Oxidation-induced surface deposition of tannic acid: towards molecular gates on porous nanocarriers for acid-responsive drug delivery

Abstract

Simple construction of porous nanocarriers with surface coatings for gated release is of great significance in the current drug delivery research. In the present study, tannic acid (TA) derived metal–phenolic network (MPN) and boronate–phenolic network (BPN) coatings on the surfaces of 70 nm-sized mesoporous silica nanoparticles (MSN) were developed by separating the TA deposition and the crosslinking into two steps. Surprisingly, the highly efficient deposition of TA on MSN by surface interactions was observed by UV-vis spectra, thermogravimetric analysis, and nitrogen sorption. The interactions stem from the oxidation of TA by superoxide anions from MSN suspension, as evidenced by fluorescence spectra and HPLC. The pore openings (∼3.9 nm) were partially occluded by TA deposition (at a weight ratio of 6 wt% TA), which circumvented the requirement of multi-layer coating for sufficient pore sealing. Moreover, it played significant roles in constructing the MPN or BPN structures after TA deposition, as well as regulating the release of a hydrophilic model drug (DOX) from the mesopores. The MPN-containing MSN particles (MSN@TA-Zn) possessed higher pH sensitivity as compared with its counterpart incorporating BPN (MSN@TA-Zn-BDBA), which gave rise to more pronounced pH-enhanced drug release kinetics. The particles were shown to deliver DOX to cancer cells for inducing efficient cytotoxicity. Therefore, these results would inspire simple controlled release systems through responsive pore gating.

---

1. Introduction

Among nanoscopic therapeutic systems, mesoporous silica nanoparticles (MSNs) have emerged as robust nanocarriers for drug delivery in the past decades.^1–4 The characteristic mesoscopically ordered pore structure accompanied by the consequential high surface areas and pore volumes can not only provide the materials with a high loading capacity but also lead to efficient retention of the cargo molecules, especially hydrophobic drugs.^5,6 For hydrophilic drugs loaded by electrostatic interactions or physical adsorption, metabolites/ions in the body fluid can displace the drugs, resulting in premature drug release. Therefore, a stimuli-responsive drug delivery system (DDS) is of high interest in clinical medicine to enhance therapeutic efficiency by controlling drug release and minimizing the side effects.

A lot of stimuli-responsive drug delivery systems based on MSN have been proposed up to now.⁷ Typical strategies to offer controlled release and/or stimuli responsive release, involve capping of MSN pores by an outer-surface organic/inorganic layer tethered by organic linkers via covalent or noncovalent pathways.^8,9 However, the synthesis of these DDS usually requires multi-step strategies along with multiple phase transitions, especially for the covalent pathways.^7,10,11 Consequently, it is highly interesting to develop composite MSN nanocarriers integrating capping/sealing materials with high affinity towards MSN surfaces, as well as sensitive structure change under environmental stimuli, to make pore sealing and controlled release easier.

Novel synthetic biomaterials through mimicking nature can be integrated on surfaces to serve as templates and building blocks for new generations of biocompatible coatings. Well-known for their antioxidant properties, polyphenols are naturally occurring compounds containing one or more phenol groups.^12,13 The excellent bio-adhesive property of polyphenols has been shown in the coating technology on surfaces of varying materials,^14–16 as well as the preparation of composite particles^17–19 facilitated by their catechol or pyrogallol groups. Taking advantage of the ability in forming vulnerable films in a short preparation period, metal–phenolic network (MPN)^20–22 and boronate–phenolic network (BPN)²³ on the basis of a representative polyphenol, i.e. tannic acid (TA), have attracted strong interests recently. To date, however, most of the research efforts have been focused predominantly on exploiting the MPN or BPN structures in forms of microscale capsules for the delivery of different drugs,^21,23–25 while very few on the integration of the structures on surfaces of small nanocarriers for DDS. Considering the possibility of discontinuous films resulting from the chain rigidity on nanoparticles with high surface curvature, a multi-layer coating strategy was generally employed to realize thick TA coatings.^14,15,22,26 Whereas no attention was paid to the interaction between tannic acid and the underlying supporting particles for a simple construction of DDS.

In the current study, the interactions between MSN surfaces and tannic acid were first investigated before the construction of MPN or BPN on MSN for controlled drug release. The efficient surface deposition of tannic acid into the mesopores was explored by thermogravimetric analysis (TGA), HPLC, and nitrogen sorption measurements. Subsequently, MPN and BPN as responsive drug release switches were constructed in situ on drug loaded MSN by separating the TA deposition and the crosslinking between tannic acid and Zn²⁺, benzene-1,4-diboronic acid (BDBA) into two steps under aqueous conditions. As a proof-of-concept application, doxorubicin hydrochloride (DOX) as a model anti-cancer drug was utilized for drug release experiments. Responsive drug release was observed on both systems with different sensitivity towards pH stimuli. The potential of the constructed nanocarrier system in drug delivery was demonstrated on Hep-G2 cancer cells by in vitro experiments.

2. Experimental

2.1 Materials

Unless otherwise noted, all reagent-grade chemicals were used as received, and distilled water was used in the preparation of all aqueous solutions. Cetylmethylammonium bromide (CTAB, AR), ethylene glycol (AR) were purchased from Fluka. Benzene-1,4-diboronic acid (BDBA) was purchased from ACROS. Fluorescein isothiocyanate (FITC, AR) was purchased from Alfa Aesar. Tannic acid (TA, M_w = 1701.23 Da), epigallocatechin gallate (EGCG), tetraethyl orthosilicate (TEOS, AR), 3-aminopropyltriethoxysilane (APTES, AR), NH₄OH (30 wt%, AR), and dihydroethidium (DHE, 99%) were purchased from Sigma. Dopamine hydrochloride (98%), zinc nitrate hexahydrate (Zn(NO₃)₂·6H₂O, AR), doxorubicin hydrochloride (DOX, 98%), and 2-[4-(2-hydroxyethyl)-1-piperazinyl]ethanesulfonic acid (HEPES) were purchased from Aladdin Industrial Inc. Counting Kit-8 (CCK-8) was purchased from Dojindo Chemical (Shanghai). 4′,6-Diamidino-2-phenylindole (DAPI) was purchased from Leagene Biotechnology (Beijing).

2.2 Synthesis of mesoporous silica nanoparticles (MSN)

The starting particles were prepared by a typical template synthesis procedure using CTAB and TEOS as surfactant and silica source, according to a recipe reported by us.²⁷ The molar ratio used in the synthesis was 1 TEOS : 0.18 CTAB : 5.9 NH₃ : 88.5 ethylene glycol : 1249 H₂O. The reaction was allowed to proceed for 3 h at 70 °C. Then, the heating was stopped and the as-synthesized colloidal suspension was aged at 70 °C without stirring overnight. After cooled to room temperature, the suspension was separated by centrifugation. The template removal was performed by a highly efficient ion-exchange method.²⁸ The final product was suspended in acetone for further use. The FITC labelled MSN particles were prepared by the addition of a APTES–FITC conjugate together with TEOS according to a previous work.²⁹

2.3 Investigation of the interaction between MSN and TA

Typically, 2 mg MSN particles was re-dispersed in 185 μL of distilled water, followed by the addition of certain amount of tannic acid solution (40 mg mL⁻¹ in water). Then, the mixture solution with varying concentration of TA (0–1 mg mL⁻¹) was stirred for 2 h at room temperature. After that, the particles were retrieved by centrifugation for characterizations like TGA, nitrogen sorption, etc. While the supernatant was taken for studies of UV-vis absorption spectra and HPLC.

2.4 Preparation of MPN incorporated MSN composite particles (MSN@TA-Zn)

In our study, MPN and BPN as responsive drug release switches was constructed on drug loaded MSN by separating the TA deposition and the crosslinking process into two steps. In a typical procedure, 2 mg blank or drug loaded MSN particles was re-dispersed in 185 μL of distilled water, followed by the addition of 5 μL of tannic acid solution (40 mg mL⁻¹ in water, 10 wt% with respect to MSN). Then, the mixture solution was stirred for 2 h at room temperature. Thereafter, 10 μL of Zn(NO₃)₂·6H₂O solution (10.8 mg mL⁻¹ in water, 36.4 mM) was added to the suspension. The pH of the suspension was raised by adding 800 μL of HEPES buffer (pH 7.4, 25 mM) to initiate the coordination reaction between the metal ions and tannic acid.²⁰ The suspension was then stirred at room temperature for another 2 h. Finally, the particles were recovered by centrifugation and denoted as MSN@TA-Zn.

2.5 Preparation of BPN incorporated MSN composite particles (MSN@TA-BDBA)

Briefly, the first step of TA deposition was the same as the one described in the preparation of MSN@TA-Zn. Thereafter, 25 μL of BDBA solution (4 mg mL⁻¹ in DMSO) was added to the suspension. The pH of the suspension was raised by adding 800 μL of TRIS buffer (pH 8.4, 25 mM) to initiate the coordination reaction between TA and BDBA.²³ The suspension was then stirred at room temperature for another 2 h. Finally, the particles were recovered by centrifugation and denoted as MSN@TA-BDBA.

2.6 Drug loading and release experiments

For drug loading, 2 mg of MSN was ultrasonically dispersed in 1 mL of a drug solution with a concentration of 0.2 mg mL⁻¹ in HEPES buffer (25 mM, pH 7.4). The mixture was stirred at room temperature for 2 h, and then centrifuged to collect the drug-loaded nanoparticles. The amount of drug loaded into MSN was calculated by subtracting the mass of drug in the supernatant from the total mass of drug in the initial solution.

The release study was conducted as follows. First, 2 mg of drug-loaded nanoparticles was dispersed in 2 mL of PBS buffer (pH 7.4), or in sodium acetate buffer solutions (20 mM, pH 4.4, 5.5) with the same ionic strength as the PBS solution (150 mM). At the predetermined time intervals, 0.2 mL of solution was withdrawn from the solution and the amount of released drug was analyzed by UV-vis. For keeping a constant volume, 0.2 mL of fresh medium was added after each sampling. All drugs release results were averaged with three measurements.

Calculation of the corrected concentration of released DOX is based on the following equation:

where C_c is the corrected concentration at time t, C_t is the apparent concentration at time t, v is the volume of sample taken (0.2 mL), and V is the total volume of the release fluid (2 mL).

2.7 Cell uptake (confocal microscopy)

The human hepatoma (Hep-G2) cell line purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China) was cultured at 37 °C under a humidified atmosphere containing 5% CO₂ in RPMI 1640 medium (Hyclone) supplemented with 10% fetal bovine serum (FBS, Gibco), 100 U mL⁻¹ penicillin (Hyclone), and 100 U mL⁻¹ streptomycin (Hyclone). The culture medium was routinely changed every two days and the cells were separated by trypsinization before reaching confluency. Hep-G2 cells were seeded in a confocal microscopy dish at a density of 1 × 10⁵ cells per dish in 1 mL of RPMI 1640 and cultured for 24 h at 37 °C prior to cell uptake assay. Next, the original medium was replaced with 1 mL of FBS-free medium in the presence of FITC labeled MSN, MSN@TA-Zn, or MSN@TA-BDBA particles (100 μg mL⁻¹). Subsequently, the cells were washed by PBS for three times to remove the residual nanoparticles. The cells were treated with 500 μL of 4% paraformaldehyde PBS solution for 30 min and their nuclei were stained with DAPI for 5 min. Finally, a confocal laser scanning microscope (CLSM, Leica SP5, Germany) was utilized to observe the cells.

2.8 In vitro cytotoxicity assay

The in vitro cytotoxicity of the particles and drug-loaded particles was tested by examining the viability of HeLa cells after the treatment of particles using the CCK-8 assay. The Hep-G2 cells were seeded in 96-well plates at a density of 8.0 × 10³ cells per well and then incubated at 37 °C for 24 h before the cytotoxicity assay. Afterwards, the culture medium was replaced with fresh cell culture medium containing different concentrations of drug-free particles, DOX loaded particles, or free DOX. After incubation for 24 h, the medium was removed and the wells were washed with PBS. Subsequently, 20 μL of CCK-8 together with 200 μL of fresh medium was added, and the cells were further incubated for 4 h. Finally, the absorbance of each well at 450 nm was measured using a microplate reader (Multiskan Spectrum, Thermo Fisher, USA).

3. Results and discussion

3.1 Surface interaction between TA and MSN

The starting MSN particles were prepared by a typical procedure of base catalyzed template synthesis using CTAB and TEOS as the template surfactant and the silica source, respectively.²⁷ The molar ratio used in the synthesis was 1 TEOS : 0.18 CTAB : 5.9 NH₃ : 88.5 ethylene glycol : 1249 H₂O. Typical scanning electron microscope (SEM) image and the small angle X-ray diffraction (SAXRD) pattern of MSNs are shown in Fig. 1. The sample consists of uniform spherical particles with an average size of ∼70 nm. Besides, the SAXRD pattern shows a single broad peak at 2.05° which should be indexed to the (1 0 0) diffraction plane of a hexagonal unit cell.

Fig. 1 SEM image and the corresponding small angle X-ray diffraction pattern of the MSN particles.

The interaction between tannic acid and MSN particles was firstly investigated. TA solutions with varying concentrations (0.02–1 mg mL⁻¹) were incubated with MSN for evaluating the loading capacity of TA on the surfaces of MSN. UV-vis absorption spectrophotometry was used to follow the changes in the concentration of TA, and absorption spectra of tannic acid solutions before and after incubation with MSN were measured.

Typical absorption spectra of TA solutions (0.2 mg mL⁻¹) before and after incubation with MSN are shown in Fig. 2a. The UV-vis absorption spectrum of TA in solution exhibited two peaks, centered at ∼225 nm and ∼277 nm. These peaks can be assigned to the neutral (i.e., protonated) form of TA which has a pK_a of around 8.5.²² Surprisingly, the absorbance intensities at 277 nm increased significantly after incubation of TA with MSN for 120 min, where a maximum increment up to 300% of the original TA absorbance was obtained at the lowest TA concentration of 0.02 mg mL⁻¹ (Fig. 2b). These results may first imply that there was not any TA loading after the adsorption in the aqueous solutions, which is not in agreement with the previous findings of remarkable TA adsorption by porous silica materials.^30–32 For example, two previous studies using amino-functionalized magnetic mesoporous silica³¹ and surfactant modified zeolites³⁰ as adsorbents showed adsorption capacities of 100–500 mg g⁻¹ by a similar quantification method via UV-vis absorbance. Moreover, the second peak was blue-shifted to 270 nm for the TA solutions after the incubation with MSN. It was reported that the blue shift to a stronger absorption band is associated with the oxidation of TA to various conjugated quinone derivatives.¹⁵ Consequently, the possibility of TA loading by adsorption can not be easily ruled out, because the degree of absorption strengthening is unknown by oxidation in the case of MSN incubation. Further characterizations were thus needed to investigate the interaction between MSN and TA.

Fig. 2 (a) UV-vis absorption spectra of tannic acid solution (0.2 mg mL⁻¹) in water before and after incubation with MSN for 120 min. (b) Comparison of the absorbance for TA solutions with varying concentrations before and after the addition of MSN. (c) TGA curves for MSN, MSN@TA, MSN@TA-Zn, and MSN@TA-BDBA.

Thermogravimetric analysis (TGA) was carried out to understand the weight loss characteristics of the components (TA) present in the MSN@TA sample. The TGA profile along synthesized by using 0.2 mg mL⁻¹ TA is shown in Fig. 2c. From the profile, it can be seen that the initial weight loss in the temperature range up to 130 °C is due to the evolution of adsorbed and pyrolysis water. The major weight loss region begins at 180 °C and witness a total weight loss of about 6 wt% between 180 °C and 250 °C, which should be attributed to the decomposition of TA.³³ Therefore, the loading degree of TA into MSN is around 6 wt%.

Having confirmed that TA was loaded into MSN, the oxidative fluorescent dye dihydroethidium (DHE) was used to evaluate in situ production of superoxide anion (O₂⁻) by MSN, and in turn, to explain the abnormal changes in the UV-vis spectra of TA. DHE has been used as a fluorescent probe for detecting O₂⁻ due to its relative specificity for this ROS. The fluorescence emission spectra of DHE before and after the introduction of the supernatant of MSN or MSN@TA suspensions are shown in Fig. 3a. The typical emission peak of DHE at 515 nm, was remarkably shifted to a red-region band at 650 nm and 600 nm for the samples from suspending MSN in water for 10 min and 120 min, respectively. The red fluorescence (also referred to as the “ethidium fluorescence”) that arises from the two-electron oxidation product of DHE (Fig. S1†) was attributed to superoxide anions in solutions.³⁴ No further change in the spectra of the red fluorescence can be observed for the samples from suspending MSN in water for longer times (data not shown), indicative of an equilibrium level of O₂⁻ generation after 120 min. The oxygen radicals released from silica were reported to result in the oxidation of polyphenols which is accelerated more significantly by higher surface areas of mesoporous silica materials.³⁵ The oxidation products are in part firmly adsorbed on silica due to the reduced solubility, as confirmed by the shifting in the retention time of the elution peak from 0.47 min to 4.7 min in HPLC analysis (Fig. 3b, measurement details are described in the ESI†), which is in agreement with the findings on the study of the polyphenol coatings developed recently.³⁶ Hence, the peak shift to the stronger absorption band in the UV-vis spectra might result from the free TA molecules which were not adsorbed onto the silica surface but oxidized by MSN.

Fig. 3 (a) Fluorescence emission spectra of dihydroethidium (DHE, 25 μM) before and after the introduction of supernatant solutions from suspending MSN in water for different time periods (10 min, 120 min). (b) Results of HPLC studies used to demonstrate the decrease in the solubility of TA after incubating with MSN.

Nitrogen sorption was then used to find out the change in the porous structure and the surfaces of MSN after TA loading (6 wt% by TGA at a feeding ratio of 10 wt%). The results for MSN after DOX loading (at a loading degree of 90 μg mg⁻¹)³⁷ were used for comparison, in order to elucidate the location of the deposited TA molecules. The nitrogen sorption isotherms and the corresponding pore size distributions of MSN, MSN@TA, and DOX loaded MSN are displayed in Fig. 4. All samples exhibit type IV isotherms with type H1 hysteresis loops (Fig. 4a), characteristic of a mesoporous material with 1D cylindrical channels. For MSN after DOX loading, the change is the reduction in the pore volume (from 0.86 cm³ g⁻¹ to 0.72 cm³ g⁻¹), indicating the loading of the drugs with small molecular dimensions. Interestingly, the isotherm changes more considerably after the loading of TA on MSN. First, the adsorption capacity decreased significantly, resulting in a remarkable reduction of the pore volume (from 0.86 cm³ g⁻¹ to 0.47 cm³ g⁻¹, as shown in Table S1†), as well as a 23% reduction in the specific surface area (from 963 m² g⁻¹ to 742 m² g⁻¹). Second, the capillary condensation (i.e., mesopore filling) step at relative pressures of 0.25–0.45 diminished dramatically, which was also reflected in the reduction of peak intensities, as well as the shifting of peaks to lower pore sizes (2.3–3.0 nm) in the corresponding pore size distribution (Fig. 4b). In the DFT pore-size distribution, the mesopores of 3.9 nm are no longer detectable after TA uptake.

Fig. 4 Comparison of nitrogen sorption isotherms (a and c) and the corresponding pore size distributions (b and d), derived from the desorption branch of the isotherms for MSN particles before and after tannic acid or DOX loading at a loading degree of 6 wt% and 9 wt%, respectively.

It was not possible to establish if TA fills completely the mesopores or if it simply blocks them at the entrance. In the following, an estimation of the mesoporous volume in MSN is correlated to the volume that the deposition amount of TA molecules can occupy and then this amount is compared with the adsorbed amount of TA after its uptake.³⁸ Considering the density of TA (2.12 g cm⁻³), 1 g of MSN sample, having a free mesoporous volume of 0.86 cm³ g⁻¹, can incorporate a maximum TA amount of 1.82 g. Since the amount of TA uptaken was 60 mg g⁻¹ (only 3% of the maximum amount of the adsorbable TA), it is proposed that the large size of the molecule prevents efficient packing when diffusing into the narrow mesopores of MSN. Considering the reported molecular dimensions of 1.85 nm × 1.65 nm × 1.01 nm for TA,³² it is highly possible that TA molecules may block the 3.9 nm sized pore space when settling on the external particle surface by adsorption, leaving significantly narrowed/occluded pore openings which would be beneficial to the easy construction of MPN or BPN in the next step.

3.2 Construction of MPN or BPN on MSN surfaces

In light of the occlusion of the pores by deposition of TA on the external surfaces of MSN, the construction of metal–phenolic network (MPN) and boronate–phenolic network (BPN) was carried out by the well-established procedures on the basis of cross-linking coordination or covalent bonds between TA and Zn²⁺ or benzene-1,4-diboronic acid (BDBA).^23,24 In the TGA curves of MSN@TA-Zn and MSN@TA-BDBA shown in Fig. 2c, the weight loss step for the organic matters shifted to higher (250–450 °C) and lower (110–260 °C) temperature ranges as compared with that for MSN@TA, implying the role of cross-linking in the thermo-stability of TA molecules on MSN surfaces. A total weight loss of 10 wt% was found for MSN@TA-Zn, revealing that there were more TA molecules in solution incorporated onto the particles during the cross-linking step. In comparison, MSN@TA-BDBA possesses a total weight loss of 18 wt%, indicative of BDBA incorporation in the BPN structure.

Typical TEM images for MSN before and after the construction of MPN or BPN are displayed in Fig. 5a–c. As shown, the MSN particles are composed of uniform particles with radially aligned pore channels. There is a reduction in the contrast of MSN@TA-Zn and MSN@TA-BDBA particles while the mesopores can still be observed, suggesting the presence of MPN or BPN on the surfaces of the particles. However, no continuous layer or shell was formed on these particles, suggesting the absence of a MPN or BPN film. The hydrodynamic diameters of the MSNs (Fig. S2†) were measured by using DLS. The hydrodynamic diameters are just slightly larger than those observed in the corresponding TEM images. The average particle diameter did not significantly change after DOX was loaded (112 nm, PDI = 0.087 for MSN and 113 nm, PDI = 0.091 for DOX loaded MSN). However, the diameters increased after MPN or BPN coating (124 nm, PDI = 0.177 for MSN@TA-Zn and 135 nm, PDI = 0.203 for MSN–TA–BDBA). This result is attributed to some aggregation of the MSNs in solution. It should be noted that further increase of MPN or BPN amount by using 20–40 wt% initial TA amount resulted in severe particle agglomerations and connections, together with MSN-free organic particles (Fig. S3†), revealing excess amounts of TA.

Fig. 5 TEM images of MSN (a), and its counterparts with constructed MPN or BPN on surfaces, i.e. MSN@TA-Zn (b) MSN@TA-BDBA (c). Comparison of FTIR spectra for varying particles (d): MSN, MSN@TA, MSN@TA-Zn, and MSN@TA-BDBA. The scale bars in (a)–(c) represents 50 nm.

To further verify the MPN or BPN structure, FTIR spectra of TA-based MSN@TA, MSN@TA-Zn, MSN@TA-BDBA materials were measured and shown in Fig. 5d, where MSN was used as a control. The typical absorption peaks for silica were found at 800 cm⁻¹ for the symmetric stretching vibration ν_s (Si–O–Si), at 1098 cm⁻¹ for the asymmetric stretching vibration ν_as (Si–O–Si), and at 464 cm⁻¹ for the Si–O–Si bending mode. In addition, the Si–OH stretching vibration was characterized at 953 cm⁻¹ and 3450 cm⁻¹. After TA loading, all particles show characteristic vibration peaks at 1635 cm⁻¹ corresponding to C=O groups of TA. The other small peaks between 1400 cm⁻¹ and 1600 cm⁻¹ can be assigned to aromatic compound stretching and the vibration of substituted benzene rings.³³ Notably, MSN@TA has a broad peak centered around 3300 cm⁻¹ coming from OH groups of polyphenols, suggesting that the silica surface composed by silanols (Si–OH) was covered with TA molecules. However, the band was shifted to 3470 cm⁻¹ in the case of MSN@TA-Zn and 3260 cm⁻¹ in the case of MSN@TA-BDBA. In addition, an emerging new peak at 1318 cm⁻¹ shows in the case of MSN@TA-BDBA, which should be attributed to the B–O stretching signal originated from the boronate structure in BPN.³⁹ These spectra results support the successful crosslinking of the adsorbed TA molecules by Zn²⁺ and BDBA.

3.3 Evaluation of controlled drug release

Drug release properties of MSN@TA-Zn and MSN@TA-BDBA were examined by using DOX as a model drug. DOX was loaded into MSN by absorption from solutions and the loading degree utilized for drug release study is 90 μg mg⁻¹. Fig. 6 depicts the cumulative DOX release profiles. To elucidate the effect of MPN or BPN incorporation, pristine MSN without any surface modifications was also employed in drug release evaluations as a reference.

Fig. 6 Comparison of cumulative release profiles of DOX from MSN (a), MSN@TA-Zn (b), and MSN@TA-BDBA (c). An initial amount of 10% tannic acid was used in the preparation process of MSN@TA-Zn and MSN@TA-BDBA. Mannitol was utilized at a concentration of 50 mM for the profiles in (c).

In the case of MSN, an uncontrollable release was clearly observed, as shown in Fig. 6a. The increased hydrophilicity and higher solubility of DOX at lower pH caused by increased protonation of the NH₂ group on DOX, resulted in release plateaus at 95%, 80%, and 50% at pH 4.4, 5.5, and 7.4, respectively, after 4 h of burst release.⁴⁰ However, the DOX release reduced substantially in the case of MSN@TA-Zn and MSN@TA-BDBA, as evidenced by a slow release up to ∼7% after 72 h of incubation at pH 7.4 (Fig. 6b and c). These results indicate that the physically adsorbed DOX was trapped inside the mesopores by the incorporation of MPN or BPN.

Regarding the responsive-cleavage property of MPN and BPN, lower incubation pHs of 4.4 and 5.5 were employed to evaluate the behaviors of MSN@TA-Zn and MSN@TA-BDBA in controlled drug release. In the case of MSN@TA-Zn, the cumulative release curves of DOX contain a fast release step up to 45% (pH 4.4) or 20% (pH 5.5) in the first 3 h, as well as a slower but steadily increasing step reaching 80% (pH 4.4) or 30% (pH 5.5) in 72 h. In comparison, the release amount levels off quickly after 5 h and a final DOX release plateau at only 30% was reached afterwards for MSN@TA-BDBA at the lowest pH of 4.4. In addition, total DOX release amounts of 50% (pH 4.4), 30% (pH 5.5), and 10% (pH 7.4) for MSN@TA-BDBA was caused by the addition of mannitol at a concentration of 50 mM, which was reported to induce an efficient breakage of the BPN structure.²³

When the initial TA amount with respect to particles in the TA deposition step was increased to 20 wt% and 40 wt%, the plateau levels of DOX release were reduced significantly for MSN@TA-Zn at pH 5.5 and pH 7.4, as well as MSN@TA-BDBA at all pH values (Fig. S4†). The decreased DOX release is most likely related to the particle agglomerations and inter-connections observed at higher TA amounts (Fig. S3†). However, no significant change in the release profiles was found for MSN@TA-Zn at pH 4.4, except for a delayed fast release step to 12 h for the particles prepared at higher TA amounts. The results above demonstrated that the pore occlusion induced by deposition of TA with the molecular dimensions apparently paved a way for complete sealing of the pore openings (Fig. 7). Besides, the metal ion–TA coordination might be more labile than the bononic acid induced covalent bonding in BPN.

Fig. 7 Schematic representation on the process of constructing MSN@TA-Zn and MSN@TA-BDBA for responsive drug release by separating the TA deposition and the crosslinking into two steps.

To further explore the contribution of TA deposition on pore sealing, it would be interesting to investigate whether other polyphenol analogues can generate a similar effect in controlled the drug release. Epigallocatechin gallate (EGCG) with two pyrogallol moieties and a low molecular weight of 458 g mol⁻¹ is a typical polyphenol substitute in generating polyphenol coatings.⁴¹ The same procedure was repeated for the generation of MSN@MPN particles, except for the replacement of TA with EGCG. Fig. 8 displays the cumulative release curves of DOX from the constructed MSN@EGCG-Zn particles. As shown, the release ratio of DOX was remarkably lower than that in the case of MSN@TA-Zn at pH 4.4. Only 20 wt% was released in the first 54 h, although the release amount at pH 7.4 was comparable to that in MSN@TA-Zn. The implied less vulnerable MPN network from EGCG is possibly related to the easy diffusion of EGCG inside mesopores during the polyphenol deposition step, considering the small molecular dimensions of EGCG. Therefore, there might be a possibility of complicated MPN structures inside the mesopores which needs further exploration.

Fig. 8 Cumulative release curves for DOX from MSN@EGCG-Zn particles (a) and the molecular structure of EGCG (b).

3.4 In vitro drug delivery by the nanocarriers

To study the feasibility of using the obtained particles for drug delivery, MSN@TA-Zn and MSN@TA-BDBA at a concentration of 100 μg mL⁻¹ were incubated with Hep-G2 at 37 °C for 24 h. The cells were then washed extensively and imaged via confocal fluorescence microscopy, as shown in Fig. 9. To determine if these particles are able to enter cells, we covalently labeled MSN with fluorescein isothiocyanate (FITC). Overlay of the confocal images at the same z-distance showed both the nucleus (blue) and the MSN@TA-Zn or MSN@TA-BDBA (green) indicating that both particles were inside of the cells due to the close planar proximity to the nuclei.

Fig. 9 Comparison of the confocal microscopy images for Hep-G2 cells after 24 h of incubation with 100 μg mL⁻¹ of MSN@TA-Zn (a–c) and MSN@TA-BDBA (d–f) particles. Cell nuclei were stained with DAPI, while MSN particles were labeled with FITC, in order to examine the internalization and location of the particles inside cells. The scale bar represents 25 μm.

Next, the biocompatibility of the MSN@TA-Zn and MSN@TA-BDBA particles was examined by incubating Hep-G2 cells with the nanoparticles at different concentrations for 24 h, as shown in Fig. 10a. As revealed by the CCK-8 assay, the viability of the cells was kept above 95% up to a very high concentration of 150 μg mL⁻¹ for both particles, indicating high biocompatibility of these nanocarriers.

Fig. 10 Comparison of Hep-G2 cells viability after incubated for 24 h with varying concentrations of MSN@TA-Zn and MSN@TA-BDBA particles (a); Hep-G2 cell viabilities after 24 h of incubation with varying concentrations of DOX (b).

To confirm the cell killing activity of the drug by the MSN@TA-Zn and MSN@TA-BDBA nanocarriers, Hep-G2 cells were incubated with free DOX and DOX-loaded nanoparticles at different concentrations up to 40 μg mL⁻¹ for 24 h. As shown in Fig. 10b, the DOX loaded MSN@TA-Zn and MSN@TA-BDBA particles demonstrate a dose-dependent cytotoxic effect against Hep-G2 cells and cytotoxicity is almost equivalent to the free drug at concentrations bellow 20 μg mL⁻¹. While free DOX shows an obvious cytotoxicity to Hep-G2 cells cancer cells at a lower concentration (half-maximal inhibitory concentration (IC50) = 5 μg mL⁻¹) after incubation for 24 h, MSN@TA-Zn and MSN@TA-BDBA show a higher IC50 of ∼10 μg mL⁻¹ after the same incubation time. It is worth noting that the cytotoxicity is lower for DOX loaded MSN@TA-Zn and MSN@TA-BDBA, especially at high drug concentrations, which should be attributed to the controlled release of DOX from these MSN particles with loaded MPN or BPN nanobarriers on the particle surface.

4. Conclusions

In conclusion, we have demonstrated that the superoxide anions released from MSN in aqueous solution can lead to efficient deposition of TA molecules, resulting in the occlusion of the pore openings, and subsequent easy construction of MPN or BPN for regulating drug release in the following step. By using DOX as a model drug, pH responsive release, as well as successful intracellular drug delivery for the inhibition of cancer cells on the basis of the MSN@TA-Zn and MSN@TA-BDBA particles, were demonstrated. The MSN@TA-Zn particles have a higher pH sensitivity, which gives rise to pH-enhanced drug release kinetics. This class of composite nanocarriers is further expected to pave a new avenue for the simple generation of controlled release mechanics on the basis of MSN and polyphenol.

Acknowledgements

This work was supported in part by the National Natural Science Foundation of China (NSFC, Grant No. 51502027, 21274169), Basic Advanced Research Project of Chongqing (Grant No. cstc2015jcyjA10051), and 100 Talents Program of Chongqing University (J. Z.). National Engineering Research Center for Nanotechnology (Shanghai) is greatly acknowledged for the help for TEM characterization.

Notes and references

1. Y. Wang, Q. Zhao, N. Han, L. Bai, J. Li, J. Liu, E. Che, L. Hu, Q. Zhang, T. Jiang and S. Wang, Nanotechnol. Biol. Med., 2015, 11, 313–327.
2. F. Figueira, J. A. S. Cavaleiro and J. P. C. Tomé, J. Porphyrins Phthalocyanines, 2011, 15, 517–533.
3. A. Baeza, M. Manzano, M. Colilla and M. Vallet-Regi, Biomater. Sci., 2016, 4, 803–813.
4. J. G. Croissant, X. Cattoen, M. Wong Chi Man, J.-O. Durand and N. M. Khashab, Nanoscale, 2015, 7, 20318–20334.
5. W. Xu, J. Riikonen and V.-P. Lehto, Int. J. Pharm., 2013, 453, 181–197.
6. V. Mamaeva, C. Sahlgren and M. Lindén, Adv. Drug Delivery Rev., 2013, 65, 689–702.
7. E. Aznar, M. Oroval, L. Pascual, J. R. Murguía, R. Martínez-Máñez and F. Sancenón, Chem. Rev., 2016, 116, 561–718.
8. A. Baeza, M. Colilla and M. Vallet-Regí, Expert Opin. Drug Delivery, 2015, 12, 319–337.
9. S. Alberti, G. J. A. A. Soler-Illia and O. Azzaroni, Chem. Commun., 2015, 51, 6050–6075.
10. S. Mura, J. Nicolas and P. Couvreur, Nat. Mater., 2013, 12, 991–1003.
11. Q. Zhang, F. Liu, K. T. Nguyen, X. Ma, X. Wang, B. Xing and Y. Zhao, Adv. Funct. Mater., 2012, 22, 5144–5156.
12. D. G. Barrett, T. S. Sileika and P. B. Messersmith, Chem. Commun., 2014, 50, 7265–7268.
13. Y. Liu, K. Ai and L. Lu, Chem. Rev., 2014, 114, 5057–5115.
14. V. Kozlovskaya, B. Xue, W. Lei, L. E. Padgett, H. M. Tse and E. Kharlampieva, Adv. Healthcare Mater., 2015, 4, 686–694.
15. F. Liu, V. Kozlovskaya, O. Zavgorodnya, C. Martinez-Lopez, S. Catledge and E. Kharlampieva, Soft Matter, 2014, 10, 9237–9247.
16. M. V. Lomova, A. I. Brichkina, M. V. Kiryukhin, E. N. Vasina, A. M. Pavlov, D. A. Gorin, G. B. Sukhorukov and M. N. Antipina, ACS Appl. Mater. Interfaces, 2015, 7, 11732–11740.
17. Z. Gao and I. Zharov, Chem. Mater., 2014, 26, 2030–2037.
18. N. Sahiner, S. Sagbas and N. Aktas, Microporous Mesoporous Mater., 2016, 226, 316–324.
19. F. Liu, X. He, H. Chen, J. Zhang, H. Zhang and Z. Wang, Nat. Commun., 2015, 6 DOI:10.1038/ncomms9003.
20. H. Ejima, J. J. Richardson, K. Liang, J. P. Best, M. P. van Koeverden, G. K. Such, J. Cui and F. Caruso, Science, 2013, 341, 154–157.
21. Y. Ping, J. Guo, H. Ejima, X. Chen, J. J. Richardson, H. Sun and F. Caruso, Small, 2015, 11, 2032–2036.
22. M. A. Rahim, H. Ejima, K. L. Cho, K. Kempe, M. Müllner, J. P. Best and F. Caruso, Chem. Mater., 2014, 26, 1645–1653.
23. J. Guo, H. Sun, K. Alt, B. L. Tardy, J. J. Richardson, T. Suma, H. Ejima, J. Cui, C. E. Hagemeyer and F. Caruso, Adv. Healthcare Mater., 2015, 4, 1796–1801.
24. H. Liang, J. Li, Y. He, W. Xu, S. Liu, Y. Li, Y. Chen and B. Li, ACS Biomater. Sci. Eng., 2016, 2, 317–325.
25. Y. Ju, J. Cui, M. Müllner, T. Suma, M. Hu and F. Caruso, Biomacromolecules, 2015, 16, 807–814.
26. S. Kim, S. Philippot, S. Fontanay, R. E. Duval, E. Lamouroux, N. Canilho and A. Pasc, RSC Adv., 2015, 5, 90550–90558.
27. J. Zhang, M. Niemela, J. Westermarck and J. M. Rosenholm, Dalton Trans., 2014, 43, 4115–4126.
28. N. Lang and A. Tuel, Chem. Mater., 2004, 16, 1961–1966.
29. J. Zhang, J. M. Rosenholm and H. Gu, ChemPhysChem, 2012, 13, 2016–2019.
30. J. Lin, Y. Zhan, Z. Zhu and Y. Xing, J. Hazard. Mater., 2011, 193, 102–111.
31. J. Wang, S. Zheng, J. Liu and Z. Xu, Chem. Eng. J., 2010, 165, 10–16.
32. V. Ball and F. Meyer, Colloids Surf., A, 2016, 491, 12–17.
33. N. Sahiner, S. Sagbas and N. Aktas, Mater. Sci. Eng., C, 2015, 49, 824–834.
34. M. Misawa and J. Takahashi, Nanotechnol. Biol. Med., 2011, 7, 604–614.
35. Z. Tao, G. Wang, J. Goodisman and T. Asefa, Langmuir, 2009, 25, 10183–10188.
36. T. S. Sileika, D. G. Barrett, R. Zhang, K. H. A. Lau and P. B. Messersmith, Angew. Chem., Int. Ed., 2013, 52, 10766–10770.
37. X. Zheng, J. Zhang, J. Wang, X. Qi, J. M. Rosenholm and K. Cai, J. Phys. Chem. C, 2015, 119, 24512–24521.
38. V. Cauda, B. Onida, B. Platschek, L. Muhlstein and T. Bein, J. Mater. Chem., 2008, 18, 5888–5899.
39. M. H. Todd, S. Balasubramanian and C. Abell, Tetrahedron Lett., 1997, 38, 6781–6784.
40. X. Zheng, F. Chen, J. Zhang and K. Cai, J. Mater. Chem. B, 2016, 4, 2435–2443.
41. S. Hong, J. Yeom, I. T. Song, S. M. Kang, H. Lee and H. Lee, Adv. Mater. Interfaces, 2014, 1, 1400113.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra15773d

---