pH-Degradable antioxidant nanoparticles based on hydrogen-bonded tannic acid assembly

Abstract

Hydrogen-bonded polyphenol-based assemblies have attracted increasing interest for biomedical applications. Polyphenolic drug-loaded films can be coated on various devices with different shapes and sizes. Here, we report a novel versatile pH-responsive system based on hydrogen-bonded poly(ethylene glycol) (PEG)/tannin acid (TA) coatings on zein/TA colloidal nanoparticles (zein/TA/PEG NPs). Hydrogen bonding was considered to be the driving force for the coating buildup between PEG and TA, which was confirmed by Fourier transform infrared (FTIR) spectra. Because of the reversible/dynamic nature of hydrogen bonding, the release profile of TA was directed by pH value, temperature, ionic strength and TA concentration. The release rate of TA increased with increasing pH and temperature, but decreased with increasing ionic strength. This new drug delivery vehicle could also be used to load hydrophobic and unstable molecules through interacting with zein by hydrophobic interaction to achieve efficient protection. We demonstrated that the hydrophobic molecular nutrient, VD3, can be successfully loaded into zein/TA/PEG NPs with high encapsulation efficiency. Photostability against UV light was significantly improved after encapsulation. The encapsulated VD3 could be regulated not only by the pH of the solutions but also by TA concentration.

---

1. Introduction

Surface engineering with natural or synthetic compounds as coatings is considered an efficient way to modify and functionalize surfaces or to prepare free-standing polymers with novel and advanced properties.^1–4 One important application of such coatings is in smart cargo-delivery systems. The targeted release of a cargo can be achieved based on the coatings which are response to external stimuli such as pH,^5–7 temperature,^8,9 and redox potential,^10,11 etc. Among the stimuli-responsive deliver systems, two general strategies have been developed based on such multifunctional coatings. The first application is that drugs are encapsulated into micro- or nano-materials, and the coatings at surfaces act as diffusion barriers.⁷ As the disruption of the coatings, the loaded drugs can be controlled released. In the second strategy, the coatings were used not only as a reservoir for the active therapeutic cargo but also a coating to modulate surface, response to small external stimuli.^12–15

Self-assembled coatings guided by hydrogen bonding interactions present new opportunities in deliver system since it has been considered to be response to small external stimuli, namely changes in temperature, pH, ionic strength thus providing a variety of promising applications in biomedical and bioengineering fields.^16,17 Tannic acid (TA), which contains numerous terminal hydroxyl groups, has unique structural properties that facilitate interactions with a variety of materials (Scheme 1).^16,18,19 Like many other plant polyphenols, TA with the high biological activity is used as an antimutagenic, anticarcinogenic, antimicrobial, antioxidant, and antibacterial agent.^20–22 A large number of carbonyl and phenolic functional groups are responsible for different types of bonding with various molecules via multiple reaction pathways, including electrostatic interaction, hydrogen bonding, hydrophobic interactions, covalent bonding.^18,23,24 With pH-/temperature-/salt-responsive properties, hydrogen-bonded coatings based on TA/poly(N-vinylcaprolactam) (PVCL),²⁵ TA/poly(ethylene oxide) (PEO),^26,27 TA/poly(N-vinylpyrrolidone) (PVPON),^12,17 have been reported previously. TA–protein interactions are essentially physical: hydrophobic and hydrogen bond-mediated.²⁸ The pentagalloyl glucose of TA was believed to form hydrophobic interactions with the pyrrolidine ring of proline in protein.²⁹ In addition, hydrogen bonds between the hydroxyl groups of tannins and the carbonyl groups of proteins are believed to be involved in the complexation of the proteins and tannins. Zein, a proline-rich alcohol-soluble protein, has been extensively investigated in the encapsulation of bioactive compounds because of its capability to form self-assembled NPs.^30,31 Due to the binding affinity to neutral polymers¹⁶ and proteins³² via hydrogen bonding and hydrophobic interactions, TA was used as an efficient tool to control the self-assembly behavior of zein, so as to fabricate stable zein/TA colloidal particles.²⁹

Scheme 1 Illustration of the synthesis and structures of TA/PEG coated zein NPs and chemical structure of TA and PEG.

Most of hydrogen-bonded systems studied to date have been demonstrated to be hollow capsules³³ or coatings on inorganic materials.¹⁷ Few studies focus on the applications of hydrogen-bonded coatings on self-assembled biodegradable NPs from natural polymers. Here we first prepared zein/TA colloidal nanoparticles (zein/TA NPs) to achieve a high loading of TA. Then TA in surface of the colloidal particles were hydrogen-bonded with poly(ethylene glycol) (PEG) forming coatings to achieve the sustained release of polyphenolic drugs.³⁴ As we know, coating the NPs with a hydrophilic polymer brush, e.g. PEG, prolongs circulation time and avoids clearance by the mononuclear phagocytic system.^35–37 So in our study we used TA as the hydrogen donor, which could interact with the ether group (–O–) of PEG to form TA/PEG coatings. PEG here was used as hydrogen acceptor, to fabricate dynamic, hydrogen-bonded coatings with polyphenolic drugs.³⁴ Additionally, because of the hydrophobic property of zein, we encapsulated vitamin D3 (VD3), as a model hydrophobic and unstable compound, into zein NPs, and then coated with TA/PEG coatings. Such system could not only achieve the controlled release of TA and VD3, but provided efficient protection of VD3 against UV light.

2. Materials and methods

2.1. Materials

Zein (Z0001) was purchased from Tokyo Chemical Industry, Co., Ltd. (Tokyo, Japan).

VD3 (99%), tannic acid (TA; M_w 1701), 1,1-diphenyl-2-picrylhydrazyl (DPPH) and 2,2′-azinobis-(3-ethylbenzthiazoline-6-sulphonate) (ABTS) was purchased from Aladdin Chemistry Co., Ltd. (Shanghai, China). Poly(ethylene glycol) (PEG; M_w 6 kDa) was obtained from the Sinopharm Chemical Reagents Co., Ltd. (Shanghai, China). Other chemicals used were of analytical grade. All the solutions used in the experiments were prepared using ultrapure water through a Millipore (Millipore, Milford, MA, USA) Milli-Q water purification system.

2.2. Formation of PEG/TA coated NPs

All solutions were freshly prepared for immediate use. The standard preparation process was described as follows: zein and TA was dissolved in aqueous ethanol solutions (75% v/v) to obtain a stock solution with final concentration of zein 10 mg mL⁻¹ and TA 10 mg mL⁻¹, 20 mg mL⁻¹ and 30 mg mL⁻¹ respectively. Then, 1 mL of the above zein/TA solution was rapidly poured into 9 mL of water, leading to different weight ratios of zein : TA at 1 : 1, 1 : 2, 1 : 3 respectively. Next, 100 μL of PEG solution (50 mg mL⁻¹) was added and the dispersion was vigorous stirred. The obtained opaque single phase solution was then freeze-dried for 48 h. The control NPs without PEG were also prepared in parallel.

2.3. Characterizations of PEG/TA coated NPs

Dynamic laser scattering (DLS) and zeta potential. Dynamic Light Scattering (DLS) and zeta potential measurement were performed on a commercial laser light scattering instrument (Nano-ZS90, Malvern, UK).

Stability of PEG/TA coated NPs. The change in particle size, PDI and zeta potential of NPs upon dilution (10–100 fold) were monitored for stability tests. Similarly the NPs were diluted 10-fold in DMEM and incubated at 37 °C for 24 h. The particle size and PDI was measured as a function of time.

Influence of pH on PEG/TA coated NPs. The pH of the aqueous phase was adjusted using 0.1 N HCl or 0.1 N NaOH. The effect of pH on the aggregation of NPs was studied by dispersing the pre-formed PEG/TA coated NPs in water adjusted to pH from 1.5 to 8 using 0.1 N HCl or 0.1 N NaOH. The turbidities of complex solutions were analyzed using UV-vis spectrophotometer (UV-1100, MAPADA) at 600 nm.

Transmission electron microscope (TEM). Images were taken on a JEM-2100F (JEOL, Japan). The samples were prepared by dropping solution onto copper grids coated with carbon and then dried naturally.

Fourier transform infrared spectroscopy (FT-IR). FT-IR spectra were obtained with a Jasco 4100 series with an attenuated total reflection cell (Jasco Inc., Easton, MO). All NPs samples after purified by successive dialysis (MWCO 3500) against deionized water and freeze-dried were prepared as KBr pellets and were scanned against a blank KBr pellet background.^31,38

2.4. TA release

To measure the release kinetics of TA, an aliquot of PEG/TA coated NPs was taken into a dialysis bag (3500 MWCO) and suspended in 100 mL release medium with different pH (pH 8, pH 7.4, pH 1.2, 0.1 M buffer) and gently shaken at 100 rpm in an air bath. Temperature (4 °C, 25 °C, 37 °C or 45 °C) was controlled with a refrigerated circulator. At predetermined intervals, the release media was removed, and the same volume of fresh media with the same temperature was added. Concentration of TA in the release media was determined using UV-vis spectroscopy at a wavelength of 276 nm.

2.5. Antioxidant activity of the released TA

The antioxidant activity of NPs was measured according to the DPPH method with minor modification.^12,39 Briefly, macroporous adsorption resin (N101) was treated with diluted DPPH˙ solution for a period of 12 h. It was then washed with water thoroughly. The DPPH˙-loaded resin was then soaked in ethanol. The freeze-dried NPs were dipped into the medium. The decoloration of the resin with time was recorded with a digital camera.

ABTS radical scavenging activity was assayed as per the method of previous study with a slight modification.⁴⁰ The stock solutions included 7.4 mM ABTS solution and 2.6 mM potassium per sulphate solution. The working solution was prepared by mixing the two stock solutions in equal quantities and allowing them to react for 12 h at room temperature in dark. Resin was treated with diluted ABTS⁺˙ solution for a period of 12 h and then washed with water thoroughly. The ABTS⁺˙-loaded resin was then soaked in pH 7.4 phosphate buffer. The lyophilized NPs were put in diluted ABTS⁺˙ solution and the mixture were left at room temperature for certain period preservation. The decoloration of the resin with time was recorded with a digital camera.

2.6. VD3 encapsulation

VD3 (5 mg mL⁻¹) was dissolved in pure ethanol as stock solution. VD3 solution (5 mg mL⁻¹) was added dropwise into zein/TA solution with mild stirring for 30 min. Then, 1 mL of the above mentioned solution was rapidly poured into 9 mL of water. Next, 100 μL of PEG solution (50 mg mL⁻¹) was added and the dispersion was vigorous stirred. The obtained opaque single phase solution was then freeze-dried for 48 h. The final concentration of VD3 was 50 μg mL⁻¹.

Encapsulation efficiency was measured according to the method of Wang et al. with minor modifications.⁴¹ Briefly, the freeze-dried NPs were washed with hexane, and the suspension was filtrated through a Whatman no. 1 filter paper. This procedure was repeated for several times until the last filtrate did not exhibit any absorbance. Then all the filtrates were combined and measured for its absorbance at 264 nm with a UV-vis spectrophotometer (UV-1100, MAPADA). The encapsulation efficiency was defined as the drug content that was entrapped into zein/TA/PEG NPs and calculated as follows:

EE (%) = (total VD3 − free VD3)/total VD3 × 100%

2.7. Characterization of VD3-encapsulated NPs

The particle size and zeta potential of VD3-encapsulated NPs were conducted on a commercial laser light scattering instrument (Nano-ZS90, Malvern, UK). TEM was conducted on a JEM-2100F (JEOL, Japan). X-ray diffraction (XRD) was carried out using a diffract meter type D/max-rA (Rigaku Co., Japan) with Cu target and Kα radiation (λ = 0.154 nm).

2.8. Photochemical stability measurement

The freshly prepared samples (VD3-encapsulated NPs), together with VD3 dispersion in water as control, were used for VD3 stability measurement. VD3 dispersion was prepared by dissolving a small amount of VD3 in ethanol followed by dispersion into water, with the final concentration of VD3 of 50 μg mL⁻¹. Samples in transparent glass vials were placed in a light-proof cabinet and exposed to two 352 nm UV light bulbs (24 W × 2) for up to 8 h. At exposure time intervals, 200 μL of sample was withdrawn from each treatment and then VD3 was extracted and measured according to the method described above. All measurements were performed in triplicate.

2.9. VD3 release

The lyophilized samples were used for in vitro kinetic release test in different conditions. For a kinetic release test at pH 7.4 and pH 1.2, a certain amount of sample was re-suspended in buffer containing 0.2% Tween 20 to provide sink condition. The experiment was carried out under the water bath at 37 °C with shaken speed of 100 rpm. At each predetermined time interval, samples were centrifuged at 20 000g and 4 °C for 15 min, the supernatant containing the released VD3 was collected and freeze-dried, while the precipitate was re-dispersed and equivalent fresh medium was added in. The VD3 percentage released was calculated as a function of time (up to 12 h).

The accumulated release profiles of the NPs in the SGI with digestive enzymes were obtained using the method as previously reported.⁴¹ The samples were first incubated in 30 mL of simulated gastric fluid (SGF) with 0.1% pepsin (w/v) for 0.5 h. Digestion was stopped by raising the pH to 7.4 using NaOH, and the sample was then centrifuged to separate aggregates from supernatant. The supernatant containing the released VD3 was collected and freeze-dried, while the precipitate was re-dispersed using 30 mL of simulated intestinal fluid (SIF) with 1.0% pancreatin (w/v) at 37 °C and digested for 6 h under mild stirring. After digestion, the supernatant was collected by centrifugation and used for VD3 measurement. All measurements were performed in three replicates.

3. Results and discussion

3.1. Preparation of PEG/TA coated NPs

Here, we reported a simple, fast and green approach to fabricate a novel core–shell deliver system by applying one-step assembly of PEG/TA complex around zein/TA colloidal particles surface. Since TA is rich in phenolic component, –OH units in TA intermolecularly interact with –O– in PEG by hydrogen bonding.^41,42 Chemical structures of polymers used in this study were shown in Scheme 1. Here hydrogen-bonded coatings on NPs were fabricated from TA and PEG (Scheme 1). Because of the reversibility of hydrogen bonding, the coating gradually disassembled and thus released TA into the media.

The influence of zein to TA ratios and pH on particle size, PDI and zeta potential in different formulations were summarized in Table 1. Without PEG coating, the particle size of zein/TA NPs was less than 100 nm (without adjusting pH). Additionally, the higher the TA concentration added, the larger the particle size was. However, the PDI was the smallest at higher concentration of TA (Z/TA₃ sample). As a result of being coated by PEG, the particle size varied with TA added, reaching 144.4 nm, 134.3 and 200.2 nm respectively with relatively small PDI for Z/TA₁/PEG sample, Z/TA₂/PEG sample, Z/TA₃/PEG sample without changing pH. At pH 7.4, the interaction between TA and PEG decreased, because the hydrogen bonding, which was its major driving force, was weakened in neutral pH due to the deprotonation. So the particle size increased at pH 7.4 for zein/TA/PEG samples. As more TA was added, the zeta potential of zein/TA/PEG samples became more negative, ranging from −38.2 to −51.1 mV due to the acidic nature of the galloyl groups in TA.

Table 1 Particle size, polydispersity index (PDI), zeta potential of NPs in different formulations^a

Sample	Particle size	PDI	Zeta potential (mV)
a Z, zein; TA, tannic acid; PEG, poly(ethylene glycol). Z/TA1, Z/TA2, Z/TA3 represented formulations with different mass ratios of zein/TA as 1 : 1, 1 : 2, 1 : 3 respectively. Z/TA1/PEG, Z/TA2/PEG and Z/TA3/PEG represented formulations with different mass ratios of zein/TA as 1 : 1, 1 : 2, 1 : 3 with PEG coating.
Z/TA1	62.0 ± 1.0	0.39 ± 0.01	28.8 ± 3.8
Z/TA2	68.4 ± 1.0	0.28 ± 0.02	29.5 ± 0.7
Z/TA3	10.0 ± 1.1	0.22 ± 0.01	23.5 ± 3.8
Z/TA1/PEG (pH 3.6)	144.4 ± 1.5	0.16 ± 0.02	15.0 ± 0.5
Z/TA1/PEG (pH 7.4)	239.5 ± 4.0	0.25 ± 0.02	−38.2 ± 1.8
Z/TA2/PEG (pH 3.5)	134.3 ± 1.9	0.15 ± 0.03	16.1 ± 0.74
Z/TA2/PEG (pH 7.4)	199.8 ± 2.3	0.19 ± 0.03	−48.8 ± 2.0
Z/TA3/PEG (pH 3.5)	200.2 ± 1.7	0.26 ± 0.03	14.4 ± 1.1
Z/TA3/PEG (pH 7.4)	219.9 ± 4.0	0.21 ± 0.01	−51.1 ± 2.0

---

3.2. Stability of PEG/TA coated NPs

As shown in Fig. 1a, the size of Z/TA₂/PEG NPs (pH 7.4) did not change on 10-fold dilution. While after 50–100-fold dilution, the particle size showed a slight decrease which could also be found in previous study.⁴³ In order to accurately perform the cellular evaluation, the NPs should be stable and able to maintain particle size during incubation at 37 °C in culture media. 10-fold diluted NPs in DMEM, did not significantly change the particle size of Z/TA₂/PEG NPs (Fig. 1b).

Fig. 1 Influence of dilution on particle size, PDI and zeta potential of pre-formed NPs at pH 7.4 (a); particle size and PDI in culture media as a function of time (b); influence of the pH on particle size, PDI and zeta potential of pre-formed NPs (c); turbidity of the NPs as a function of pH value with a digital photograph inset (d). The zein/TA/PEG NPs were prepared with mass ratio zein/TA as 1 : 2.

Hydrogen bonding is sensitive to pH value, so we would like to study the effect of pH on zein/TA/PEG samples (Fig. 1c). The particle size and PDI of Z/TA₂/PEG NPs were high at pH 1.5, reaching 664.3 nm. At pH 4, larger NPs with a higher PDI were formed and the particles have tended to aggregate. However, it kept small particle size and PDI for Z/TA₂/PEG NPs in basic pH. The pH-response of NPs was linked to the hydrogen bonding between TA and PEG under various pH values. In acidic pH, hydrogen bonding strengthened, weakened in neutral pH and even vanished in basic pH. But in pH 1.5, extremes of pH, because of the stronger hydrogen bonding, the NPs seemed to aggregate, leading to a higher particle size, which could also see from the turbidity in Fig. 1d.

Furthermore, the effect of pH on particle aggregation was studied by measuring their zeta potential and turbidity varying in pH from 1.5 to 8. The zeta potential was positive at pH < 4 while negative at pH > 4. There was a very high variation of zeta potential at highly alkaline pH (pH 8). In pH 4, the NPs precipitated, leading to a low transmittance. In our experiments, we added TA and PEG, which influenced the surface potential of the NPs, leading to the precipitation of NPs at pH 4, not pH 6.⁴⁴

3.3. Characterizations of PEG/TA coated NPs

Fig. 2 showed the TEM images and typical size distribution profiles of zein/TA NPs and zein/TA/PEG NPs. As can be seen, TEM images revealed that zein/TA NPs without PEG coatings shared features of a near-spherical shape, but most of the particles were clumped and connected to each other (Fig. 2a–c). After zein/TA NPs were coated by PEG, it was possible to see individual NPs clearly with well-defined spherical shape (Fig. 2d–f). In addition, Z/TA₂/PEG NPs (Fig. 2g) and Z/TA₃/PEG NPs (Fig. 2h) have homogeneous distribution compared with Z/TA₁/PEG NPs (Fig. 2i).

Fig. 2 TEM image of zein/TA₁ NPs (a), zein/TA₂ NPs (b), zein/TA₃ NPs (c), zein/TA₁/PEG NPs at pH 7.4 (d), zein/TA₂/PEG NPs at pH 7.4 (e), and zein/TA₃/PEG NPs at pH 7.4 (f). Size distribution of zein/TA₁/PEG NPs at pH 7.4 (g), zein/TA₂/PEG NPs at pH 7.4 (h), and zein/TA₃/PEG NPs at pH 7.4 (i). Z/TA₁, Z/TA₂, Z/TA₃ represented formulations with different mass ratios of zein/TA as 1 : 1, 1 : 2, 1 : 3 respectively. Z/TA₁/PEG, Z/TA₂/PEG and Z/TA₃/PEG represented formulations with different mass ratios of zein/TA as 1 : 1, 1 : 2, 1 : 3 with PEG coating.

Then, FT-IR method was used to verify the existence of the specific intermolecular interactions between PEG, TA and zein components in the PEG/TA coated NPs. The representative spectra of zein, TA, PEG, zein/TA NPs and PEG/TA coated NPs were shown in Fig. 3. In the infrared spectra, a characterization peak was in the range of 3200–3400 cm⁻¹, indicating the hydrogen bonding.⁴⁵ The O–H stretching band of the hydroxyl groups in zein and TA was at 3420 cm⁻¹ and 3396 cm⁻¹, respectively, and shifted to 3374 cm⁻¹ and 3365 cm⁻¹ in zein/TA NPs and PEG/TA coated NPs respectively, suggesting the hydrogen bonding was formed between TA and zein, PEG and zein/TA. Compared with zein, the bands of amide I and amide II groups shifted to 1656 and 1535 cm⁻¹ respectively, in zein/TA and zein/TA/PEG samples, indicating the electrostatic interactions between zein and TA.⁶ Meanwhile the stretching band of the carbonyl groups in TA shifts from 1714 to 1720 cm⁻¹ in zein/TA/PEG sample. The result suggests the carbonyl groups in TA are originally bonded with hydroxyl groups via intramolecular hydrogen bonds.⁴⁶ The peak at 1110 cm⁻¹ is associated with of the C–O–C peak of PEG.⁴⁷ Seen from the FT-IR spectrum of zein/TA/PEG NPs, the C–O–C peak merged with the absorption bands of TA, suggesting again the formation of the existence of hydrogen bonding between the ether group of PEG and hydroxyl group of TA in zein/TA/PEG NPs.

Fig. 3 Fourier transform infrared spectroscopy (FT-IR) spectra of different samples. Z, zein powder; TA, tannic acid powder; PEG, poly(ethylene glycol) powder; zein/TA₂, zein/TA NPs prepared with mass ratio zein/TA as 1 : 2; zein/TA₂/PEG, PEG coated NPs prepared with mass ratio zein/TA as 1 : 2.

3.4. Release of TA from PEG/TA coated NPs

Hydrogen bonding was the main interaction between PEG and TA in NPs, which is reversible in nature, when immersed in different aqueous solutions, such as different pH value, different temperature or different ionic strength, the coating disintegrated gradually, thus releasing TA into the media.^17,18,48

To study the effect of pH on TA release, the samples were monitored under different pH conditions, i.e., pH 1.2, 7.4 and 8.0. As illustrated in Fig. 4a–c, at pH 1.2, all the samples showed a relative slower release profile in the release media and about 20%, 30%, 40% of the TA was released from Z/TA₁/PEG NPs (Fig. 4a), Z/TA₂/PEG NPs (Fig. 4b) and Z/TA₃/PEG NPs (Fig. 4c) respectively, after 12 h incubation. This was because phenolic hydroxyl groups were strongly hydrogen-bonded with PEG, as we described above. In contrast, at pH 7.4, the breakage of hydrogen bonds resulted in a fast release of TA. Release of TA was readily accelerated under basic condition. 85%, 90%, 95% of the TA was released from Z/TA₁/PEG NPs, Z/TA₂/PEG NPs and Z/TA₃/PEG NPs respectively, after 12 h incubation at pH 8. The faster TA release at higher pHs should be attributed to the increased dissociation degree of TA. Compared the release profile of Z/TA₁/PEG NPs, Z/TA₂/PEG NPs and Z/TA₃/PEG NPs, it was clear to see that more TA led to more apparent TA release tendency under different pH conditions.

Fig. 4 Release of TA from zein/TA₁/PEG NPs (a), zein/TA₂/PEG NPs (b), and zein/TA₃/PEG NPs (c) in 0.1 M buffers of various pHs. T = 37 °C.

TA is the weak acid (pK_a ≈ 10) and it is certain that the release of TA will be sensitive to pH.¹ Although TA ionization was suppressed by interaction with PEG, it remained pH sensitive.¹⁷ Bound to PEG, TA can still be reversibly ionized by varying solution pH. In neutral and basic pH, the dissociation of the phenolic hydroxyl groups weakened the hydrogen bonding between TA and PEG and hence increased the rate of TA released. In other words, the hydrogen bonds between TA and PEG largely remain intact at relatively lower pHs. As can be seen in the previously published research,⁴⁶ the release of TA from PVPON/TA films showed the same trend and the films disassembled under basic conditions, resulting from the breakage of the hydrogen bonds.

Next, we studied the effect of temperature on TA release. Here all the samples were immersed in 0.1 M pH 7.4 phosphate buffers at various temperatures. Seen from Fig. 5, the release rate of TA increased dramatically as temperature increased from 4 °C to 45 °C. This tendency was more noteworthy for Z/TA₃/PEG NPs (Fig. 5c). It can be seen that only about 10%, 15%, 20% of TA was released at 4 °C from Z/TA₁/PEG NPs (Fig. 5a), Z/TA₂/PEG NPs (Fig. 5b) and Z/TA₃/PEG NPs (Fig. 5c) respectively, into bulk within 12 h. The amount increases to 45%, 50%, 75% at 37 °C and even to 50%, 80% and 90% at 45 °C for Z/TA₁/PEG NPs, Z/TA₂/PEG NPs and Z/TA₃/PEG NPs respectively. As temperature increased, the hydrogen bonding between PEG and TA weakened leading to the dissociation of the hydrogen bonds and release of TA at a faster rate. This was in agreement with previous studies.^49,50

Fig. 5 Release of TA from zein/TA₁/PEG NPs (a), zein/TA₂/PEG NPs (b), and zein/TA₃/PEG NPs (c) in 0.1 M phosphate buffer (pH 7.4) of various temperatures. Release of TA from zein/TA₂/PEG NPs in 0.1 M phosphate buffer (pH 7.4) of various concentrations of NaCl (d). T = 37 °C.

Then we explored the effect of ionic strength on the release property of Z/TA₂/PEG NPs. The sample was soaked in 0.1 M pH 7.4 phosphate buffers containing various amounts of NaCl at 37 °C. Compared with the release rate in control sample (NaCl-free), the release rate decreased dramatically when the [NaCl] is 100 mM. The added NaCl may screen the electrostatic repulsions among the charged TA molecules, resulting in a reduced TA release. In fact, there existed two tendencies in this system. First, the electrostatic screening effect and the decrease of solubility of both TA and PEG with increasing ionic strength may all delay the release of TA. Second, the added NaCl may also enhance the dissociation of the phenolic hydroxyl groups in TA, leading to the breakage of more hydrogen bonds and therefore a faster TA release. When [NaCl] is 100 mM, this resulted in a significantly depressed TA release, suggesting the prevailing effect of salt in this case was to delay the release of TA. If further increased the amount of TA, the change in the release rate was negligible.

3.5. Antioxidant activity of the released TA

Free radicals, which, if overproduced, indiscriminately attacks on DNA, proteins, and other important biomolecules and they were associated with many degenerative conditions including the aging process.^39,51 TA, a kind of polyphenols, can scavenge free radicals effectively and thus provide protection to the body.^52,53 Antioxidant activity of the NPs were evaluated by determining the scavenging activities of DPPH˙ radical in ethanol system and ABTS⁺˙ cation radicals in aqueous solution system.

We employed an intuitive method to confirm the antioxidant activity of the NPs. The DPPH˙ radicals or ABTS⁺˙ cation radicals were immobilized onto macroporous adsorption resin. After adsorbed the DPPH˙ radicals or ABTS⁺˙ cation radicals, the originally white resin particles turned purple/blue. For DPPH radical scavenging activity assay, the resin particles were then soaked in ethanol together with the NPs. While the resin particles were soaked in phosphate buffer (pH = 7.4), after the adsorption of the ABTS˙⁺ radical cation radicals. As shown in Fig. 6a and b, the resin particles were decolorized gradually from bottom to top. But in ethanol solution, the speed of change in color was faster, which indicated the faster release of TA in ethanol (Fig. 6a). Synthesize the above results, we known that the released TA could still react with the radical, indicating it remained active as an antioxidant to scavenge free radicals not only in ethanol system but in aqueous solution system.

Fig. 6 Decolorization of DPPH˙-loaded (a) and ABTS˙⁺-loaded (b) macroporous adsorption resin with the lyophilized zein/TA₂/PEG NPs in the bottom. T = 25 °C.

3.6. Characterization of VD3-encapsulated NPs

VD3 is a kind of lipophilic functional component to human health. Unfortunately, it is not soluble in water and very sensitive to various environmental factors i.e., light, heat, and oxygen which could rapidly induce isomerization or oxidation of VD3 and then adversely affect its physiological benefits.^45,54 Because zein NPs were promising drug deliver systems for hydrophobic molecules, VD3 could be loaded through interaction with zein by hydrophobic interactions, providing remarkable protection of this compound.⁴⁴

As illustrated in Table 2, the particle size of VD3-encapsulated NPs varied with the weight ratios of zein and TA. The smallest mean particle diameter (221.3 nm) was obtained and the encapsulation efficiency (EE) reached 85.6% at a zein/TA ratio of 1 : 2. The surface charge of the resulting NPs was TA-concentration-dependent and ranged from −41.6 to −52.3 mV. The TEM images revealed that the NPs were dispersed as individual NPs with well-defined spherical shape (Fig. 7a–c). The addition of VD3 did not instigate morphological changed in the NPs.

Table 2 Particle size, polydispersity index (PDI), zeta potential and encapsulation efficiency (EE) of NPs in different formulations^a

Samples	Particle size	PDI	Zeta potential (mV)	EE (%)
a Z, zein; TA, tannic acid; PEG, poly(ethylene glycol); VD3, vitamin D3. Z/TA1/PEG (VD3 pH 7.4), Z/TA2/PEG (VD3 pH 7.4) and Z/TA3/PEG (VD3 pH 7.4) represented formulations with different mass ratios of zein/TA as 1 : 1, 1 : 2, 1 : 3 with PEG coatings VD3 encapsulated NPs and the pH of the solutions was 7.4.
Z/TA1/PEG (VD3 pH 7.4)	247.8 ± 4.4	0.28 ± 0.01	−41.6 ± 0.01	81.8 ± 1.4
Z/TA2/PEG (VD3 pH 7.4)	221.3 ± 3.2	0.31 ± 0.05	−51.2 ± 0.05	85.6 ± 1.5
Z/TA3/PEG (VD3 pH 7.4)	275.7 ± 7.8	0.28 ± 0.01	−52.3 ± 0.01	88.6 ± 0.2

---

Fig. 7 TEM images of VD3 loaded zein/TA₁/PEG NPs (a), zein/TA₂/PEG NPs (b), and zein/TA₃/PEG NPs (c) at pH 7.4. XRD patterns of different samples (d). Z, zein powder; TA, tannic acid powder; PEG, poly(ethylene glycol) powder; VD3, vitamin D3 powder; Z/TA₂/PEG, PEG coated NPs prepared with mass ratio zein/TA as 1 : 2. Z/TA₂/PEG (VD3), VD3 loaded PEG coated NPs prepared with mass ratio zein/TA as 1 : 2. Photochemical stability of different samples against UV light (e). Control, VD3 dispersion in water; Z/TA₁/PEG, Z/TA₂/PEG and Z/TA₃/PEG represented formulations with different mass ratios of zein/TA as 1 : 1, 1 : 2, 1 : 3 with PEG coating. Kinetic release of VD3 from Z/TA/PEG NPs in pH 1.2 media (f) and pH 7.4 media (g). Release of VD3 in SGF and SIF from Z/TA/PEG NPs (h).

The XRD patterns of the NPs and pure VD3 were shown in Fig. 7d. The peaks of VD3 indicated the highly crystalline nature. In contrast, zein showed two flatter humps instead of sharp peaks, indicating the amorphous nature of the protein.^55,56 VD3 specific peaks disappeared in all the formations of suggesting that VD3 in the NPs or the complex did not manifest in crystal form which provided additional evidence of encapsulation.

3.7. Photochemical stability against UV light

As shown in Fig. 7e, the control sample VD3 underwent a photochemical degradation very quickly when exposed to 352 nm UV light. It was observed that 20% of control sample VD3 remained after 8 h of UV light exposure. Compared with control sample at the same treatments, all of the NPs were able to provide great protection again UV light-induced degradation, with >75% of VD3 remaining in samples after 8 h of UV light exposure. One possible protection mechanism of Z/TA/PEG NPs against photochemical degradation of VD3 was that TA was an effective photo chemopreventive agent protecting the VD3 against adverse effects of UV radiation.^57–59 In addition, proteins with aromatic side groups and double bonds can absorb UV light and hence reduce the absorption of UV light by VD3.⁴⁵

3.8. Release profiles

The releasing profiles of VD3-loaded NPs were depicted in Fig. 7. At pH 1.2, PEG/TA coatings were strongly hydrogen-bonded under this condition, resulting in a slow and sustained release of VD3 for all samples (Fig. 7f). In contrast, the coating disintegrated quickly when soaked in aqueous solutions with pH 7.4 (Fig. 7g). In this case, the coatings disassembly via the breakage of hydrogen bonds, as we showed above, should be attributed to a fast VD3 release.^60,61

Under the SGI condition with digestive enzymes, after incubation at 37 °C in SGF for 30 min, over 45% of the VD3 was released from the Z/TA₁/PEG NP. In contrast, 30% and even 20% of the VD3 was detected in the releasing medium in Z/TA₂/PEG NPs and Z/TA₃/PEG NPs samples. The release profile of VD3 in gastric fluid was significantly decreased after more TA was added into the system. When transferred to the SIF (pH 7.4), at most 95% of the remaining VD3 was released from the polymeric matrix within 6 h for Z/TA₁/PEG NPs and Z/TA₂/PEG NPs samples. In addition, among the three tested complex formulas, the one with a zein : TA mass ratio of 1 : 3 resulted in a least amount of released VD3. This was probably attributed to the greater number of particles achieved by this formula compared with the other two, and more TA was hydrogen bonded with PEG, leading to a thicker coating.

4. Conclusion

In this work, we achieved the sustained release of TA, a model polyphenolic drug, from hydrogen-bonded PEG/TA coating on zein/TA colloidal particles. We exploited there reversible/dynamic nature of the hydrogen-bonded coatings for sustained drug delivery by exposing to external stimuli. The disassembly of PEG/TA coatings was controlled by changing the pH value, temperature, ionic strength and TA concentration. Additionally, the released TA retained its ability to scavenge harmful radicals. In the FT-IR spectrum, a wavenumber shift related to the hydroxyl groups (–OH) and the carbonyl groups (C=O) in TA was observed revealing that complex formation was conducted by hydrogen bond interactions.

This new drug delivery vehicle could also be used to load hydrophobic molecules into the core which could interact with zein by hydrophobic interactions. The PEG/TA coatings could protect the loadings against various environmental factors. In addition, the loadings release rate can be tuned by both temperature and pH because the hydrogen bonded coating was highly sensitive to these external stimuli.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (No. 31371841). The authors would like to express their sincere gratitude to many conveniences offered by colleagues of Key Laboratory of Environment Correlative Dietology of Huazhong Agricultural University.

References

1. V. Kozlovskaya, S. Harbaugh, I. Drachuk, O. Shchepelina, N. Kelley-Loughnane, M. Stone and V. V. Tsukruk, Soft Matter, 2011, 7, 2364–2372.
2. Y. Teramura, Y. Kaneda and H. Iwata, Biomaterials, 2007, 28, 4818–4825.
3. P. R. LeDuc, M. S. Wong, P. M. Ferreira, R. E. Groff, K. Haslinger, M. P. Koonce, W. Y. Lee, J. C. Love, J. A. McCammon and N. A. Monteiro-Riviere, Nat. Nanotechnol., 2007, 2, 3–7.
4. 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.
5. J. Guo, Y. Ping, H. Ejima, K. Alt, M. Meissner, J. J. Richardson, Y. Yan, K. Peter, D. von Elverfeldt, C. E. Hagemeyer and F. Caruso, Angew. Chem., Int. Ed. Engl., 2014, 53, 5546–5551.
6. H. Zheng, L. Xing, Y. Cao and S. Che, Coord. Chem. Rev., 2013, 257, 1933–1944.
7. Y. Ping, J. Guo, H. Ejima, X. Chen, J. J. Richardson, H. Sun and F. Caruso, Small, 2015, 17, 2032–2036.
8. S. W. Choi, Y. Zhang and Y. Xia, Angew. Chem., 2010, 49, 7904–7908.
9. Z. Zou, D. He, X. He, K. Wang, X. Yang, Z. Qing and Q. Zhou, Langmuir, 2013, 29, 12804–12810.
10. D. Li, J. He, W. Cheng, Y. Wu, Z. Hu, H. Tian and Y. Huang, J. Mater. Chem. B, 2014, 2, 6089–6096.
11. H. Li, Y. Guo, L. Xiao and B. Chen, Biosens. Bioelectron., 2014, 59, 289–292.
12. L. Zhou, M. Chen, L. Tian, Y. Guan and Y. Zhang, ACS Appl. Mater. Interfaces, 2013, 5, 3541–3548.
13. X. Chen, W. Wu, Z. Guo, J. Xin and J. Li, Biomaterials, 2011, 32, 1759–1766.
14. R. M. Flessner, C. M. Jewell, D. G. Anderson and D. M. Lynn, Langmuir, 2011, 27, 7868–7876.
15. J. Blacklock, H. Handa, D. S. Manickam, G. Mao, A. Mukhopadhyay and D. Oupický, Biomaterials, 2007, 28, 117–124.
16. I. Erel-Unal and S. A. Sukhishvili, Macromolecules, 2008, 41, 3962–3970.
17. F. Liu, V. Kozlovskaya, O. Zavgorodnya, C. Martinez-Lopez, S. Catledge and E. Kharlampieva, Soft Matter, 2014, 10, 9237–9247.
18. E. Costa, M. Coelho, L. M. Ilharco, A. Aguiar-Ricardo and P. T. Hammond, Macromolecules, 2011, 44, 612–621.
19. T. S. Sileika, D. G. Barrett, R. Zhang, K. H. Lau and P. B. Messersmith, Angew. Chem., Int. Ed. Engl., 2013, 52, 10766–10770.
20. M. Akagawa and K. Suyama, Eur. J. Biochem., 2001, 268, 1953–1963.
21. T. G. Shutava, S. S. Balkundi, P. Vangala, J. J. Steffan, R. L. Bigelow, J. A. Cardelli, D. P. O'Neal and Y. M. Lvov, ACS Nano, 2009, 3, 1877–1885.
22. M. Daglia, Curr. Opin. Biotechnol., 2012, 23, 174–181.
23. K. T. Chung, T. Y. Wong, C. I. Wei, Y. W. Huang and Y. Lin, Crit. Rev. Food Sci. Nutr., 1998, 38, 421–464.
24. T. Shutava, M. Prouty, D. Kommireddy and Y. Lvov, Macromolecules, 2005, 38, 2850–2858.
25. V. Kozlovskaya, E. Kharlampieva, I. Drachuk, D. Cheng and V. V. Tsukruk, Soft Matter, 2010, 6, 3596.
26. K. C. Yen and E. M. Woo, Polym. Bull., 2008, 62, 225–235.
27. B. S. Kim, H. i. Lee, Y. Min, Z. Poon and P. T. Hammond, Chem. Commun., 2009, 28, 4194–4196.
28. J. P. Van Buren and W. B. Robinson, J. Agric. Food Chem., 1969, 17, 772–777.
29. Y. Zou, J. Guo, S. W. Yin, J. M. Wang and X. Q. Yang, J. Agric. Food Chem., 2015, 63, 7405–7414.
30. T. Zou, Z. Li, S. S. Percival, S. Bonard and L. Gu, Food Hydrocolloids, 2012, 27, 293–300.
31. H. Liang, B. Zhou, J. Li, W. Xu, S. Liu, Y. Li, Y. Chen and B. Li, Colloids Surf., B, 2015, 136, 1224–1233.
32. A. Shukla, J. C. Fang, S. Puranam, F. R. Jensen and P. T. Hammond, Adv. Mater., 2012, 24, 492–496.
33. J. A. Hiller and M. F. Rubner, Macromolecules, 2003, 36, 4078–4083.
34. Y. N. Zhao, J. Gu, S. Jia, Y. Guan and Y. Zhang, Soft Matter, 2016, 12, 1085–1092.
35. C. Tang, D. Amin, P. B. Messersmith, J. E. Anthony and R. K. Prud'homme, Langmuir, 2015, 31, 3612–3620.
36. P. Caliceti and F. M. Veronese, Adv. Drug Delivery Rev., 2003, 55, 1261–1277.
37. S. N. Alconcel, A. S. Baas and H. D. Maynard, Polym. Chem., 2011, 2, 1442–1448.
38. 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.
39. B. Zhou, X. Jin, J. Li, W. Xu, S. Liu, Y. Li and B. Li, RSC Adv., 2014, 4, 54517–54524.
40. Y. N. Zhao, J. Gu, S. Jia, Y. Guan and Y. Zhang, Soft Matter, 2016, 12, 1085–1092.
41. Y. Luo, T. T. Wang, Z. Teng, P. Chen, J. Sun and Q. Wang, Food Chem., 2013, 139, 224–230.
42. K. C. Yen, T. K. Mandal and E. M. Woo, J. Biomed. Mater. Res., Part A, 2008, 86, 701–712.
43. S. Podaralla, R. Averineni, M. Alqahtani and O. Perumal, Mol. Pharmaceutics, 2012, 9, 2778–2786.
44. H. Liang, B. Zhou, L. He, Y. An, L. Lin, Y. Li, S. Liu, Y. Chen and B. Li, RSC Adv., 2015, 5, 13891–13900.
45. Y. Luo, Z. Teng and Q. Wang, J. Agric. Food Chem., 2012, 60, 836–843.
46. L. Zhou, M. Chen, L. Tian, Y. Guan and Y. Zhang, ACS Appl. Mater. Interfaces, 2013, 5, 3541–3548.
47. K. C. Yen and E. M. Woo, Polym. Bull., 2009, 62, 225–235.
48. I. Erel, H. Schlaad and A. L. Demirel, J. Colloid Interface Sci., 2011, 361, 477–482.
49. W. Zhang, A. Zhang, Y. Guan, Y. Zhang and X. Zhu, J. Mater. Chem., 2011, 21, 548–555.
50. S. Yang, Y. Zhang, G. Yuan, X. Zhang and J. Xu, Macromolecules, 2004, 37, 10059–10062.
51. B. Nemzer, T. Chang, Z. Xie, Z. Pietrzkowski, T. Reyes and B. Ou, Food Sci. Nutr., 2014, 2, 647–654.
52. V. Kozlovskaya, B. Xue, W. Lei, L. E. Padgett, H. M. Tse and E. Kharlampieva, Adv. Healthcare Mater., 2015, 4, 686–694.
53. W. Y. Huang, Y. Z. Cai and Y. Zhang, Nutr. Cancer, 2009, 62, 1–20.
54. J. M. Ballard, L. Zhu, E. D. Nelson and R. A. Seburg, J. Pharm. Biomed. Anal., 2007, 43, 142–150.
55. Y. Luo, T. T. Wang, Z. Teng, P. Chen, J. Sun and Q. Wang, Food Chem., 2013, 139, 224–230.
56. H. Liang, Q. Huang, B. Zhou, L. He, L. Lin, Y. An, Y. Li, S. Liu, Y. Chen and B. Li, J. Mater. Chem. B, 2015, 3, 3242–3253.
57. J. Y. Bae, J. S. Choi, S. W. Kang, Y. J. Lee, J. Park and Y. H. Kang, Exp. Dermatol., 2010, 19, e182–e190.
58. J. Y. Bae, J. S. Choi, Y. J. Choi, S. Y. Shin, S. W. Kang, S. J. Han and Y. H. Kang, Food Chem. Toxicol., 2008, 46, 1298–1307.
59. F. Afaq and H. Mukhtar, Exp. Dermatol., 2006, 15, 678–684.
60. S. E. Burke and C. J. Barrett, Langmuir, 2003, 19, 3297–3303.
61. S. A. Sukhishvili and S. Granick, Macromolecules, 2002, 35, 301–310.

---