Co-delivery of fucoxanthin and Twist siRNA using hydroxyethyl starch-cholesterol self-assembled polymer nanoparticles for triple-negative breast cancer synergistic therapy

Graphical abstract

(A) Schematic diagram of the synthesis of hydroxyethyl starch-cholesterol (HES-CH) monomers. Schematic illustration of the initial construction (B) and subsequent administration (C) of siTwist/FX@HES-CH NPs for co-delivery of FX and siTwist to enhance anti-tumor efficacy.

Highlights

• •The combination of fucoxanthin and Twist siRNA can be implemented in a single HES-CH nanovector.
• •The dual-functional siTwist/FX@HES-CH nanoparticles target both tumor cells and tumor microenvironment. .
• •siTwist/FX@HES-CH were able to penetrate deep into tumor parenchyma.
• •siTwist/FX@HES-CH remarkably alleviated the TNBC orthotopic tumor burden and inhibited lung metastasis.
• •This combinative strategy provides a new avenue for the treatment of stroma-rich tumor.

Abstract

Introduction

Triple-negative breast cancer (TNBC) represents the most aggressive subtype of breast cancer with an extremely dismal prognosis and few treatment options. As a desmoplastic tumor, TNBC tumor cells are girdled by stroma composed of cancer-associated fibroblasts (CAFs) and their secreted stromal components. The rapidly proliferating tumor cells, together with the tumor stroma, exert additional solid tissue pressure on tumor vasculature and surrounding tissues, severely obstructing therapeutic agent from deep intratumoral penetration, and resulting in tumor metastasis and treatment resistance.

Objectives

Fucoxanthin (FX), a xanthophyll carotenoid abundant in marine algae, has attracted widespread attention as a promising alternative candidate for tumor prevention and treatment. Twist is a pivotal regulator of epithelial to mesenchymal transition, and its depletion has proven to sensitize antitumor drugs, inhibit metastasis, reduce CAFs activation and the following interstitial deposition, and increase tumor perfusion. The nanodrug delivery system co-encapsulating FX and nucleic acid drug Twist siRNA (siTwist) was expected to form a potent anti-TNBC therapeutic cyclical feedback loop.

Methods and results

Herein, our studies constituted a novel self-assembled polymer nanomedicine (siTwist/FX@HES-CH) based on the amino-modified hydroxyethyl starch (HES-NH₂) grafted with hydrophobic segment cholesterol (CH). The MTT assay, flow cytometry apoptosis analysis, transwell assay, western blot, and 3D multicellular tumor spheroids growth inhibition assay all showed that siTwist/FX@HES-CH could kill tumor cells and inhibit their metastasis in a synergistic manner. The in vivo anti-TNBC efficacy was demonstrated that siTwist/FX@HES-CH remodeled tumor microenvironment, facilitated interstitial barrier crossing, killed tumor cells synergistically, drastically reduced TNBC orthotopic tumor burden and inhibited lung metastasis.

Conclusion

Systematic studies revealed that this dual-functional nanomedicine that targets both tumor cells and tumor microenvironment significantly alleviates TNBC orthotopic tumor burden and inhibits lung metastasis, establishing a new paradigm for TNBC therapy.

Introduction

As per the report on global cancer statistics for 2021, breast cancer has overtaken lung cancer as the most frequently diagnosed malignant carcinoma [1]. Triple-negative breast cancer (TNBC) represents approximately 15–20 % of all cases of breast cancer. It is the most aggressive subtype of the disease and is characterized by deficiencies in the hormone receptor (HR) and human epidermal growth factor receptor 2 (HER2) [2], [3]. Because of its high molecular heterogeneity, metastatic potential, and relapse proneness, TNBC has a poor prognosis, with a mortality rate of around 40 % within the first five years after diagnosis [4], [5]. Furthermore, distant metastasis occurs in most of TNBC patients, which typically happens in the lungs, bones, and brain [4], [6]. Chemotherapy, the standard systemic treatment for TNBC, exhibits limited efficacy and high recurrence rates mainly due to the treatment resistance [6], [7], [8]. Moreover, owing to the lack of expression of all three receptors, TNBC was unable to effectively respond to endocrine therapy or trastuzumab treatment [9], [10], [11]. Therefore, the development of advanced strategies to improve the therapeutic effect of TNBC is highly desirable.

Natural compounds with tumor-suppressive capabilities are regarded to be a promising candidate to expand the TNBC therapeutic choices [12], [13]. Fucoxanthin (FX), a natural xanthophyll carotenoid that could be isolated broadly from marine brown algae and other macro-/microalgae, shows negligibly adverse effects in rodents or humans at therapeutic dosages [14], [15], [16], [17], [18], [19], [20]. FX's substantial anti-cancer potential is possibly due to the structure of the abundant allenic bond, which has been demonstrated in a number of malignancies, including breast cancer [21], [22], [23], [24], [25], [26], [27], [28], [29]. Mechanism studies confirmed that FX exerts a multi-target anti-tumor effect via classical pathways implicated in apoptosis induction, cell cycle arrest, autophagy regulation, and anti-metastasis [30], [31]. Our previous studies also proved that FX could improve gefitinib sensitivity in non-small cell lung cancer (NSCLC) [32]. Besides, FX was able to act as an antilymphangiogenic agent to alleviate the tumor burden and lymphatic metastasis in the TNBC xenograft model [33]. Collectively, it is believed that FX with widespread sources and low side effects is expected to be used as a novel anti-cancer agent. Despite its promise, the special allenic bond structure is quite brittle and readily destroyed by heat, air exposure, or light [34], [35], [36]. Furthermore, sufficient accumulation and high penetration of agents into tumor tissue to achieve lethal effects is always challenged. The utilization of free FX revealed limited distribution, poor solubility and low bioavailability, all of which compromised its clinical application [29], [36], [37], [38], [39]. Therefore, encapsulating FX in drug delivery systems could achieve substantial tumor accumulation while also improving FX efficacy.

Another important obstacle limiting the efficacy of conventional TNBC treatment regimens is the complicated tumor microenvironment (TME). Numerous evidence suggested that the TME is not a “bystander” but rather a “active participant” in tumor growth and anticancer treatment resistance [40], [41], [42]. TNBC is distinguished from other malignancies by the presence of a dense tumor stroma composed primarily of α-SMA⁺ cancer-associated fibroblasts (CAFs) and their derived stromal components [43]. By secreting chemokines, cytokines, and growth factors (e.g., FGF, CXCL-10, IL-6/8, and VEGF), CAFs remodel the tumor microenvironment, mediate interstitial deposition, induce abnormal tumor angiogenesis, trigger epithelial to mesenchymal transition (EMT), and finally promote tumor growth, metastasis, and drug resistance [44]. In addition, CAFs produce an abundance of matrix proteins, which are then bundled into fibrils to form a three-dimensional stroma network, and together with rapidly proliferating tumor cells and other stromal cells raise solid tissue pressure (STP), compressing perfusion vessels and limiting drug access to some tumors while allowing them to regenerate [45], [46], [47]. Consequently, remodeling the pathological TME could be an advantageous strategy for boosting TNBC therapeutic outcomes. Twist, a central regulator of EMT, is specificity overexpressed in TNBC and other tumor tissues [48], [49], [50]. Twist expression is a necessary and prerequisite condition for the transdifferentiation of quiescent normal fibroblasts to cancer-associated fibroblasts in breast cancers, as well as for the deposition of dense tumor stroma. Concurrently, activated CAFs also promoted Twist expression and activated EMT, invasion, and metastasis [51], [52]. Besides, Li et al. found that Twist depletion using RNA interference technology not only completely blocked the EMT but also partially reversed multidrug resistance in breast cancers [53]. Therefore, we hypothesized that depletion of Twist using small interfering RNA (siRNA) and combining it with FX treatment, all of which present dual effects on both tumor cells and TME, may serve as a promising strategy.

Nanotechnology-based drug delivery systems display superior advantages in improving the solubility and bioavailability of natural products, overcoming the in vivo delivery barrier of RNAi molecules, and enabling co-delivery of therapeutic agents owing to their encapsulation capability and simple modification strategies [54], [55], [56]. In this work, we constituted a FX and siTwist co-delivery nanomedicine (siTwist/FX@HES-CH) by grafting biocompatible hydrophobic segment cholesterol (CH) onto hydroxyethyl starch (HES). Intriguingly, we incorporated amino groups on HES composed of sugar units to expand the loading capability of negatively charged siRNA while also endowing HES-CH NPs with pH-sensitive properties, triggering an acidic tumor environment-responsive release. The systematic in vitro and in vivo therapeutic test verified that 100–200 nm siTwist/FX@HES-CH NPs could effortlessly accumulate in tumor tissue. Following the phenomenon, the co-delivered FX and siTwist not only eradicated TNBC cells synergistically but also effectively acted on fibroblasts, inhibiting their transdifferentiation into CAFs and remodeling the TME. This nanodrug delivery strategy forms a virtuous cyclic feedback loop in the treatment of TNBC. Tumor cell removal, CAFs inhibition, and ECM suppression result in brilliant STP reduction in the TME. Tumor blood vessel decompression allows more nanoparticles to be transported across the blood vessels, readily across the tumor stroma, and bind to tumor intracellular targets. Tumor cell destruction remodels the TME for deep drug penetration, resulting in remarkable anti-TNBC efficacy in primary tumor growth and the following lung metastasis.

Materials and methods

Materials

HES (average molecular weight (Mw) = 130 kDa; hydroxyethyl molar substitution (MS) = 0.4) was purchased from Wuhan HUST life Sci & Tech Co., Ltd (Wuhan, China). Bromopro pylamine hydrobromide, succinic anhydride, cholesterol, EDCI, and polycaprolactone were purchased from Aladdin Reagent Inc. (Shanghai, China). FX was purchased from MedChemExpress Co., Ltd. (Shanghai, China). Small interfering RNA duplexes against Twist (siTwist), negative control RNA (siNC), and Cy5-labeled siNC were all ordered from GenePharm Co., Ltd. (Shanghai, China). The live/dead cell double staining kit, and apoptosis detection Kit were all obtained from Dojindo China Co., Ltd. (Shanghai, China). The DEPC water, MTT cytotoxicity assay kit, DAPI and Lyso-Tracker green were obtained from Beyotime Biotechnology Co., Ltd. (Shanghai, China). TGF-β was obtained from PeproTech, Inc. (Jiangsu, China). DiR iodide were purchased from AAT Bioquest, Inc. (CA, USA). The anti-Twist antibodies were purchased from Abcam (Cambridge, UK). The anti-α-SMA antibodies were purchased from Abmart (Shanghai, China). The anti-GAPDH antibodies were purchased from Proteintech Group, Inc. (Wuhan, China). All other chemicals were of analytical grade and used as received.

Synthesis of aminopropyl hydroxyethyl starch (HES-NH₂)

Hydroxyethyl starch was dissolved in 6.2 mol NaOH solution at 0.5 g/L, cooled for 40 min at 4 °C, and 0.96 g 3-bromopropylamine hydrobromide was added for the reaction. Subsequently, adjust the pH to 7.0 by adding 37 % HCl solution. The grafted polysaccharide was recovered by cooling precipitation in anhydrous ethanol (4 °C). The precipitate was then freeze-dried after being redissolved in distilled water for dialysis. The amino group content was determined using hydrogen nuclear magnetic resonance spectroscopy (1H NMR) characteristic spectra.

Synthesis of carboxylated CH (CH-COOH)

0.5 g of cholesterol and 0.5 g of succinic anhydride were added in 10.0 mL of pyridine solution and stirred for 3 h at 70 °C. The precipitation obtained is dissolved in ethanol and rotated again, and the white final product is the carboxylated modified cholesterol (CH-COOH), which is freeze-dried.

Synthesis of hydroxyethyl starch-cholesterol polymer (HES-CH)

CH-COOH (1.0 g) and HES-NH₂ (1.0 g) were added in 10 mL DMSO, then HOBT (0.5 g) and EDC (0.5 g) were added. The mixture was heated and reacted to obtain hydroxyethyl starch grafted cholesterol polymer (HES-CH), and unreacted was removed by dialysis impurities. The reactant solution was freeze-dried to obtain hydroxyethyl starch coupled cholesterol polymer (HES-CH) powder.

Preparation of FX@HES-CH self-assembled NPs

Emulsion solvent evaporation was used to produce FX@HES-CH NPs. In a nutshell, HES-CH (0.1 g) was dissolved in deionized water (100 mL), and FX (0.01 g) was dissolved in chloroform mixture (10 mL). A cell crushing apparatus was used for ultrasound during emulsification. Subsequently, the chloroform solvent in the emulsion was removed using the rotary evaporation method to obtain the suspension of FX@HES-CH. The prepared FX@HES-CH NPs were lyophilized to obtain FX@HES-CH freeze-dried powder. The drug loading content and entrapment efficiency of FX were calculated using the formula below.

$$\mathit{DL}\left(\mathit{Wt}\%\right)=\frac{\mathit{Wt}\left(\mathit{FX}\right)}{\mathit{Wt}\left(\mathit{FX}@HES-CH\right)}\times 100$$

$$\mathit{EE}\left(\mathit{Wt}\%\right)=\frac{\mathit{Wt}\left(\mathit{FX}\right)}{\mathit{Weight}ofthefeedingFX}\times 100$$

Where, Wt (FX) is the amount of FX entrapped by the HES-CH NPs and Wt (FX@HES-CH) is the total weight of FX and HES-CH NPs.

Loading siRNA onto FX@HES-CH NPs

FX@HES-CH was mixed with 1 mg/mL siRNA (dissolve with DEPC water) in different weight ratios (2.5, 5, 10, 20, 40, 60, 80, 100, w/w). After magnetically stirring on ice bath, the binding of FX@HES-CH to siRNA was evaluated via agarose gel (2 %) electrophoresis. siRNA was loaded onto the surface of the FX@HES-CH NPs through electrostatic interactions to prepare siRNA/FX@HES-CH NPs, in which the feeding of siRNA to FX@HES-CH was calculated based on FX loading in FX@HES-CH, i.e., FX (mol): siRNA (mol) = 200:1.

The free siRNA solution and siRNA/FX@HES-CH NPs solution were co-incubated with FBS (50 %, v/v) at 37 °C for the serum degradation stability assay. The sample was obtained at each time point for agarose gel (2 %) electrophoresis analysis.

Measurement of HES-CH pKa by titration assay

HES-CH titration experiment was used to determine the pKa of HES-CH. The pKa of HES-CH depended mainly on the ionization of amino group on HES. The method is roughly as follows: 2 mg HES-CH polymer is dispersed in 4 mL of deionized water, the pH is adjusted to 5 with 0.1 mol/L HCl, then 0.01 mol/L NaOH is slowly added to the HES-CH polymer suspension, adding one drop of NaOH solution per drop, and the changed pH is recorded with a pH meter. Then the results are counted and the function corresponding to the pH value and the volume of NaOH added is drawn. The pKa value of HES-CH can be obtained by differentiating the obtained function.

Characterization

1H NMR spectra of HES-NH₂, CH-COOH, and HES-CH were recorded on a nuclear magnetic resonance spectrometer (AscendTM 600 MHZ, Bruker) using tetramethylsilane (TMS) as an internal reference. (FT-IR) spectra of HES-NH₂ and HES-CH were recorded on a Flourier transform infrared spectrometer (Vertex70, Bruker) with an attenuate total reflection (ATR) accessory. The molecular weight of CH-COOH was measured by the HRMS instrument (Agilent Technologies 6200 series TOF/6500 series Q-TOF B.06.01 (B6157)). The sample morphologies were observed using transmission electron microscopy (ITACHI, HT7700) at an operating voltage of 100 kV. All samples dispersed in water (0.1 mg/mL, 10 µL) were dropped onto carbon-coated copper grids (200 mesh) and dried overnight at room temperature. The samples were stained with a phosphotungstic acid solution (0.2 %, w/v). The samples' hydrodynamic sizes and zeta potential were measured by using Zetasizer (Zetasizer Nano-ZS, Malvern). To investigate the stability of HES-CH and FX@HES-CH, the particle size was measured by dynamic light scattering (DLS) using Zetasizer (Zetasizer Nano-ZS, Malvern).

To explore the pH-triggered FX release profile, FX@HES-CH was immersed in different PBS solutions (pH = 7.4 and 5.5) with gentle shaking at 37 ℃. The concentration was set to 1 mg/mL. At 1, 2, 4, 8, 12, 24, and 48 h, 2 mL of FX@HES-CH solution was taken out. The unreleased FX@HES-CH was removed by ultrafiltration. The FX release profile was determined according to the procedure described by Yue Sui, et al [57]. The siRNA release profile was assessed with a Hitachi F-4600 fluorescence spectrophotometer (Hitachi, Japan). Cy5-siRNA@HES-CH was immersed in various PBS solutions (pH = 7.4 and 5.5) at 37 °C with light shaking. At 0.5, 1, 2, 4, 6, 8, 12, 24, 36, and 48 h, the supernatant's fluorescence spectra was then analyzed after the solution was centrifuged for 10 min at 15,000 rpm.

Cell culture

Mouse triple-negative breast cancer cell line (4T1) and mouse embryonic fibroblasts line (NIH/3T3) were purchased from the Chinese Academy of Sciences Cell Bank (Shanghai, China). 4T1 and MCF 10A cells were cultured at 37 °C in a humidified atmosphere of 5 % CO₂ in DMEM medium, and NIH/3T3 cells were cultured at 37 °C in a humidified atmosphere of 5 % CO₂ in RPMI-1640 media. Both media contained 10 % fetal bovine serum (FBS), 100 U/mL penicillin, and 0.1 g/mL streptomycin.

Cellular uptake of nanoparticles

To understand the cell internalization of nanoparticles or free siRNA, 4T1 cells treated with Cy5-siNC@HES-CH was imaged using CLSM (Olympus, Japan). Cells were first seeded in glass-bottom dishes and cultured for 24 h, and then treated with Cy5-siNC@HES-CH (100 nM Cy5-siNC) for another 6 or 12 h. PBS, free Cy5-siNC and HES-CH (empty vectors) were used as control. After incubation with NPs, cells were washed, and then DAPI (1 μg/mL) and Lyso-Tracker Green (100 nM) were added and fixed in 2 % paraformaldehyde. Red fluorescence (λex/em, 650 nm/670 nm) of Cy5-siNC, blue fluorescence of DAPI (λex/em, 340 nm/488 nm) and green fluorescence of LysoTracker Green (λex/em, 504 nm/511 nm) were observed using CLSM.

Quantitative analysis of nanoparticles internalized by 4T1 cells was performed using flow cytometry. Cells were incubated using the same dosing regimen as described above. Subsequently, all cells were harvested and washed with ice-cold PBS. Fluorescence was acquired by flow cytometer (BD FACSAriaTM Ⅲ, USA), and analyzed by FlowJo_V10 software.

Cytotoxicity assay

Briefly, 3 × 10³ 4T1 cells were seeded in a 96-well cell culture plate and cultured overnight, and then treated with siTwist/FX@HES-CH for 24, 48, and 72 h. After that, the cytotoxicity of the nanomedicine on 4T1 cells was detected using the protocol provided by the MTT assay kit. The absorbance at 570 nm was measured with a microplate reader (Multiskan FC, Thermo Scientific, USA).

The MTT assay was also used to assess the cytotoxicity of various formulations. In a 96-well cell culture plate, 3 × 10³ 4T1 cells were seeded and cultured overnight. After that, The culture media was then replaced with new media containing different drugs, including PBS, free FX, HES-CH, FX@HES-CH, siNC/FX@HES-CH, and siTwist/FX@HES-CH (FX concentration of 20 μM, siTwist concentration of 100 nM). After another 48 h, cell viability was determined by MTT assay kit.

Apoptosis assay

4T1 cells were seeded at a density of 1 × 10⁵ cells per well in 6-well cell culture plates and cultured overnight. The culture media was then replaced with fresh media containing different drugs, including PBS, free FX, HES-CH, FX@HES-CH, siNC/FX@HES-CH, and siTwist/FX@HES-CH (FX concentration of 20 μM, siTwist concentration of 100 nM). After another 36 h, the cells were harvested, and stained with Annexin V-FITC/PI for 20 min at 4 °C in the dark. Fluorescence was quantified by flow cytometer (BD FACSAriaTM Ⅲ, USA).

Viable and dead cells observations

In order to intuitively assess cellular viability after treating with the various formulations, the live/dead cell double staining kit was used according to the manufacturer’s protocol. Cells were then incubated with different formulations for 48 h, and washed with PBS. Following that, cells were incubated with calcein-AM and PI solutions for 15 min at 37 °C, then washed with cold PBS and imaged with fluorescence microscope.

Wound-healing assay

4T1 cells were seeded at a density of 2 × 10⁵ cells per well in 6-well cell culture plates. When cells grew to 80 % confluence, the cell monolayer was scraped by sterile 10 μL pipette tips, and incubated with PBS, free FX, HES-CH, FX@HES-CH, siNC/FX@HES-CH, and siTwist/FX@HES-CH for another 36 h. Cells were imaged under an inverted microscope.

Transwell migration and invasion assays

For the migration assay, the upper chamber was filled with 2 × 10⁵ 4T1 cells in 100 μL serum-free medium, and the lower chamber was filled with 600 μL of medium containing 10 % FBS. PBS, free FX, HES-CH, FX@HES-CH, siNC/FX@HES-CH, and siTwist/FX@HES-CH were added in both chambers. Cells were cultured and given 24 h to migrate through the insert membrane.

For the invasion assay, the upper chamber was coated with diluted Matrigel by serum-free medium (1:10, 100 μL/well, BD Biocoat), adding 100 μL of serum-free medium containing 2 × 10⁵ 4T1 cells to the upper chamber, and filling the lower chamber with 600 μL of medium containing 10 % FBS. PBS, free FX, HES-CH, FX@HES-CH, siNC/FX@HES-CH, and siTwist/FX@HES-CH were also added in both chambers, and the cells were incubated for 36 h.

Cells that had not migrated (on the upper chamber) were eliminated. Cells that had migrated or invaded were fixed with methanol and stained with 0.1 % crystal violet before being photographed and counted using an inverted microscope.

Western blot assay

NIH/3T3 cells were seeded in 6-well cell culture plates and cultured overnight. Subsequently, then changed out for fresh media that contained TGF-β (10 ng/mL) and pre-incubated at 37 °C. Following that, PBS, free FX, siTwist, free FX + siTwist, and siTwist/FX@HES-CH were then used to treat the cells for 48 h (FX concentration of 20 μM, siTwist concentration of 100 nM). After harvesting cellular proteins, the western-blot assay was used to evaluate the expression levels of α-SMA.

In vitro penetration and inhibition evaluation in MCTSs

In order to evaluate the effect of inhibiting the activity of cancer associated fibroblasts (CAFs) on drug penetration, we simulated the solid tumor microenvironment to establish hybrid multicellular tumor spheroids (MCTSs) model co-cultured with tumor cells (4T1) and fibroblasts (NIH/3T3), and then evaluated permeation of siTwist/FX@HES-CH NPs in vitro. In brief, 80 μL of agarose solution (1.5 %, w/v) were pre-plated in a 96-well plate. After cooling to room temperature for solidification, 4T1 cells and NIH/3T3 cells were mixed in a 96-well plate and cultured into hybrid MCTSs in a 2:1 ratio. After MCTSs grew to appropriate sizes, siTwist@HES-CH (containing 100 nM siTwist) and siTwist/FX@HES-CH (containing 100 nM siTwist and 20 μM FX) nanoparticles were incubated with MCTSs, respectively. At the same time, Cy5-NC/C6@HES-CH was also added to each group. After 48 h incubation, MCTSs were washed using PBS, and fixed with 4 % paraformaldehyde. Subsequently, spheroids were scanned by CLSM from the top to the middle of the MCTSs. Each 4T1 & NIH/3T3 MCTS has a scan layer of 15 μm.

Emulsion solvent evaporation was also used to produce C6@HES-CH NPs. In a nutshell, HES-CH (0.1 g) was dissolved in deionized water (100 mL), and C6 (0.01 g) was dissolved in a chloroform mixture (10 mL). A cell-crushing apparatus was used for ultrasound during emulsification. Subsequently, the chloroform solvent was removed using rotary evaporation to obtain the suspension of C6@HES-CH. The solution was freeze-dried to obtain C6@HES-CH powder. Cy5-siNC was loaded onto the surface of the C6@HES-CH NPs through electrostatic interactions to prepare Cy5-siNC/C6@HES-CH NPs, in which the feeding of Cy5-siNC to C6@HES-CH was C6 (mol): Cy5-siNC (mol) = 200:1. The hydrodynamic sizes and zeta potential of samples were measured by using Zetasizer (Zetasizer Nano-ZS, Malvern).

The construction scheme of MCTSs for tumor spheroids inhibition is the same as described above. After the spheroids reached to appropriate sizes, they were divided into six groups, and PBS, FX, HES-CH, siTwist@HES-CH, FX@HES-CH, siTwist/FX@HES-CH (FX concentration of 20 μM, siTwist concentration of 100 nM) were each implemented to the medium, respectively. An inverted light microscope was used to image and record the diameter changes in MCTSs on a daily basis.

Ethics statement

All animal experiments followed the National Institute Guide for the Care and Use of Laboratory Animals. All experimental protocols were approved by the Institutional Animal Care and Use Committee, Tongji Medical College, Huazhong University of Science and Technology (Approval no. S2139). The ARRIVE guidelines are followed for reporting on animal studies.

In vivo biodistribution of NPs

Emulsion solvent evaporation was also used to produce DiR@HES-CH NPs. In a nutshell, HES-CH (0.1 g) was dissolved in deionized water (100 mL), and DiR (0.01 g) was dissolved in a chloroform mixture (10 mL). A cell-crushing apparatus was used for ultrasound during emulsification. Subsequently, the chloroform solvent was removed using rotary evaporation to obtain the suspension of DiR@HES-CH. The solution was freeze-dried to obtain DiR@HES-CH powder. siRNA was loaded onto the surface of the DiR@HES-CH NPs through electrostatic interactions to prepare siRNA/DiR@HES-CH NPs, in which the feeding of siRNA to DiR@HES-CH was DiR (mol): siRNA (mol) = 200:1. The hydrodynamic sizes and zeta potential of samples were measured by using Zetasizer (Zetasizer Nano-ZS, Malvern).

Breast tumors were orthotopically implanted in female BALB/c mice by injecting 2 × 10⁵ 4T1 cells into the mammary fat pad (GemPharmatech, JiangSu, China). When the tumor volumes reached around 200 mm³, mice were randomized into three treatment groups and injected via the tail vein with PBS, free DiR iodide, and DiR@HES-CH (1 mg/kg DiR). After 48 h, the mice were euthanized, and their primary organs (heart, liver, spleen, lung, and kidney) and tumors were collected. All of the organs were washed with cold PBS and photographed using an in vivo imaging system (Pearl Trilogy, LI-COR, USA).

In addition, tumor-bearing mice were randomized into eight groups when tumor sizes reached around 200 mm³, and injected through the tail vein with DiR@HES-CH NPs at a DiR dose of 1 mg/kg. All mice were euthanized and imaged at 1, 6, 12, 24, 36, 48, 72, and 96 h after injection. After that, all mice were then sacrificed, and tumors along with other organs were removed for ex vivo imaging.

Orthotopic mouse tumor model and treatment

1 × 10⁵ 4T1 cells into the mammary fat pad of female BALB/c mice on day −7 to establish orthotopic mouse tumor model. When tumors grew to a volume of roughly 70–100 mm³, mice were randomized into six groups (n = 6) and given saline, free FX (15 mg/kg), HES-CH, siTwist@HES-CH (1 mg/kg siTwist), FX@HES-CH (15 mg/kg FX) or siTwist/FX@HES-CH (15 mg/kg FX and 1 mg/ kg siTwist) intravenously on days 0, 3, 6, 9, 12, and 15. Body weight and tumor size were monitored every 3 days. Tumor volume was computed as follows: volume (mm³) = length × width²/2. The mice were euthanized eighteen days following treatment, and the locally developed tumors were removed, rinsed with cold PBS, and imaged. Subsequently, the completely peeled tumors were fixed in 4 % formaldehyde. Tumor tissues and major organs slices were stained with H&E for routine histology assay. Immunohistochemistry revealed TUNEL assay, Ki67, and CD31. Immunofluorescence was used to detect Twist and α-SMA expression in primary tumors, as well as Masson staining for collagen analyses.

siTwist/FX@HES-CH NPs suppress lung metastasis of TNBC

TNBC spontaneous metastasis mouse model was established followed by the same procedures as the previous model. When tumors growing reached approximately 100 mm³, the 4T1 tumor-bearing mice were randomized into six groups (n = 5) and given saline, free FX (15 mg/kg), HES-CH, siTwist@HES-CH (1 mg/kg siTwist), FX@HES-CH (15 mg/kg FX) or siTwist/FX@HES-CH (15 m/kg FX or 1 mg/kg siTwist) on day 0, 3, 6, 9, 12, 15 and 18. All the animals were euthanized on day 35. The lungs were photographed after they were removed. The metastasis-suppressive effects of NPs were assessed by counting the number of metastatic foci on the lung and H&E staining.

Statistical analysis

All the experiments were repeated at least in triplicate. Data were expressed as mean ± standard deviation (SD). Results were analyzed by two-tailed Student's t-test for two groups and one-way ANOVA for multiple groups. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 were considered to be statistically significant.

Results and discussion

Preparation and characterization of siRNA/FX@HES-CH NPs

Because of HES's hydrophilicity, biocompatibility, and PEG-like long-circulation properties, it is frequently employed to fabricate drug carriers that increase the hydrophobic drugs’ solubility and bioavailability in vivo [58], [59]. Scheme S1 characterized the chemical synthesis of HES-CH monomers. The incorporation of amino groups on HES composed of sugar units has proven to improve the loading of negative drugs such as siRNA (amino groups were detected by 1H NMR characteristic). As shown in Fig. S1A, 1H NMR was used to confirm the effective synthesis of HES-NH₂. The methylene response signals at 1.8 (a) and 3.0 (b) in propylamine are characteristic of HES-NH₂ and show that propylamine was successfully grafted onto HES. CH-COOH was synthesized by grafting succinic acid onto cholesterol. Molecular weight was determined to be 509.36 by high-resolution mass spectrometer (HRMS) (Fig. S2), and new peaks corresponding to methylene protons in succinic acid appeared between 2.4 and 2.5 in 1H NMR spectra (Fig. S1B). These results indicated the successful synthesis of CH-COOH.

The HES-CH achievement was validated by 1H NMR and FT-IR. The protons of cholesterol between 0 and 2.5 appear in the 1H NMR spectrum of HES-CH (Fig. S1C). The FT-IR spectra of HES-CH show the characteristic band of C O stretching vibration of ester bond at 1731 cm⁻¹, C O stretching vibration of amide bond and C C stretching vibration at 1600–1700 cm⁻¹ and N–H bending vibration of amide bond at 1533 cm⁻¹, indicating the successful conjugating of cholesterol onto HES (Fig. S1D).

The titration result confirmed that the pKa of aminated HES-CH is around 8.1, and below this pH, aminated HES-CH was mainly positively charged. siRNA is negatively charged at pH 7.4 and thus can interact with aminated HES-CH electrostatically (Fig. S3, A and B). HES-CH NPs had a zeta potential of +21.20 mV at pH 7.3, +22.90 mV at pH 6.8, and +24.73 mV at pH 5.4 (Fig. S4A). Amusingly, as the pH goes down from 7.3 to 5.4, the diameter of HES-CH NPs slightly increases from 115.1 nm to 136.4 nm (Fig. S4B). It is primarily attributed to the protonation of the amino group, thus loosening the structure of HES-CH NPs. This property contributes to the accelerated release of agents in the acidic environment.

The nanostructure of HES-CH, FX@HES-CH, and siTwist/FX@HES-CH NPs was imaged using transmission electron microscopy (TEM) (Fig. 1A, (a)). The emulsion solvent evaporation method was used to load FX into HES-CH. As shown in Fig. 1A ((a) and (b)), FX@HES-CH shows a larger particle size of 138.7 ± 0.9 nm compared to the empty HES-CH NPs 119.2 ± 1.9 nm, indicating a successful FX assembly. Since FX is electrically neutral, the zeta potentials of FX@HES-CH NPs (+21.40 ± 0.17 mV) and HES-CH NPs (+21.43 ± 1.1 mV) are relatively closed (Fig. 1A, (c)). Subsequently, siRNA was loaded onto the surface of the FX@HES-CH NPs through electrostatic interactions to prepare siRNA/FX@HES-CH NPs. A significant outer corona was observed on the siRNA/FX@HES-CH NPs. Encapsulation of siRNA resulted in a lower negative zeta potential (−0.41 ± 0.06 mV) and larger sizes (172.2 ± 2.2 nm), which is believed to reduce the interaction of the nanoparticles with the stromal components of the TME and making it more accessible to target tumor cells (Fig. 1A). HES-CH NPs, FX@HES-CH NPs, and siTwist/FX@HES-CH NPs have relatively uniform size distribution, with polydispersity index (PDI) values of 0.18 ± 0.01, 0.16 ± 0.05, and 0.14 ± 0.04, respectively. The calculated drug loading (DL%) of FX in FX@HES-CH NPs was 6.72 ± 0.3 %, and the drug entrapment efficiency (EE%) of FX in FX@HES-CH NPs and siRNA/FX@HES-CH NPs was greater than 90 %. As shown in the agarose gel electrophoresis, when the HES-CH:siRNA (w/w) ratio was 80:1, about 96 % of siRNA would effectively adhere to the HES-CH nanocarriers surface (Fig. 1B). Moreover, compared with the almost complete degradation of free siRNA in serum (6 h), the 48 h serum stability assay revealed that loading siRNA with nanocarriers would effectively avoid its nucleases degradation in the serum (Fig. 1C). Nanoparticles remained stable in a 10 % FBS medium (Fig. 1D) and exhibited faster FX and siRNA release in acidic environments (pH = 5.5) (Fig. 1, E and F).

Fig. 1.

Characterization of different nanomaterials. (A) TEM images of different nanoparticles (a). Scare bar, 200 nm. DLS analysis (b) and zeta potential of HES-CH, FX@HES-CH, and siRNA/FX@HES-CH NPs (c). (B) Loading content of siRNA in HES-CH NPs assessed by agarose gel assay. (C) Stability of siRNA@HES-CH NPs in serum. (D) Colloidal stability of nanoparticles in 10 % serum-containing phosphate-balanced saline (pH = 7.4) after 6 days incubation. In vitro releasing profiles of FX (E) and siRNA (F) in two different media after incubation. *p < 0.05 and **p < 0.01.

In vitro cellular uptake and intracellular trafficking

Although RNAi therapy has been identified as a promising cancer therapeutic technique, in vivo siRNA administration has proven to be a significant hurdle, restricting its practical applicability [60], [61]. Because of their anionic property, siRNAs would not spontaneously diffuse across cell membranes. Moreover, siRNA is highly unstable in the systemic circulation, which is rapidly destroyed by nucleases and could elicit undesirable immunological responses [62]. Furthermore, aiming to exert functions effectively, siRNA should penetrate the cell membrane, then escape from the endosome, and uncomplex to interact with the target [60]. The high efficiency of the payload motivates us to investigate the cellular uptake of Cy5-labeled siRNA. As illustrated in Fig. 2A, cells treated with Cy5-siNC@HES-CH for 1 h exhibited a slight visible red signal as opposed to PBS and HES-CH treated cells (negligible fluorescence). With prolonged incubation, the red fluorescence of intracellular Cy5-siNC gradually increased, indicating that HES-CH-loaded siRNA could be effectively uptaken by 4T1 cells. Furthermore, the colocalization analysis displayed a low Pearson's correlation coefficient (R = 0.283) at 1 h of incubation versus a high one (R = 0.592) at 3 h, suggesting that with cellular uptake increases, Cy5-siNC@HES-CH was first localized in lysosomes. After 6 or even 12 h of incubation, the red fluorescence of intracellular Cy5-siNC was separated from the green fluorescence (Lyso-tracker Green), displaying a declining Pearson's correlation coefficient (R = 0.450 and 0.403, respectively), indicating that Cy5-siNC@HES-CH could induce siRNA escape from the lysosome (Fig. 2, A and B).

Fig. 2.

In vitro cellular uptake and lysosome-escaping evaluation of nanoparticles. (A) Representative CLSM images of 4T1 cells incubated with PBS, HES-CH, or Cy5-siNC@HES-CH. DAPI and Lyso-tracker Green were used to stain the nuclei and lysosomes, respectively. Scale bar, 50 μm. (B) Colocalization analysis of green and red fluorescent signals by the Pearson’s correlation coefficient. (C) Flow cytometry analysis was carried out to evaluate the cellular uptake efficiency of Cy5-labeled siNC by incubating 4T1 cells with PBS, Cy5-siNC, HES-CH, or Cy5-siNC@HES-CH NPs. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Moreover, quantitative detection of the cellular uptake of HES-CH NPs was achieved by the use of flow cytometry analysis (Fig. 2C). The cellular uptake efficiency of siRNA after Cy5-siNC@HES-CH treatment was up to 99.6 %, which was approximately 4.45 times greater than that of free Cy5-siNC (22.4 %). The non-specific cellular internalization of free siRNA may be ascribed to its aqueous solubility and lower molecular weight, but CLMS did not show visible red fluorescence under the same observation conditions. on the whole, it is concluded that effective encapsulation of siRNA by HES-CH vectors could remarkably improve cellular uptake efficiency.

In vitro cytotoxicity

MTT assays

The MTT assay was used to measure the in vitro cytotoxicity of free FX to 4T1 cells and MCF 10A cells to evaluate its anti-tumor activity. According to Fig. S5A, the IC₅₀ of free FX was 34.43, 17.87, and 9.61 μM at 24, 48, and 72 h, respectively. On the other hand, Fig. S5B showed that 0–20 μM free FX did not cause any damage to MCF 10A cells. siTwist/FX@HES-CH was also used to assess the cytotoxicity of 4T1 cells. As shown in Fig. 3A, the IC₅₀ of siTwist/FX@HES-CH was 26.00, 11.25, and 6.68 μM at 24, 48, and 72 h, respectively. siTwist/FX@HES-CH exhibited significantly improved cytotoxicity to 4T1 cells along with a typical time- and dose-dependent cytotoxicity.

Fig. 3.

(A) The cytotoxicity of siTwist/FX@HES-CH against 4T1 cells after 24, 48, and 72 h of incubation at varied FX doses. (B) Viability of 4T1 cells after 48 h treated with PBS, free FX, HES-CH, FX@HES-CH, siNC/FX@HES-CH, and siTwist/FX@HES-CH. (C) Live/death cell analysis of 4T1 cells treated with PBS, free FX, HES-CH, FX@HES-CH, siNC/FX@HES-CH, and siTwist/FX@HES-CH and (D) quantitative analysis. Scale bar, 50 μm. (E) Apoptosis analysis of 4T1 cells treated with PBS, free FX, HES-CH, FX@HES-CH, siNC/FX@HES-CH, and siTwist/FX@HES-CH after 24 h and (F) quantitative analysis. All data were expressed as means ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

In addition, we also evaluated the viability of 4T1 cells and MCF 10A cells following treatment with various therapies. As shown in Fig. 3B, when compared to the control group (PBS), 20 μM free FX displayed obvious anti-tumor efficacy, and this cytotoxicity was enhanced by siTwist/FX@HES-CH, which was associated with the synergistic anti-tumor effect of the dual components. The cytotoxicity difference between siNC/FX@HES-CH and siTwist/FX@HES-CH revealed that targeting the Twist gene could improve FX sensitivity. Furthermore, no significant cytotoxicity occurs following HES-CH nanovector treatment, suggesting that HES-CH has high biocompatibility. Encouragingly, MCF 10A cells showed no cytotoxicity in all of the treatments, indicating that normal breast epithelial cells were unharmed by siTwist/FX@HES-CH (Fig. S5C).

Viable and dead cells assay

The proportion of live and dead cells in 4T1 cells was directly observed after staining with calcein-AM and PI solution. The live cells were stained with green fluorescence from hydrolyzed calcein-AM, whereas the dead cells were stained with red fluorescence from PI. Notably, free FX had only modest red and definite green fluorescence, suggesting that FX alone possesses cytotoxicity. Nonetheless, after the siTwist/FX@HES-CH combination therapy, the residual green fluorescence was switched off and replaced by red fluorescence, indicating that the cytotoxicity of FX was greatly enhanced by the co-delivery of siTwist (Fig. 3, C and D).

Apoptosis analysis

Furthermore, the pro-apoptotic performance of siTwist/FX@HES-CH on TNBC cells was assessed systemically. As shown in Fig. 3 (E and F), groups of free FX, FX@HES-CH, and siNC/FX@HES-CH resulted in 12.27 %, 21.78 %, and 16.53 % of apoptotic cells (in early or late apoptotic stages). This value for PBS control and HES-CH were 4.15 % and 4.62 %, respectively, suggesting low toxicities of materials and siTwist. However, siTwist/FX@HES-CH presented higher apoptosis rates (44.90 %), which was ascribed to the synergistic therapeutic impact of the siTwist and FX, as well as the enhanced internalization of dual components by the nano systems.

siTwist/FX@HES-CH inhibits TNBC cell migration and invasion

The wound healing and transwell assays were used to assess the migration and invasion abilities of 4T1 cells in vitro. As illustrated in Fig. 4 (A and B), the control and HES-CH groups healed the cell wound by about 67.06 ± 5.97 % and 67.67 ± 10.51 % after 36 h of incubation, respectively. FX group with healing rates of 54.23 ± 2.89 %, intimating that free FX delayed migration to the certain extent. due to the successful encapsulation of HES-CH nanovectors, the healing rates of FX@HES-CH, siNC/FX@HES-CH and siTwist/FX@HES-CH were 40.47 ± 5.74 %, 43.69 ± 6.57 %, and 22.36 ± 2.65 %, respectively. In both transwell migration (Fig. 4, C and D) and matrigel invasion assays (Fig. 4, E and F), 4T1 cells consistently exhibited a significant reduction in migration and invasion in the presence of siTwist/FX@HES-CH. The HES-CH group's migration, invasion, and wound healing rates were comparable to those of the control group, indicating that the vector materials had little impact on cell metastasis. In general, tumor metastasis could be obviously inhibited by delivering FX and siTwist together via HES-CH nanovectors.

Fig. 4.

The anti-metastasis effects of siTwist/FX@HES-CH on 4T1 cells in vitro. (A) wound-healing assay and (B) quantitative analysis. Scale bar, 100 μm. (C) Transwell migration assay and (D) quantitative analysis. Scale bar, 100 μm. (E) Transwell invasion assay and (F) quantitative analysis. Scale bar, 100 μm. *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001.

To verify that the synergistic inhibitory effect of siTwist/FX@HES-CH on TNBC cell migration and invasion achieved by interfering with Twist protein, we evaluated the expression of Twist protein in 4T1 cells after treatment with siTwist/FX@HES-CH. As shown in Fig. S6 (A and B), the FX group showed a slight down-regulation of the Twist protein. In contrast, the FX@HES-CH group and the siNC/FX@HES-CH group inhibited Twist expression more effectively than free FX, which was caused by the improved cellular uptake of FX. Furthermore, compared to the other groups, the siTwist/FX@HES-CH NPs treatment significantly down-regulated the expression of Twist protein.

In vitro penetration and growth inhibition evaluation in hybrid multicellular tumor spheroids (MCTSs)

Co-cultured multicellular tumor spheroids of 4T1 cells and NIH/3T3 cells were produced as the in vitro model to investigate nanoparticle solid tumor penetration. To evaluate the tumor parenchyma penetration ability of siTwist/FX@HES-CH nanoparticles, we synthesized Cy5-siNC/C6@HES-CH NPs with Cy5-siNC (red) and coumarin 6 (C6, green) as fluorescent tracers instead of siTwist and FX, respectively. Cy5-siNC/C6@HES-CH is similar in size and zeta potential to siTwist/FX@HES-CH, measuring 171.30 ± 6.12 nm and −0.26 ± 0.27 mV, respectively (Fig. S7, A and B). As shown in Figs. 5 and S8, after 48 h incubation of Cy5-siNC/C6@HES-CH with PBS (Figs. 5A, B and S8A), siTwist@HES-CH (Figs. 5C, D and S8B) and siTwist/FX@HES-CH (Figs. 5E, F and S8C) respectively, CLSM was employed to observe the deep penetration of nanoparticles into MCTSs. The results indicated that while Cy5-siNC/C6@HES-CH alone could penetrate the tumor spheroid, the depth of penetration was significantly increased by siTwist@HES-CH and was even more remarkable after siTwist/FX@HES-CH treatment.

Fig. 5.

In vitro three-dimensional MCTSs penetration of Cy5-siNC/C6@HES-CH and quantification of maximum cross-section fluorescence intensity. Tumor spheroids were incubated with Cy5-siNC/C6@HES-CH for 48 h, concurrently administered with PBS (A and B), siTwist@HES-CH (containing 100 nM siTwist) (C and D) and siTwist/FX@HES-CH (containing 100 nM siTwist and 20 μM FX) (E and F).

The Twist's role in mediating TME remodeling has been well known. Considering the synergistic inhibition of Twist protein expression by siTwist/FX@HES-CH, we investigated whether this consequence would substantially modulate the TME. The TME of TNBC is primarily composed of α-SMA⁺ cancer-associated fibroblasts (CAFs) and stromal components derived from them [43]. Growth factors (PDG, TGF-β, etc) secreted by tumor cells and immune infiltrating cells have been shown to induce fibroblasts to transdifferentiate into α-SMA⁺ CAFs [63], [64], [65]. For assessing the effect of siTwist/FX@HES-CH NPs on CAFs, α-SMA expression was measured in activated NIH/3T3 cells treated with various preparations [66]. As shown in Fig. S9 (A and B), Free FX exhibited typical dose-dependent cytotoxicity to NIH-3T3 cells, and siTwist/FX@HES-CH augmented the cytotoxicity due to the dual components' synergistic anti-tumor action. The Blank HES-CH vector showed no obvious cytotoxicity to NIH-3T3 cells. Furthermore, the expressions of α-SMA decreased to a certain extent after FX and siTwist treatment, and the most significantly down-regulated occurred after administration of siTwist/FX@HES-CH NPs (Fig. S9C). In combination with the penetration research, it is speculated that the increased infiltration of tumor spheroids is induced by the reduction of solid tissue pressure and tumor microenvironment remodeling following siTwist and FX inhibition of tumor cell proliferation and CAFs activation.

In addition, we also evaluated the growth inhibition of three-dimensional multicellular tumor spheroids by different treatments. PBS, FX, HES-CH, siTwist@HES-CH, FX@HES-CH, and siTwist/FX@HES-CH were incubated with tumor spheroids for four days, respectively. As shown in Fig. S10, empty nanovectors and siTwist@HES-CH-treated spheroids performed similarly to the control group, with no significant growth inhibition. The free FX group inhibited the growth of tumor spheroids, but this effect was remarkably enhanced by FX@HES-CH and siTwist/FX@HES-CH. The siTwist/FX@HES-CH group and FX@HES-CH group even showed tumor spheroids depolymerization on day 3 and day 4, respectively.

In vivo biodistribution of nanoparticles

The drug accumulation in the pathological site is critical for the evaluation of targeting efficiency of siTwist/FX@HES-CH. Aiming for qualitative observation, FX was substituted in the HES-CH with DiR iodide, DiR@HES-CH has a size of 136.30 ± 0.95 nm and a zeta potential of + 20.57 ± 0.15 mV, similar to that of FX@HES-CH (Fig. S11, A and B). At 48 h, free DiR fluoresced intensely in the liver and spleen, while the signal was significantly weaker in the tumor. The DiR@HES-CH was also nonspecifically distributed in the liver and spleen, but the intratumoral accumulation of DiR was much greater in this group than in the others (Fig. 6A). Furthermore, the time-dependent biodistribution of HES-CH was assessed to predict accumulation and destination. After intravenously injecting DiR@HES-CH NPs, the accumulated amount of DiR in tumor reached a maximum at 72 h after counting its relative average fluorescence intensity (Fig. 6, B and C). Isolated organ imaging revealed a consistent time-dependence profile (Fig. 6, D and E). The results revealed that HES-CH could target tumors and that therapeutic agents in HES-CH could remain in tumors for more than 96 h. Furthermore, considering the surface charge difference between DiR@HES-CH and siRNA/FX@HES-CH, we loaded siRNA onto DiR@HES-CH to prepare siRNA/DiR@HES-CH, which may better reflect the in vivo biodistribution profile of siRNA/FX@HES-CH. As shown in Fig. S11, siRNA/DiR@HES-CH has a size of 173.90 ± 0.87 nm and a zeta potential of −0.62 ± 0.44 mV, similar to siTwist/FX@HES-CH. Compared to DiR@HES-CH, the accumulated amount of siTwist/FX@HES-CH in tumors peaked at 48 h and persisted for more than 96 h (Fig. S12, A and B). The fluorescence signal was diffused throughout the tumor core, implying that HES-CH would penetrate deeper into the tumor and be more effective for solid tumor therapy.

Fig. 6.

(A) Ex vivo fluorescence images of tumor and organs in 4T1 tumor-bearing mice at 48 h after intravenously administrated with PBS, free DiR, and DiR@HES-CH. (B) Representative images of tumor-bearing mice at 1, 6, 12, 24, 36, 48, 72, and 96 h after DiR@HES-CH injection. (C) Quantitative analysis of the bioluminescence signal in tumor-bearing mice. (D) Representative ex vivo bioluminescence images of isolated organs at 1, 6, 12, 24, 36, 48, 72, and 96 h after DiR@HES-CH injection. (E) Quantitative analysis of the bioluminescence signal in isolated organs.

In vivo antitumor therapy of siTwist/FX@HES-CH

Previous studies demonstrated that siTwist/FX@HES-CH could efficiently accumulate in the tumor tissues, have high permeability, and are readily uptaken by tumor cells. Encouraged by these results, we examined the efficacy of siTwist/FX@HES-CH-mediated combination strategy in vivo. Fig. 7A showed the schematic design of orthotopic tumor therapy model. For preclinical animal models closest to human TNBC is the 4T1 model [67]. 4T1 tumor-bearing mice model was established one week later after cells were implanted into mouse mammary pads. The mice were randomly divided into six groups (n = 6), and each group was treated by injecting formulations into the tail vein once every three days for six times. As shown in Fig. 7(B–D), the control group (saline) showed continued tumor growth, confirming the successful model establishment. Tumor growth in the empty HES-CH group was similar to that seen in the control group, suggesting that the HES-CH had no therapeutic effect. Silencing of Twist gene expression by siTwist@HES-CH exhibited negligible effects on tumor growth. Besides, free FX also presented limited tumor suppression in the later stages of treatment. This may be attributed to its poor solubility and non-specific biodistribution. FX@HES-CH showed better anti-tumor effect compared with free FX due to the increased FX accumulation inside tumor cells. Moreover, siTwist/FX@HES-CH outperforms FX@HES-CH in tumor suppression because twist depletion might well overcome therapeutic resistance to antitumor drugs. In addition, compared to the weight loss of mice after repeated administration of FX (dissolved in DMSO), the siTwist/FX@HES-CH group had no obviously weight loss despite incorporating two therapeutic modalities into the HES-CH vectors (Fig. 7E). These results confirmed that the advantage of combinative treatment in alleviating the growth of the primary tumor.

Fig. 7.

The anti-tumor effect in 4T1 tumor-bearing mice. (A) Schematic design of orthotopic tumor therapy model. (B) Representative image of excised orthotopic TNBC tumors from mice treated with saline (Control), free FX, HES-CH, siTwist@HES-CH, FX@HES-CH, or siTwist/FX@HES-CH, respectively (n = 6 per group). (C) Tumor volume measurement with a caliper was used to monitor tumor development. (D) Tumor weights at the end of the treatment period. (E) Body weight during the therapy period. (F) Tumor apoptosis, proliferation and angiogenesis evaluated by H&E (scale bar, 100 μm), TUNEL, Ki67 and CD31 immunohistochemistry staining (scale bar, 50 μm) of primary tumors. All data are shown as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001.

To further evaluate the potential anti-tumor mechanisms of siTwist/FX@HES-CH, tumor tissues were examined using H&E and immunohistochemical staining. As shown in Fig. 7F, extensive nuclear pyknosis and cancer necrosis occurred in the siTwist/FX@HES-CH-treated group, demonstrating the synergistic anti-tumor effect of siTwist/FX@HES-CH. Compared to the other groups, siTwist/FX@HES-CH treatment resulted in the highest TUNEL and lowest Ki67 expression in tissue sections, indicating an increase in apoptotic cell death and a significant decrease in tumor proliferation. Furthermore, FX inhibits the differentiation of endothelial progenitor cells into endothelial cells, reducing the generation of new blood vessels [68], [69]. Twist overexpression has the potential to cause aberrant angiogenesis [70], [71]. Therefore, CD31 staining assays were performed to assess angiogenesis in tumors. Results indicating that single-agent treatment with free FX, siTwist@HES-CH or FX@HES-CH inhibited abnormal angiogenesis, which was enhanced by siTwist/FX@HES-CH combinative treatments. Subsequently, we investigated Twist expression in the orthotopic tumor tissue. Immunofluorescence revealed the the lowest expression in siTwist/FX@HES-CH treatment (Fig. 8, A and D). These results indicated that siTwist/FX@HES-CH nanoparticles could significantly inhibit the growth of primary tumors by synergistically inhibiting tumor cell proliferation, promoting apoptosis and anti-abnormal angiogenesis.

Fig. 8.

Representative images of Twist (A), collagen (B) and α-SMA (C) staining of 4T1 tumor after treatments. Twist-positive cells were stained red fluorescence. Scale bar, 40 μm. Masson's trichrome was used to stain the collagen. The blue lines are collagen. Scale bar, 50 μm. α-SMA-positive cells were stained green fluorescence. Scale bar, 20 μm. (D) Quantitative analysis of Twist fluorescence intensity and (E) α-SMA fluorescence intensity in 4T1 orthotopic tumor. All data are shown as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

In vivo distribution revealed that HES-CH nanoparticles circulated for a relatively long time and accumulated at tumor sites. However, in the case of fibrous tumors, such as breast and colon cancer, deep penetration into the tumor parenchyma was blocked by the dense tumor stroma. Inhibiting the activation of α-SMA⁺ CAFs could reduce the secretion of CAFs-derived stroma [72], [73]. Mitigating solid stress in the tumor mass decompresses the tumor blood vessels further, allowing for more intratumoral drug delivery and penetration [74]. Therefore, we investigated the collagen content and α-SMA expression in the tumor microenvironment of each group. α-SMA is a CAF activation marker that could boost collagen deposition and cross-linking [75]. According to representative immunofluorescence images, siTwist/FX@HES-CH remarkably decreased the expression of α-SMA (Fig. 8, C and E). The collagen content was consistently lower in the siTwist/FX@HES-CH group than that in the Control, FX, HES-CH, siTwist@HES-CH, and FX@HES-CH groups (Fig. 8B). To summarize, siTwist/FX@HES-CH not only fight against the tumor synergistically, but it also inhibits CAFs activation and collagen expression in the tumor microenvironment. Therefore, siTwist/FX@HES-CH presented a greater inhibitory effect on tumor growth than FX@HES-CH. Additionally, other major organs' H&E staining revealed no obvious histopathological damage (Fig. S13).

In vivo anti-metastasis evaluation of siTwist/FX@HES-CH NPs

In vitro experiments confirmed that co-delivery of FX and siTwist using HES-CH vectors significant inhibition tumor cells invasion and migration. As consequently, the anti-metastasis impact was evaluated in the highly metastatic orthotopic 4T1 tumor model utilizing the treatment schedule indicated in Fig. 9A. Day 0 was designated as the beginning day of treatment. Based on the number of lung metastases nodules observed after 35 days of treatment, we discovered that the siTwist/FX@HES-CH treated mice had a considerably lower number of lung metastases (Fig. 9, B and C). In summary, the co-delivery of FX and siTwist via the pH-responsive, deeply penetrating tumor HES-CH nanoparticles is a successful therapeutic strategy for metastatic TNBC.

Fig. 9.

siTwist/FX@HES-CH inhibit lung metastasis of TNBC. (a) Schematic design of TNBC lung metastasis therapy. (b) Photographs of metastatic nodules in lungs. THE yellow arrow denotes the metastatic nodules. (c) Numbers of the lung metastatic nodules. All data are shown as mean ± SD. **p < 0.01, ***p < 0.001. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

TNBC is the most complex and aggressive type of breast cancer [76]. Several nanomedicines have been approved to treat TNBC as regular therapeutics but they have not drastically increased overall survival [77]. Panagi et al. found that Doxil monotherapy did not induce any significant delay in tumor growth compared to the untreated group in 4T1 mice, on the other hand, a combination of therapeutic drugs has shown substantial improvement in results and has proven to be an effective strategy for TNBC treatment [78]. The application of liposome or Poly (lactic-co-glycolic acid) for therapeutic agent co-delivery demonstrates more excellent 4T1 tumor treatment effects [79], [80]. Recent evidence has uncovered that the TME, which consists of blood vessels, various cell types, ECM, interstitial fluid, and other tumor-associated components, would provide a supportive environment for TNBC, and could result in therapy resistance [81]. A more contemporary approach focuses on regulating the dynamic interplay of the tumor microenvironment components. Li et al. incorporated celastrol and tiny micelles containing betulinic acid in liposomes and were more effective in reducing tumor growth in 4T1 tumor-bearing mice by disrupting CAFs [82]. This study supports our findings and confirms the positive outcomes of the co-delivery strategies that break through the tumor microenvironment barrier in anti-TNBC.

Conclusion

In this study, we reported an ingenious nanotherapeutic strategy that incorporates the co-delivery of the natural active antitumor product FX and the nucleic acid molecule siTwist. On one hand, HES-CH nanovectors can carry two completely different therapeutic agents and effective accumulation in tumor tissue due to the positive charge, long systemic circulation, and tumor tissue-responsive drug release of aminated-modified HES. Additionally, the co-delivery strategy acts on both tumor cells and TME, forming a powerful anti-tumor cyclic feedback loop. Tumor cell killing, CAFs inhibition, and extracellular matrix synthesis blockade together result in the solid tissue pressure decrease in the tumor microenvironment. Tumor blood vessel decompression facilitates the transport of the nanomedicines across the blood vessels and stroma, and their uptake by tumor cells. Nanomedicines destroy the tumor cells after drug release and alter the tumor microenvironment to facilitate deep tumor penetration, primary tumor burden reduction and lung metastasis prevention. This combinative strategy targeting both tumor cells and tumor stroma provides a novel avenue for the treatment of TNBC, and also holds great promise in other stroma-rich tumor therapy.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article: