dual-drug nanocarrier was constructed for the effective delivery of a commonly used chemotherapeutic drug PTX, and a COA-modified 4-arm PEG polymer (4PSC) was synthesized. COA, an active component in oleanolic acid that exerts strong antitumor activity by downregulating Hsp90 expression, was used as a structural and functional element to endow 4PSC with redox responsiveness and Hsp90 inhibitory activity. Our results showed that 4PSC/PTX nanomicelles efficiently delivered PTX and COA to tumor locations without inducing systemic toxicity. By blocking the Hsp90 signaling pathway, 4PSC significantly enhanced the antitumor effect of PTX, inhibiting tumor proliferation and invasiveness as well as chemotherapy-induced resistance in vitro. Remarkable results were further confirmed in vivo with two preclinical tumor models. These findings demonstrate that the COA-modified 4PSC drug delivery nanosystem provides a potential platform for enhancing the efficacy of chemotherapies.

Graphical abstract

The schematic illustration of the chemotherapy failure (left) and chemotherapy therapeutic strategy of 4 PSC/PTX nanomicelles (right).

1. Introduction

Chemotherapy is still one of the main therapeutic modalities used in clinical tumor treatment. Although properly administered chemotherapy can kill cancer cells, recurrence is nearly inevitable¹. Many studies have shown that tumor recurrence is closely related to drug resistance and metastasis. A key mechanism of these chemotherapeutic agents involves the activation of the heat shock protein 90 (Hsp90)-related pathway^2,3. Hsp90 is a heat shock family protein that is highly expressed in tumor cells, such as lung cancer, ovarian cancer and breast cancer, and tumor cells with sustained high expression of Hsp90 can become resistant to paclitaxel (PTX)⁴, cisplatin⁵, doxorubicin⁴ and other chemotherapeutic drugs^3,6, reducing the sensitivity of cells to chemotherapy⁷. High expression of Hsp90 in leukemia cells have been found to be involved in the generation of drug resistance in tumors⁸. In contrast, inhibition of Hsp90 expression can inhibit the proliferation of tumor cells⁹. Therefore, Hsp90 has become the target of antitumor therapy. In addition, the pharmacological inhibition of Hsp90 may limit the molecular pathways involved in the epithelial-to-mesenchymal transition (EMT), migration, and metastasis in vitro and in xenograft transplantation of different cancer cell lines from different tissue types into animals^10,11. Hsp90 has been reported to mobilize cancer cells and promote their spread through several key molecular pathways of cell migration and invasion¹¹. Moreover, Hsp90 increases the expression of p-glycoprotein (P-gp), an important protein that mediates multidrug resistance in cancer^4,12. In summary, high expression of hsp 90 contributes to drug resistance and cancer cell migration, contributing to tumor recurrence¹³. Therefore, targeting Hsp90 is regarded as an effective chemotherapeutic strategy for tumor elimination.

Many Hsp90 inhibitors played a key role in anti-cancer, cancer resistance and metastasis progress. For example, 17-(demethoxy)-17-allylamino geldanamycin (17-AAG), a less toxic analog of geldanamycin, displays high binding affinity with N-terminal domain of Hsp90, which prevents ATP binding through steric inhibition¹⁴. The second generation of Hsp90 inhibitors, such as NVP-AUY922 inhibits proliferation of various human cancer cell lines¹⁵. Other Hsp90 inhibitors, such as SNX-2112 inhibits cancer growth and induces apoptosis through inactivating the unfolded protein response in Huh 7 xenograft models¹⁶. However, few Hsp90 inhibitors have been developed into new drugs.

Our previous study reported that 3-O-(Z)-coumaroyl oleanolic acid (COA), an active component of oleanolic acid, not only exerts strong antitumor activity against A549 human lung carcinoma cells¹⁷ but also overcomes lung cancer drug resistance by inhibiting the Hsp90 and MEK pathways in vitro and in vivo¹⁸. Importantly, combining COA with PTX enhances anti-lung cancer drug resistance^18,19. However, COA is insoluble in water, and its low biocompatibility limits its future clinical application. Drug delivery systems can be used to solve biocompatibility problems and thus enhance treatments for lung cancer resistance and metastasis.

To inhibit tumor recurrence and eliminate tumors, we hypothesized that a nanodrug carrier system that codelivers chemotherapeutic agents and hydrophobic COA to the tumor site can be created to target Hsp90 and thus improve chemotherapeutic effects. Constructed through disulfide bonds (SS), 4-arm polyethylene glycol (PEG)–SS–COA (4PSC) was prepared as an amphiphilic copolymer, and hydrophobic PTX was encapsulated to yield 4-arm-PEG–SS–COA/PTX (4PSC/PTX) nanomicelles. These carriers have the following advantages: 1.4-arm PEG not only can increase the hydrophilicity of COA, making it conducive to long-term circulation and biosecurity in vivo, but can also increase the amount of COA carried to an increase in the amount of COA successfully delivered. 2. The increase in the hydrophobicity of the system not only can reduce the leakage caused by excessive CMC but can also increase the capacity for packing PTX. 3. COA, a hydrophobic component, is a biopatible material when combined with a hydrophilic material by chemical bonding to obtain an amphiphilic copolymer; therefore, it can reduce the toxicity of biomaterials in nanoparticles. 4. High levels of glutathione (GSH) in cancer cells promote the release of COA and PTX from the 4PSC/PTX micelles through the rapid dissolution of disulfide bonds, which is beneficial for killing tumor-resistant cells²⁰.

In this study, we pioneered redox-responsive COA-modified 4-arm PEG polymer generated through disulfide linkages (4PSC) to use as a drug nanocarrier in tumor therapy. This 4PSC nanosystem, which can effectively deliver two drug molecules, COA and PTX, into local tumors, enters cancer cells through the enhanced permeability and retention (EPR) effect. COA and PTX are quickly released in the presence of high GSH levels, and COA decreases P-gp, ATP, E-cadherin, and β-catenin levels by targeting Hsp90, further enhancing the inhibition of cancer cell proliferation, metastasis, and drug resistance and cooperating with PTX to suppress cancer recurrence (Scheme 1). Hence, 4PSC containing COA segments is a promising drug delivery nanosystem for use in tumor eradication.

Scheme 1.

The schematic illustration of the chemotherapy failure (left) and chemotherapy therapeutic strategy of 4 PSC/PTX nanomicelles (right).

2. Materials and methods

2.1. Materials

4-Arm PEG (M_w = 5000 Da) was purchased from Macklin Co., Ltd. (Guangzhou, China). COA (batch number, 130506XF; purity >99%) was purchased from Shanghai KANGLANG Biomedical Ltd (Shanghai, China). 4-Dimethylaminopyridine (DMAP), N,N′-dicyclohexylcarbodiimide (DCC), coumarine-6 (C6), amiloride, genistein, chlorpromazine, GSH (reduced), and PTX were purchased from Sigma–Aldrich (Shanghai, China). Bis(2-hydroxyethyl) disulfide was obtained from TCI Development Co., Ltd. (Shanghai, China). FBS and RPMI 1640 medium were purchased from HyClone and Gibco, respectively (Thermo Fisher Scientific Co., Ltd., Shanghai, China). The near-infrared (NIR) fluorescent dye DIR was supplied by Keygen Biotech (KGMP0026, China). Hematoxylin and eosin (H&E) staining reagents were supplied by Leagene Biology Technology Co., Ltd. (Beijing, China). LysoTracker® Red was purchased from Yeasen Biotech Co., Ltd (Shanghai, China). All other reagents were of analytical grade.

2.2. Synthesis of PPA-CDI

N,N-Carboxycarbonyl diimidazole (CDI, 0.18 mmol, 29.79 g) and 200 mL of anhydrous CH₂Cl₂ and propargyl alcohol (PPA, 5.82 mL) were added dropwise into a dry round-bottom flask and stirred for 1 h at room temperature. At the end of the reaction, 200 mL of water was added, and the excess CDI reacted with the water to generate imidazole and CO₂. The organic phase was separated, and the solvent was removed via rotating evaporation to obtain a colorless liquid, which was placed in the refrigerator as a white solid; at 28.97 g, and the yield was 85%.

2.3. Synthesis of PPA-Cys

Cysteamine (4.40 g, 28.52 mmoL) and PPA-CDI (3.47 g, 23.1 mmoL) were dissolved in 50 mL of anhydrous CH₂Cl₂ and reacted at room temperature for 24 h CH₂Cl₂ was evaporated and added to NaH₂PO₄ (100 mL, 1.0 mol, pH 4.0) in solution, and it immediately became viscous. After stirring for 1 h, 40 mL of diethyl ether was added and stirred for 20 min. After stratification, the water layer was extracted with diethyl ether. After separating the water layer, the pH was adjusted to 9.0 with NaOH (60 mL, 1.0 mol, Na₂H₂PO4) solution. The yield was 73.60%.

2.4. Synthesis of COA-PPA

COA (100.50 mg, 0.22 mmol) and PYBOP (171.73 mg, 0.33 mmol) were dissolved in 3 mL of a dimethylformamide (DMF) solution, and the reaction was completed in 10 min. PPA-Cys (18.17 μL, 0.33 mmol), DIPEA (85.29 μL, 0.66 mmol) and 2 mL of DMF were added under N₂ for 24 h. After the reaction, DMF was dried by an oil pump and extracted 3 times with saturated salt water. After drying, anhydrous MgSO₄ was purified by column chromatography with petroleum ether/ethyl acetate (4/1, v/v). The yield of the COA-PPA was 70.56%.

2.5. Synthesis of 4 PSC

COA-PPA (1.04 g, 2.09 mmoL), 4-arm PEG-N₃ (0.84 g, 0.35 mmoL), CuSO₄·5H₂O (72.99 mg, 0.29 mmoL), and Vitamin c (62.05 mg, 0.31 mmoL) were dissolved in dimethyl sulfoxide (DMSO), and refluxed at 50 °C for 2 days under N₂, and the products were put into dialysis bags (MWCO, 3000 Da). Distilled water with EDTA-2Na was used for dialysis (2 days), and the products were collected after freeze-drying. The yield of 4 PSC was 65.45% (Supporting Information Scheme S1).

2.6. Preparation of PTX-loaded 4PSC (4PSC/PTX) nanomicelles

PTX-loaded nanomicelles were prepared using a previously described nanoprecipitation method²¹. In brief, PTX (10 mg) was dissolved in DMF (10 mL). Then, 4PSC (100 mg) was added to the solution with magnetic stirring for 30 min. The mixed solution was transferred into a dialysis bag (MWCO 1000 Da) and dialyzed against organic solvent at room temperature. 4PSC/C6, 4PSC/DIR, 4PC/PTX, and 4PSB/PTX were prepared using the same procedure. When PTX-loaded micelles were finally obtained, the size, potential, and shape of these micelles were assessed by Malvern Zetasizer Nano ZS90 and transmission electron microscopy (TEM). The amount of drug released from the PTX-loaded nanomicelles was measured under different conditions in vitro by HPLC. The drug loading capacity (DLC) and DLE were calculated as described previously²². The percentage engrafted was calculated as followingEq. (1)^;23:

A (%) = (B–C) / C × 100 (1)

where A is grafting, B is the mass of 4 PSC, C is the mass of 4-arm PEG.

The stability of the 4PSC/PTX nanomicelles in PBS containing 10% FBS at 37 °C was examined using the protocol described in reference²⁴.

2.7. In vitro drug release

PTX-loaded nanomicelles (0.5 mg) were incubated in the PBS buffer (pH 7.4, 1.5 mL) or acetate buffer (pH 5.0 or 6.0, 1.5 mL) with or without GSH (5 × 10⁻³ mol/L or 10 × 10⁻³ mol/L) under stirring (100 rpm). At predetermined time intervals, 2 mL of medium was withdrawn and replaced with the same volume of release medium. The release profile of PTX and COA was measured by the HPLC.

2.8. Cell culture

Human lung adenocarcinoma cells (A549 cells and A549/ADR cells) and mouse breast cancer cells (4T1 and 4T1/ADR cells) were obtained from the College of Pharmaceutical Science, Guangzhou Medical University (Guangzhou, China). The cells were cultured in RPMI 1640 (with 10% FBS, 100 unit/mL penicillin and 100 μg/mL streptomycin). All cell-based protocols were performed according to the guidelines of the Institutional Animal Care Committee and Ethics Committee at Guangzhou Medical University (GZMUC 10-05010).

2.9. Intracellular drug concentration

A cell suspension (containing 1 × 10⁶ cells with Taxol, 4PC/PTX, 4PSC/PTX treatment) was placed in a homogenizer and ground repeatedly with tissue homogenizer until cell fragmentation was achieved. The cell lysate was transferred to a 15 mL centrifuge tube, 5 mL of tert-butylmethyl ether was added as the extraction reagent, and the upper organic phase was centrifuged at 3500 rpm (Allegra 64r, BECKMEN COULTER, Germany) for 15 min after vortexing for 5 min. The residue was redissolved with 50 μL of methanol and centrifuged at 3500 rpm (Allegra 64r, BECKMEN COULTER, Germany) for 5 min, and 20 μL of supernatant was taken for sample injection and HPLC determination.

2.10. Mechanism of cellular uptake and endocytosis

A549 cells were seeded in 6-well plates (10⁵ cell per well) and cultured for 24 h. Next, the medium was replaced with fresh 1640 containing free C6, 4PC/C6, 4PSB/C6, or 4PSC/C6 at a COA concentration of 5 μg/mL, respectively. After incubation for 2 h, the cells were washed, harvested, and analyzed by flow cytometry (Canto II, BD Company, USA).

A549 cells were seeded on 6-well plates (10⁵ cells per well) for 12 h, followed by washing with PBS. The medium was replaced with fresh RPMI 1640 containing 1 mmol/L 5-(N-ethyl-N-isopropyl)-amiloride, 10 μg/mL chlorpromazine, and 200 μmol/L genistein for 30 min. The cells were treated with 4PSC/C6 (5 μg/mL) nanomicelles at 37 °C for 2 h. The cells were rinsed 3 times with cold PBS. Finally, the A549/ADR cells were imaged by confocal laser scanning microscopy (CLSM, Olympus, FV1000, Central Valley, PA, USA).

2.11. Lysosome escape

A549 and A549/ADR cells were seeded in a special confocal microscopy dish (NEST) at a density of 5 × 10⁴ cells/well. After 24 h, the cells incubated with free C6 or C6-loaded nanomicelles for 2, 4, 7, or 10 h were stained with 1 μmol/L LysoTracker Red for 30 min and 2 μg/mL DAPI for 10 min at 37 °C in the dark. Finally, these cells were washed with cold PBS and imaged by CLSM (Olympus).

2.12. Cytotoxicity and hemo-compatibility of nanomicelles in vitro

A549, A549/ADR, 4T1, and 4T1/ADR cells were seeded in a 96-well plate at 1 × 10⁴ cells per well. Twelve hours later, the cell culture medium was replaced with fresh medium containing various nanomicelles. The blank control consisted of culture medium alone. After 48 h, cell viability was determined with Cell Counting Kit-8 (CCK-8) assay²⁵. The half-maximal inhibitory concentration (IC₅₀) values of different drug formulations were calculated by SPSS 16.0 software (IBM, New York State, USA). The synergistic effects between COA and PTX in the 4PSC/PTX nanomicelles was evaluated by the following Eq. (2)^;26:

CI = COA 1/COA 0 + PTX 1/PTX 0 (2)

where CI is combination index, COA 1 and PTX 1 represent the IC₅₀ values of drugs used in the combination treatment, and COA 0 and PTX 0 represent the IC₅₀ values of the single drugs. CI < 1 indicates drug synergism, whereas CI >1 shows an antagonistic effect.

The hemo-compatibility of nanomicelles was examined using the protocol described in reference²⁴.

2.13. Establishment of a hyperthermic cell model

Growing human lung cancer cells were seeded into a 60-mm cell culture dish at a density of 1 × 10⁶ cells/mL. When the cell confluence was approximately 80%–90%, the Petri dish was placed in a water bath for heat shock treatment (43 °C for 0–3 min) and then placed in a CO₂ incubator at 37 °C. Fresh medium containing various PTX-loaded nanomicelles was added and incubated for 48 h when the temperature had decreased to room temperature, the relative viability of cells was examined by Cell Counting Kit-8 (CCK-8) assay as described above.

2.14. Transwell migration assay

Corning transwell 3422 24-well plates (8-μm pore size) were used for a cell migration assay. A549 or A549/ADR cells (10⁵ cells per well, 100 μL) were seeded in the upper chamber in medium without FBS. Each bottom chamber was supplemented with medium (containing 10% FBS) containing various PTX-loaded nanomicelles (6 μg/mL), and the plates were incubated for 24 h. The cells that had migrated to the bottom surface of the membranes were stained with crystal violet, and the cells on the upper surface of the membrane were removed. The cells were imaged, and the percentage of cells that migrated was calculated by ImageJ-Pro Plus 6.0 software (National Institutes of Health, Bethesda, MD, USA).

2.15. Scratch test

A549 or A549/ADR cells at logarithmic growth stage were inoculated in 6-well plates (1 × 10⁶ cells/well) and cultured in a 5%CO₂ incubator at 37 °C. When the cells reached 80%–90% confluence, the medium was replaced with serum-free medium, and the cells were starved for 24 h. A scratch perpendicular to the horizontal line was made with a 10-μL pipette tip. Sterile PBS was used as a wash to remove floating cells, and then, the cells were cultured in 3 mL of medium containing Taxol, Taxol + COA, 4PSB/PTX, or 4PSC/PTX at a PTX concentration of 6 μg/mL for 24 h at 37 °C in 5% CO₂ and saturated humidity. A microscope and photographs were used to observe the condition of the healing cells, and then, the migration distance between the scratches was measured.

2.16. Western blotting and quantitative RT-PCR

Human lung cancer cells were seeded and cultured in a 6-well plate at 4 × 10⁵ cells/well at 37 °C in 5% CO₂ for 48 h and treated with Taxol, Taxol + COA and PTX-loaded nanomicelles at a final PTX concentration of 10 μg/mL for 12 h. Western blotting was used to determine the protein levels in human lung cancer cells as described previously²⁹. Antibodies against Hsp90, P-gp, E-cadherin, β-catenin, and β-actin, obtained from Cell Signaling Technology (Beverly, MA, USA), were used in this experiment. The membranes were rinsed, and the signals were visualized using an enhanced chemiluminescence detection kit. The total intensity of the band for each protein was calculated with ImageJ and normalized to that of β-actin²⁷.

Real-time PCR assays were performed to analyze the expression of P-gp, E-cadherin, β-catenin and related genes²⁸. In summary, RNA was extracted using a TRIzol® Plus RNA Purification kit (Invitrogen, ThermoFisher, CA, USA). The extracted and purified RNA samples (500 ng) were reverse transcribed into cDNA using SuperScript™ III First-Strand Synthesis SuperMix (Invitrogen). cDNA samples were used as templates to perform a standard PCR analysis with Power SYBR® Green PCR Master Mix (Invitrogen), and the PCR products of human lung cancer cells that had been subjected to different treatments were detected with a Real-Time PCR Detection System (CFX384, Bio-Rad, CA, USA).

2.17. In vivo imaging assay

To establish an 4T1/ADR xenograft tumor model, 4T1/ADR cells (1 × 10⁷ cells in 200 μL of PBS) were subcutaneously injected into the right flank of BALB/c mice (5 weeks old, weighing 16–18 g). When the volumes of the tumors reached approximately 200–230 mm³, the mice were intravenously administered 100 μL of free DIR or DIR-loaded nanomicelles, which were prepared using a similar procedure as 4PSC/PTX nanomicelles. At predetermined time points (2, 4, 8, 12 and 24 h), images were obtained with an NIR fluorescence imaging system²⁴. Finally, 24 h post-injection, the mice were sacrificed, and the main organs were imaged by assessing the DIR signals. All animal experiments were carried out in accordance with the guidelines of the Animal Care and Use Institution of Guangzhou Medical University Committee. The fluorescence intensity was further analyzed by indiGOTM software (Berthold Technologies, Germany).

2.18. In vivo antitumor activity assay

BALB/C nude mice bearing A549/ADR tumors were randomly assigned to six groups when their tumors reached a volume of approximately 90 mm³ (five mice per group): PBS, Taxol, Taxol + COA, 4PC/PTX, 4PSB/PTX, or 4PSC/PTX nanomicelles (4 mg/kg PTX or 13.5 mg/kg COA) were administered through intravenous injection 4 times at intervals of three days. Tumor growth and body weight were assayed with a caliper every two days, and the tumor volume was calculated as Eq. (3):

V = 0.5 × (Width)² × (Length) (3)

where V is volume.

The tumor volume inhibition rate was calculated on Day 28 using Eq. (4)

Rv = 100% − (V_drug/V_saline) × 100% (4)

where V_drug indicated the tumor volume after drug treatment, and V_saline indicated the tumor volume after treatment with PBS.

BALB/c mice were sacrificed, and the tumors and primary organs were collected. For RT–PCR assays, the tumors were fixed in 10% formalin and embedded in paraffin blocks to prepare tissue sections at a thickness of 5 μm. After deparaffinization, the tissue sections were stained with H&E and visualized with an optical microscope (Shenzhen, China).

Twenty-five BALB/C nude mice (female, 6–8weeks old) were administered 4T1/ADR tumor cells (0.5 × 10⁷ cells in 200 μL of PBS) by intravenous injection to generate a breast cancer-bearing mouse model with lung metastasis²⁴. PBS, Taxol, 4PSB/PTX nanomicelles, or 4PSC/PTX nanomicelles were administered to the mice on Day 6, as described above. Tumors and vital organs were isolated, weighed, imaged and fixed after the mice were sacrificed on Day 28. The mouse lungs were dissected, and metastatic nodules on the surface of the lungs were counted²⁴. Lung tissue was used for the Western blot assay.

2.19. Statistical analysis

GraphPad Prism 5.0 software (San Diego, CA, USA) was used for statistical analyses. All experiments were performed in triplicate and quantitatively assessed. Quantitative values are expressed as the means ± standard deviation (SD). One-way ANOVA was performed to determine the significance of the differences. The significance of the results is reported with the following annotations: ns, P < 0.05, P < 0.01, or P < 0.001. ImageJ software was applied for quantitative analysis.

3. Results

3.1. Characteristics of 4PSC/PTX nanomicelles

The synthesis of 4PSC is divided into two steps. The first step involves synthesizing 4-arm PEG-N₃ and PPA-COA. The second step involves producing 4PSC through click chemistry with COA-PPA:4-armPEG-N₃ at a mole ratio of 6:1.¹H NMR spectra revealed a signal corresponding to 4-arm PEG at 3.6–3.7 ppm (‒[CH2CH2O]_n‒), and the signal at 5.25 ppm (‒CH=CH–C‒) was assigned to COA; interestingly, the integral signal ratio of these components was approximately 110:1, leading to the inference that 4 COA molecules were added to 4-arm PEG through disulfide bonds. In addition, the characteristic peak of the aromatic ring in COA appeared at 7.36–7.65 ppm (Supporting Information Fig. S1).

In the infrared image of 4PSC, we observed an absorption peak at 1100 cm⁻¹ and attributed it to the C–O–C stretching vibration in 4-arm PEG‒N₃. In contrast to 4-arm PEG‒N₃, 4PSC showed characteristic carbonyl (C=O stretching vibration in COA-PPA) and amide (C=O stretching vibrations in COA‒PPA, I band) peaks at 1730 cm⁻¹ and 1630 cm⁻¹, respectively. Importantly, the absorption peak at 2100 cm⁻¹ (the azide group of PEG‒N₃) was significantly smaller and ultimately disappeared, which indicated alkynyl and azide base reactions between COA‒PPA and 4-arm PEG‒N₃, confirming that 4PSC was successfully prepared (Supporting Information Fig. S2). DSC detection results showed some difference among of 4PSC, COA, 4-arm PEG, and COA + 4-arm PEG (Supporting Information Fig. S3). 4PSC has absorption peaks at approximately 280 and 410 °C, respectively. Due to the influence of covalent bonds, these absorption peaks in 4 PSC had obvious change compared with other groups, indicating the success of 4PSC synthesis. In addition, to study the specific effects of COA, we replaced COA in 4PSC with phenyl groups because of the similar structure. Thus, 4PSB was used as a control (Supporting Information Scheme S2). Similarly, 4PC was also used as a control; it was synthesized following the process shown in Supporting Information Scheme S3.

Dynamic light scattering (DLS) showed that the average diameter of the 4PSC/PTX nanomicelles was 140 ± 2.6 nm, which was approximately 14 nm larger than that of the 4 PSC nanomicelles. The zeta potential was −14.9 ± 2.3 mV. Furthermore, high PTX-loading capacity (LC) of the 4PSC/PTX nanomicelles (5.3%) was established at a drug-loading efficiency (DLE) of 82.7%. On the basis of the percentage engrafted, the amount of loaded COA onto 4PSC was determined to be 20.9% (Table 1). In other words, a mass ratio of PTX:COA = 1:3.2 was obtained. These results indicated that when the proportion of hydrophobic ends in a 4PSC nanomicelles was increased, the amount of PTX encapsulated and the particle size were increased. Using a pyrene fluorescent probe, CMC of 4PSC was found to be 8.35 mg/L, which is relatively low in comparison to the CMC values of other PEGylated polymeric micelle NPs; this is an advantage because NPs with low CMC values are suitable for pharmaceutical applications²⁹. As shown in the transmission electron microscopy (TEM) images presented in Fig. 1A, 4PC/PTX, 4PSB/PTX, and 4PSC/PTX were uniformly sized spheres. In addition, the particle size and PDI did not change significantly after a 96-h incubation in medium containing 10% fetal bovine serum (FBS), indicating that the structures of these three PTX-loaded nanomicelles showed good stability (Fig. 1B, Supporting Information Fig. S4).

Table 1.

Characterization of every nanomicelles.

Formulation	Particle size (nm)a	PDI	Zeta potential (mV)a	DLC (%)	DLE (%)
4PC/PTX	115 ± 1.02	0.215	‒‒10.12 ± 0.32	4.12	83.65
4PSB/PTX	135 ± 1.21	0.164	‒12.35 ± 0.25	5.12	79.56
4PSC/PTX	140 ± 0.89	0.135	‒14.95 ± 0.15	5.30	82.70
4PC	98 ± 0.56	0.142	‒11.25 ± 0.26	N/A	N/A
4PSB	122 ± 1.32	0.156	‒13.24 ± 0.35	N/A	N/A
4PSC	126 ± 1.34	0.104	‒15.85 ± 0.18	N/A	N/A

^aData represented as mean ± SD, n = 3. N/A indicated nothing.

Figure 1.

Characterization of 4PSC/PTX nanomicelles and stimulus-responsive release of PTX (A) TEM images of 4PC/PTX, 4PSB/PTX, and 4 PSC/PTX nanomicelles. Scale bar = 200 μm (B) The size changes of 4PSC/PTX nanomicelles in during incubation in PBS containing 10% FBS at 37 °C for 120 h (C) In vitro COA (1) and PTX (2) release from 4PSC/PTX nanomicelles in PBS (with various GSH concentrations) (D) Flow cytometry analysis of A549 cells incubated with free C6 or C6-loaded nanomicelles for 2 h (E) In vitro COA (1) and PTX (2) release from 4PSC/PTX nanomicelles in PBS (with different pH) (F) Retention of COA (1) and PTX (2) in A549/ADR cells after preincubation with 4PSC/PTX, 4PC/PTX, and taxol for different time periods. All data are presented as the means ± standard deviations (n = 3); ∗P < 0.05, ∗∗P < 0.01.

As presented in Fig. 1C, without GSH, the amount of COA and PTX released from the 4PSC/PTX nanomicelles was approximately 20% in 12 h. In contrast, in the presence of 5 or 10 mmol/L GSH for 96 h, the 4PSC/PTX nanomicelles released 78.65% and 87.52% of the loaded PTX, respectively, which was more than the amount of COA released, although disulfide bonds were rapidly broken in the presence of the corresponding concentrations of GSH. The reason may be that the dissociation of forces depending on physical adsorption is less difficult than that the dissociation of covalent bonds in 4PSC. Buthionine sulphoximine (BSO), which is capable of downregulating the intracellular glutathione level by inhibiting γ-glutamylcysteine synthetase, was used to verify the redox-responsive dual release of PTX and COA (Fig. 1D). Due to the high GSH concentrations in lung cancer cells, no significant difference in PTX release from 4PSC/PTX and 4PSB/PTX nanomicelles was found (Supporting Information Fig. S5).

In addition to the disulfide bond, 4PSC contains an amide bond and carbamate linkages. It is known that the amide bond is stable at relatively low pH and in the presence of esterase. However, carbamate linkages are stable in the presence of esterase but not at low pH³⁰. In other words, drug release from 4PSC increased when the pH was gradually lowered. As shown in Fig. 1E, 4PSC/PTX showed sustained release of PTX at pH 7.4: only 23% was released in 96 h. In the same time period, 4PSC/PTX micelles released 45% of PTX at pH 6.5 and 60% at pH 5.0, indicating that the acidic tumor environment has an impact on drug release from 4PSC. Although this increase is not as large as that of 10 mmol/L GSH, the release of PTX in 4PSC/PTX reached 72% at pH 5.0 + 5 mmol/L GSH for 96 h, which was higher than that under 5 mmol/L GSH alone (Supporting Information Fig. S6). The GSH concentration in lysosomes is almost 1–5 mmol/L, which is almost 2 times lower than that in the cytoplasm of tumor cells. This is because the acidic lysosome increases the standard redox potential between GSH and its oxidized form glutathione disulfide, thus, significantly slowing the reduction rate of disulfides related to GSH³¹. Within contrast to PTX release, COA release in 4PSC/PTX is similar at different GSH concentrations, indicating that the dual drug release from 4PSC/PTX depended partly on the pH difference.

High expression of P-gp protein results in PTX efflux. Therefore, the retention of 4PSC/PTX nanomicelles in A549/ADR cells was assayed. To prevent an initial concentration effect, 5.5 μg/mL PTX-loaded nanomicelles or 42 μg/mL Taxol was added and incubated with A549/ADR cells for 4 h²⁰. After the cells were washed with PBS and incubated in fresh medium for different periods of time, HPLC was used to quantitatively examine the intracellular PTX concentrations. The results showed that 69.9% of the PTX and 61.5% of COA from 4PSC/PTX nanomicelles were retained in A549/ADR cells, and both values were significantly higher than those obtained by using 4PC/PTX nanomicelles (53.5% for PTX, 48.5% for COA) and Taxol (8.6%), indicating that the 4PSC/PTX nanomicelles had a clearly lower rate of drug efflux, contributing in redox-responsive nanomicelles, which enhanced intracellular accumulation and retention of PTX in MDR cancer cells (Fig. 1F).

3.2. 4 PSC/PTX mechanism of endocytosis and lysosome escape

To investigate the mechanism of endocytosis and lysosome escape of the 4PSC/PTX nanomicelles in human lung cancer cells, which is important for improved treatment, 4PSC was specifically prepared with coumarine-6 (C6), a fluorescence probe, because the polarity of the probe was similar to that of PTX²¹. Fig. 2A shows that cells pretreated with the endocytosis inhibitor (amiloride) internalized the 4PSC/C6 nanomicelles at a lower rate, as indicated by the weak green fluorescence and dramatic inhibitory effects, compared with rate of genistein and chlorpromazine uptake, indicating that the 4PSC/C6 nanomicelles entered these cancer cells through the main macropinocytosis-mediated endocytosis pathway.

Figure 2.

Cellular uptake and lysosomal escape of 4PSC/PTX nanomicelles (A) CLSM images of A549 cells pretreated with chlorpromazine, genistein, and amiloride and then treated with 4 PSC/C6 nanomicelles. Scale bar = 20 μm (B) CLSM images of lysosomal escape of nanomicelles in A549 and A549/ADR cells. Green fluorescence signal corresponds to C6 at 2, 4, 6, or 8 h, respectively. The late endosomes and lysosomes were labeled by LysoTracker Red. Yellow fluorescence signal corresponds to overlay of the C6 and lysosomal signals. Scale bar = 50 μm. (1) CLSM; (2) The intensity of the fluorescence signal in the corresponding images or Pearson coefficient and Mander coefficient. All data are presented as the means ± standard deviations (n = 3); ∗∗P < 0.01, ∗∗∗P < 0.001.

To investigate the machinery involved in lysosome escape, LysoTracker Red was assayed in the experiment. The colocalization of the red fluorescence of lysosomes and green fluorescence of C6 resulted in a merged yellow signal. Fig. 2B shows that A549 cells endocytosed nanomicelles and emitted a combination of signals. The strong green fluorescence from 4PSC/C6 was maintained from 2 to 8 h. The yellow fluorescence from 4PSC/C6 gradually increased from 2 to 4 h, but gradually decreased from 6 to 8 h, and the signal became dark red, suggesting that a large amount of C6 had escaped from the lysosomes. Compared with the 4PSC/C6 nanomicelles in A549, a weak yellow signal of 4PSC/C6 in A549/ADR was obtained at 4 h. This signal decreased over time and was mostly dark red at 8 h, suggesting that C6 eventually escaped the lysosome and was distributed in the cytoplasm. This result is demonstrated in section 3.5. It is possible that the release of the drugs from 4PSC/PTX was partly attributable to the large amounts of amidase and GSH in lysosomes, but the lipid solubility of PTX and COA helped them escape by fusion with lysosomes³¹.

3.3. 4 PSC improves PTX antitumor activity via COA-induced inhibition of Hsp90

PTX-sensitive cancer cells (A549 and 4T1 cells) and PTX-resistant cancer cells (A549/ADR and 4T1/ADR cells) were used to assess 4 PSC/PTX nanomicelles antitumor activity (Supporting Information Fig. S7). According to the mass ratio of COA and PTX in 4PSC/PTX nanomicelles, we established PTX/COA = 1/3 (m/m?) in the control group. As shown in Fig. 3A, the cytotoxic effect on cancer cells in all treatment groups significantly depended on the drug concentrations. However, the survival rate of A549/ADR and 4T1/ADR cells was only slightly decreased after incubation with Taxol and 4PSB/PTX nanomicelles, suggesting that A549/ADR and 4T1/ADR cells were relatively resistant to PTX even at high concentrations. However, 4PSC/PTX had higher antitumor activity against A549/ADR and 4T1/ADR cells. The IC₅₀ value of 4PSC/PTX nanomicelles was approximately 3- to 15- fold lower than that of Taxol or 4PSB/PTX and 1- to 3-fold lower than that of Taxol + COA or 4PC/PTX nanomicelles, indicating that COA promoted the PTX antitumor effect (Table 2). This finding revealed that 4PSC/PTX nanomicelles that co-delivered two different anticancer drugs (COA and PTX) into cancer cells resulted in an excellent anticancer effect and synergistic effect in vitro. The best synergistic ratio (COA/PTX = 1/3, m/m) was consistent with our previous data (Supporting Information Fig. S8). The synergistic effect was also demonstrated by the calculated combination index (CI) (0.17/0.68, for A549/A549ADR) and (0.21/0.74, for 4T1/4T1/ADR), which were smaller than 1. In fact, this increased cytotoxic activity was associated with a decrease in Hsp90 expression, and it has previously been shown that antitumor activity is reduced in PTX-based therapies in a time-dependent manner when heat-induced Hsp90 expression is increased (Fig. 3B).

Figure 3.

The enhanced antitumor activity of 4PSC/PTX (A) Cytotoxicity of Taxol and PTX-loaded nanomicelles in human lung adenocarcinoma cells (A549 cells and A549/ADR) and mouse breast cancer cells (4T1 and 4T1/ADR) cells after incubation for 48 h (B) Cytotoxicity of 4PSC/PTX in (A549 and A549/ADR) and (4T1 and 4T1/ADR) after heat for different time (2.5 μg/mL of PTX in PTX-sensitive cancer cells; 8.0 μg/mL of PTX in PTX-resistant cancer cells). The data are presented as the mean ± standard deviation (n = 3). ∗P < 0.05, ∗∗P < 0.01. NS indicated no significant difference.

Table 2.

The IC₅₀ values in human lung cancer and mouse breast cancer cell incubated in the presence of PTX and PTX loaded nanomicelles for 48 h (μg/mL).

Cells/drug	A549a	A549/ADRa	4T1a	4T1/ADRa
Taxol	7.73 ± 0.32	83.25 ± 5.21	11.44 ± 2.12	156.07 ± 10.35
Taxol + COA	2.17 ± 0.12	20.27 ± 2.32	2.68 ± 0.08	30.67 ± 5.21
4PC/PTX	5.18 ± 0.05	16.15 ± 1.33	5.29 ± 0.35	25.59 ± 2.38
4PSB/PTX	31.22 ± 2.52	38.18 ± 4.65	40.93 ± 4.23	97.83 ± 5.41
4PSC/PTX	2.94 ± 0.06	8.12 ± 1.56	3.51 ± 0.25	11.12 ± 0.88

^aData represented as mean ± SD, n = 5.

3.4. 4 PSC/PTX inhibits cancer invasiveness and chemoresistance in vitro

To evaluate the antimetastatic effects of the 4PSC/PTX nanomicelles. Transwell chamber assays and scratch migration experiments were performed to examine cancer cell ability to migrate through the extracellular matrix. E-cadherin is a typical cadherin that is also plays a key functional role in epithelial cell adhesion and in inhibiting the growth and adherence of invading tumor cells and in maintaining cell polarity and differentiation³². As a soluble cytosolic protein, β-catenin is an intercellular adhesion molecule involved in Wnt signaling and transcriptional activation³³. E-cadherin can bind β-catenin to form a complex that participates in a cell anchorage mechanism, and the abnormal expression of E-cadherin is closely related to the invasion, metastasis and prognosis of various malignant tumors^34,35. Fig. 4A shows that compared with the other treatment groups, the 4PSC/PTX nanomicelles effectively suppressed the migration of A549 and A549/ADR cells. The promigratory ability of 4PSC/PTX nanomicelles was further confirmed by the Transwell assay (Fig. 4B), which is another widely used method for evaluating cell migration. The expression of E-cadherin (β-catenin) was 17.55 (or 43.26), 13.45 (or 49.32), 38.35 (22.42), 72.36 (12.38), 35.21 (40.38), 70.25 (15.25), and 31.68 (37.43)% in A549 cells and 13.40 (76.23),10.58 (80.25), 17.36 (70.23), 29.32 (55.36), 22.21 (60.56), 41.45 (32.40), and 29.48 (56.62)% in A549/ADR cells treated with PBS, heat, Taxol, Taxol + COA, 4PSB/PTX, 4PSC/PTX, and 4PSC/PTX + heat, respectively, as determined by Western blot analysis (Fig. 4C). Similar mRNA expression results suggested that 4PSC/PTX markedly inhibited the migration of A549 and A549/ADR cells (Fig. 4D). Moreover, increased heat-induced Hsp90 expression was inhibited by 4PSC/PTX, demonstrating that COA enhanced the antimigration effects of PTX treatment. Our previous data showed COA can target and downregulate hsp 90 in drug resistance lung cancer. PTX can downregulate Hsp90 in A549, but or not A549/ADR due to PTX efflux pump. COA plays synergistic effect in 4PSC/PTX group (Supporting Information Fig. S9). COA inhibits Hsp90 activity like Hsp90 inhibitor PU-H71, the levels of Hsp 70 were assayed. Interestingly, expression of Hsp70 was significantly increased in both COA and PU-H71 treatments (Supporting Information Fig. S10), suggesting an inhibition of Hsp90 activity by COA. These provided strong evidence again that 4PSC/PTX inhibited A549/ADR through targeting Hsp90.

Figure 4.

4PSC/PTX suppressed cancer migration and drug resistance induced in vitro (A) The scratch wound healing assays of A549 and A549/ADR cells with heat, taxol, taxol + COA, and PTX-loaded nanomicelles treatment (B) The migration of A549 and A549/ADR cells with heat, taxol, taxol + COA, and PTX-loaded nanomicelles treatment. The levels of Hsp90, P-gp, E-cadherin, β-catenin protein (C) and mRNA (D) in A549 and A549/ADR cells treated with various agents. (1) Images; (2) quantitative analysis of cell scratch assay, migrating, or relative expression of proteins. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ^&P < 0.001 vs. 4PSC/PTX. Lane 1–7 is PBS, Heat, taxol, taxol + COA, 4PSB/PTX, 4PSC/PTX, 4PSC/PTX + heat, respectively. Scale bar = 200 μm.

A striking feature of tumor drug resistance is the high expression of P-gp, which indicates that the drug is pumped out of cells³⁶. Many studies have reported that drug-resistant tumor cells highly express Hsp90, and reducing Hsp90 expression can effectively reverse tumor drug resistance and induce cell apoptosis³⁷. In our study, compared with other treatment groups, 4PSC/PTX nanomicelles decreased P-gp protein and mRNA expression in A549 and A549/ADR cells. Interestingly, the inhibition rate of 4PSC/PTX nanomicelles in A549/ADR cells was close to that in A549 cells, indicating that 4PSC/PTX nanomicelles can overcome drug resistance by decreasing P-gp expression. Importantly, high P-gp expression in the cells treated with 4PSC/PTX nanomicelles had been activated through heat pretreatment because high Hsp90 expression promotes high P-gp expression, indicating that the COA released from the nanomicelles targets Hsp90 to suppress P-gp upregulation. Ultimately, the 4PSC/PTX nanomicelles enhanced anticancer activity. Although the expression of P-gp in the Taxol + COA-treated A549/ADR cells was lower than that in the Taxol-treated cells, the effect of treatment was not ideal (Figure 3, Figure 6A). Compared with 4PSC/PTX nanomicelles, redox-responsive nanotechnology-enhanced PTX more effectively inhibited invasiveness and the acquisition of drug resistance.

Figure 6.

Anticancer efficacy of 4PSC/PTX in A549/ADR tumor-bearing BALB/c-nude mice (A) Representative tumors isolated from the euthanized mice receiving treatment. 1, Image; 2, tumor growth. Tumor weight (B) and TIR (C) for A549/ADR tumor-bearing mice treated with PBS, Taxol, Taxol + COA, 4PBS/PTX, 4PC/PTX, or 4PSC/PTX over 28 d (n = 5 per group). Red arrows indicate the time points for treatment (3 d apart, 4 times, consecutively). TIR (%) = [1−X/Y] × 100%. X indicated the average weights of the tumors from the experimental groups; Y indicated the average weights of the tumors from control groups). The mRNA levels of Hsp90 (D), P-gp (E), E-cadherin (F), β-catenin (G) was assayed by RT-PCR. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

3.5. 4 PSC/PTX nanomicelles in vivo images of drug-resistant cancer xenografts in mice

To further demonstrate that 4PSC/PTX targeted cancer cells, the whole-body distribution of 4PSC/PTX nanomicelles was assayed. Fig. 5A shows the distribution of DIR-loaded 4PSCs (4 PSCs/DIR) in mice. Compared with 4PC/DIR, the 4PSC/DIR nanomicelles showed stronger fluorescence after 8 h possibly because they escaped lysosomes and entered the cytoplasm. In addition, DIR was quickly released in the presence of high GSH levels, and the gradual increase in the signal remained strong for 24 h.

Figure 5.

Biodistribution analysis of every formulations in A549/ADR cells xenografts in BALB/c nude mice after i. v. Of free DIR, 4PC/DIR, 4PSC/DIR nanomicelles (A) Real-time non-invasive whole-body imaging in vivo (B) Fluorescence images of tumor and organs collected. (1) Images. (2) Fluorescence intensity of image. Data are presented as the mean ± SD (n = 3). ∗P < 0.05, ∗∗P < 0.01.

The quantification of fluorescence intensity in major organs and tumors in mice was determined by region-of-interest (ROI) analysis. As shown in Fig. 5B, among all groups, the 4PSC/DIR nanomicelles group showed the most intense fluorescence signal in tumors. The fluorescence intensity was 5.76-fold and 2.04-fold higher than that of free DIR and 4 PC/DIR nanomicelles, respectively, indicating that the 4 PSC/DIR nanomicelles exhibited a longer circulation time and passive targeting.

3.6. 4PSC enhances PTX antitumor activity by inhibiting cancer metastasis and drug resistance in vivo

To investigate the antitumor ability of the 4PSC/PTX nanomicelles in vivo, two tumor models were assayed. First, resistant A549/ADR cell xenografts were established in nude mice. As shown in Fig. 6A, compared with PBS (control group), 4PSC/PTX nanomicelles effectively reduced A549/ADR tumor growth, and it showed the highest tumor inhibition rate (TIR, 74.75%), which was 9.21-, 3.63-, 2.92-, and 1.94-fold higher than that achieved with Taxol, Taxol + COA, 4PSB/PTX, and 4PC/PTX nanomicelles, respectively (Fig. 6B). This result was obtained by measuring the weight of tumors in each group (Fig. 6C) and is described in our preliminary findings (Supporting Information Fig. S11). 4PSC did not exhibit higher antitumor activity than 4PSC/PTX nanomicelles (data are not shown), suggesting that COA released from in 4PSC nanomicelles in vivo enhanced the PTX antitumor effect, which is consistent with the 4PSC-mediated enhancement of PTX antitumor activity observed in vitro.

COA has been previously shown to inhibit the proliferation of lung cancer cells by targeting Hsp90 and to act in synergy to enhance PTX-treated resistant lung cancer cells¹⁸. In this study, we found that compared with other treatment groups, 4PSC/PTX nanomicelles significantly decreased Hsp90, P-gp, and β-catenin expression and increased E-cadherin mRNA expression in A549/ADR tumor tissues, as determined by RT–qPCR (Fig. 6D–G). These findings indicated that combining COA and PTX effectively inhibited A549/ADR tumors at the transcriptional level.

Metastasis is a key factor in cancer recurrence, and therefore, the effect of 4PSC/PTX nanomicelles on distant metastases in a mouse model of breast cancer in situ with most commonly occurring metastases, i.e., in the lung, was investigated (the second model to be tested in vivo, Fig. 7A). The number of metastatic foci in the lungs treated with 4PSC/PTX nanomicelles was significantly reduced, by 2.05- and 1.78-fold, compared with that in the lungs treated with Taxol or 4PSB/PTX nanomicelles, respectively (Fig. 7B). Similar differences were also found for the protein expression in 4PSC/PTX nanomicelles-treated mouse lungs (Fig. 7C).

Figure 7.

Anticancer metastatic efficacy of 4PSC/PTX in 4T1 tumor-bearing BALB/c mice (A) Representative images of the metastatic lung cancer excised from each group. The metastatic nodules in each lung tissue by red arrow were showed. The levels of Hsp90, P-gp, E-cadherin, and β-catenin protein (B) in tumor treated with various agents. Representative images of H&E (C) stained organ section and body weight (D) of A549/ADR tumor-bearing mice with PBS, Taxol, 4PSB/PTX, 4PSC/PTX treatment in vivo. 1, Images; 2, quantitative analysis of relative expression of metastatic nodules, or proteins. ∗P < 0.05, ∗∗P < 0.01.

In addition to inhibiting cancer drug resistance and metastasis, the 4PSB/PTX nanomicelles exhibited outstanding biocompatibility and nonsystemic toxicity (Table 3, Fig. 7D). Supporting Information Fig. S12 shows the biocompatibility of 4PSC polymer in vitro. Although 4PSC contains COA, a much higher concentration of 4PSC (for example, 1 mg/mL) than that in the treated mice showed little effect on blood. In addition, there was no significant difference in body weight (Fig. 7D) or the results of routine blood examinations in any group (Table 3). Furthermore, the PEG modifications of 4PSCs resulted in beneficial long-term circulation in the body, and passive targeting of nanoparticles in the tumor through the EPR effect. Moreover, due to the addition of COA, targeting Hsp90 led to greater inhibition of tumor drug resistance and metastasis, truly preventing tumor recurrence in the future.

Table 3.

Hematological parameters of the mice with every form treatment on Day 28.

Parameter	PBSa	Taxola	4PSB/PTXa	4PSC/PTXa	Reference range
WBC (109/L)	8.20 ± 2.65	7.85 ± 3.25	7.85 ± 1.65	5.65 ± 0.56	0.80–6.80
Lymph (109/L)	7.36 ± 4.52	6.60 ± 2.32	6.15 ± 2.58	3.58 ± 0.14	0.70–5.70
Mon (%)	2.33 ± 0.06	1.45 ± 0.25	1.42 ± 0.32	0.22 ± 0.08	0.00–0.30
Gran (109/L)	9.07 ± 3.02	2.15 ± 0.32	1.65 ± 0.25	0.14 ± 0.02	0.10–1.80
RBC (1012/L)	7.39 ± 2.32	7.65 ± 1.54	7.88 ± 1.65	7.35 ± 2.01	6.36–9.42
HGB (g/L)	102.34 ± 40.12	125.32 ± 54.21	135.62 ± 43.11	123.25 ± 28.85	110.00–143.00
HCT (%)	32.65 ± 5.32	34.85 ± 7.65	40.25 ± 10.32	43.15 ± 6.25	34.60–44.60
MCV (fl.)	45.32 ± 3.65	40.25 ± 6.36	44.56 ± 2.85	40.65 ± 1.65	48.20–58.30
MCH (pg)	13.65 ± 1.25	16.85 ± 2.30	18.52 ± 2.30	18.35 ± 2.22	15.80–19.00
MCHC (g/L)	300.25 ± 25.3	332.25 ± 25.11	348.65 ± 30.12	348.36 ± 32.50	302.00–353.00
RDW (%)	13.56 ± 3.88	18.25 ± 3.25	15.45 ± 4.21	14.65 ± 2.02	13.00–17.00
MPV (fL)	6.32 ± 0.52	6.35 ± 0.62	5.25 ± 0.75	5.08 ± 0.87	3.80–6.00

^aData represented as mean ± SD, n = 3.

4. Conclusions

This study presented resolutions to the limitations of traditional chemotherapy regimens by reducing tumor volume. By targeting Hsp90 to block cancer metastasis and drug resistance, a smart drug delivery system (4PSC/PTX) was designed to inhibit lung cancer cell proliferation, drug resistance and metastasis. 4PSC/PTX nanomicelles were successfully prepared, and COA and PTX were effectively released into tumor sites in response to the intracellular redox environment. In vitro and in vivo cell biological, biochemical, and molecular biological data have previously demonstrated that conventional and nanomedicine-based chemotherapy regimens do not significantly inhibit cancer proliferation, metastasis or drug resistance; however, the 4PSC nanomicelles in this study sufficiently enhance PTX inhibition through COA targeting of Hsp90, which promotes the therapeutic efficacy of PTX-based chemotherapy. Thus, we describe a new and promising dual-drug delivery platform for the design and engineering of an intelligent nanoplatform that can be used to reverse tumor drug resistance and inhibit cancer cell proliferation and metastasis.

Author contributions

Jianhai Chen, Yi Zhou, Xiyong Yu conceived the work. He Wang, Qiudi Huang, Jiacui Xie, Jiachan Lin participated in the design and implementation of the experiment. Yi Zhou, Qiudi Huang, Linhao Qin, Yingling Miao, Yuan Qin, and Pei Huang carried out data processing and analysis. Yi Zhou wrote and revised the paper. All authors discussed and approved the manuscript.

Conflicts of interest

The authors declare no conflicts of interest.

Appendix ASupplementary data

The following is the Supplementary data to this article: