Improving quercetin dissolution and bioaccessibility with reduced crystallite sizes through media milling technique

Journal of Functional Foods 37 (2017) 138–146

Contents lists available at ScienceDirect

Journal of Functional Foods journal homepage: www.elsevier.com/locate/jff

Improving quercetin dissolution and bioaccessibility with reduced crystallite sizes through media milling technique Muwen Lu, Chi-Tang Ho, Qingrong Huang ⇑ Department of Food Science, Rutgers University, 65 Dudley Road, New Brunswick, NJ 08901, USA

a r t i c l e

i n f o

Article history: Received 30 April 2017 Received in revised form 22 July 2017 Accepted 22 July 2017 Available online 20 September 2017 Keywords: Quercetin Nanoparticle Wet milling Dissolution Bioaccessibility

a b s t r a c t Quercetin (QC) is a common bioflavonoid with low water solubility, which limits its oral bioavailability and in vivo beneficial functions. To enhance QC bioaccessibility, QC nanoparticles were produced using the media milling technology. Hydrophobically modified starch (HMS) was added with the ratio of 1:1 to QC as a stabilizer to prevent the agglomeration of QC particles. The QC nanodispersions were either spray-dried or freeze-dried after media milling process. Physicochemical characteristics of dried QC powders were measured through dynamic light scattering (DLS), Fourier Transform-Infrared spectroscopy (FTIR), dissolution test and X-ray diffraction (XRD). The detailed crystallite structures were carefully analyzed. The TNO dynamic gastro-intestinal model-1 (TIM-1) was utilized to study the in vitro QC bioaccessibility by simulation of the digestive processes in the upper GI tract. This study suggests that media milling technique combined with spray/freeze-drying treatment is an efficient processing method for the development of crystalline nutraceuticals-based functional food products with reduced crystallite sizes and enhanced metastable equilibrium solubility, dissolution and bioaccessibility. Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction Flavonoids were polyphenolic compounds rich in fruits, vegetables and plant-derived beverages (Jacobs et al., 2011). Quercetin (QC, 3,30 ,40 ,5,7-pentahydroxyflavone) as one of the abundantly consumed flavonoids is widely distributed in onions, tea, apples and red wine and has many beneficial bioactivities, including antioxidative (Azuma, Ippoushi, & Terao, 2010), anticarcinogenic (Ramos, Lima, Pereira, Fernandes-Ferreira, & Pereira-Wilson, 2008), anti-arthritic (Jeyadevi et al., 2013), anti-inflammatory and anti-atherosclerotic effects (Kleemann et al., 2011). Recent researches have demonstrated its anti-proliferative effects in multiple human cancer cell types, such as lung (Lee, Han, Yun, & Kim, 2015), prostate (Paller et al., 2015), colon (Lu et al., 2015), gastric

Abbreviations: ATR, attenuated total reflection; CSD, Cambridge Structural Database; DI water, deionized water; DLS, dynamic light scattering; DMSO, dimethyl sulfoxide; FTIR spectroscopy, Fourier Transform-Infrared spectroscopy; GI tract, gastrointestinal tract; HMS, Hydrophobically modified starch; HPLC, high performance liquid chromatography; n-OSA, n-octenyl succinic anhydride; QC, quercetin; UV–vis, ultravioletvisible; XRD, X-ray diffraction; TIM-1, TNO gastrointestinal model-1. ⇑ Corresponding author. E-mail address: [email protected] (Q. Huang). http://dx.doi.org/10.1016/j.jff.2017.07.047 1756-4646/Ó 2017 Elsevier Ltd. All rights reserved.

(Du, Feng, Fang, & Yang, 2015), breast (Ruotolo et al., 2014), cervical (Lin et al., 2015), and pancreatic cancer cells (Lee et al., 2015). However, the poor solubility of QC in water (0.17–7 lg/mL), artificial gastric juice (5.5 lg/mL), and artificial intestinal juice (28.9 lg/mL) greatly limits its bioavailability after oral administration (Gao et al., 2009). Bioavailability studies of QC in humans revealed that the mean plasma values ranged between 15 and 24 lg/L in subjects consuming their habitual diets (Erlund et al., 2002; Freese et al., 2002), and was 42 lg/L after a diet containing 815 g/d of vegetables, fruits, and berries (Erlund, 2004). The urinary excretion of QC was also investigated in several studies, which showed that urinary recovery was lower than 3% (Erlund, 2004). Therefore, methods that can increase the solubility and bioavailability of QC are highly desired to better perform its beneficial bioactivities in vivo. In order to improve the oral bioavailability of QC, various delivery systems, such as nanodispersions (Karadag, Ozcelik, & Huang, 2014; Karadag, Yang, Ozcelik, & Huang, 2013), liposomes (Wong & Chiu, 2010), polymeric micelles (Dian et al., 2014), nanoemulsions (Dario et al., 2015), microemulsions (Vicentini, Vaz, Fonseca, Bentley, & Fonseca, 2011), solid-lipid nanoparticles (Li et al., 2009) and self-emulsifying drug delivery systems (Ting, Jiang, Ho, & Huang, 2014) have been developed and evaluated.

M. Lu et al. / Journal of Functional Foods 37 (2017) 138–146

Among all delivery systems, nanodispersions have been confirmed as one of the most efficient systems to increase the gastric lumen solubility and dissolution rate of QC (Shegokar & Muller, 2010), resulting in a higher membrane concentration gradient and better intestinal uptakes. Moreover, for materials with relatively simple compositions, nanodispersions can be transformed into powder after spray-drying processing, prolonging the storage time for nutraceuticals. For the production of nanoparticles, technologies such as media milling technique, precipitation, spray/freeze dry, supercritical antisolvent technique, high-pressure homogenization, etc., have been applied to improve metastable equilibrium solubility, bioaccessibility and bioavailability of nutraceuticals (Chattopadhyay & Gupta, 2001; Chen & Wang, 2007; Kakran et al., 2010; Sahoo et al., 2011). According to Kakran (Kakran et al., 2012), QC nanocrystals were prepared by three processing methods, including highpressure homogenization, wet bead milling, and caviprecipitation. The smallest nanocrystals were fabricated by wet bead milling method stabilized by Tween 80, which had mean particle sizes of 276.7 ± 0.3 nm and saturation solubility of 25.59 ± 1.11 lg/mL, about nine times higher than coarse QC. Furthermore, the zeta potential value and stability of the QC nanosuspensions in distilled water were the highest for wetmilled particles among all produced nanocrystals. By reducing the particle sizes through media milling processing, the surface area of crystals was increased, resulting in faster dissolution rates compared with the original coarse QC. Niwa (Niwa & Danjo, 2013) reported that the technique of media milling and spray & freeze-drying processes could enhance the dissolution for poorly water-soluble drugs. Moreover, media milling technique could also produce uniformed particle sizes of nanocrystals and prevent the agglomeration and precipitations of processed nanoparticles by providing a liquid environment (Jung, Sohn, Sung, Hyun, & Shin, 2015). Therefore, the media milling technique was adopted in this research to reduce the particle sizes and increase the dissolution rate of QC. Until now, the exact mechanism of how milled quercetin particles improved their water solubility remains unclear. In this paper, hydrophobically modified starch (HMS) from waxy maize and n-octenyl succinic anhydride (n-OSA), which has been widely used as wall material to encapsulate flavors in spraydrying process (Shaikh, Bhosale, & Singhal, 2006), was added as a stabilizer to prepare the QC formulation due to its hydrophobic property. By adsorbing around the QC particles, HMS could function as steric barriers to prevent the agglomeration of the primary particles prepared by ground mixture effectively. The combined treatment of media milling and spray/freeze-drying method was applied in the development of QC nanoparticles. The produced QC nanoparticles in powdered form could be immediately re-dispersed in water. The X-ray diffraction profiles of QC before and after media milling process were analyzed, and their crystallite sizes were calculated for the first time. Physiochemical properties, including metastable equilibrium solubility, dissolution rate and bioaccessibility of formulated QC nanocrystals were also characterized.

139

Aldrich Company (St. Louis, MO, USA). Milli-Q water (18.3 MX) was used in all experiments. 2.2. Preparation of formulated QC dispersions QC formulations were prepared by dispersing 15 g of QC dihydrate and 15 g of HMS in 300 mL deionized water. The suspension was under agitation overnight at room temperature. 2.3. Media milling processing The formulated QC suspension was fed to a milling chamber before being milled. The media-milling machine (MiniCer, NETZSCH-Feinmahltechnik GmbH, Staufen, Germany) was used in this research to disintegrate the microsized QC crystals into nanoparticles. During the milling process, Yttrium-stabilized zirconium oxide grinding beads with a diameter of 0.8–1.0 mm performed as the milling media. As the mill started running, the formulation was under agitation of grinding beads at the speed of 3100 rpm and recycled by a pump at the speed of 100 rpm. The milling temperature was controlled at 4 °C by a cooling system. Samples were collected every 5 min for particle size analysis. 2.4. Spray drying processing 150 mL of wet-milled QC nanodispersions were spray dried in the model Pulvis GB22 fluidized bed spray dryer (Yamato, Santa Clara, CA). The dryer conditions were: drying air flow 0.43– 0.45 m3/min, inlet temperature 110 °C; outlet temperature 200 °C; air pressure 6 bar; aspirator 100% (Sansone et al., 2011). All spray-dried QC powders were then collected from the filter chamber and stored at 4 °C. 2.5. Freeze drying processing The remaining 150 mL of media milled QC nanodispersions were frozen overnight and freeze-dried using the freeze dryer (Freezone 4.5, Labconco, Kansas City, MO, USA), open to the system and under vacuum until all moisture was removed and a dry solid mass remained (Pralhad & Rajendrakumar, 2004). Then dried powders were stored in 50 mL sterile centrifuge tubes at 4 °C until further evaluation. 2.6. Particle size measurements The particle size of the wet-milled QC dispersions were measure by dynamic light scattering based Particle Size Analyzer (Model 90 Plus, Brookhaven Instrument Corp., Holtsville, NY, USA) at a fixed scattering angle of 90° and temperature of 25.0 °C. Each time, 5 lL of dispersed samples were collected from the recirculation chamber during the media milling process within a certain time interval and diluted by 5 mL distilled water (diluted by 1000 times). All samples were measured in triplicate to ensure accuracy. 2.7. Fourier-transform Infrared spectroscopy (FT-IR)

2. Materials and methods 2.1. Materials QC dihydrate (purity >90%) was purchased from VWR Scientific (Seattle, WA, USA). HMS with the brand name of Hi-Cap 100 was obtained from National Starch and Chemical Company (Bridgewater, NJ, USA). High performance liquid chromatography (HPLC) grade-methanol and acetonitrile were purchased from Sigma-

Infrared absorption spectra of pure QC, HMS and formulated QC powders were recorded using a Thermo Nicolet Nexus 670 FT-IR system with attenuated total reflection set-up (ATR-FTIR) (Thermo Fisher Scientific, Waltham, MA, USA) (Gagos, 2008). Absorption spectra at a resolution of one data point every 4 cm1 were obtained in the region between 4000 and 600 cm1 using the air as the background. 512 repeated scans for each sample were collected. All experiments were carried out at room temperature. The absorption spectra were averaged and smoothed, and their

140

M. Lu et al. / Journal of Functional Foods 37 (2017) 138–146

baselines were calibrated with the Spectra Manager software. The ATR crystals were cleaned with organic solvent (acetone). Measurements were carried out in triplicate. 2.8. Loading capacity 20 mg of QC-loaded formulation powders were dissolved in 10 ml water and diluted by 10 times. Then the saturated dispersion was filtered by a 0.45 lm filter and the concentration was measured using HPLC at the wavelength of 370 nm. Measurements were carried out in triplicate. Results were used to calculate the QC loading capacity according to the following formulae (Hao et al., 2013):

% Loading Capacity ¼

Total mass of QC in the formulation Total mass of the formulation  100%:

2.9. Solubility studies The metastable equilibrium solubility of QC formulations in Milli-Q water was determined by using an incubating orbital shaker (VWR Scientific, Cornelius, OR, USA). Saturated dispersion of QC formulation was prepared by dispersing an excessive amount of formulated QC powders in water. Then solutions were collected in 40 mL capped vials to avoid evaporation and shaken at 25 °C for 24 h. After the equilibrium was reached, suspensions were filtered through a 0.45 lm filter and analyzed by HPLC. Experiments were carried out in triplicate. 2.10. Dissolution studies The dissolution rates of pure QC dihydrate powders, wet-milled pure QC dihydrate powders, QC/HMS (1:1) physical mixtures and wet-milled QC formulations were measured using flow-through cell USP apparatus 4 (SOTAX Corporation, Westborough, MA, USA). 10 mg of powdered samples were introduced into the dissolution medium containing 400 mL of phosphate buffer solution at 37 ± 0.5 °C and a speed of 3 mL/min. Samples were collected from the reservoir every 5 min within an hour and analyzed through HPLC at 370 nm. All tests were carried out in triplicate. 2.11. X-ray diffraction on powder (XRD) The X-ray diffraction patterns of pure QC, wet-milled pure QC, spray-dried QC and freeze-dried QC were analyzed using a D/M2200T X-ray powder diffractometer (Ultima+, Rigaku) with nickel filtered Cu Ka radiation (k = 1.54056 Å). The scanned radiation was detected in the angular range of 5–40° (2h) with a step of 0.02° (2h) per second. Graphite monochromator was used and the generator power settings were at 40 kV and 40 mA (Wang, Gao, Chen, & Xiao, 2006). 2.12. TNO gastro-intestinal model-1 (TIM-1) The dynamic, computer-controlled in vitro TNO gastrointestinal model-1 (TIM-1) (TNO, Zeist, The Netherlands) was used to mimic the human upper gastrointestinal (GI) tract using four compartments, which are stomach, duodenum, jejunum, and ileum (Havenaar et al., 2013; Ribnicky et al., 2014). Water was heated to 37 ± 1 °C in order to maintain body temperature. PH conditions and secretion of digestive fluids were also controlled by computer programs to resemble the human physiological GI conditions. In this study, samples were fed into the system and tested for at least 5 h. Suspensions containing same content of QC (0.1%) were

prepared by dissolving coarse QC and formulated QC powders into the DI water respectively. The QC bioaccessibility was analyzed by collecting dialysates from jejunal and ileal filtrates at 30, 60, 90, 120, 180, 240 and 300 min. Bioaccessible QC nanoparticles were able to pass through the hollow fiber filtration device (Spectrum Milikros modules M80S-300-01P, with 0.05 lm pore size) and got absorbed in the jejunum and ileum. In the meantime, ileal efflux samples were also collected, which represented the percentage of compound that would theoretically be delivered to the colon. Samples obtained from TIM-1 were immediately stored at 20 °C until HPLC analysis. The experiments were performed in duplicate and samples were analyzed in triplicate. For HPLC analysis, 400 lL of sample was inoculated with an internal standard (Morin, 10 lg/mL) was then extracted by mixing with 600 lL of ethyl acetate and centrifuge at 16000g for 30 min at room temperature. After centrifuge, the 400 lL of supernatant was obtained and mixed with equal amount of dimethyl sulfoxide (DMSO) for use in HPLC analysis. The percent cumulative bioaccessibility of QC was calculated using equation below:

%Bioaccessibility ¼

Total mass of solubilized QC  100% Total mass of QC in original samples

2.13. Quantitative analysis of QC using high-performance liquid chromatography (HPLC) QC concentration was quantified by an automated highperformance liquid chromatograph system (Dionex, Sunnyvale. CA, USA). The system consists of a quaternary solvent delivery system, an ultraviolet–visible (UV–vis) diode array detector and an automated injection system. The column used in this study is Supelco’s RP-Amide C18, 15 cm  64.6 mm, 3 lm (Bellefonte, PA, USA). The mobile phase consists of (A) acetonitrile and (B) 0.2% acetic acid in water (HPLC grade). 10 lL samples were injected each time and then eluted under gradient conditions: 0–2 min, 40% A and 60% B; 2–8 min, linear gradient from 40 to 55% A; 9– 12 min, linear gradient from 55 to 70% A; 12–13 min, linear gradient from 70 to 90% A; 14–6 min, held at 90% A; 17–18 min, A went back to 40% linearly; 19–20 min, held at 40% to balance the column. The flow rate was set at 1.0 mL/min and the eluent was detected with UV wavelength at 370 nm. Different standard concentrations (5, 10, 15, 20, 40, 60, 80 and 100 lg/mL) of QC were dissolved in methanol and analyzed using the HPLC to generate the calibration curve. Each measurement was carried out in triplicate. 3. Results and discussion 3.1. Effect of the media milling process on particle sizes of QC In this research, the media milling technique was used to produce QC nanodispersions. According to Fig. 1, the average particle size decreased from 1.1 ± 0.2 lm to 340 ± 20 nm after the processing. The reduction of the particle size was mainly due to the intense shear forces during the media milling process. Apart from that, the modified starch also functioned as the stabilizer to prevent the agglomeration of QC particles due to its hydrophobic property. In 2008, Pongpeerapat proposed a formation mechanism of colloidal nanoparticles, suggesting that the adsorption behavior between two hydrophobic molecules helped the stabilization of the ground mixture (Pongpeerapat, Wanawongthai, Tozuka, Moribe, & Yamamoto, 2008). Therefore, the HMS would adsorb to the hydrophobic surface of dispersed QC nanoparticles, reducing the contact between QC themselves. Therefore, the HMS might effec-

M. Lu et al. / Journal of Functional Foods 37 (2017) 138–146

141

would be no covalent interactions between the pure QC and HMS powders. These results indicated that neither media milling process nor addition of modified starch would significantly influence the chemical properties of QC. 3.3. Loading capacity and solubility studies The excess amount of the media-milled QC powder was dispersed into distilled water and gently agitated for 1 day at room temperature to reach full saturation. According to the HPLC analysis, the water dispersity of the formulated QC was 28.8 ± 0.3 lg/ mL. Compared with the pure QC’s water solubility of 2.6 lg/mL at 25 °C (Srinivas, King, Howard, & Monrad, 2010), this formulation can enhance the water solubility by approximately eleven times. Besides, the loading capacity of the formulation powder was 48.7% after calculation, which theoretically should be 50%. This result demonstrated that QC and HMS were well mixed, and there was negligible loss of QC after processing. 3.4. Dissolution studies Fig. 1. Change in particle size and polydispersity of formulated QC dispersion as a function of time.

The dissolution rate of pure QC powders, wet-milled pure QC powders, physical mixtures of QC and HMS with the ratio of 1:1, and the formulated QC powders are shown in Fig. 3. The wetmilled QC referred to powders that only went through the media milling and drying process without the addition of HMS. For the dissolution processes of different solutions, only about 1.1% of the coarse QC was dissolved after one hour. While the dissolution rate of wet-milled pure QC was stable around 5.4%, which was slightly higher than the original one. Compared with pure QC, both physical mixtures and formulated powders had higher dissolution rates of 36.9% and 98.8%, respectively. Therefore, the addition of modified starch can function as a stabilizer to increase the solubility. The significant increase of the formulated QC powder in dissolution rate might be attributed to the formation of modified starch stabilized QC nanosuspensions. 3.5. XRD analysis

Fig. 2. FTIR spectra of hydrophobically modified starch (HMS) (a), pure QC powder (b) and formulated QC powder (c).

tively contribute to the reduction of particle size of the nanodispersion after the milling process.

The powder X-ray diffractogram of pure QC, wet-milled pure QC and formulated ones are illustrated in Fig. 4A. The characteristic peaks of pure QC samples corresponded to the one in the Cambridge Structural Database (CSD), which included some characteristic peaks of 2h at 10.68°, 12.42°, 16.15°, 24.46°, etc.

3.2. FTIR spectroscopy The FTIR spectra of HMS, pure QC dihydrate powder and formulated QC powder were shown in Fig. 2. There were many FTIR bands in the fingerprint region from 1700 to 700 cm1: 1618, 1339, 1664, 1450, 1262–1168 and 1130–1014 cm1 (Lu, Ross, Powers, & Rasco, 2011) (Sun-Waterhouse, Wadhwa, & Waterhouse, 2012). The band at 3100–3500 cm1 of the pure QC powder was attributed to QC-OH stretching. In the physical mixture this band shape has minor change and peaks at 3323 and 3405 cm1 could still be identified. However, the formulated powders showed broad peaks at 3000–3600 cm1 that were similar to the bands of HMS powders. It might be the weak intermolecular hydrogen bonding between QC and the HMS that led to the broadening of this peak, which may disrupt the QC crystalline structure. Besides, most of the characteristic peaks of both HMS and pure QC powders could be found in formulated ones, though with less intensity. Apart from that, no new peak or band shift was found in the FTIR spectra for formulated ones, suggesting that there

Fig. 3. Dissolution profiles of coarse pure QC, wet-milled pure QC, physical mixture (QC and HMS) and wet-milled (QC and HMS) formulations.

142

M. Lu et al. / Journal of Functional Foods 37 (2017) 138–146

Fig. 4. (A) The XRD diffractogram of pure QC powder, wet-milled pure QC powder, spray-dried & wet-milled QC powder and freeze-dried & wet-milled QC powder; (B) Subtraction of pure QC pattern from normalized starch containing pattern; (C) Comparison of pure QC with the background-subtracted curve of the freeze-dried QC samples; (D) The Williamson and Hall analysis of QC crystals in each group. Fitted to the data, the strain is extracted from the slope and the crystallite size is extracted from the yintercept of the fit.

The percent crystallinity of spray-dried and freeze-dried samples containing amorphous starch was determined by subtraction of pure QC pattern from normalized starch containing pattern. As shown in Fig. 4B, the area under the difference curve between the pure QC pattern and freeze-dried QC pattern was proportional to the total weight ratio of amorphous starch, which was measured by the Peak Paint Cursor function in MDI program JADE7. The pure QC curve compared very well with the amorphous starch background-subtracted curve of the freezedried QC samples, as shown in Fig. 4C. The area under the difference curve between the background-subtracted pattern and

manually determined background-containing pattern was proportional to the weight ratio of QC crystals in the freeze-dried samples. Using this method, the percent crystallinity of freezedried and spray-dried QC samples calculated were 50.2% and 52.3%, respectively, which was consistent with the weight ratio of QC in the formulated samples. This result implied that the 1-h media milling process did not reduce the level of crystallinity of QC particles. Furthermore, based upon the peak widths b for 12 intense low angle peaks on different h positions corresponding to single reflections, the crystallite sizes and strain information were computed

M. Lu et al. / Journal of Functional Foods 37 (2017) 138–146

143

Fig. 4 (continued)

using the Fit Peak Profile routine in JADE7. During this process, we used the Williamson-Hall plot (Prabhu, Rao, Kumar, & Kumari, 2014):

b cos h ¼

Kk þ 4e sin h D

where b is the peak width; h is the Bragg angle; K is the Scherrer Constant; k is the X-ray wavelength; D is the crystallite size; and e is the strain. Β⁄cos h was plotted with respect to sinh for the peaks of QC in each samples (Fig. 4D). Strain in crystals created by dislocations, domain boundaries and deformations, etc., was defined as

144

M. Lu et al. / Journal of Functional Foods 37 (2017) 138–146

Table 1 The structure parameters of QC crystals in each group. Samples

Crystallite size (nm)

Strain (%)

Pure QC Wet-milled pure QC Spray-dried & wet-milled QC Freeze-dried & wet-milled QC

49.4 ± 0.5 39.5 ± 0.8 24.3 ± 0.3 20.4 ± 0.3

0.39 ± 0.03 0.15 ± 0.01 0.04 ± 0.01 -0.10 ± 0.02

relative lattice displacement (Maniammal, Madhu, & Biju, 2017). Crystallite sizes and strain were calculated from the y-intercept and slope of the fitted line. Results are shown in Table 1. According to calculations, crystallite sizes for pure QC, wet-milled QC, spraydried formulated QC and freeze-dried formulated QC were 49.4, 39.5, 24.3 and 20.4 nm respectively, implying that combination of media milling process and addition of modified starch could help reduce the QC crystallite sizes. Moreover, freeze-drying process was more efficient in producing QC dried powders with smaller crystallite sizes than spray-drying process. Therefore, the increase in the solubility of formulated QC nanoparticles could be explained by the reduction in crystallite sizes, instead of changes in the level of crystallinity. 3.6. Comparison of in-vitro digestion between pure QC suspension and QC formulations The bioaccessibility of QC for both pure QC suspension and formulation was studied through the TIM-1 system. After the HPLC analysis, the bioaccessible QC content in jejunum dialysate, ileum dialysate and ileum effluent was calculated and shown in Fig. 5.

As was demonstrated from Fig. 5A, for pure samples, the QC digested in jejunum gradually increased during first 2 h, and reached the maximum concentration during 90–120 min. Then bioaccessible QC content started to decrease until the end of the 5 h period. While for QC formulations, the maximum bioaccessibility was achieved in 120–180 min, meaning that the formulation continued to be digested and absorbed in jejunum and ileum after 2 h. The sustained digestion process can be explained by the reduced particle size and stabilization function of modified starch in the formulated suspensions. Besides, comparing Fig. 5A with Fig. 5B, the bioaccessible QC in jejunum dialysate was higher than the amount detected in ileum, which demonstrated that Jejunum was the major location for QC absorption. Moreover, the overall bioaccessibility was defined as the combined bioaccessible QC content in jejunum and ileum dialysates. According to Fig. 5C, compared with the coarse samples, the formulated ones had an increased amount of overall bioaccessibility during almost every time interval. Meanwhile, a cumulative bioaccessibility profile of QC in the jejunum, ileum and jej+ile was analyzed and plotted in Fig. 6. According to the cumulative bioaccessibility in jejunum dialysate, the amount of QC derived from formulations (34.0%) was about twice the amount from pure suspensions (15.1%) after the 5 h TIM-1 test. Same trend was observed for the ileum dialysate, where 12.4% of its original input was produced from the formulation through digestion and absorption; while only 6.16% of input was obtained from the unformulated sample. Combining results from jejunum and ileum together, we had observed that the cumulative bioaccessibility of QC doubled for the formulation (46.4%) compared with the pure sample (22.3%). Therefore, we can come

Fig. 5. Bioaccessible QC (% of input) accumulated in every 30 min during first 2 h and every 60 min during 2–5 h from different parts of the TIM-1 model. (A) Bioaccessible QC in jejunum dialysate from pure QC suspension and QC formulation; (B) Bioaccessible QC in ileum dialysate from pure QC suspension and QC formulation; (C) Total bioaccessible QC in both jejunum and ileum dialysate from pure QC suspension and QC formulation.

M. Lu et al. / Journal of Functional Foods 37 (2017) 138–146

145

Fig. 6. Cumulative bioaccessibility profile of QC (% of input) accumulated in every 30 min during first 2 h and every 60 min during 2–5 h from different parts of the TIM-1 model. (A) Cumulative bioaccessibility of QC in jejunum dialysate from pure QC suspension and QC formulation; (B) Cumulative bioaccessibility of QC in ileum dialysate from pure QC suspension and QC formulation; (C) Cumulative bioaccessibility of QC in both jejunum and ileum dialysate from pure QC suspension and QC formulation.

to the conclusion that media milling process combined with the addition of modified starch could effectively increase the bioaccessibility of QC through the dynamic in vitro gastrointestinal digestion model. 4. Conclusion In this study, QC nanocrystals were successfully fabricated by media milling and spray & freeze drying process. The particle size was reduced to 340 nm with a saturation solubility of 28.78 ± 0.31 lg/mL, about eleven times higher than pure QC. The addition of HMS could help enhance the metastable equilibrium solubility by working as a stabilizer to prevent the agglomeration of QC nanosuspensions after media milling. FTIR spectroscopy results illustrated that neither media milling nor drying process would affect the chemical properties of QC. Dynamic light scattering and XRD results revealed that media milling treatment not only reduced the QC particle size to nano-range, but also reduced the crystallite sizes, enhancing the solubility and dissolution rate. And TIM-1 in vitro digestion model, which simulated the digestion process in upper GI tract, had determined an increased bioaccessibility for the formulated nanoparticles. In summary, the QC nanoparticles were proved to have enhanced water solubility, dissolution rate and bioaccessibility with reduced crystallite sizes. Acknowledgements We thank Dr. Yaqi Lan for the technical support and use of TNO TIM-1 system. This work was supported by Advanced Orthomolec-

ular Research Inc. (AOR). Financial support from Chinese Scholarship Council (CSC) was also acknowledged (ML) (File Number 201206850004, 2012). Conflict of interest The authors declare no competing financial interest. References Azuma, K., Ippoushi, K., & Terao, J. (2010). Evaluation of tolerable levels of dietary quercetin for exerting its antioxidative effect in high cholesterol-fed rats. Food and Chemical Toxicology, 48(4), 1117–1122. Chattopadhyay, P., & Gupta, R. B. (2001). Production of griseofulvin nanoparticles using supercritical CO2 antisolvent with enhanced mass transfer. International Journal of Pharmaceutics, 228(1), 19–31. Chen, G., & Wang, W. (2007). Role of freeze drying in nanotechnology. Drying Technology, 25(1), 29–35. Dario, M. F., Santos, M. S. C. S., Viana, A. S., Arêas, E. P. G., Bou-Chacra, N. A., Oliveira, M. C., ... Velasco, M. V. R. et al. (2015). A high loaded cationic nanoemulsion for quercetin delivery obtained by sub-PIT method. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 489, 256–264. Dian, L., Yu, E., Chen, X., Wen, X., Zhang, Z., Qin, L., ... Wu, C. et al. (2014). Enhancing oral bioavailability of quercetin using novel soluplus polymeric micelles. Nanoscale Research Letters, 9, 684–695. Du, F., Feng, Y., Fang, J., & Yang, M. (2015). MicroRNA-143 enhances chemosensitivity of Quercetin through autophagy inhibition via target GABARAPL1 in gastric cancer cells. Biomedicine and Pharmacotherapy, 74, 169–177. Erlund, I. (2004). Review of the flavonoids quercetin, hesperetin, and naringenin. Dietary sources, bioactivities, bioavailability, and epidemiology. Nutrition Research, 24(10), 851–874. Erlund, I., Silaste, M. L., Alfthan, G., Rantala, M., Kesaniemi, Y. A., & Aro, A. (2002). Plasma concentrations of the flavonoids hesperetin, naringenin and quercetin in

146

M. Lu et al. / Journal of Functional Foods 37 (2017) 138–146

human subjects following their habitual diets, and diets high or low in fruit and vegetables. European Journal of Clinical Nutrition, 56(9), 891–898. Freese, R., Alfthan, G., Jauhiainen, M., Basu, S., Erlund, I., Salminen, I., ... Mutanen, M. (2002). High intakes of vegetables, berries, and apples combined with a high intake of linoleic or oleic acid only slightly affect markers of lipid peroxidation and lipoprotein metabolism in healthy subjects. The American Journal of Clinical Nutrition, 76, 950–960. Gagos, M. (2008). Molecular organization of 2-(2,4-dihydroxylphenyl)-5,6-dichlor 1,3-benzothiazole in monomolecular layers formed with diphytanoylphosphatidylcholine: A linear dichroism-FTIR study. Biochimica et Biophysica Acta, 1778(11), 2520–2525. Gao, Y., Wang, Y., Ma, Y., Yu, A., Cai, F., Shao, W., et al. (2009). Formulation optimization and in situ absorption in rat intestinal tract of quercetin-loaded microemulsion. Colloids and Surfaces B: Biointerfaces, 71(2), 306–314. Hao, S., Wang, Y., Wang, B., Deng, J., Liu, X., & Liu, J. (2013). Rapid preparation of pHsensitive polymeric nanoparticle with high loading capacity using electrospray for oral drug delivery. Materials Science & Engineering. C, Materials for Biological Applications, 33(8), 4562–4567. Havenaar, R., Anneveld, B., Hanff, L. M., de Wildt, S. N., de Koning, B. A., Mooij, M. G., ... Minekus, M. (2013). In vitro gastrointestinal model (TIM) with predictive power, even for infants and children? International Journal of Pharmaceutics, 457 (1), 327–332. Jacobs, H., Moalin, M., van Gisbergen, M. W., Bast, A., van der Vijgh, W. J., & Haenen, G. R. (2011). An essential difference in the reactivity of the glutathione adducts of the structurally closely related flavonoids monoHER and quercetin. Free Radical Biology and Medicine, 51(11), 2118–2123. Jeyadevi, R., Sivasudha, T., Rameshkumar, A., Ananth, D. A., Aseervatham, G. S., Kumaresan, K., ... Renganathan, R. (2013). Enhancement of anti arthritic effect of quercetin using thioglycolic acid-capped cadmium telluride quantum dots as nanocarrier in adjuvant induced arthritic Wistar rats. Colloids and Surfaces B: Biointerfaces, 112, 255–263. Jung, H. J., Sohn, Y., Sung, H. G., Hyun, H. S., & Shin, W. G. (2015). Physicochemical properties of ball milled boron particles: Dry vs. wet ball milling process. Powder Technology, 269, 548–553. Kakran, M., Sahoo, N. G., Li, L., Judeh, Z., Wang, Y., Chong, K., et al. (2010). Fabrication of drug nanoparticles by evaporative precipitation of nanosuspension. International Journal of Pharmaceutics, 383(1–2), 285–292. Kakran, M., Shegokar, R., Sahoo, N. G., Shaal, L. A., Li, L., & Muller, R. H. (2012). Fabrication of quercetin nanocrystals: Comparison of different methods. European Journal of Pharmaceutics and Biopharmaceutics, 80(1), 113–121. Karadag, A., Ozcelik, B., & Huang, Q. (2014). Quercetin nanosuspensions produced by high-pressure homogenization. Journal of Agricultural and Food Chemistry, 62, 1852–1859. Karadag, A., Yang, X., Ozcelik, B., & Huang, Q. (2013). Optimization of preparation conditions for quercetin nanoemulsions using response surface methodology. Journal of Agricultural and Food Chemistry, 61, 2130–2139. Kleemann, R., Verschuren, L., Morrison, M., Zadelaar, S., van Erk, M. J., Wielinga, P. Y., et al. (2011). Anti-inflammatory, anti-proliferative and anti-atherosclerotic effects of quercetin in human in vitro and in vivo models. Atherosclerosis, 218(1), 44–52. Lee, J., Han, S. I., Yun, J. H., & Kim, J. H. (2015a). Quercetin 3-O-glucoside suppresses epidermal growth factor-induced migration by inhibiting EGFR signaling in pancreatic cancer cells. Tumour Biology, 36(12), 9385–9393. Lee, S. H., Lee, E. J., Min, K. H., Hur, G. Y., Lee, S. Y., Kim, J. H., ... Kang, K. H. et al. (e et al., 2015 b). Quercetin enhances chemosensitivity to gemcitabine in lung cancer cells by inhibiting heat shock protein 70 expression. Clinical Lung Cancer, 16(6), 235–243. Li, H., Zhao, X., Ma, Y., Zhai, G., Li, L., & Lou, H. (2009). Enhancement of gastrointestinal absorption of quercetin by solid lipid nanoparticles. Journal of Controlled Release, 133(3), 238–244. Lin, C. W., Lai, G. M., Chen, K. C., Lin, T. H., Fan, J. J., Hsu, R. L., ... Cheng, C. H. (2015). RPS12 increases the invasiveness in cervical cancer activated by c-Myc and inhibited by the dietary flavonoids luteolin and quercetin. Journal of Functional Foods, 19, 236–247. Lu, H. Y., Peng, T. S., Hu, X. D., Li, S. J., Luo, M., He, Y. H., et al. (2015). Quercetin potentiates the effect of cd T cells via modulating the expressions of Granzyme B, perforin and IFN-c and also regulates the Wnt/b-catenin signalling pathway in human colon cancer cells. Bangladesh Journal of Pharmacology, 10(2), 251–259.

Lu, X., Ross, C. F., Powers, J. R., & Rasco, B. A. (2011). Determination of quercetins in onion (Allium cepa) using infrared spectroscopy. Journal of Agricultural and Food Chemistry, 59(12), 6376–6382. Maniammal, K., Madhu, G., & Biju, V. (2017). X-ray diffraction line profile analysis of nanostructured nickel oxide: Shape factor and convolution of crystallite size and microstrain contributions. Physica E: Low-dimensional Systems and Nanostructures, 85, 214–222. Niwa, T., & Danjo, K. (2013). Design of self-dispersible dry nanosuspension through wet milling and spray freeze-drying for poorly water-soluble drugs. European Journal of Pharmaceutical Sciences, 50(3–4), 272–281. Paller, C. J., Kanaan, Y. M., Beyene, D. A., Naab, T. J., Copeland, R. L., Tsai, H. L., ... Hudson, T. S. (2015). Risk of prostate cancer in African-American men: Evidence of mixed effects of dietary quercetin by serum vitamin D status. The Prostate, 75 (13), 1376–1383. Pongpeerapat, A., Wanawongthai, C., Tozuka, Y., Moribe, K., & Yamamoto, K. (2008). Formation mechanism of colloidal nanoparticles obtained from probucol/PVP/ SDS ternary ground mixture. International Journal of Pharmaceutics, 352(1–2), 309–316. Prabhu, Y. T., Rao, K. V., Kumar, V. S. S., & Kumari, B. S. (2014). X-ray analysis by Williamson-Hall and size-strain plot methods of ZnO nanoparticles with fuel variation. World Journal of Nano Science and Engineering, 4(1), 21–28. Pralhad, T., & Rajendrakumar, K. (2004). Study of freeze-dried quercetin– cyclodextrin binary systems by DSC, FT-IR, X-ray diffraction and SEM analysis. Journal of Pharmaceutical and Biomedical Analysis, 34(2), 333–339. Ramos, A. A., Lima, C. F., Pereira, M. L., Fernandes-Ferreira, M., & Pereira-Wilson, C. (2008). Antigenotoxic effects of quercetin, rutin and ursolic acid on HepG2 cells: evaluation by the comet assay. Toxicology Letters, 177(1), 66–73. Ribnicky, D. M., Roopchand, D. E., Oren, A., Grace, M., Poulev, A., Lila, M. A., ... Raskin, I. (2014). Effects of a high fat meal matrix and protein complexation on the bioaccessibility of blueberry anthocyanins using the TNO gastrointestinal model (TIM-1). Food Chemistry, 142, 349–357. Ruotolo, R., Calani, L., Brighenti, F., Crozier, A., Ottonello, S., & Del Rio, D. (2014). Glucuronidation does not suppress the estrogenic activity of quercetin in yeast and human breast cancer cell model systems. Archives of Biochemistry and Biophysics, 559, 62–67. Sahoo, N. G., Kakran, M., Shaal, L. A., Li, L., Muller, R. H., Pal, M., et al. (2011). Preparation and characterization of quercetin nanocrystals. Journal of Pharmaceutical Sciences, 100(6), 2379–2390. Sansone, F., Picerno, P., Mencherini, T., Villecco, F., D’Ursi, A. M., Aquino, R. P., et al. (2011). Flavonoid microparticles by spray-drying: Influence of enhancers of the dissolution rate on properties and stability. Journal of Food Engineering, 103(2), 188–196. Shaikh, J., Bhosale, R., & Singhal, R. (2006). Microencapsulation of black pepper oleoresin. Food Chemistry, 94(1), 105–110. Shegokar, R., & Muller, R. H. (2010). Nanocrystals: industrially feasible multifunctional formulation technology for poorly soluble actives. International Journal of Pharmaceutics, 399(1–2), 129–139. Srinivas, K., King, J. W., Howard, L. R., & Monrad, J. K. (2010). Solubility and solution thermodynamic properties of quercetin and quercetin dihydrate in subcritical water. Journal of Food Engineering, 100(2), 208–218. Sun-Waterhouse, D., Wadhwa, S. S., & Waterhouse, G. I. N. (2012). Spray-drying microencapsulation of polyphenol bioactives: A comparative study using different natural fibre polymers as encapsulants. Food and Bioprocess Technology, 6(9), 2376–2388. Ting, Y., Jiang, Y., Ho, C. T., & Huang, Q. (2014). Common delivery systems for enhancing in vivo bioavailability and biological efficacy of nutraceuticals. Journal of Functional Foods, 7, 112–128. Vicentini, F. T., Vaz, M. M., Fonseca, Y. M., Bentley, M. V., & Fonseca, M. J. (2011). Characterization and stability study of a water-in-oil microemulsion incorporating quercetin. Drug Development and Industrial Pharmacy, 37(1), 47–55. Wang, S., Gao, W., Chen, H., & Xiao, P. (2006). Studies on the morphological, thermal and crystalline properties of starches separated from medicinal plants. Journal of Food Engineering, 76(3), 420–426. Wong, M. Y., & Chiu, G. N. (2010). Simultaneous liposomal delivery of quercetin and vincristine for enhanced estrogen-receptor-negative breast cancer treatment. Anti-Cancer Drugs, 21(4), 401–410.