Dopamine hydrochloride (DA), ammonia hydroxide (NH₄OH, 28–30%), ethanol (EtOH), 2-propanol (iPrOH), tetraethyl orthosilicate (TEOS), cetyltrimethylammonium bromide (CTAB), Pluronic F127, 3-(aminopropyl) triethoxysilane (APTES), catalase (C30), fluorescein isothiocyanate (FITC), hydrogen peroxide (H₂O₂), dimethyl sulfoxide (DMSO), and sodium azide (NaN₃) were purchased from Sigma-Aldrich Australia. Phosphate buffer (PB) solutions with different pH were made by mixing 0.2 M NaH₂PO₄ with Na₂HPO₄ and adjusted to appropriate pH using 2.0 M HCl and 2.0 M NaOH solutions. Milli-Q water (18.2 MΩ cm) was utilized in all the experiments.

MSN particles were prepared by a sol-gel process with binary surfactant system previously described by Kim et al. (2010). In brief, 0.5 g of CTAB and 2.05 g of Pluronic F127 were dissolved in a mixture of water (96 mL), EtOH (43.1 mL), and concentrated NH₄OH solution (28 wt. %) (11.2 mL) by stirring at room temperature (RT). After obtaining a well-dispersed homogenous solution, 1.93 mL of TEOS was added quickly into the mixture under vigorous stirring for 90 s. The mixture was kept static for the next 24 h at RT for further condensation. The white precipitate was then collected, washed three times with water, and calcined at 550°C for different time intervals (2, 4, and 6 h) to remove the surfactants.

The functionalization of bare MSN with amine groups followed previous work described by Huang et al. (2003) with slight modifications. Briefly, 0.2 g of MSN was refluxed with 10 mL of toluene and 1 mL of APTES under stirring at 70°C for 18 h. The white precipitate was then collected via centrifugation and washed with toluene and pentane, respectively.

A previous procedure of PDA coating (Nador et al., 2017) was followed with slight modifications. Briefly, 57 mg DA was dissolved in 39 mL iPrOH. Then, 2.25 mL NH₄OH and 30 mg MSN were added to the solution. The coating process was left for 18 or 24 h under magnetic stirring at RT. The MSN@PDA particles were obtained by centrifugation and washed several times with EtOH.

For MSN core removal, MSN@PDA particles were dispersed in different water volumes. The vial was kept under stirring for 48 h in an incubator with set temperature at 37°C. The PDA hollow capsules were obtained via centrifugation and washed 3 times by water.

The conjugation of FITC and catalase followed the previous procedure described by Yu et al. (2005) with slight modifications. In brief, 300 μl of FITC/DMSO (1 mg/mL) was added to 4 mL of 1 mg/mL catalase solution in PB (pH 8.0). The mixture was allowed to stand for 50 min, followed by the dialysis against PB (pH 7.0) for 72 h, changing the solution every 24 h.

Approximately 4 mg of amine-grafted MSN particles were added and incubated with 4 mL of 1 mg/mL FITC-catalase solution at room temperature for 24 h. The FITC-catalase loaded MSN were then washed with PB (pH 7.4) to remove unbound enzyme. These particles were later subjected to PDA coating at RT and core removal process as described above.

An amount of 50 μL of free PDA capsule and catalase@PDA capsule solution (5 mg/mL) was added to a solution of 10 mM H₂O₂ in PB pH 7. The activity of catalase was examined based on the absorbance change at 240 nm of the H₂O₂ solution, in response to the addition of free PDA capsules and catalase@PDA capsules.

The UV-Visible spectroscopy was carried out with a Cary 5000 UV-Vis spectrophotometer (Australia). FTIR was performed on MSN and PDA@MSN to check the silanol bonds in the silica network and success of PDA coating. It was performed using a Thermo Scientific Nicolet iS5 FT-IR Spectrometer, with 8 scans being performed per test.

Morphology of MSN and PDA hollow capsules were examined by using a field emission scanning electron microscope (FeSEM, ZEISS SUPRA 40VP, Germany) at an acceleration voltage of 3 kV.

Additionally, transmission electron microscopy (TEM) and energy dispersive X-ray spectroscopy (EDS) was utilized to observe of the interior structure of MSN@PDA after being dispersed in water as well as to quantify the remaining Si element. TEM and EDS were performed by using a JEM 2100F (JEOL Ltd., Japan) at room temperature at a voltage of 200 kV.

To confirm the successful loading of FITC-catalase into PDA capsules, visualization of FITC fluorescence were performed using Olympus Fluoview 1000 IX81 (Japan) confocal laser scanning microscope, operated using 100× oil-immersion objective combined with 5× optical zoom.

The preparation of monodispersed MSN was reproduced by a sol-gel process using a binary surfactant system with the aid of triblock copolymer Pluronic F127 and CTAB described by Kim et al. (2010). In this process, TEOS plays as the main source of silica precursor. As could be seen in Figure 1, the obtained particles were quite uniform, with a spherical shape, and their size distribution was calculated to be 158 ± 13 nm based on an average of 100 particles. It was found that the calcination time at 550°C (2, 4, or 6 h) does not significantly affect the size of synthesized MSN. However, as per a previous report, calcination time might affect the surface area and rigidity of particles due to the condensation of the Si-O-Si network (Mokaya, 1999; Ruckdeschel et al., 2015).

The PDA coating on MSN took place in a DA solution under mild basic conditions, with the assistance of oxygen. It has been suggested that the formation of PDA has consecutive steps, initiated by the oxidization of DA monomer, followed by further oxidization, cyclization, and rearrangement to form 5, 6-dihydroxyindole (Hong et al., 2012). The obtained 5, 6-dihydroxyindole products are then adsorbed onto MSN particles via covalent oxidative polymerization and non-covalent interactions, such as II-II interactions, electrostatic interactions and hydrogen bonding (Hong et al., 2012; García et al., 2014). The visual color change of MSN from white to dark-brown gives an indication of the successful coating of PDA (Figure 2A). To support for this, the FT-IR spectrum of MSN before coating shows two peaks in the range of 963–1,054 cm⁻¹, which is attributed to the silanol bonding (Si-O-Si) in the silica network. After PDA coating, two additional peaks are observed in the range of 1,495–1,609 cm⁻¹ due to the C-C stretching of benzene rings and the N-H bending of PDA (Figure 2B).

It was observed that the average size of MSN significantly increased after PDA coating, and the coating thickness was strongly affected by the coating conditions such as polymerization time. For example, the average size of MSN increased from 158 to 178 and 226 nm after being coated with PDA for 18 and 24 h, respectively. This equals to PDA coating thickness of 20 nm (18 h) and 68 nm (24 h). The PDA coating thickness was also affected by the coating temperature. For example, for the polymerization carried out at 45°C, the average sizes of PDA particles were 221 and 232 nm for 18 and 24 h coating condition, respectively. It could be reasoned that the deposition rate and the polymerization degree of the PDA layer were influenced by the coating temperature and the duration of the coating process (Jiang et al., 2011; Cho and Kim, 2015). Compared to the coating condition at ambient temperature, the polymerization of DA occurs at a higher rate, which results in thicker PDA coating. Of note, there was no aggregation of PDA-coated MSN regardless of different coating conditions, as shown in Figures 2C,D.

PDA hollow capsules were obtained by the removal of silica core templates by simply dispersing the MSN@PDA particles in water. However, we found that the silica dissipation process was strongly influenced by a few factors including dissolution conditions and the properties of MSN@PDA particle itself, such as the thickness of PDA coating layer and/or the rigidity of the interior core of MSN; see below for detailed studies.

PDA hollow capsules were then assessed for their capacity to encapsulate macro-biomolecules. For this purpose, catalase was chosen as a model enzyme to be encapsulated into PDA hollow capsules using the pre-loading strategy that we developed previously (Yu et al., 2005). The encapsulation was carried out by first loading catalase onto MSN, followed by PDA coating and silica dissolution in water. Before catalase adsorption, MSN was grafted with amino groups in order to enhance the electrostatic interaction with catalase. The amount of catalase adsorbed onto amine-functionalized MSN was measured the UV-Vis absorbance of catalase before and after loading. It was determined that the amount of catalase adsorbed onto MSN was approximately 302 ± 78 mg/g of MSN. This amount was significantly higher than that for bare MSN (~75 mg/g). The catalase loaded MSN was subjected to PDA coating and silica core removal. The successful loading of catalase into PDA capsules was confirmed by the uniform fluorescence within the hollow capsule under CLSM when FITC-catalase was used (Figure 7A). It was previously reported that the illumination of protein with a fluorescent tag could be quenched by PDA (~80% quenching effectiveness) (Qiang et al., 2015; Xing et al., 2018); however, the fluorescence of FITC-catalase entrapped within PDA capsules could be observed to be quite vibrant. This is likely due to the high concentration of FITC-labeled protein being encapsulated.

One drawback of previous work on using silica nanoparticles for hollow capsule preparation is that harsh conditions such HF are needed to dissolve silica templates, which might affect the bioactivity of biomacromolecules. The advantage of the present work is that we are able to dissolve the silica core using water. Whether the encapsulated catalase retains its activity was examined by its capacity to interact with and decompose H₂O_2. This was carried out by measuring the absorbance of H₂O₂ at 240 nm upon the addition of catalase@PDA capsule. Empty PDA capsules at the same concentration was used as a control. As can be seen in Figure 7B, the catalytic activity of encapsulated catalase was prominent upon the addition to H₂O₂ solution, and the rate of absorbance change of H₂O₂ solution was calculated to be 0.006 per minute. This result suggests that the catalytic activity of entrapped catalase is still retained after the encapsulation and silica removal process, even though the breakdown of H₂O₂ occurred slowly and gradually since the interaction between encapsulated catalase and H₂O₂ molecules was reduced by the PDA layer. From the obtained results, it suggested that the PDA capsule wall was likely porous, which served as a pathway for H₂O₂ molecule penetration.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.