Functionalized biological metal–organic framework with nanosized coronal structure and hierarchical wrapping pattern for enhanced targeting therapy

Abstract

Inefficient tumor-targeted delivery and uncontrolled drug release are the major obstacles in cancer chemotherapy. Herein, inspired by the targeting advantage of coronavirus from its size and coronal structure, a coronal biological metal–organic framework nanovehicle (named as corona-BioMOF) is constructed for improving its precise cancer targeting ability. The designed corona-BioMOF is constructed as the carriers-encapsulated carrier model by inner coated with abundant protein-nanocaged doxorubicin particles and external decorated with high-affinity apoferritin proteins to form the spiky surface for constructing the specific coronal structure. The corona-BioMOF shows a higher affinity and an enhanced targeting ability towards receptor-positive cancer cells compared to that of MOF-drug composites without spiky surface. It also exhibits the hierarchical wrapping pattern-endowed controlled lysosome-specific drug release and remarkable tumor lethality in vivo. Moreover, water-induced surface defect-based protein handle mechanism is first proposed to shape the coronal-BioMOF. This work will provide a better inspiration for nanovehicle construction and be broadly useful for clinical precision nanomedicine.

1. Introduction

Nanoscale and nanostructure-based therapeutic agents have been playing important roles in cancer therapy [1]. Recently, various nanoagents have been designed as drug delivery platforms for cancer chemotherapy [2], [3], [4], [5]. However, engineering an advanced delivery system capable of both precise targeting and controlled drug release to prevent premature drug leakage and obtain better treatment outcomes remains a tremendous challenge.

It has been reported that the coronal structure of virus is beneficial to enhance the targeting ability with host cell [6]. For instance, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) shows high infectivity toward its host cells [7], [8], [9]. The spike proteins on the CoV surface significantly increase the adhesion and infection abilities partly due to the fact that the spiky shape is beneficial to the interaction between species [9], [10]. Recent studies have also confirmed that spiky surfaces enhanced the interaction of nanomaterials with bacteria and cancer cells [11], [12], [13]. Moreover, it has been reported the diameter of the CoV ranges between 50 nm and 140 nm, and the size affects its encounter efficiency with host cells and subsequent infection ability [14], [15]. The large nanocarriers (50–200 nm) can spontaneously accumulate around leaky regions of the tumor vasculature via the “enhanced permeability and retention” (EPR) effect, and the size-related advantages on enhancing targeting ability have emerging in drug delivery and cancer therapy [16]. Therefore, inspired from the CoV, engineering a nanovehicle with an appropriate size and a coronal structrue might be a workable approach to obtain enhanced targeting ability towards cancer cells.

Apart from targeting, controlled and efficient drug release is another key point in precise cancer treatment. Many artificial stimuli-responsive nanocarriers have been applied for controlled drug release through the following stimulation ways: light, pH, hypoxia, and multiple stimulations [17], [18], [19], [20], [21], [22]. Thereinto, metal–organic frameworks (MOFs), have been widely applied in the controlled delivery of various drugs, owing to their distinct stimulus responsiveness [23], [24], [25], [26], [27], [28], [29]. For example, zeolitic imidazolate framework-8 (ZIF-8) showed the properties of acid-induced disintegration; Zr-MOF showed high phosphate concentration induced collapse [30], [31], [32], [33]. In recent years, MOFs have been found to show protective effect toward encapsulated biomolecules [34], [35], [36], [37], [38]. Also, the encapsulated biomolecules might increase the functionality of MOFs. For example, some biomolecules, such as silk fibroin and ferritin, have distinct pH-sensitive hydrolytic degradation property different from MOFs [39], [40]. The composite of MOFs and biomolecules might produce a new pH-responsive performance. By considering the pH difference in tumor microenvironment (pH≈6.5) and subcellular organelles (pH≈5), we wonder whether a new biological MOF nanovehicle could be constructed to display a hierarchical pH-responsive property for achieving precise controlled drug release intracellularly.

In this work, we designed a coronal biological MOF nanovehicle (named as corona-BioMOF) to achieve both precise targeting and controlled drug release in cancer therapy (Scheme 1 ). Herein, ZIF-8 was introduced as the host material in corona-BioMOF, serving as the nanocarrier and tumor microenvironment (weak acid)-responsive drug nanocontrol. Then, apoferritin (AFt) was chosen as one of the important constituents for constructing corona-BioMOF due to the following three aspects: (1) It can specifically bind to the transferrin receptor 1 (TfR1), a cell membrane receptor up-regulated in malignant proliferating cells like MDA-MB-231 triple-negative breast cancer (TNBC) cells [41]; (2) It can serve as another nanocarrier to load small molecular chemotherapy drugs due to its structure of protein “cage”[42] (Scheme 1A) and can biomineralize MOF simultaneously (Scheme 1B), thus constructing the carriers-encapsulated carrier model in corona-BioMOF and decorating the surface of MOF to form spiky shape; (3) It can be disassembled into subunits at low pH and reassembled at near-neutral pH [43], making the deep design in controlled loading and release of drugs to become possible. Another constituent, glucose oxidase (GOx), was utilized to assist the biomineralization synthesis and famish cancer cells. By the following two-stage process, corona-BioMOF has shown its advantages in targeting ability and hierarchical pH-responsive controlled drug release: First, doxorubicin (DOX)-loaded AFt proteins (DOX@AFt) on the surface of corona-BioMOF endow the nanovehicle strong target activity. Second, weak-acidic (pH≈6.5), as well as the GOx-catalyzed gluconic acid production, will accelerate the degradation of ZIF-8, thus corona-BioMOF can slowly disintegrate accompanied by the exposure of more inclusions (DOX@AFt and GOx) and produce more binding sites towards cancer cells. After interaction, the decrease of pH during the cellular endocytosis drove the further disintegration of corona-BioMOF and the continued release of encapsulated DOX@AFt. Finally, explosive DOX release from DOX@AFt was achieved with the disassembly of AFt at low pH of lysosomes (pH≈5). Thus, the lysosome-specific DOX delivery was accomplished with commendably preventing premature drug leakage, and corona-BioMOF displayed the outstanding antitumor activity. We highly expect that the corona-BioMOF nanovehicle is of benefit to provide inspiration for the development of novel nanoagents in precise medicine.

Scheme 1.

Schematic illustration of the design and assembly process of the proposed corona-BioMOF nanovehicle and its programmed therapy against breast cancer. (A) Assembly process of DOX@AFt. (B) Preparation steps of the corona-BioMOF nanovehicle. (C) Therapy process of corona-BioMOF in vivo.

2. Materials and methods

2.1. Materials and reagents

AFt from equine spleen (0.2 μm filtered), 2-methylimidazole (2-MIM), zinc nitrate hexahydrate, thiazolyl blue tetrazolium bromide (MTT), and GOx were purchased from Sigma-Aldrich. DOX was obtained from Tokyo Chemical Industry Company. MDA-MB-231 and MCF-10A cells were obtained from Chinese National Collection of Authenticated Cell Cultures. MX-1 cells were kindly sponsored by Professor Yuqing Chen. LysoSensor Blue DND-167 was obtained from Thermo Fisher Scientific Incorporated. TfR1 (CD71) mouse monoclonal antibody and CoraLite488-conjugated goat anti-mouse lgG(H + L) were purchased from Proteintech Incorporated. Goat anti-Mouse lgG H&L (Alexa FluorR 555) preadsorbed was purchased from Abcam. Blocking buffer, antibody dilution buffer for immunofluorescence, antifade mounting medium with DAPI, and PBS were purchased from Beyotime Biotechnology Incorporated. Fetal bovine serum (FBS), Dulbecco’s modified eagle medium (DMEM), RPMI 1640 medium, penicillin/streptomycin, and 0.25 % trypsin-EDTA solution were purchased from Gibco. BCA protein assay kit was purchased from CoWin Biotech Co., ltd. Amicon Ultra 3 K filters were purchased from Merck Millipore Company. All the chemicals were of analytical grade and used without further purification.

2.2. Measurements and characterizations

TEM images were carried out on JEM-2100 at an accelerating voltage of 200 kV (JEOL, Japan). HAADF-STEM and ColorSTEM were measured with Apreo 2S (Thermo Scientific Apreo 2S, Czech). XRD patterns were recorded on a model D/max-RC X-ray diffractometer (Ragaku, Japan). N₂ adsorption − desorption isotherms were performed with ASAP-2050 automated sorption analyzer (Micromeritics, USA). DRIFTS studies were conducted by a VERTEX 70 FT-IR spectrometer (Bruker ltd., Germany) equipped with a DRIFTS accessory. Ultraviolet–visible (UV-vis) absorption spectroscopy was obtained with a Varian Cary 60 spectrophotometer (Agilent, USA). MTT assay was determined with a microplate reader (Thermo Scientific Multiskan GO, USA) at 490 nm. CLSM images were performed on a Nikon A1 confocal microscope (Nikon, Japan). Fluorescence imaging of animals was taken by IVIS spectrum (PerkinElmer, USA). The content of Zn elements in tumors, organs, and blood were examined by ICP-AES (Leeman Labs, USA).

2.3. Synthesis of ZIF-8

Typically, 2-MIM (17.5 mmol) was dissolved in 3.0 mL of methanol at 30 °C. Then, zinc nitrate hexahydrate (0.25 mmol, 0.5 mL) was mixed and stirred for 30 min. The white product was collected by centrifuging at 3500 rpm for 20 min and washed twice.

2.4. Synthesis of DOX@AFt

In detail, 2.5 mL of 0.1 mM DOX and 1.0 mL of 1.0 mg mL⁻¹ AFt were mixed and then agitated for 5 min. Then, 1.0 M hydrochloric acid was added to decrease the pH of solution to 2.5, which could disassemble the structure of AFt. The solution was stirred for 15 min. Then the pH of solution was adjusted to 6.5 with 1.0 M sodium hydroxide. The mixture was continuously stirred for 15 min to entrap DOX inside AFt cavity. Water exchange was performed several times through Amicon Ultra 3 K filters to remove the non-encapsulated DOX. Finally, DOX@AFt was collected and stored at 4 °C. To investigate the encapsulation efficiency of DOX in DOX@AFt protein cage, the filtrates of reaction mixture in different pH, assembly time, or concentrations of DOX were measured by fluorescence spectrum.

2.5. Synthesis of GOx@ZIF-8, DOX@AFt@ZIF-8, and (DOX@AFt + GOx)@ZIF-8

According to the previous reports with slight changes [34], [37], GOx@ZIF-8, DOX@AFt@ZIF-8, and (DOX@AFt + GOx)@ZIF-8 were prepared with the following procedures. For the synthesis of GOx@ZIF-8, GOx (1.5 mg) was first dispersed in an aqueous solution of 2-MIM (35 mmol, 6.3 mL) and stirred for 10 min at 30 °C. Afterward, 1.0 mL of 0.5 mmol zinc nitrate hexahydrate aqueous solution was added in the above mixture and stirred for another 30 min at 30 °C. The production was collected by centrifuging at 3500 rpm for 20 min, washed twice, and freeze-dried. For the synthesis of DOX@AFt@ZIF-8, DOX@AFt (7.5 mg) was dispersed in an aqueous solution of 2-MIM (35 mmol, 6.3 mL) and stirred for 10 min at 30 °C. Afterward, 1.0 mL of 0.5 mmol zinc nitrate hexahydrate aqueous solution was added in the above mixture and stirred for another 30 min at 30 °C. The production was collected by centrifuging at 3500 rpm for 20 min, washed twice, and freeze-dried. For the synthesis of (DOX@AFt + GOx)@ZIF-8, GOx (1.5 mg) and DOX@AFt (7.5 mg) were dispersed in an aqueous solution of 2-MIM (35 mmol, 6.3 mL) and stirred for 10 min at 30 °C. Afterward, 1.0 mL of 0.5 mmol zinc nitrate hexahydrate aqueous solution was added in the above mixture and stirred for another 30 min at 30 °C. The production was collected by centrifuging at 3500 rpm for 20 min, washed twice, and freeze-dried.

2.6. Synthesis of BioMOF

Firstly, GOx (1.5 mg) and DOX@AFt (1.5 mg) were dispersed in an aqueous solution of 2-MIM (35 mmol, 5.8 mL) with stirring for 10 min at 30 °C. Next, an aqueous solution of zinc nitrate hexahydrate (0.5 mmol, 1.0 mL) was added and the mixture turned milky promptly. Then it was stirred for 30 min at 30 °C. The mixture was precipitated by centrifugation at 3500 rpm for 20 min at 4 °C, washed and resuspended in water to obtain BioMOF without coronal surface.

2.7. Synthesis of corona-BioMOF

Firstly, GOx (1.5 mg) and DOX@AFt (1.5 mg) were dispersed in an aqueous solution of 2-MIM (35 mmol, 5.8 mL) with stirring for 10 min at 30 °C. Next, an aqueous solution of zinc nitrate hexahydrate (0.5 mmol, 1.0 mL) was added and the mixture turned milky promptly. Then it was stirred for 30 min at 30 °C. The mixture was precipitated by centrifugation at 3500 rpm for 20 min at 4 °C, washed, and resuspended in water to obtain BioMOF, an intermediate product without coronal surface. Immediately, additional DOX@AFt (6.0 mg) was added into the above aqueous solution and stirred for another 30 min at 30 °C. The corona-BioMOF was collected by centrifuging at 3500 rpm for 20 min at 4 °C and freeze-dried. The prepared pink corona-BioMOF was stored at − 20 °C for future use.

2.8. Synthesis of BioMOF composites-1

Zinc nitrate hexahydrate (0.15 mmol, 0.3 mL) was added into the solution of BioMOF without coronal surface and stirred for 15 min. Immediately, additional DOX@AFt (6.0 mg) was added into the above aqueous solution and stirred for another 30 min at 30 °C. The BioMOF composites-1 was collected by centrifuging at 3500 rpm for 20 min at 4 °C.

2.9. Synthesis of BioMOF composites-2

2-MIM (1.5 mmol, 0.3 mL) was added into the solution of BioMOF without coronal surface and stirred for 15 min. Immediately, additional DOX@AFt (6.0 mg) was added into the above aqueous solution and stirred for another 30 min at 30 °C. The BioMOF composites-2 was collected by centrifuging at 3500 rpm for 20 min at 4 °C.

2.10. Simulated performance assessment of drug release kinetics

To assess the drug release kinetics of DOX from corona-BioMOF in different microenvironment, corona-BioMOF was dispersed in 0.9 % NaCl at pH 7.4, 6.5, and 5.0 to simulate the pH values of normal physiological environment, tumor microenvironment, and lysosome, respectively. After incubation for a given time at 37 °C, the mixture was centrifuged at 10000 rpm for 10 min. The fluorescence intensities of the supernatants were measured by fluorescence spectrophotometry at 595 nm with an excitation wavelength of 484 nm.

2.11. Determination of the loading capacity of DOX

The loading capacity of DOX in DOX@AFt was determined upon the centrifuged supernatant of reaction mixture. The fluorescence intensity at 595 nm of DOX was measured by fluorescence spectrophotometry. The loading capacity of DOX was calculated as (fluorescence intensity of total added DOX − fluorescence intensity of supernatant DOX) / fluorescence intensity of total added DOX × 100 %.

2.12. Cell culture

Breast MCF-10A cells and breast adenocarcinoma MX-1 cells were cultured in DMEM medium containing 10 % FBS and 1 % penicillin–streptomycin. Breast adenocarcinoma MDA-MB-231 cells were cultured in RPMI 1640 medium supplemented with 10 % FBS and 1 % penicillin–streptomycin. All cell lines were incubated at 37 °C containing 5 % CO₂ in humidified atmosphere. MCF-10A, MX-1, or MDA-MB-231 cells were digested with trypsin and resuspended in fresh cell culture medium before planting.

2.13. In vitro evaluation of targeting behavior of corona-BioMOF nanovehicle

Characterization of TfR1 in MDA-MB-231 was performed by immunofluorescence analysis. Cells were fixed with ice ethanol for 15 min, and then blocked with blocking buffer for 10 min. Primary antibodies were incubated for 1.5 h at RT. The following secondary antibodies were incubated for 1.5 h at RT in the dark. The nuclei were counterstained by incubating with DAPI for 10 min.

Targeting behavior of the corona-BioMOF nanovehicle was investigated as follows: MDA-MB-231 or MX-1 cells were seeded on a 35 mm Petri dish with a 10 mm well at the density of 5 × 10⁵ per well. The cells were incubated for 24 h in a water-jacket humidified incubator under 37 °C and 5 % CO₂ condition. Afterward, 1.0 mL of corona-BioMOF dispersed in culture medium (100 μg mL⁻¹) was added and incubated for 0.5, 2, 6, or 12 h. Subsequently, the cells were washed with PBS and stained with LysoSensor Blue probe for cell imaging. For flow cytometry analysis, MCF-10A, MDA-MB-231, and MX-1 cells were seeded on a 6-well plate at the density of 1 × 10⁶ per well. The cells were incubated for 24 h in a water-jacket humidified incubator under 37 °C and 5 % CO₂ condition. Afterward, 1.0 mL of corona-BioMOF dispersed in culture medium (100 μg mL⁻¹) was added and incubated for 2 h. Subsequently, the cells were washed with PBS and detached for flow cytometry analysis.

To confirm that corona-BioMOF binding to TfR1 of MDA-MB-231 cancer cells plays a role in killing cancer cells, an antibody blocking study was carried out by incubating the MDA-MB-231 cells with TfR1 antibody for 2 h followed by treatment with corona-BioMOF for 0.5, 2, and 6 h. Then CLSM imaging and flow cytometry analysis was done in the same way as described above.

2.14. In vitro cytotoxicity study

MCF-10A, MDA-MB-231, and MX-1 cells were planted on 96-well plates at the density of 1 × 10⁴ cells, and were allowed to adhere for 24 h. Next, the cells were incubated with culture medium containing ZIF-8, GOx@ZIF-8, DOX, DOX@AFt@ZIF-8, DOX@AFt, or corona-BioMOF for 12 h. To determine cytotoxicity, 20 μL of 5.0 mg mL⁻¹ MTT was added into each well and incubated sequentially for 4 h. After that, the supernatant was replaced by 150 μL of DMSO. Absorbance values of formazan were measured through microplate reader at 490 nm.

2.15. Antitumor experiment in vivo

All animal experiments were conducted in accordance with protocol No. SYXK (Su) 2020–0047 approved by Jiangsu Provincial Department of Science and Technology, and carried out in accordance with the institutional animal use and care regulations approved by the Animal Ethical and Welfare Committee of Nanjing Normal University (Approval Number: IACUC-20200601). Female BALB/c mice models were established by subcutaneously inoculating 9 × 10⁶ MDA-MB-231 cells in 100 μL Matrigel/PBS. When the tumor reached approximately 90 mm³, the MDA-MB-231 tumor-bearing mice were divided into eight groups (n = 5 mice for each group) at randomly and intravenously injected with PBS, ZIF-8, GOx@ZIF-8, DOX, DOX@AFt, DOX@AFt@ZIF-8, or corona-BioMOF (2.5 mg mL⁻¹, 200 μL) every 2 days with a total of five injections per mouse. The medium of injection was PBS. During the therapy, mice body weights and tumor volumes were recorded every other day. The tumor volumes were quantified by formula Length × Width² / 2. All tumor samples were weighted before fixation. H&E staining of major organs and tumors, immunofluorescence, and TUNEL staining of tumors were performed by the third-party company.

2.16. Statistical analysis

All data are expressed as means ± SD. Statistical differences were determined by two-tailed Student’s t test; **P < 0.01, ***P < 0.001, and ****P < 0.0001.

3. Results and discussion

3.1. Synthesis and characterization of the corona-BioMOF nanovehicle

The UV-vis adsorption peaks of DOX@AFt indicated that DOX had been embedded inside the AFt (Fig. S1). Besides, DOX@AFt and corona-BioMOF remained close to DOX in the fluorescence emission spectrum at 595 nm (Fig. S2, Fig. 1 A), implying the encapsulation of DOX in AFt and the successful construction of corona-BioMOF. As the assembly of DOX@AFt mainly relied on the pH-dependent disassembly and reassembly of AFt, the pH conditions of AFt reassembly for efficient loading of DOX into DOX@AFt were investigated (Fig. S3). Finally, pH 6.5, 0.10 mM of DOX, and 15 min of the assembly time were selected as the optimized conditions for preparation of DOX@AFt (Fig. S4 and S5). A maximum percentage of drugs (78 %) loaded into AFt proteins (Fig. 1B) was obtained, which might attributed to the electrostatic gradient and dynamics of pore residues [44].

Fig. 1.

Characterization, drug loading yields, and drug release ability of corona-BioMOF. (A) Fluorescence spectra of ZIF-8, DOX@AFt, and corona-BioMOF. (B) The loading yields of DOX in DOX@AFt at different pH values in synthesis process of DOX@AFt. The data are presented as the mean ± SD, n = 3. (C) XRD patterns of simulated ZIF-8, GOx@ZIF-8, DOX@AFt@ZIF-8, and corona-BioMOF. TEM images of (D) DOX@AFt and (E) corona-BioMOF. (F) High-magnification TEM image of corona-BioMOF. (G) HAADF-STEM image and (H) Color STEM image of corona-BioMOF. (I) Time-dependent dynamic releasing profiles of DOX in corona-BioMOF at different conditions. The error bars represented the standard deviation of three measurements. The data are presented as the mean ± SD, n = 3.

The crystal structures of the prepared ZIF-8-based nanomaterials were investigated by the powder X-ray diffraction (XRD) measurements (Fig. 1C), indicating that they retained almost the same crystal structure as the simulated ZIF-8 and the incorporation of GOx and DOX@AFt had little influence on the crystalline form of ZIF-8. Meanwhile, the XRD peaks shifted slightly to the low angle, reflecting the larger lattice constant, which is likely attributed to defects in composite [45]. Besides, the coronal morphology and element distribution of the corona-BioMOF were investigated by transmission electron microscopy (TEM), high angle angular dark field-scanning transmission electron microscopy scanning electron microscopy (HAADF-STEM), and ColorSTEM. As shown in Fig. 1D, DOX@AFt exhibited a spherical morphology with approximately 8 nm in diameter. The corona-BioMOF was uniformly coated with a large number of black spheres protruding from the surface, featuring a distinctive coronal structure (Fig. 1E). Magnified view in Fig. 1F showed more clearly that the coronal morphology and size of corona-BioMOF. HAADF-STEM image and corresponding ColorSTEM image (Fig. 1G and H) demonstrated that the outside small spheres were rich in N element and lacking in Zn element, confirming that the coronal periphery was indeed AFt proteins rather than small-sized MOFs. The above results demonstrated the satisfactory fabrication of nanomaterials with a specific coronal morphology. Furthermore, the Brunauer–Emmett–Teller (BET) surface area of corona-BioMOF (266.4 m²/g) was much smaller than ZIF-8 (1099 m²/g), DOX@AFt@ZIF-8 (911.1 m²/g), and GOx@ZIF-8 (423.8 m²/g), confirming the encapsulation of GOx and DOX@AFt in corona-BioMOF (Fig. S6). The loading capacity of total proteins in the corona-BioMOF was determined to be 57.6 μg mg⁻¹ from the BCA protein assay kit (Fig. S7), where the amounts of GOx and AFt were 12.1 and 45.5 μg mg⁻¹, respectively. This loading capacity was superior to previous most reports, which might attribute to the fully utilization of both the internal cavity and external surface of MOF [34], [46]. Besides, for assessing the pH controllability of corona-BioMOF towards DOX release, pH 7.4, 6.5, and 5.0 were chosen to simulate the pH values of normal physiological environment, tumor microenvironment, and lysosome, respectively. The time-dependent dynamic releasing profiles of DOX in corona-BioMOF (Fig. 1I) showed that little drug released upon exposure in pH 7.4 condition within 5 h. While at pH 5.0, a burst release of DOX within 1 h was found, implying a significant pH-responsive property of corona-BioMOF attributed to the combination of the degradation ability of the ZIF-8 structure in weak acid and the disassembly performance of the AFt protein at low pH. Ultimately a maximum DOX release amount (59.8 %) was achieved at pH 5.0 with glucose, which was attributed to the fact that GOx could catalase the oxidation of glucose into gluconic acid, which raised the acidity, further accelerating the degradation of ZIF-8, the disassembly of AFt and the drug release.

Different from the spiky surface of corona-BioMOF, the TEM images of ZIF-8 (Fig. 2 A), GOx@ZIF-8 (Fig. 2B), and BioMOF (Fig. 2C), the intermediate without spiky surface, showed smooth surfaces without black spheres. Moreover, we found that partial AFt proteins (indicated by arrows and dashed ovals) inside BioMOF began surfacing in pH 6.5 with glucose (Fig. 2D), which confirmed the inclusion of DOX@AFt and acid-induced disintegration of ZIF-8. Furthermore, the statistical comparisons of the amounts of spiky spheres on the surface of corona-BioMOF (Fig. 2E and F) and corona-BioMOF treated with weak acid medium containing glucose (Fig. 2G and H) were performed. A 45.1 % increase of exposed DOX@AFt proteins was found in the latter, implying the capability of tumor microenvironment-boosted more favorable targeting behavior of corona-BioMOF.

Fig. 2.

Performance assessment in simulated tumor microenvironment and possible formation mechanism of corona-BioMOF. TEM images of (A) ZIF-8, (B) GOx@ZIF-8 NPs, (C) BioMOF, (D) BioMOF in pH 6.5 containing glucose, and (E) corona-BioMOF. (F) Corresponding quantitative statistics of DOX@AFt on the surface of corona-BioMOF in Fig. 2E. (G) TEM image of corona-BioMOF in pH 6.5 containing glucose. (H) Corresponding quantitative statistics of DOX@AFt on the surface of materials in Fig. 2F. (I) DRIFTS of ZIF-8 synthesized in methanol, BioMOF, and corona-BioMOF. (J) Water-induced defect-based handle mechanistic illustration for the binding of DOX@AFt on the surface of BioMOF.

To elucidate the formation mechanism of the special topography of corona-BioMOF, additional metal ions or ligand molecules were introduced into the BioMOF without coronal surface. We noticed that the later added DOX@AFt was difficult to grow onto the above BioMOF composite surface with the additional introduction of Zn²⁺ or 2-MIM. The additional introduction of zinc ions resulted in visible voids on the surface of the material (Fig. S8), implying that large-scale defects are not conducive to the binding between composites and proteins. Also, extra ligand molecules are not benificial to the formation of coronal structures of corona-BioMOF (Fig. S9). Subsequently, diffuse reflectance infrared Fourier-transform spectra (DRIFTS) were conducted to achieve a deeper insight into the interaction between protein and BioMOF [47]. As shown in Fig. 2I, the adsorption bands at 600–1500 cm⁻¹ were assigned to the bending vibration of the imidazole ring, while the bands at 1620–1730 cm⁻¹ were amide I stretching characteristic of proteins and the band at 409–440 cm⁻¹ were related to the stretch of Zn–N and Zn–O [37], [48], [49], [50]. The DRIFTS of the corona-BioMOF exhibited a red shift of the amide-I stretch (1644 cm⁻¹) compared to BioMOF (1666 cm⁻¹) and ZIF-8 (1703 cm⁻¹), indicating stronger chemical interaction between proteins and MOFs, which was consistent with previous report [37]. The band at 422 cm⁻¹ was assigned to the stretching vibration of Zn–N and the band at 410 cm⁻¹ was assigned to the bond of Zn–O. As shown in partial enlarged view of Fig. 2I, from ZIF to BioMOF, the peak intensity of Zn–N bond gradually weakened at 428 cm⁻¹ and the Zn–O bond increased at 410 cm⁻¹, indicated that the Zn–N bonds were broken to form Zn–O bonds owing to the water-induced defect. When the protein is attached on the surface of the BioMOF, the peak intensity of Zn–O (410 cm⁻¹) weakened and the peak at 422 cm⁻¹ increased as the N of the protein bound to the Zn site of the BioMOF to form a new Zn–N bond. The possible formation mechanism for the binding of DOX@AFt on the surface of BioMOF is illustrated in Fig. 2J. The involvement of water cleaved the Zn–N bond of ZIF-8 and induced a point defect to expose the Zn active site [51], [52], [53]. This facilitates coordination between N-terminal of the later added protein and Zn active site, driving the self-assembly of the protein on the BioMOF surface. These results show that the protein binding occurred via water-induced linker defect mechanism on the surface of BioMOF, where the binding could directly cooperate with the active site of the metal ion.

3.2. Targeting investigation of the corona-BioMOF

To confirm the TfR1 targeting in cancer cells, the characterization result of TfR1 in TfR1-positive MDA-MB-231 cancer cells by immunofluorescence staining was shown in Fig. S10. By antibody blocking assay towards TfR1, the cellular interaction mediated by AFt and TfR1 was further proven (Fig. S11 and S12). The CLSM imaging and flow cytometry analysis all showed that the increase of fluorescence intensity is not obvious after incubated with corona-BioMOF, indicating that the interaction was mediated by TfR1. Then the precise targeting ability of corona-BioMOF was investigated (Fig. 3 A). The TfR1-negative MX-1 cancer cells and normal breast MCF-10A cells were utilized as the control [54]. After incubation with corona-BioMOF, MDA-MB-231 cells presented an 8-fold higher fluorescence intensity over MX-1 and MCF-10A cells. Meanwhile, after co-culturing with corona-BioMOF, the fluorescence intensity of the MDA-MB-231 cells was 5-fold higher than that of (DOX@AFt + GOx)@ZIF-8 without spiky surface. The results all indicated the higher cellular uptake ability of corona-BioMOF compared to (DOX@AFt + GOx)@ZIF-8, revealing the superiority of the DOX@AFt-based spiky surface on corona-BioMOF for improving the targeting ability. Also, the dynamic cellular uptake behaviors of corona-BioMOF were investigated. As shown in Fig. 3B, the fluorescence signals of corona-BioMOF and LysoSensor blue (a blue fluorescent marker of lysosome) in MDA-MB-231 cells overlapped gradually from 0.5 h to 6 h, confirming that corona-BioMOF was uptaken by lysosome. Also, it can be seen that some red fluorescence signals appeared in the nucleus at 6 h, implying that chemotherapeutic drug, DOX, achieved lysosomal escape and entered the nucleus due to the imidazole protonation resulted lysosomal rupture [46]. After 6 h, the intensity of blue fluorescence decreased obviously, verifying that some lysosomes have been degraded due to the drug-induced cell apoptosis (Fig. S13). Correspondingly, after 2 h of co-culturing, Confocal laser scanning microscope (CLSM) images showed that MDA-MB-231 cells emitted bright red fluorescence (DOX fluorescence) while MX-1 cells had poor fluorescence, indicating that the corona-BioMOF was rapidly internalized into the TfR1-positive targeted cells within 2 h. Furthermore, the content of Zn element in MDA-MB-231 cells after incubation with corona-BioMOF was 2-fold more than (DOX@AFt + GOx)@ZIF-8 (Fig. 3C). The above results showed that the spiky DOX@AFt coating promoted the cellular uptake, and corona-BioMOF possessed distinct precise targeting and intracellular lysosome-specific drug delivery abilities.

Fig. 3.

Evaluation on targeting behavior and killing efficacy of the corona-BioMOF in vitro. (A) Flow cytometry analysis of corona-BioMOF and (DOX@AFt + GOx)@ZIF-8 after incubation in TfR1-positive MDA-MB-231 cancer cells, TfR1-negative MX-1 cancer cells, and normal breast MCF-10A cells. (B) CLSM imaging and quantitative fluorescence analysis incubated with 100 μg mL⁻¹ corona-BioMOF in MDA-MB-231 cells for 0.5, 2, 6, 12 h, and in MX-1 cells for 2 h. The red fluorescence was associated with DOX in corona-BioMOF, and the blue fluorescence was expressed by lysosome localized LysoSensor Blue probe. The locations in yellow shadow represent the nucleus. Scale bar represents 40 μm. (C) Quantitative analysis of Zn element by ICP-AES in MDA-MB-231 cells after incubated with (a) (DOX@AFt + GOx)@ZIF-8 and (b) corona-BioMOF. The data are presented as the mean ± SD, n = 3. **P < 0.01, analyzed by Student’s t test.

3.3. Therapeutic efficacy in vitro

Through standard methyl thiazolyl tetrazolium (MTT) assay, corona-BioMOF showed strong cytotoxicity towards MDA-MB-231 (IC50 = 2.49 μg mL⁻¹) and good security for the normal breast cell line MCF-10A (IC50 = 64.1 μg mL⁻¹), which was ascribed to the difference of cellular microenvironment and the receptor interactions (Fig. 4 A and Fig. S14). In contrast to this, the BioMOF displayed weaker cytotoxicity towards MDA-MB-231 (IC50 = 5.28 μg mL⁻¹) and better security for MCF-10A (IC50 = 98.3 μg mL⁻¹) (Fig. S15), which further proved the therapeutic superiority of the corona structure. Fig. 4B showed that corona-BioMOF exhibited excellent cell killing efficiency towards MDA-MB-231 cells. Moreover, to inspect the possible apoptosis mechanism induced by corona-BioMOF, FITC-Annexin V/PI flow cytometry analysis was conducted. The cell apoptotic rate of the corona-BioMOF treated group (31.5 %) was much higher than that of groups treated by ZIF-8 (10.8 %), GOx@ZIF-8 (16.8 %), DOX (18.3 %), DOX@AFt@ZIF-8 (18.6 %), and DOX@AFt (26.0 %), suggesting a most efficient therapy performance with corona-BioMOF (Fig. 4C). Meanwhile, no severe necrosis was exhibited for the MCF-10A cells treated with corona-BioMOF, which meant its minimized inflammation and damage to normal cells (Fig. S16). Therefore, these results demonstrated the therapeutic superiority of the corona-BioMOF.

Fig. 4.

(A) Cell viability on MCF-10A and MDA-MB-231 cells in the presence of corona-BioMOF for 12 h (n = 3). (B) Cell viability after various treatment in MCF-10A and MDA-MB-231 cells (n = 3). ***P < 0.001, analyzed by Student’s t test. (C) Flow-cytogram representing apoptosis assay based on Annexin V-FITC and PI staining of MDA-MB-231 cells after various treatment for 8 h.

3.4. Antitumor activity of corona-BioMOF in vivo

We further evaluated its antitumor activity in subcutaneous MDA-MB-231 tumor-bearing nude mice. According to the treatment schedule (Fig. 5 A), after the tumor size reached approximately 90 mm³, MDA-MB-231 tumor-bearing nude mice were intravenously injected with PBS, ZIF-8, GOx@ZIF-8, DOX, DOX@AFt, DOX@AFt@ZIF-8, or corona-BioMOF. Immunofluorescence staining assay of tissue section was performed and verified the TfR1 expression on the tumor cells at first (Fig. 5B). Before injection, the stability of corona-BioMOF nanovehicle in serum was characterized by TEM (Fig. S17 and S18). The corona-BioMOF remained its original coronal morphology in serum after 1 h incubation and the MOF backbone showed little change, while BioMOF has undergone an obvious degradation. It might be attributed to the protection from the cage structure of spiky proteins on the external surface. After injected with corona-BioMOF, the fluorescence signal of DOX from corona-BioMOF accumulated at the tumor sites and peaked at 8 h, and a strong signal was found in the bladder at 24 h (Fig. 5C), indicating the gradual accumulation pathway of corona-BioMOF in tumor site and its final excretion through kidney metabolism. The biodistribution profiles of drugs were also investigated through ex vivo fluorescence images (Fig. 5D). With prolonged metabolism in vivo, similar to most materials, corona-BioMOF might aggregate in the liver and kidneys in the short time. However, the fluorescence intensities in the liver and kidneys decreased obviously at 24 h, demonstrating the good biodegradability and biosafety of the corona-BioMOF nanovehicle, which might be related to the reassuring composition including biological proteins, Zn²⁺, and 2-MIM. In stark contrast, the fluorescence intensity was strong in the tumor but much weaker in other major organs, confirming the good targeting effect of the corona-BioMOF in tumor site. The most significant inhibition of tumor growth was achieved with the corona-BioMOF treated MDA-MB-231 tumor-bearing mice (Fig. S19). The therapy performance of corona-BioMOF was further evaluated by tumor sizes, tumor weights, relative tumor volumes, and tumor growth inhibition (TGI) rate (Fig. 5E–G and Fig. S20). Significantly, the TGI rate in the corona-BioMOF group was increased to 86.5 % at 22 days, displaying the most favorable treatment results. It is worth to be highlighted that the therapeutic effect of corona-BioMOF was significantly better than DOX@AFt and slightly higher than DOX@AFt@ZIF-8, which is in contrast to the minor difference in the cell cytotoxicity assay between them, further reflecting the unique size-induced enhanced permeability and retention (EPR) effect and practical application potential of corona-BioMOF. Besides, the body weights of mice also showed negligible changes affected by various treatments compared to PBS group, which further verified their prominent therapeutic outcomes (Fig. 5H).

Fig. 5.

In vivo antitumor activity of the corona-BioMOF. (A) Schematic illustration of the therapeutic schedule of the corona-BioMOF. (B) Immunofluorescence image of tumor section from MDA-MB-231 tumor-bearing mice. Scale bar represents 20 μm. (C) In vivo fluorescence imaging of MDA-MB-231 tumor-bearing mice after intravenous injection of corona-BioMOF. (D) Distribution of corona-BioMOF in tumor (Tu) and major viscera (He, heart; Li, liver; Sp, spleen; Lu, lung; and Ki, kidney) of MDA-MB-231 tumor-bearing mice at (a) 8 h and (b) 24 h post intravenous injection with corona-BioMOF. (E) Photographs of tumor dissection. (F) Analysis of tumor weights after 22 days of different treatments at a dosing frequency via intravenous injection with (a) PBS, (b) ZIF-8, (c) GOx@ZIF-8, (d) DOX, (e) DOX@AFt, (f) DOX@AFt@ZIF-8, and (g) corona-BioMOF. (G) Analysis of relative tumor volumes. (H) Body weights of mice. (I) Histological H&E analysis and TUNEL staining of tumors collected from MDA-MB-231 tumor-bearing mice after 22 days with various treatments. Scale bars represent 100 μm. Data above are presented as the mean ± SD, n = 5. **P < 0.01, ***P < 0.001, and ****P < 0.0001, analyzed by Student’s t test.

Moreover, we subsequently investigated the histological damage and apoptosis levels of tumors by using anatomical and histological observation, hematoxylin and eosin (H&E) staining, and terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay, respectively (Fig. 5I, Fig. S21 and S22, Supporting Information). The results demonstrated no obvious lesions or abnormalities occurred in major organs compared with PBS group during treatment periods, confirming high biocompatibility of corona-BioMOF. The results showed that severe apoptosis happened in tumor cells after the treatments with corona-BioMOF, whereas only moderate damage was examined for those tumors treated with GOx@ZIF-8, DOX, DOX@AFt, or DOX@AFt@ZIF-8, further demonstrating the effectiveness of the proposed corona-BioMOF nanovehicle. Finally, quantitative results obtained from inductively coupled plasma-atomic emission spectrometry (ICP-AES) showed that Zn²⁺, the degradation product of the corona-BioMOF, could be accumulated in liver through the blood circulation (Fig. S23 and Table S1), which was consistent with previous reports [55]. The above results indicated that corona-BioMOF exhibited decent tumor-targeting ability, effective therapeutic effect, and biosecurity, fully indicating its application performance in cancer therapy.

4. Conclusion

In summary, through mimicking the coronal structure and size of CoV, we designed a corona-BioMOF nanovehicle fabricated with a recognition proteins-based spiky surface and a hierarchical wrapping pattern for precise targeting and controlled drug release in cancer therapy. The formation mechanism of the special topography was also studied. The corona-BioMOF displayed remarkable specificity and lethality towards receptor-positive cancer cells in vitro and in vivo. These findings remind us of an interesting news, which reported that SARS-CoV-2 can induce remission of Hodgkin lymphoma [56]. We believe that although viruses are vicious, we can learn from their slyness and take advantages of it for human health. Therefore, the unique design of corona-BioMOF can provide an inspirational biomimetic strategy to offer an efficient tool for precise cargo delivery in diagnosis and treatment, and open the door for application potentials in precision medicine, especially in vivo tracing, individualized interaction diagnostics, and tailored target therapies.

CRediT authorship contribution statement

Huafeng Wang: Conceptualization, Methodology, Validation, Formal analysis, Investigation, Data curation, Writing – original draft, Visualization. Shi Li: Methodology, Investigation, Data curation. Lei Wang: Investigation, Data curation. Zimei Liao: Methodology, Investigation. Hang Zhang: Investigation, Supervision. Tianxiang Wei: Conceptualization, Methodology, Formal analysis, Writing – original draft, Writing – review & editing, Data curation, Supervision, Funding acquisition. Zhihui Dai: Writing – review & editing, Supervision, Project administration, Funding acquisition.

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:

Data availability

Data will be made available on request.