ZIF-8/doxorubicin nanoparticles camouflaged with Cucurbita-derived exosomes for targeted prostate cancer therapy

Abstract

Development of biomimetic drug delivery systems (DDSs) holds great promise to overcome various nanoparticle-associated hindrances in cancer therapy. However, producing biomimetic nanoparticles camouflaged by animal cell-secreted exosomes presents several challenges, including low yield and some ethical considerations. Herein, we designed a biomimetic nanocarrier composed of zeolitic imidazolate framework-8 (ZIF-8) encapsulating doxorubicin (DOX) as the core and a shell of exosome-like nanoparticles (EXO) derived from Cucurbita moschata (CEXO). This design enhances safety and addresses some exosome limitations. The CEXO@ZIF-8/DOX platform was further functionalized with an epithelial cell adhesion molecule (EpCAM) aptamer (Apt-CEXO@ZIF-8/DOX) for targeted delivery to prostate cancer (PC) cells. After investigating the anticancer activity of CEXOs on PC-3 cells, the exosomes were utilized to coat ZIF-8/DOX. The immune evasion capability, cellular uptake, and anticancer effects of nanoplatforms were assessed. Moreover, the in vivo effectiveness of the targeted platform in inhibiting tumor growth and minimizing the adverse effects, was assessed using immunocompromised C57BL/6 mice bearing human PC-3 tumors. Cucurbita exosomes decreased cell viability and induced cell cycle arrest and apoptosis in PC-3 cells without affecting the normal cells. The biomimetic CEXO@ZIF-8/DOX improved immune escaping ability and hemocompatibility. The targeted nanocarrier, with augmented uptake and cellular toxicity in EpCAM-positive PC-3 cells, indicated active targeting efficacy mediated by the EpCAM aptamer. These results were supported by animal experiments that implied the effectiveness of Apt-CEXO@ZIF-8/DOX in inhibiting tumor growth without adverse side effects. This study introduces a novel functional nanocarrier that could potentially revolutionize DDSs by utilizing safer and more biocompatible plant exosomes.

---

1. Introduction

Advancements in nanotechnology have revolutionized cancer therapy by introducing innovative treatment strategies. The use of diverse nanoparticles as drug delivery systems (DDSs) mitigates the detrimental effects of chemotherapy drugs, including systemic toxicity and multidrug resistance. This improvement is partially due to the heightened localized drug accumulation at the tumor site, attributed to passive targeting achieved by the enhanced permeability and retention (EPR) effect of tumors.¹ The emergence of DDSs such as liposomes, polymeric nanoparticles, and dendrimers—offering advantages including enhanced drug solubility, controlled release, and multivalent drug conjugation—yet facing issues like toxicity and limited drug loading capacity, spurred the development of more efficient delivery systems.^2,3 Among DDSs, metal–organic frameworks (MOFs), constructed from metal ions and organic linkers, have attracted considerable interest because of their large surface area, high porosity, adjustable pore size, and adaptable surface functionalities. These properties enable a high loading capacity of therapeutic agents, which improves the DDSs efficiency.^4,5 Zeolitic imidazolate framework-8 (ZIF-8), a subgroup of MOFs made of zinc ions and imidazole ligands, is one of the most studied nanocarriers for drug delivery. The inherent biodegradability of ZIF-8 in acidic environments positions it as an excellent drug carrier for intelligent drug release in the acidic pH typically present at tumor sites.^6,7

Given that nanoparticles interact with biological systems at the surface level, optimizing their interface properties—such as surface chemistry, mechanical characteristics, and structural features—is essential for improving biocompatibility, immune evasion, circulation time, stability, and cargo delivery.⁸ Although polyethylene glycol (PEG) coating has long been the preferred method for surface functionalization of NPs, providing NP protection from serum protein adsorption and thus reducing macrophage recognition, recent clinical research has revealed the existence of anti-PEG immunity and rapid clearance from the circulation after multiple administrations, hindering its efficient in vivo performance.^9–11 At present, studies have focused on using nature-inspired and biomimetic technologies to overcome the limitations of traditional methods for modifying the surface of NPs. This approach utilizes a cell membrane coating strategy to develop NPs with an efficient biological interface on their surface, which involves the use of a variety of biomembranes from natural/cancer cell membranes to extracellular vesicles (EVs), especially the membrane of exosomes.^5,12–14 As extracellular nanovesicles, cell-secreted exosomes hold great promise for constructing biomimetic nanocarriers. NPs coated with exosome membrane offer a combination of the physicochemical advantages of nanoparticles, such as higher capacity for drug loading, ease of surface modification, and on-demand drug release, along with the biological benefits of exosomes, including strong biocompatibility, immune evasion ability, and prolonged circulation. It is worth noting that some drawbacks are associated with naturally cell-secreted exosomes, including ethical concerns related to existing contaminations in cell cultures, such as endotoxins, mycoplasma, viruses, and prion proteins. Moreover, low production and isolation yields, as well as difficulties in drug loading, are significant obstacles that need to be addressed for clinical applications.^15–17 Plant-derived exosome-like nanoparticles (ELNs) exhibit lower immunogenicity, superior biocompatibility, renewability, mass production capability, and cost-effectiveness. Additionally, these ELNs provide therapeutic effects due to their inherent bioactivity. Several studies have reported their anti-proliferative and anticancer properties. For instance, EVs derived from lemon, garlic, and ginger have shown anti-proliferative effects and induced apoptosis in different types of cancer cells.^18–20 In addition to their inherent biological activities, plant ELNs can act as outstanding nanovectors for the intercellular delivery of therapeutic compounds or agents with poor solubility. A phase I clinical trial (https://ClinicalTrials.gov Identifier: NCT01294072) is currently underway to assess the delivery capabilities of plant ELNs conjugated with curcumin to healthy and colon cancer tissues.²¹ The application of plant ELNs in the fabrication of biomimetic NPs represents an emerging trend in nanomedicine.²² In this context, Mao et al. introduced a new method for inflammatory bowel disease therapy using a biomimetic nanocomposite made of ginger-derived exosomes (GE) and a mesoporous silicon nanoparticle (LMSN). This nanocomposite (LMSN@GE) demonstrated benefits for oral administration of the antibody infliximab (INF), including stability in the gastrointestinal tract, targeted delivery to the colon, and enhanced permeability through the intestinal epithelium. Remarkably, in addition to its role in delivery, GE also exhibited an anti-inflammatory effect by inhibiting the NLRP3 inflammasome.²³

Understanding the unique characteristics of cancer cells within the tumor microenvironment (TME) can facilitate the design of DDSs specifically customized to target cancer cells. This active targeting strategy potentially improves the effectiveness of anticancer drugs and minimizes potential side effects. Cancer cells exhibit a unique characteristic, which is the pronounced expression of certain surface antigens, a feature that is not as prominent in normal tissues. This differential expression presents a valuable opportunity for the development of targeted therapies.²⁴ Epithelial cell adhesion molecule (EpCAM or CD326) as an overexpressed surface receptor in most carcinomas and cancer stem cells, with low expression levels in normal tissues, has been used as a useful target for cancer therapy.²⁵ Nucleic acid aptamers, as small strands of DNA or RNA with superior traits over other ligands, such as smaller size, simplicity of production, diminished immunogenicity, increased in vivo stability, and high affinity and selectivity towards target molecules, are valuable alternatives for cancer cell detection. The design of aptamer-conjugated nanocarriers to bind to overexpressed receptors on cancer cells ensures efficient tumor targeting and optimizes chemotherapeutic delivery.²⁶

Prostate cancer, accounting for the second most common cancer in men after skin cancer, is one of the major causes of death worldwide. Based on the statistics, the estimated annual number of new prostate cancer cases is forecasted to double from 1.4 million in 2020 to 2.9 million by 2040.²⁷ Pumpkin is being used as a traditional medicine in treating prostate-related problems, including prostate cancer. Considering the anti-inflammatory and anticancer properties of the extract and bioactive compounds present in Cucurbita moschata,^28–30 as well as the lack of studies on its exosomes, our study focused on isolating exosomes from the fruit of Cucurbita moschata (denoted as CEXO).

We aimed to demonstrate two functional applications of CEXOs: (1) their potential anticancer efficacy against prostate cancer (PC) cells and (2) their role as an exosome-mimicking functionalization source for ZIF-8/DOX nanoparticle delivery to PC cells. To this end, we first assessed the effect of characterized CEXOs on cell survival, cell cycle arrest, and apoptosis induction. In the following step, we utilized a biomimetic strategy to enhance the targeted delivery of doxorubicin (DOX), a widely used chemotherapy drug, to PC cells. Specifically, DOX was encapsulated within porous, acid-sensitive ZIF-8 NPs for smart release in the tumor area. These ZIF-8/DOX NPs were then coated with CEXOs (CEXO@ZIF-8/DOX) for improved biocompatibility and immune evasion, and functionalized with an EpCAM aptamer (Apt-CEXO@ZIF-8/DOX) to boost targeting toward PC cells that overexpress EpCAM receptors (Scheme 1). The fully characterized nanocarriers were used for in vitro experiments to assess cellular uptake, cell toxicity, immune evasion capability, and apoptosis induction. Immunocompromised C57BL/6 mice bearing PC-3 tumors were further used to evaluate tumor targeting, therapeutic efficacy, biosafety, and biodistribution of nanocarriers in vivo. To the best of our knowledge, this is the first study to report the use of edible plant ELNs to coat ZIF-8 NPs in cancer treatment.

Scheme 1 Schematic illustration of Apt-CEXO@ZIF-8/DOX nanocarrier preparation for prostate cancer targeted therapy.

2. Experimental section

2.1. Materials

Zinc nitrate hexahydrate Zn(NO₃)₂·6H₂O, 2-methylimidazole (2-MIM), 4′,6-diamidino-2-phenylindole (DAPI), 1-ethyl-3-(3-di-methylaminopropyl)carbodiimide hydrochloride (EDC), and N-hydroxysuccinimide (NHS) were purchased from Sigma-Aldrich. Dulbecco's modified Eagle's medium (DMEM), Roswell Park Memorial Institute 1640 (RPMI 1640) medium, penicillin/streptomycin, and fetal bovine serum (FBS) were provided by Gibco (Scotland). Trypsin and 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) were bought from Sigma-Aldrich. An apoptosis detection kit containing FITC-annexin V with propidium iodide (PI) was obtained from Mahboub Bioresearch (Iran). The BCA (bicinchoninic acid) protein quantification kit was purchased from Parstous (Iran). The 48-mer EpCAM DNA aptamer (5′-amine CAC TAC AGA GGT TGC GTC TGT CCC ACG TTG TCA TGG GGG GTT GGC CTG-3′) was procured from MicroSynth (Switzerland). The tris-acetate-EDTA buffer, DNA marker (50 bp), and agarose powder were purchased from DENAzist Asia (Iran). Additionally, ethidium bromide was obtained from SinaClon (Iran).

This study used three cell lines: human prostate cancer cells (PC-3), Chinese hamster ovary (CHO) cells, and murine macrophage-like RAW 264.7 cells, sourced from Ferdowsi University of Mashhad, Iran. RPMI 1640 medium supplemented with 10% (v/v) FBS and 1% penicillin/streptomycin was used for PC-3 and CHO cell culture, while the RAW 264.7 cells were cultured in DMEM-high glucose with 10% FBS and 1% penicillin–streptomycin. The cells were incubated in a humidified atmosphere with 5% CO₂ at 37 °C.

2.2. Isolation and characterization of Cucurbita exosomes

Cucurbita moschata was purchased from the local market. Exosome isolation was carried out using the ultrafiltration technique. The pumpkin was washed, peeled, and blended, followed by differential centrifugation (3000g for 10 min, 6000g for 30 min, and 10 000g for 60 min) to remove debris. The supernatant was filtered through a 0.22 μm membrane and processed using an ultrafiltration device. The ultrafiltration device being applied in the current project was registered in the Intellectual Property Center of Iran under Declaration Number 140050140003008723 on 7 February 2022 (Patent No. 108718, International Classification C12M 1/12; C12M 1/00). The device consists of two separate chambers: one for biological sample loading and the other for a salt washing solution. The biological sample is loaded in the lower chamber, and washing solution is placed in the upper one. A paper filter with 30-nanometer pores is installed at the bottom, and nitrogen gas pressure forces the sample through the filter. Particles smaller than 30 nm pass through the filter by repeated washing and exit the device, while larger particles (30–200 nm) are trapped on the surface of the filter. These particles can be collected and stored at −80 °C until further use.

The size distribution and zeta potential of isolated exosomes were measured using a dynamic light scattering instrument (DLS; VASCO-3, France) and a zeta potential analyzer (Zeta Compact, France). Morphology and structure of Cucurbita exosomes were examined by transmission electron microscopy (TEM; Zeiss LEO 912AB, Germany) after negative staining with uranyl acetate as well as atomic force microscopy (AFM; BRUKER, USA). The exosome quantification was conducted by determining the total protein content using a BCA protein assay kit.

2.3. Western blot analysis

To analyze the expression of exosome-relevant proteins including CD63 and TSG101, western blot analysis was performed. Briefly, the protein preparations were separated using 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and then transferred to polyvinylidene fluoride membrane. The membrane was blocked with 5% non-fat milk at room temperature. This was followed by an overnight incubation at 4 °C with the primary antibodies including anti-CD63 (sc-5275, Santa Cruz Biotechnology, USA; 1 : 300) and anti-TSG101 (sc-7964, Santa Cruz Biotechnology; 1 : 300). After washing, the membrane was incubated with the secondary antibody (sc-516102, Santa Cruz Biotechnology; 1 : 6000) at room temperature for 1 h. Finally, signals were visualized using chemiluminescent reagents.^31,32

2.4. Evaluation of anti-proliferative activity of Cucurbita exosomes

The cytotoxic effects of the exosomes was investigated on PC-3 and CHO cells using the MTT assay.³³ Cells were seeded at a density of 6 × 10³ cells per well in a 96-well plate. 24 h post incubation, the cells were treated with various concentrations of exosomes (15–400 μg ml⁻¹) for 24, 48 and 72 h. Subsequently, the supernatant was replaced with MTT solution (5 mg ml⁻¹) and incubated for 4 h. Finally, 150 μl of DMSO was added to dissolve the purple crystals, and the optical density (OD) was measured at 545 nm using an enzyme linked immunosorbent assay reader (ELISA reader; Awareness Technology, Inc., USA).

2.5. Cell cycle analysis

The effect of Cucurbita exosomes on cell cycle distribution was analyzed in PC-3 cells cultured at a density of 2 × 10⁵ cells per well in 6-well plates. The cells were treated with 150 and 200 μg ml⁻¹ exosomes for 48 h, then trypsinized, rinsed twice with cold phosphate-buffered saline (PBS) and fixed in 70% cold ethanol at 4 °C for 2 h. Following ethanol removal and two additional PBS washes, the cells were stained with propidium iodide and cell cycle distribution was analyzed using flow cytometry (BD Accuri C6, USA).³⁴

2.6. Apoptosis assay

To understand the mode of cell death induced by Cucurbita exosomes, an apoptosis assay was conducted. PC-3 and CHO cells (2 × 10⁵ cells per well) were treated with 150 and 200 μg ml⁻¹ exosomes for 48 h. The cells were then collected and stained with the annexin V-FITC kit with PI and analyzed using flow cytometry. The apoptotic rates of the cells were determined using FlowJo V10 software.

2.7. Synthesis of ZIF-8/DOX

ZIF-8/DOX was prepared using a one-pot method in accordance with the procedure previously described by Zheng et al., with some modifications.³⁵

In brief, 2 ml of DOX solution (2 mg ml⁻¹) was stirred with Zn(NO₃)₂·6H₂O solution (0.84 M) for 30 min. Then, 2-methylimidazole (2-MIM) solution (3 M) was slowly added dropwise to the above mixture and agitated for 15 min. The ZIF-8/DOX was precipitated by high-speed centrifugation and washed three times with deionized water. Finally, the resulting product was freeze-dried, and the obtained powder was used for characterization studies. Bare ZIF-8 was synthesized using a similar method with 2 ml of deionized water for comparison with ZIF-8/DOX. The supernatant collected from ZIF-8/DOX was analyzed to determine the amount of free DOX. This was done by measuring its absorbance at 480 nm using ultraviolet-visible spectrophotometry (UV/vis; Eppendorf, Germany) and referencing the DOX calibration curve. By substituting the resulting amount in eqn (1) and (2), drug encapsulation efficiency (DEE%) and drug loading capacity (DLC) were estimated.³⁶

2.8. Preparation of CEXO@ZIF-8/DOX

To prepare biomimetic nanoparticles, isolated exosomes were mixed with ZIF-8/DOX nanoparticles in a 2 : 1 mass ratio and incubated for 2 h. The coated NPs were then collected by centrifugation at 7000g for 5 min and washed with PBS tree times. The CEXO@ZIF-8/DOX NPs were dispersed in PBS for subsequent use.^13,37,38

2.9. Conjugation of aptamer onto CEXO@ZIF-8/DOX

The prepared CEXO@ZIF-8/DOX NPs were used to covalently conjugate the 5′-amine EpCAM aptamer to the carboxyl-containing molecules on the exosome surface via EDC/NHS chemistry. To this end, EDC (46 mg) and NHS (35 mg) were separately dissolved in PBS and then mixed with CEXO@ZIF-8/DOX suspension for 30 min to activate the carboxylic groups on the exosome surface. EpCAM aptamer (20 μl, 10 μM) was then added to the aforementioned solution and incubated overnight at room temperature while stirring. Finally, Apt-CEXO@ZIF-8/DOX NPs were separated from free aptamers by centrifugation at 7000g for 10 min.

2.10. Physical characterization of nanocarriers

Nanocarriers were synthesized according to the mentioned procedures and characterized by different methods. Ultraviolet-visible absorption spectra of samples were obtained using a UV-Vis spectrophotometer (Thermo Scientific 2000c, USA). The Fourier-transform infrared (FTIR) spectra of the prepared samples were recorded by an AVATAR 370 FTIR spectrometer (Thermo Nicolet spectrometer, USA). In order to determine the crystal structure of nanoparticles, an X-ray diffraction (XRD) device was used (PHILIPS_PW1730, Netherlands). The elemental composition of synthesized nanoparticles (C, N, O, Zn, and P) was evaluated by energy-dispersive X-ray analysis (EDX; TESCAN MIRA, Czech Republic). The morphology, size, and surface characteristics of the synthesized nanocarriers were further evaluated by AFM, field emission-scanning electron microscopy (FE-SEM; TESCAN MIRA, Czech Republic), and TEM. Exosome-coated NPs were negatively stained and then applied for TEM imaging. The size distribution and zeta potential of nanocarriers before and after coating were measured using DLS equipment and zeta potential analyzer. To verify the nanocarrier coating by exosome, 20 μg (quantified by BCA kit) of Cucurbita exosomes and CEXO@ZIF-8/DOX samples were subjected to 10% SDS-PAGE for membrane protein analysis. Gel staining was performed by Coomassie blue for 2 h. The conjugation of aptamer on the CEXO@ZIF-8/DOX surface was verified using an agarose gel electrophoresis assay. The DNA ladder, free aptamer, CEXO@ZIF-8/DOX, and Apt-CEXO@ZIF-8/DOX were loaded on a 2% (w/v) agarose gel and stained with ethidium bromide to visualize the aptamer by a gel documentation system (Major Science, USA).

2.11. In vitro drug release measurements

The pH-dependent release pattern of DOX from CEXO@ZIF-8/DOX was investigated by incubating 2.0 mg of CEXO@ZIF-8/DOX in 1 ml of PBS (pH 5.4, 6.4, and 7.4) at 37 °C with shaking at 100 rpm for 48 h. At various time intervals, the supernatants were collected through centrifugation, and the released DOX was quantified using a UV/Vis spectrophotometer at 480 nm. Finally, the cumulative DOX release was calculated over time.

2.12. Hemolysis assay

To evaluate the blood compatibility and practicability of the intravenous injection of nanocarriers, a hemolysis assay was performed. First, to isolate red blood cells (RBCs), human blood obtained from a healthy volunteer was centrifuged (1500g, 10 min at 4 °C), washed, and diluted (1 : 10) with PBS. RBC suspension (200 μl) was mixed with various concentrations of ZIF-8/DOX and CEXO@ZIF-8/DOX (25–800 μg ml⁻¹) and incubated at 37 °C for 4 and 12 h. After centrifugation of the mixtures at 2500g for 1 min, the absorbance of each supernatant was measured at 545 nm by a UV-Vis spectrophotometer. The hemolysis percentage was calculated by using the value of optical density for each sample, along with deionized water and PBS as positive and negative controls, using the eqn (3):³⁹

2.13. In vitro cytotoxicity assay

The in vitro therapeutic efficacy of nanocarriers was assessed using the MTT assay. PC-3 and CHO cells were cultured in triplicate in 96-well plates (6 × 10³ cells per well) for 24 h, then treated with varying concentrations of free DOX, CEXO@ZIF-8/DOX, and Apt-CEXO@ZIF-8/DOX (0.7–25 μg ml⁻¹; equivalent DOX concentrations) for 24, 48, and 72 h. At each time point, the supernatant was replaced by MTT solution (5 mg ml⁻¹) for 4 h. The supernatant was then replaced with DMSO to dissolve formazan crystals, and absorbance was measured at 545 nm using an ELISA microplate reader.

2.14. Assessment of in vitro cellular internalization

The cellular uptake of free DOX, CEXO@ZIF-8/DOX, and Apt-CEXO@ZIF-8/DOX was quantitatively evaluated using flow cytometry. PC-3 and CHO cells were cultured in 6-well plates (2 × 10⁵ cells per well) for 24 h, then treated with free DOX and referred nanocarriers at an equal DOX concentration (2 μg ml⁻¹) for 6 h. The cells were rinsed, detached, and centrifuged at 400g for 10 min. The final samples were re-suspended in 300 μl of cold PBS. The fluorescence signals were examined in the FL2 channel of the BD Accuri C6 flow cytometer, and the results were processed with FlowJo V10 software. For qualitative analysis, after the treatment period, the cells were washed three times with PBS and fixed in 4% paraformaldehyde for 15 min. Next, the cells were stained with DAPI and imaged using a fluorescent microscope (Olympus BX51, Japan) after three additional washes with PBS.

2.15. Assessment of cell death mechanism

The cell death mechanism activated by nanocarriers was investigated using flow cytometry. In brief, after culturing PC-3 and CHO cells in 6-well plates (2 × 10⁵ cells per well) for 24 h, the cells were subjected to free DOX, CEXO@ZIF-8/DOX, and Apt-CEXO@ZIF-8/DOX at the same DOX concentration (1 μg ml⁻¹) for 24 h. The treated cells were then collected and stained with annexin V-FITC/PI. Flow cytometry was used to quantify the fluorescence intensity in the FL1 and FL2 channels. Data analysis was performed via FlowJo V10 software.

2.16. Immune evasion

The immune-evasion capability of the nanoplatforms was evaluated through an in vitro experiment assessing nanoparticle uptake by the RAW 264.7 murine macrophage cell line. These cells were seeded in 6-well plates at a density of 2 × 10⁵ cells per well and exposed to free DOX, ZIF-8/DOX, CEXO@ZIF-8/DOX, and Apt-CEXO@ZIF-8/DOX at a DOX concentration of 5 μg ml⁻¹ for 4 h. Finally, the cells were washed, detached, and resuspended in cold PBS for flow cytometry analysis.

2.17. In vivo antitumor efficacy and biosafety

To investigate the in vivo antitumor efficacy of nanocarriers, PC-3 tumors were established in immunocompromised male C57BL/6 mice (4–6 weeks old). All animals were cared for in compliance with the Guide for the Care and Use of Laboratory Animals. The animal experiments were approved by the Animal Ethics Committee of Ferdowsi University of Mashhad (Approval No. IR.UM.REC.1401.267). After immunosuppression treatment based on previous reports,^40,41 all mice received subcutaneous injections of 1 × 10⁶ PC-3 cells into the right flank. Upon reaching tumor volumes of 60–70 mm³, the animals were divided into four groups (n = 3) and intravenously injected with PBS, free DOX, CEXO@ZIF-8/DOX, and Apt-CEXO@ZIF-8/DOX at a DOX dose of 5 mg kg⁻¹ body weight via the tail vein on days 1 and 3. Body weight and tumor volume were assessed on alternate days after treatment. Tumor volume was measured based on the (length × width × height)/2 formula. To assess possible side effects of nanocarriers, the mice were sacrificed on day 15. After isolating the main organs (heart, liver, lungs, kidneys, and spleen) and tumor, histopathological analysis was performed following hematoxylin and eosin (H&E) staining.

2.18. Ex vivo biodistribution analysis

To assess the distribution profile of nanocarriers within a biological system, free DOX, CEXO@ZIF-8/DOX, and Apt-CEXO@ZIF-8/DOX were intravenously injected at a DOX equivalent dose of 5 mg kg⁻¹ into immunocompromised mice bearing PC-3 tumor xenografts. After administration, the animals were sacrificed at 12 and 24 h intervals. Major organs (liver, kidneys, heart, lungs, and spleen) and the tumor were excised for fluorescence intensity analysis using an in vivo imaging system (IVIS; KODAK, USA).

2.19. Statistical analysis

Statistical data analysis was conducted using GraphPad Prism version 9.0.0 (GraphPad Software, San Diego, CA), based on at least three independent samples. Results are expressed as the mean ± standard deviation. Statistical significance was assessed using one-way analysis of variance (ANOVA). Significant levels were denoted by p-values: * < 0.05, ** < 0.01, *** < 0.001, and **** < 0.0001.

3. Results and discussion

3.1. Isolation and characterization of CEXOs

The CEXOs were isolated from the Cucurbita juice by differential centrifugation and ultrafiltration. CEXOs were visualized using TEM after negative staining with uranyl acetate. TEM analysis showed cup-shaped CEXOs with membrane integrity (Fig. 1A) and a size distribution consistent with that of other edible exosomes.⁴² The nanoscale size of the CEXOs was confirmed by DLS. The results from DLS analysis revealed that the average size of CEXOs was approximately 121 nm with a polydispersity index of 0.28. Additionally, the morphology and size distribution of CEXOs were further evaluated by AFM, which implied a nearly spherical shape for the CEXOs (Fig. 1B). The average zeta potential of the CEXOs was −13.34 ± 0.6 mV.

Fig. 1 Characterization of Cucurbita exosomes. Representative Cucurbita exosome images obtained by transmission electron microscopy (A); the arrows point to the exosomes, scale bar: 100 nm. Topography of Cucurbita exosomes observed under atomic force microscopy (B), scale bar: 500 nm. Immunoblotting analysis of exosome markers (TSG101 and CD63) (C).

Western blot analysis confirmed the presence of mammalian exosome markers TSG101 and CD63 on the CEXOs (Fig. 1C and Fig. S1, ESI†) which aligns with that of Momordica charantia ELNs.³² Additionally, western blotting of ELNs isolated from Arabidopsis thaliana leaves also showed surface marker tetraspanin proteins (CD9, CD63, and CD81) as well as TSG101 and ALIX proteins, which are common markers reported in mammalian exosomes.³¹ These data collectively validate the exosome-like nature of nanovesicles found in Cucurbita juice, as evidenced by their size and marker profile.¹⁹

3.2. CEXOs exerted selective toxicity on PC-3 cells

The impact of CEXOs on the viability of PC-3 cells was examined using the MTT assay. PC-3 cells were exposed to different EV concentrations for 24, 48, and 72 h. Our results illustrated a dose- and time-dependent cytotoxic effect of CEXOs on PC-3 cells (Fig. 2A). Notably, CEXOs evidently affected cell morphology within 24 h. Furthermore, the half-maximal inhibitory concentration (IC₅₀) of CEXOs against PC-3 cells was 153 and 79 μg ml⁻¹ after 48 and 72 h, respectively. Interestingly, CEXOs exhibited significantly lower toxicity on CHO cells (Fig. 2B), with IC₅₀ values of 1582 and 1580 μg ml⁻¹ after 48 and 72 h, respectively, demonstrating their tumor-selective inhibitory activity. In line with our results, Zhang and colleagues demonstrated that ACNVs isolated from Asparagus cochinchinensis exerted tumor-selective inhibition activity against hepatocellular carcinoma cells compared to normal cells, and they attributed this effect to various cellular uptake abilities.⁴³ Kim and his team hypothesized that the distinct endocytosis pathways for the internalization of plant sap-derived EVs could be the reason for their selective inhibitory effect on tumor cells, as opposed to normal cells. Specifically, cancer cells predominantly use a process called caveolae-mediated endocytosis, which helps prevent the EVs degradation by lysosomes. On the other hand, normal cells mainly use micropinocytosis to internalize these EVs, a process that results in their degradation by lysosomal compartments. However, they mentioned that this hypothesis needs further research.⁴⁴

Fig. 2 Evaluation of cytotoxicity, apoptosis, and cell cycle distribution in PC-3 and CHO cells after treatment with CEXOs. Cytotoxicity of Cucurbita exosomes against PC-3 (A) and CHO (B) cells after co-incubation with exosomes at different concentrations (31.25–500 μg ml⁻¹) for 24, 48, and 72 h. Flow cytometry assessment of apoptosis induced by Cucurbita exosomes (150 and 200 μg ml⁻¹) in PC-3 (C) and CHO (D) cells after 48 h, using FITC-annexin V/PI staining. Representative flow cytometry histograms of cell cycle distribution in exosome-treated PC-3 cells are shown after treatment with two concentrations (150 and 200 μg ml⁻¹) of CEXOs for 48 h (E). Data are expressed as mean ± standard deviation. ns non-significant, * p < 0.05, ** p < 0.01, *** p < 0.001, and **** p < 0.0001.

3.3. CEXOs specifically triggered apoptotic cell death in PC-3 cells

The annexin V-FITC/PI staining was performed to assess the effect of CEXOs on the induction of apoptosis in PC-3 cells. In comparison to untreated cells, PC-3 cells treated with CEXOs (150 μg ml⁻¹) exhibited an increased apoptotic rate (early and late apoptosis) of up to 36.5%, as shown in Fig. 2C. The apoptosis rate in PC-3 cells rapidly rose to 49.6% as the concentration increased to 200 μg ml⁻¹, highlighting that CEXOs have the ability to induce apoptosis in a dose-dependent manner. At doses of 150 or 200 μg ml⁻¹, no notable necrosis was observed. Conversely, in CHO cells, the percentage of early and late apoptotic cells (Q2 + Q3) after treatment with concentrations of 150 and 200 μg ml⁻¹ of CEXOs was similar to that of the control (Fig. 2D). ELNs derived from other plants such as garlic,¹⁹Asparagus,⁴³ tea flowers,⁴⁵ cannabis,³⁴ and Centella asiatica⁴⁶ have also been reported to induce apoptosis. The antitumor effects of ELNs are attributed to the bioactive molecules they contain.⁴⁷

3.4. CEXOs induced cell cycle arrest in PC-3 cells

To explore the mechanism of CEXOs on cell growth inhibition, PC-3 cells were treated with CEXOs at 150 and 200 μg ml⁻¹ for 48 h and analyzed using flow cytometry. The percentage of PC-3 cells in the G0/G1 phase significantly increased from 49.5 in untreated cells to 63.4 and 70.2% in CEXO-treated cells with concentrations of 150 and 200 μg ml⁻¹ for 48 h, respectively. Accordingly, the number of PC-3 cells in S phase decreased from 44.1% to 31.5% and 26%, respectively. Moreover, cell population in the G2/M phase was changed from 8.52% to 11.2% and 14.4%, respectively (Fig. 2E). These findings imply that CEXOs may hinder cancer cell proliferation, potentially by inducing G0/G1 cell cycle arrest.

All in all, the data presented here strongly suggest that CEXOs display anticancer effects on PC-3 cells, while they exerted minimum to no cytotoxic impact on normal cells. These CEXOs selectively decreased the viability by apoptosis induction and cell cycle arrest in PC-3 cells.

3.5. Characterization of biomimetic ZIF-8/DOX nanoparticles

Inspired by the cross-kingdom communication abilities and selective inhibition activity of CEXOs, and in light of the burgeoning advancements in biomimetic nanoparticles for cancer treatment, we developed a novel nanosystem using CEXOs. Our nanoplatform is structured with ZIF-8/DOX NPs forming the core and CEXO creating the shell (CEXO@ZIF-8/DOX). To further enhance targeted therapy, we equipped this platform with the EpCAM aptamer (Apt-CEXO@ZIF-8/DOX).

Accordingly, a one-pot method was used to synthesize the ZIF-8/DOX core by adding a 2-MIM solution to a mixture of zinc nitrate and DOX. Bare ZIF-8 was synthesized using a similar method. The successful synthesis of these nanocarriers was confirmed by comparing their characteristic attributes with those reported in the literature.

The color of ZIF-8/DOX turned purplish-red compared to the original ZIF-8. This color change indicates the integration of DOX into the ZIF-8 structure. As shown in Fig. 3A, the UV-Vis absorption spectra of ZIF-8/DOX further confirmed this, revealing a characteristic absorbance peak of DOX at 480 nm, which signifies successful loading of the drug into nanocarriers.⁴⁸ Moreover, the incorporation of DOX within ZIF-8 can be inferred from the obtained DLC% and DEE% with values of 28% and 88%, respectively. These values were derived using the standard curve of DOX. The FTIR spectrum of ZIF-8/DOX exhibited typical absorption bands associated with the ZIF-8 backbone (Fig. 3B). For instance, the C–N stretching bands at 1143 and 995 cm⁻¹, which correspond to the imidazole ring, as well as the Zn–N stretching vibration at 421 cm⁻¹, are some of these characteristic bands. Additionally, the presence of peaks at 1729 and 3437 cm⁻¹ was attributed to the carbonyl and hydroxyl groups of DOX.^49,50 The XRD analysis of ZIF-8/DOX exhibited characteristic peaks identical to those of ZIF-8 (Fig. 3C). The embedded DOX had a slight effect on the crystal structure of ZIF-8, suggesting that DOX is present in the cavity of ZIF-8.^12,51 EDX analysis of ZIF-8 and ZIF-8/DOX revealed an elemental composition consisting of Zn, O, C, and N, with a defined weight percentage (W%) within the nanocarrier structure, validating the successful synthesis of pure nanocarriers (Fig. 3D and E). The FE-SEM images confirmed that ZIF-8/DOX has a uniform morphology and a narrow size distribution, with a diameter of around 65 nm (Fig. 3F). DLS analysis revealed that incorporating DOX increased the average particle size of ZIF-8 from 56 to 62 nm, consistent with SEM observations. The zeta potential changed after the integration of DOX into ZIF-8, shifting from −11 mV in ZIF-8 to −8 mV in ZIF-8/DOX.

Fig. 3 Physicochemical and structural characteristics of ZIF-8/DOX nanoparticles. UV/Vis absorption of different samples (A). Fourier-transform infrared spectroscopy (FTIR) spectra of ZIF-8 and ZIF-8/DOX (B). XRD patterns of ZIF-8 and ZIF-8/DOX (C). EDX spectrum signals of ZIF-8 and ZIF-8/DOX (D) and (E). FE-SEM image of ZIF-8/DOX (F). Scale bar: 1 μm.

The use of nanoparticles, such as ZIF-8, is associated with several limitations, including rapid clearance rates by the reticuloendothelial system (RES). Surface modification of NPs with polyethylene glycol (PEG) has long been considered the gold standard for reducing RES uptake. However, emerging evidence indicates that the accelerated blood clearance (ABC) phenomenon may occur following the administration of PEGylated NPs in animal models.⁵² Biomimetic strategies offer a promising solution to improve the biointerface of NP surfaces. These strategies involve coating NPs with membranes derived from erythrocytes, neutrophils, cancer cells, stem cells, or extracellular vesicles. This coating enhances biocompatibility and stability while reducing immune responses.⁵³ Additionally, owing to their lower immunogenicity, reduced toxicity, and feasibility for scalable production, edible plant ELNs are ideal candidates for biomimetic approaches compared to EVs derived from mammalian cells.⁵⁴

To this end, the surface of ZIF-8/DOX NPs was coated with CEXOs via the incubation method. Certain interactions, including electrostatic, hydrophilic, and coordination interactions between the CEXOs and ZIF-8/DOX NPs, can play a pivotal role in the successful preparation of CEXO@ZIF-8@/DOX NPs.³⁸

TEM imaging confirmed the core–shell structure of CEXO@ZIF-8/DOX (Fig. 4A and Fig. S2, ESI†). The completely covered nanocarriers had a shell thickness of about 12 nm, which is close to the reported membrane thickness.⁵⁵ DLS analysis showed that the CEXO@ZIF-8/DOX NPs have a larger size (129 nm) than the bare ZIF-8/DOX. Furthermore, the shift in the surface charge of ZIF-8/DOX NPs from −8 to −14 mV implies the successful coating of the CEXOs membrane onto ZIF-8/DOX NPs. The CEXO coating was further corroborated by the SDS-PAGE electrophoresis results. As depicted in Fig. 4B, the CEXOs membrane protein pattern was preserved in CEXO@ZIF-8/DOX. Moreover, EDX mapping analysis indicated the existence of phosphorus (P) along with Zn, O, C, and N within the CEXO@ZIF-8/DOX nanoplatform (Fig. 4C), providing evidence of the successful membrane coating.

Fig. 4 Characterization of prepared nanoparticles. A representative TEM image of CEXO@ZIF-8/DOX (A), scale bar: 100 nm. Protein analysis of the Cucurbita exosomes and CEXO@ZIF-8 using SDS-PAGE (B). EDX mapping of CEXO@ZIF-8/DOX (C). Verification of aptamer conjugation by agarose gel electrophoresis (D). Cumulative DOX release profile from CEXO@ZIF-8/DOX at different pH values for 48 h (E). Data are displayed as mean ± standard deviation.

Passive targeting enables the accumulation of nanoparticles in cancer sites, mainly due to the EPR effect of the tumor. Active targeting, which involves attaching specific ligands like aptamers to the nanoparticles, can augment the EPR effect by specifically targeting cancer cells through ligand recognition of tumor biomarkers, enhancing selective drug delivery and therapeutic index.⁵⁶ In accordance with this concept, the synthesized CEXO@ZIF-8/DOX was functionalized by linking the amine groups of the EpCAM aptamer to carboxyl-containing molecules on the CEXO surface through an amide formation bond. The efficient binding of the EpCAM aptamer to the CEXO@ZIF-8 nanoparticles was verified using agarose gel electrophoresis (Fig. 4D). The free EpCAM aptamer, approximately 50 bp in size, moved downward in the gel and matched with the corresponding band on the DNA size marker. Apt-CEXO@ZIF-8 exhibited a bright band in the loading position, whereas CEXO@ZIF-8 generated no signals, confirming the successful fabrication of targeted nanocarriers. Aptamer binding slightly increased the size of CEXO@ZIF-8/DOX NPs to 130 nm and decreased surface charge to −16 mV.

3.6. The selective release behavior and pH sensitivity of CEXO@ZIF-8/DOX NPs was confirmed

The drug release behavior of CEXO-ZIF-8/DOX nanoplatforms was examined in PBS solutions with varying pH levels (pH 7.4, 6.4, and 5.4), which are representative of physiological status, the TME, and endosomes, respectively. As shown in Fig. 4E, only about 29.6% of DOX was released after 48 h at pH 7.4. This suggests that CEXO-ZIF-8/DOX can partially prevent DOX leakage, thereby reducing the drug side effects. However, after 48 h at acidic pH levels (6.4 and 5.4), 44.9% and 64.7% of DOX were released, respectively. This increased drug release under acidic conditions can be attributed to the acid-responsive characteristic of CEXO-ZIF-8/DOX. This characteristic results from disruption of coordination between zinc ions and the imidazole ring due to protonation of 2-methylimidazole linkers. This process facilitates ZIF-8 decomposition and subsequent DOX release in the acidic environment of the tumor, a feature that is beneficial for drug delivery.^57,58 Similarly, Gao et al. designed a biomimetic platform by coating DOX/IR-780/ZIF-8 with myeloma cell membrane. This platform demonstrated a higher cumulative DOX release in acidic pH compared to neutral pH. Additionally, the DOX release profiles from uncoated NPs were similar to cell membrane-coated NPs under both pH conditions, indicating cell membrane coating did not influence the release pattern.⁵⁹

3.7. CEXO coating of ZIF-8/DOX NPs improved their hemocompatibility

To ensure the hemocompatibility and safety of the prepared nanocarriers for intravenous administration, one of the fundamental assessments is the evaluation of their hemolytic potential.⁶⁰ The blood compatibility of ZIF-8/DOX and CEXO-ZIF-8/DOX was compared using a hemolysis assay. Released hemoglobin from RBCs incubated with different concentrations (12.5 to 800 μg ml⁻¹) of ZIF-8/DOX and CEXO@ZIF-8/DOX NPs served as an indicator of membrane disruption. Minimal hemolysis (below 1%) occurred in RBCs treated with both nanocarriers at 4 and 12 h intervals (Fig. 5A–D). It should be noted that the hemolysis rate of CEXO-ZIF-8/DOX was significantly lower than that of ZIF-8/DOX. Lebrón and colleagues reported that ZIF-8 with up to 500 μg ml⁻¹ concentration had no hemolytic effect.⁶¹ Furthermore, Han et al. demonstrated that apatinib/Ce6@ZIF-8 camouflaged by a 4T1 cell membrane (up to 500 μg ml⁻¹) had a hemolysis rate of less than 1% after both 4 and 24 h.⁶² In another study, it was shown that DOX/IR-780/ZIF-8 and its cell membrane-coated form, even at 1000 μg ml⁻¹, did not significantly change the hemolysis percentage compared to the PBS negative control.⁵⁹ Our results also indicated that these nanocarriers are non-hemolytic (0–2% hemolysis)⁶³ and CEXO-ZIF-8/DOX benefits from enhanced biocompatibility due to membrane coating.

Fig. 5 Hemocompatibility and immune evasion of exosome-coated nanoparticles. Visual observation of hemolytic activity of ZIF-8/DOX and CEXO@ZIF-8/DOX after 4 h (A) and 12 h (B) of incubation with different concentrations of nanoparticles. Comparison of hemolysis percentages of coated and uncoated nanocarriers at 4 h (C) and 12 h (D) of incubation. Flow cytometry analysis of RAW 264.7 cellular uptake after treatment with DOX, ZIF-8/DOX, CEXO@ZIF-8/DOX, and Apt-CEXO@ZIF-8/DOX nanoparticles (E). Quantification of the mean fluorescence intensity (MFI) of RAW 264.7 cellular uptake for different formulations (F). Data are displayed as means ± standard deviation. ns non-significant, and ****p < 0.0001.

3.8. Reduced immune system clearance with the CEXO coating

The capability to evade the immune system and accumulate at the desired location demonstrates the efficacy of systemic DDSs. The immune defense system relies heavily on macrophages to eliminate foreign particles from the bloodstream.⁶⁴ To compare the ability of CEXO-coated NPs with ZIF-8/DOX to evade macrophage detection, the murine macrophage-like cell line RAW 264.7 was used. The RAW 264.7 cells were incubated with ZIF-8/DOX, CEXO-ZIF-8/DOX, and Apt-CEXO@ZIF-8/DOX NPs, containing equivalent amounts of DOX, for 4 h. Flow cytometry analysis and mean fluorescence intensity values showed that RAW 264.7 cells internalized the ZIF-8/DOX approximately 1.6 times more than that of CEXO@ZIF-8/DOX (Fig. 5E and F), indicating a significant decrease in ZIF-8/DOX uptake after CEXO coating. Cheng et al. developed a biomimetic nanosystem (EMP) that integrates protein cargo (P) into ZIF-8 (M) and cloaks the nanoparticles with an EV membrane derived from MDA-MB-231 cells. They showed that the internalization of the EMP by RAW 264.7 is only 30% of that of MP.³⁸ The excellent stealth effect of CEXO could confer the ability to avoid clearance by the phagocyte system and prolong blood circulation time.

3.9. Higher uptake of aptamer-equipped NPs in PC-3 cells

Cell surface receptors play a pivotal role in controlling some cellular processes and are crucial targets for precise recognition and binding.⁶⁵ To verify the strong effect of aptamer functionalization in recognizing surface receptors on cancer cells, we compared the cellular uptake efficiency of EpCAM aptamer-conjugated nanocarriers in EpCAM⁺ (PC-3) versus EpCAM⁻ (CHO) cells using flow cytometry and fluorescent microscopy. Specifically, the fluorescent intensity of DOX in PC-3 and CHO cells treated with free DOX, CEXO@ZIF-8/DOX, and Apt-CEXO@ZIF-8/DOX was quantified using flow cytometry. PC-3 cells exhibited higher fluorescence intensity for Apt-CEXO@ZIF-8/DOX compared to CEXO@ZIF-8/DOX, indicating greater uptake of targeted nanocarriers. In contrast, CHO cells showed lower uptake of targeted nanocarriers than PC-3 cells. Notably, the internalization of free DOX was higher in both cell lines than in other groups (Fig. 6A and B), which is attributed to the fast diffusion of this small molecule. Additionally, fluorescent microscopy images confirmed the enhanced uptake of targeted nanocarriers in PC-3 cells, as evidenced by stronger red fluorescence in these cells (Fig. 6C and D). Ligand-mediated targeting can enhance the cellular internalization of NPs via receptor-mediated endocytosis.⁵⁶ In this regard, Falsafi et al. reported a higher cellular uptake of mucin-1 (MUC1)-conjugated RBC-MOF@DOX compared to RBC-MOF@DOX in MUC1 receptor-overexpressing 4T1 cells.⁶⁶ Similarly, a targeted nanoplatform (FA-EM@MnO₂/ZIF-8/ICG) designed by Li et al. demonstrated a fourfold increase in the fluorescence intensity of ICG in 4T1 cells with abundant folate receptors compared to normal GES-1 cells.⁶⁷ Our findings suggest that Apt-CEXO@ZIF-8/DOX NPs can specifically bind to EpCAM receptors, which are overexpressed on the surface of PC-3 cells, resulting in greater uptake of targeted NPs by these cells.

Fig. 6 Evaluating the uptake of prepared nanocarriers by PC-3 and CHO cells. Cellular internalization of DOX, CEXO@ZIF-8/DOX, and Apt-CEXO@ZIF-8/DOX was investigated after 4 h of treatment by flow cytometry and fluorescence microscopy in PC-3 (A) and (C), and CHO (B) and (D) cells, respectively. The blue fluorescence represents nuclei stained with DAPI. Scale bar: 100 μm.

3.10. Targeted NPs exhibited enhanced cytotoxic effects in vitro

The therapeutic efficacy of non-targeted and targeted nanocarriers was compared through cytotoxicity evaluation of free DOX, CEXO@ZIF-8/DOX, and Apt-CEXO@ZIF-8/DOX on PC-3 (EpCAM⁺) and CHO (EpCAM⁻) cells using the MTT assay. PC-3 cells exhibited significantly lower viability after exposure to Apt-EXO@ZIF-8/DOX compared to CEXO@ZIF-8/DOX in a concentration- and time-dependent manner (Fig. 7A–C), indicating the selective toxic potency of targeted NPs against EpCAM⁺ cells. The cellular toxicity of Apt-CEXO@ZIF-8/DOX was significantly higher at almost all concentrations compared to CEXO@ZIF-8/DOX in PC-3 cells. This enhanced cell growth inhibition potency can be ascribed to higher cellular uptake of targeted NPs by PC-3 cells, leading to increased DOX accumulation. Conversely, targeted and non-targeted NP treatments resulted in similar levels of cell toxicity in CHO cells (Fig. 7D–F), which was substantially milder than the toxicity observed in PC-3 cells. The validity of this claim is supported by the IC₅₀ values listed in Table 1. Specifically, the IC₅₀ values for PC-3 cells treated with targeted and non-targeted nanoparticles were 1.8 and 4.8 μg ml⁻¹ within 24 h, respectively, while those for CHO cells were 13 μg ml⁻¹ in the same time period. Free DOX induced the highest cell death compared to the other two groups in both cell lines, likely due to its rapid diffusion into cells and direct action on the nucleus, whereas the carriers release the drug in the cytoplasm after entering the cell.⁶⁸ Our results highlight the crucial role of the EpCAM aptamer in actively targeting EpCAM-overexpressing PC-3 cells. This targeting leads to increased delivery of chemotherapeutics specifically to the cancerous cells, maximizing their cytotoxic properties while minimizing off-target effects. Similarly, Lin et al. demonstrated that coating DOX@ZIF-8 NPs with an erythrocyte membrane conjugated with a tumor-targeting RGD peptide (DOX@ZIF-8@eM-cRGD) resulted in significantly higher cytotoxicity in HeLa cells compared to DOX@ZIF-8@eM. These targeted NPs actively recognized cancer cells by preferentially binding to integrin αvβ3 receptors that are overexpressed on the surface of HeLa cells.⁶⁸

Fig. 7 In vitro cytotoxicity evaluation. Cytotoxicity assessment of free DOX, CEXO@ZIF-8/DOX, and Apt-CEXO@ZIF-8/DOX on PC-3 (A)–(C) and CHO (D)–(F) cells after 24, 48, and 72 h of incubation with different concentrations. Data are expressed as mean ± standard deviation. * p < 0.05, ** p < 0.01, *** p < 0.001, and **** p < 0.0001. Apoptosis analysis of PC-3 cells by flow cytometry (G) after incubation with prepared formulations and FITC-annexin V/PI staining. Viable, early, and late apoptotic cell populations are presented as Q4, Q3, and Q2, respectively.

Table 1 IC₅₀ values of different treatment groups on PC-3 and CHO cells during 24, 48, and 72 h

Treatments	IC50 (μg ml^−1) (PC-3 cells)			IC50 (μg ml^−1) (CHO cells)
	24 h	48 h	72 h	24 h	48 h	72 h
Free DOX	0.15	0.05	0.04	5.63	0.73	0.32
CEXO@ZIF-8/DOX	4.83	2.09	2.08	13.31	3.64	1.79
Apt-CEXO@ZIF-8/DOX	1.84	1.03	0.91	13.58	4.16	1.83

---

3.11. Targeted NPs induced higher apoptotic cell death in PC-3 cells

Since apoptosis is regarded as the key mechanism associated with chemotherapy-induced cell death, this process and the extent of apoptosis in PC-3 cells across different treatment groups were investigated using an annexin V-FITC apoptosis kit and flow cytometry. As shown in Fig. 7G, the total apoptosis (early and late apoptosis) of PC-3 cells was 88.4%, 63.9%, and 89.9% following treatment with free DOX, CEXO@ZIF-8/DOX, and Apt-CEXO@ZIF-8/DOX, respectively. Targeted NPs exhibited greater potency in inducing apoptosis in PC-3 cells compared to non-targeted ones at an equal concentration of DOX. This enhanced effect can be attributed to the higher receptor-mediated uptake observed in the targeted group. Our results were in line with previous research by Lin and colleagues. They also reported the highest levels of reactive oxygen species (ROS) and apoptotic cell population in HeLa cells after treatment with targeted nanoparticles (DOX@ZIF-8@eM-cRGD), compared to other groups.⁶⁸ Our consistent findings from both the MTT and apoptosis assays suggest that Apt-CEXO@ZIF-8/DOX could induce cell death in PC-3 tumor cells through targeted therapy.

3.12. Targeted NPs exerted selective antitumor effects in vivo

Encouraged by the in vitro results, immunocompromised C57BL/6 mice with xenograft prostate tumors were used as a model to investigate the targeting potential and antitumor effects of prepared nanocarriers. Tumor volume, as an indicator of nanocarrier efficacy, was measured during the treatment period. While the control group showed an increasing trend in tumor volumes, the tumor growth was effectively suppressed in other groups (Fig. 8A and B). The potency of tumor growth inhibition by free DOX and Apt-CEXO@ZIF-8/DOX was approximately identical, and they showed significantly smaller tumor sizes than the control group. However, the CEXO@ZIF-8/DOX group exhibited lower efficacy in tumor growth inhibition compared to the free drug. Importantly, there was a significant difference in tumor size between targeted and non-targeted nanoparticles, highlighting the effect of NP functionalization using the EpCAM aptamer. Harnessing an active targeting strategy facilitated selective receptor recognition on PC-3 cells, increased the specific uptake and accumulation of chemotherapeutics in tumor cells, and ultimately enhanced the effectiveness of antitumor agents. This enhancement is evidenced by a significant reduction in tumor volume in mice treated with Apt-CEXO@ZIF-8/DOX compared to those receiving CEXO@ZIF-8/DOX. The findings were further supported by histological examination of tumor sections using H&E staining (Fig. 8C). Light microscopic observation showed a high level of tumor necrosis in free DOX and Apt-CEXO@ZIF-8/DOX groups. The significant tumor necrotic area in the targeted group, compared to the non-targeted one, suggests deeper penetration of Apt-CEXO@ZIF-8/DOX nanoparticles. These results arise from a combination of both the EPR effects and the active targeting facilitated by the EpCAM aptamer.

Fig. 8 Comparing the antitumor efficacy of various formulations in immunocompromised C57BL/6 mice bearing PC-3 tumors. Tumor volume curves (A) of mice treated with PBS, free DOX, CEXO@ZIF-8/DOX, and Apt-CEXO@ZIF-8/DOX were plotted during 15 days of treatments. Data are expressed as mean ± standard deviation. ****p < 0.0001. Images of dissected tumors (B) from various treatment groups at the end of the 15-day period. H&E staining of tumor sections from different groups (C). “N” indicates necrotic areas within the tumor tissue. Scale bar: 50 μm.

3.13. Designed NPs demonstrated excellent biosafety in vivo

H&E staining of the main organs (heart, liver, lung, kidney, and spleen) was performed in each treatment group to assess whether the nanocarriers induced any organ damage. Additionally, the in vivo biosafety of prepared formulations was evaluated by monitoring body weight fluctuations in each group. These examinations reveal the possible side effects of the nanocarriers for in vivo application. Fig. 9A shows histopathological examination of main organs in the various treatment groups. The results demonstrated that no obvious damage or pathological changes were observed in the experimental groups except for free DOX. In the free drug group, light microscopic observation of liver tissue showed mild chronic inflammation in the portal area (blue arrow). Additionally, alveolar damage with massive hemorrhage (red arrow) and mild infiltration of mononuclear inflammatory cells (blue arrow) was detected in the lung tissue obtained from the DOX-treated group. Kidney tissue also demonstrated histological damage following free DOX treatment. The mild infiltration of mononuclear inflammatory cells (blue arrow) along with mild renal congestion (red arrows) was evident from microscopic observation. It should be noted that the body weight of treated mice with free DOX was significantly reduced compared to other groups (Fig. 9B). These results indicated that Apt-CEXO@ZIF-8/DOX could not only inhibit tumor growth effectively but also did not damage the main organs.

Fig. 9 In vivo biosafety assessment in mice bearing PC-3 tumors. Representative histological images of main organs after PBS, free DOX, CEXO@ZIF-8/DOX, and Apt-CEXO@ZIF-8/DOX treatments using H&E staining (A). Scale bar: 50 μm. Body weight changes in different treatment groups during the 15-day period (B). Data are expressed as mean ± standard deviation. *p < 0.05, **p < 0.01.

3.14. Targeted NP accumulation in tumors was visualized using ex vivo fluorescence imaging

To investigate the biodistribution of free DOX and DOX-loaded targeted/non-targeted nanocarriers within major organs and tumor, the emitted fluorescent intensity of each sample was measured using an ex vivo imaging technique at 12 and 24 h post-treatment in PC-3 tumor-bearing mice. As depicted in Fig. 10A and B, the fluorescence intensity of the targeted NPs within the tumor tissue was significantly elevated compared to the non-targeted NPs. Importantly, a marked increase in the accumulation of targeted nanoparticles within the tumor was observed compared to the free DOX group after 24 h (Fig. 10B). This high tumor targeting efficiency of Apt-CEXO@ZIF-8/DOX was ascribed to the specific binding of EpCAM aptamers to receptors on the membrane of PC-3 tumor cells. In contrast, the distribution of targeted NPs within the major organs was significantly lower compared to other groups, and this difference was especially noticeable in the heart, lung, and spleen 24 h post-treatment. The non-targeted group presented fluorescence signals with low intensity in the tumor and high intensity in the main organs compared to targeted NPs, indicating that the EPR effect alone is inadequate for favorable NP accumulation at the tumor site. Nanoparticles that reach the tumor site via the EPR effect may be pushed back into the bloodstream due to elevated fluid pressure in the tumor interstitium. However, ligand-conjugated NPs can adhere to receptors on tumor cells, enhancing their internalization.⁵⁶ As shown in Fig. 10B, the fluorescence signals in some organs weakened 24 h post-injection in all groups, indicating their gradual elimination from the body. The results were supported by quantitative region-of-interest (ROI) analysis (Fig. 10C and D), which showed increased mean fluorescence intensity in the tumors of the targeted group compared to those in the non-targeted and free DOX groups at both 12 and 24 h post-injection. This observation suggests that Apt-CEXO@ZIF-8/DOX nanocarriers, in comparison with free DOX, have the capability for selective tumor targeting, which consequently leads to a reduction in systemic toxicity and biosafety enhancement.

Fig. 10 Evaluating the biodistribution of nanocarriers in C57BL/6 mice bearing PC-3 tumors. Ex vivo fluorescence images and mean fluorescence intensity (MFI) of different organs and tumor tissue at 12 h (A) and (C), and 24 h (B) and (D) post-administration of PBS, free DOX, CEXO@ZIF-8DOX, and Apt-CEXO@ZIF-8/DOX. The values are shown as mean ± standard deviation. ** p < 0.01 and **** p < 0.0001.

4. Conclusion

In this study, we designed a biomimetic drug delivery system with the dual nature of Cucurbita-derived exosomes and a metal–organic framework that integrates different functionalities (e.g., shielding, targeting, and pH responsiveness) for efficient delivery of the chemotherapeutic agent DOX to prostate tumor cells. To the best of our knowledge, this study is the first to investigate the cytotoxicity of Cucurbita exosomes and the application of these edible exosomes for the construction of biomimetic nanocarriers in prostate cancer treatment. Cucurbita exosomes exhibited significant cytotoxicity against prostate cancer cells, as evidenced by proliferation inhibition, cell cycle arrest, and apoptosis induction. We carried out a feasibility study to evaluate the efficiency of edible exosomes in coating drug-containing nanoparticles and facilitating aptamer binding to target tumor cells. The designed biomimetic nanosystem effectively evaded phagocyte-dependent clearance and selectively accumulated at the tumor site, resulting in enhanced therapeutic outcomes and reduced toxicity in healthy tissues. This study opens new avenues to overcome limitations associated with leveraging cell membranes and cell-secreted exosome membranes in the development of biomimetic drug delivery systems by substituting edible exosomes. Selecting the appropriate edible exosomes with effective bioactive contents targeting certain cancer cells can synergistically enhance the therapeutic effects of drug-loaded nanosystems. This strategy has the potential to revolutionize the safe and effective production of drug delivery systems.

Author contributions

AS: methodology; data analysis; investigation; software; visualization; writing original draft. ARB: validation; data analysis; funding acquisition. ASS: supervision; validation. MMM: supervision; conceptualization, validation; data analysis; funding acquisition; review & editing.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding authors on reasonable request.

Conflicts of interest

The authors declare that there is no conflict of interest.

Acknowledgements

This work was supported by Ferdowsi University of Mashhad, grant number: 3.59512. The authors would like to thank Mr. Mohammad Hasan Mollaei, Mr. Malaekeh Nikouei, Dr. Shaterzadeh, and Dr. Norouzpour Laboratory for their excellent support and technical assistance.

References

1. M. Shao, D. Lopes, J. Lopes, S. Yousefiasl, A. Macário-Soares, D. Peixoto, I. Ferreira-Faria, F. Veiga, J. Conde, Y. Huang, X. Chen, A. C. Paiva-Santos and P. Makvandi, Matter, 2023, 6, 761–799.
2. Z. Li, S. Tan, S. Li, Q. Shen and K. Wang, Oncol. Rep., 2017, 38, 611–624.
3. S. He, L. Wu, X. Li, H. Sun, T. Xiong, J. Liu, C. Huang, H. Xu, H. Sun, W. Chen, R. Gref and J. Zhang, Acta Pharm. Sin. B, 2021, 11, 2362–2395.
4. J. Yang and Y. Yang, Small, 2020, 16, 1–24.
5. W. Liu, Q. Yan, C. Xia, X. Wang, A. Kumar, Y. Wang, Y. Liu, Y. Pan and J. Liu, J. Mater. Chem. B, 2021, 9, 4459–4474.
6. C. Adhikari, A. Das and A. Chakraborty, Mol. Pharmaceutics, 2015, 12, 3158–3166.
7. F. Li, T. Chen, F. Wang, J. Chen, Y. Zhang, D. Song, N. Li, X.-H. Lin, L. Lin and J. Zhuang, ACS Appl. Mater. Interfaces, 2022, 14, 21860–21871.
8. C. Y. Lee, S. M. Hu, J. Christy, F. Y. Chou, T. C. Ramli and H. Y. Chen, Adv. Mater. Interfaces, 2023, 10, 2202286.
9. J. M. J. M. Ravasco, A. C. Paiva-Santos and J. Conde, Nat. Rev. Bioeng., 2023, 156–158.
10. H. Zhang, S. Dong, Z. Li, X. Feng, W. Xu, C. M. S. Tulinao, Y. Jiang and J. Ding, Asian J. Pharm. Sci., 2020, 15, 397–415.
11. H. Liu, Y. Y. Su, X. C. Jiang and J. Q. Gao, Drug Delivery Transl. Res., 2023, 13, 716–737.
12. Z. Li, H. Cai, Z. Li, L. Ren, X. Ma, H. Zhu, Q. Gong, H. Zhang, Z. Gu and K. Luo, Bioact. Mater., 2023, 21, 299–312.
13. B. Illes, P. Hirschle, S. Barnert, V. Cauda, S. Wuttke and H. Engelke, Chem. Mater., 2017, 29, 8042–8046.
14. D. Lopes, J. Lopes, M. Pereira-Silva, D. Peixoto, N. Rabiee, F. Veiga, O. Moradi, Z. H. Guo, X. D. Wang, J. Conde, P. Makvandi and A. C. Paiva-Santos, Mil. Med. Res., 2023, 10, 1–26.
15. M. Shao, D. Lopes, J. Lopes, S. Yousefiasl, A. Macário-Soares, D. Peixoto, I. Ferreira-Faria, F. Veiga, J. Conde, Y. Huang, X. Chen, A. C. Paiva-Santos and P. Makvandi, Matter, 2023, 6, 761–799.
16. M. Lu and Y. Huang, Biomaterials, 2020, 242, 119925.
17. O. Urzì, R. Olofsson Bagge and R. Crescitelli, J. Extracell. Vesicles, 2022, 11, 12271.
18. S. Raimondo, F. Naselli, S. Fontana, F. Monteleone, A. Lo Dico, L. Saieva, G. Zito, A. Flugy, M. Manno, M. A. Di Bella, G. De Leo and R. Alessandro, Oncotarget, 2015, 6, 19514–19527.
19. İ. Özkan, P. Koçak, M. Yıldırım, N. Ünsal, H. Yılmaz, D. Telci and F. Şahin, Sci. Rep., 2021, 11, 14773.
20. R. Anusha, M. Ashin and S. Priya, Food Chem. Toxicol., 2023, 182, 114102.
21. Z. Xu, Y. Xu, K. Zhang, Y. Liu, Q. Liang, A. Thakur, W. Liu and Y. Yan, J. Nanobiotechnol., 2023, 21, 1–13.
22. C. Cui, M. Du, Y. Zhao, J. Tang, M. Liu, G. Min, R. Chen, Q. Zhang, Z. Sun and H. Weng, ACS Appl. Mater. Interfaces, 2024, 16, 53460–53473.
23. Y. Mao, M. Han, C. Chen, X. Wang, J. Han, Y. Gao and S. Wang, Nanoscale, 2021, 13, 20157–20169.
24. H. Tian, T. Zhang, S. Qin, Z. Huang, L. Zhou, J. Shi, E. C. Nice, N. Xie, C. Huang and Z. Shen, J. Hematol. Oncol., 2022, 15, 132.
25. J. Zhong, J. Ding, L. Deng, Y. Xiang, D. Liu, Y. Zhang, X. Chen and Q. Yang, Drug Des., Dev. Ther., 2021, 15, 3985–3996.
26. Y. Li, Y. Duo, S. Bao, L. He, K. Ling, J. Luo, Y. Zhang, H. Huang, H. Zhang and X. Yu, Int. J. Nanomed., 2017, 12, 6239–6257.
27. S. Zhang, X. Ji, Z. Liu, Z. Xie, Y. Wang, H. Wang and D. Ni, J. Am. Chem. Soc., 2024, 146, 22530–22540.
28. M. Russo, S. Moccia, S. Bilotto, C. Spagnuolo, M. Durante, M. S. Lenucci, G. Mita, M. G. Volpe, R. P. Aquino and G. L. Russo, Oxid. Med. Cell. Longevity, 2017, 2017, 1–15.
29. W. Shen, C. Chen, Y. Guan, X. Song, Y. Jin, J. Wang, Y. Hu, T. Xin, Q. Jiang and L. Zhong, Int. J. Biol. Macromol., 2017, 104, 681–686.
30. S. Moccia, M. Russo, M. Durante, M. S. Lenucci, G. Mita and G. L. Russo, Curr. Res. Biotechnol., 2020, 2, 74–82.
31. S. Jokhio, I. Peng and C.-A. Peng, Protoplasma, 2024, 261, 1025–1033.
32. H. Cai, L.-Y. Huang, R. Hong, J.-X. Song, X.-J. Guo, W. Zhou, Z.-L. Hu, W. Wang, Y.-L. Wang, J.-G. Shen and S.-H. Qi, Front. Pharmacol., 2022, 13, 908830.
33. T. Mosmann, J. Immunol. Methods, 1983, 65, 55–63.
34. T. Tajik, K. Baghaei, V. E. Moghadam, N. Farrokhi and S. A. Salami, Biomed. Pharmacother., 2022, 152, 113209.
35. H. Zheng, Y. Zhang, L. Liu, W. Wan, P. Guo, A. M. Nyström and X. Zou, J. Am. Chem. Soc., 2016, 138, 962–968.
36. Z. Qian, N. Zhao, C. Wang and W. Yuan, J. Mater. Sci. Technol., 2022, 127, 245–255.
37. Y. Wang, D. Zhang, M. Jia, X. Zheng, Y. Liu, C. Wang, F. Lei, H. Niu and C. Li, J. Drug Targeting, 2022, 30, 1006–1016.
38. G. Cheng, W. Li, L. Ha, X. Han, S. Hao, Y. Wan, Z. Wang, F. Dong, X. Zou, Y. Mao and S. Y. Zheng, J. Am. Chem. Soc., 2018, 140, 7282–7291.
39. R. Y. Zhang, K. Cheng, X. Sun, X. Q. Yang, Y. Li, Y. G. Hu, X. S. Zhang, B. Liu, W. Chen, Y. Di Zhao and D. S. Huang, Chem. Eng. J., 2022, 450, 138337.
40. M. Jivrajani, M. V. Shaikh, N. Shrivastava and M. Nivsarkar, Anticancer Res., 2014, 34, 7177–7183.
41. M. Aghasizadeh, T. Moghaddam, A. R. Bahrami, H. Sadeghian, S. J. Alavi and M. M. Matin, Life Sci., 2022, 293, 120272.
42. X. Lu, Q. Han, J. Chen, T. Wu, Y. Cheng, F. Li and W. Xia, J. Agric. Food Chem., 2023, 71, 8413–8424.
43. L. Zhang, F. He, L. Gao, M. Cong, J. Sun, J. Xu, Y. Wang, Y. Hu, S. Asghar, L. Hu and H. Qiao, Int. J. Nanomed., 2021, 16, 1575–1586.
44. K. Kim, H. J. Yoo, J.-H. Jung, R. Lee, J.-K. Hyun, J.-H. Park, D. Na and J. H. Yeon, J. Funct. Biomater., 2020, 11, 22.
45. J. J. Rennick, A. P. R. Johnston and R. G. Parton, Nat. Nanotechnol., 2021, 16, 266–276.
46. J. Y. Huang, X. Y. Cao, W. F. Wu, L. Han and F. Y. Wang, Biomed. Pharmacother., 2024, 176, 116855.
47. Q. Chen, M. Zu, H. Gong, Y. Ma, J. Sun, S. Ran, X. Shi, J. Zhang and B. Xiao, J. Nanobiotechnol., 2023, 21, 6.
48. Z. Zhao, Z. Liu, Y. Hua, Y. Pan, G. Yi, S. Wu, C. He, Y. Zhang and Y. Yang, Front. Pharmacol., 2022, 13, 850534.
49. C. Adhikari, A. Das and A. Chakraborty, Mol. Pharmaceutics, 2015, 12, 3158–3166.
50. A. Mittal, S. Gandhi and I. Roy, Sci. Rep., 2022, 12, 1–11.
51. H. S. Bahlol, J. Zhang, Z. Song, K. Zhang, Z. Ma, W. Zhang and H. Han, ACS Appl. Nano Mater., 2024, 7, 10941–10951.
52. Q. F. Meng, L. Rao, M. Zan, M. Chen, G. T. Yu, X. Wei, Z. Wu, Y. Sun, S. S. Guo, X. Z. Zhao, F. B. Wang and W. Liu, Nanotechnology, 2018, 29, 134004.
53. Y. Wang, M. Zeng, T. Fan, M. Jia, R. Yin, J. Xue, L. Xian, P. Fan and M. Zhan, Int. J. Nanomed., 2024, 19, 5523–5544.
54. Z. Qiao, K. Zhang, J. Liu, D. Cheng, B. Yu, N. Zhao and F. J. Xu, Nat. Commun., 2022, 13, 1–16.
55. H. Cheng, J. Zhu, S. Li, J. Zeng, Q. Lei, K. Chen, C. Zhang and X. Zhang, Adv. Funct. Mater., 2016, 26, 7847–7860.
56. Z. Fu and J. Xiang, Int. J. Mol. Sci., 2020, 21, 9123.
57. C. Xu, L. Cao, T. Liu, H. Chen and Y. Li, Environ. Sci.: Nano, 2023, 10, 2578–2590.
58. L. Ren, Y. Sun, J. Zhang, L. Nie, A. Shavandi, K. E. Yunusov, U. E. Aharodnikau, S. O. Solomevich and G. Jiang, Int. J. Pharm., 2024, 652, 123811.
59. G. Gao, J. Che, P. Xu, B. Chen and Y. Zhao, Aggregate, 2024, 5, e631.
60. K. de la Harpe, P. Kondiah, Y. Choonara, T. Marimuthu, L. du Toit and V. Pillay, Cells, 2019, 8, 1209.
61. J. A. Lebrón, F. J. Ostos, M. Martínez-Santa, F. García-Moscoso, M. López-López, M. L. Moyá, E. Bernal, S. Bachiller, G. González-Ulloa, D. Rodríguez-Lucena, T. Lopes-Costa, R. Fernández-Torres, E. Ruiz-Mateos, J. M. Pedrosa, M. Rafii-El-Idrissi Benhnia and P. López-Cornejo, J. Mater. Chem. B, 2024, 12, 5220–5237.
62. X. Han, C. Zhou, X. Luo, H. Pang, C. Han, L. Tang, Z. Yang, Y. Nong and C. Lu, Curr. Mol. Med., 2023, 24, 648–666.
63. S. Malehmir, M. A. Esmaili, M. Khaksary Mahabady, A. Sobhani-Nasab, A. Atapour, M. R. Ganjali, A. Ghasemi and A. Moradi Hasan-Abad, Front. Chem., 2023, 11, 1–14.
64. Z. Liu, L. Zhang, T. Cui, M. Ma, J. Ren and X. Qu, Angew. Chem., 2021, 133, 15564–15572.
65. J. Zhang, J. Shang, X. Tang and X. Xu, ACS Omega, 2023, 8, 48975–48983.
66. M. Falsafi, M. Zahiri, A. S. Saljooghi, K. Abnous, S. M. Taghdisi, A. Sazgarnia, M. Ramezani and M. Alibolandi, Microporous Mesoporous Mater., 2021, 325, 111337.
67. X. Li, Q. Ji, C. Yan, Z. Zhu, Z. Yan, P. Chen, Y. Wang and L. Song, Nanoscale Res. Lett., 2022, 17, 103.
68. Y. Lin, Y. Zhong, Y. Chen, L. Li, G. Chen, J. Zhang, P. Li, C. Zhou, Y. Sun, Y. Ma, Z. Xie and Q. Liao, Mol. Pharmaceutics, 2020, 17, 3328–3341.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5tb00086f

---