2020-03-06 11:42:55
Choongjin Bane, Myeongsu Joa,1, Young Hyun Parka, Jae Hwan Kima, Jae Yong Hana, Ki Won Leea, Dae-Hyuk Kweonf, Young Jin Choi
1.Introduction
Curcumin, found in turmeric extracted from the rhizome of Curcuma longa, can prevent cancers (Strimpakos & Sharma, 2008) and has anti- inflammatory (Julie & Jurenka, 2009), oxygen radical-scavenging (Kunchandy & Rao, 1990), and anti-amyloidogenic effects (Ono, Hasegawa, Naiki, & Yamada, 2004). Additionally, it can be safely consumed by animals and humans based on many studies that have investigated the toxicity (Sharma et al., 2004). Therefore, curcumin could be appropriate as a constituent of functional foods. However, curcumin’s low oral bioavailability (OBa), due to its poor water solubility (Anand, Kunnumakkara, Newman, & Aggarwal, 2007), auto- xidation (Schneider, Gordon, Edwards, & Luis, 2015), instability at neutral/alkaline pH (Wang et al., 1997), active metabolism (Metzler, Pfeiffer, Schulz, & Dempe, 2013), and low permeability across the apical surface of enterocytes (Wahlang, Pawar, & Bansal, 2011) hampers its use as a bioactive material for functional foods or pharmaceu- ticals. Therefore, a strategy to overcome these weaknesses of curcumin is needed.
The low OBa of curcumin can be overcome using absorption en- hancers and formulation strategies. In preclinical and clinical studies, piperine, as an OBa enhancer, inhibited the activity of metabolic en- zymes (Shoba et al., 1998). However, the OBa improvement achieved using enhancers, including piperine, did not reach a satisfactory ther- apeutic level. Formulation strategies involving nanoparticles (Hartono, Hadisoewignyo, Yang, Meka, & Yu, 2016), curcumin/piperine-co- loaded-nanoparticles (Baspinar, Üstündas, Bayraktar, & Sezgin, 2018), liposomes (Peng et al., 2017), and nanoemulsions (Vecchione et al., 2016) have been proposed to overcome the low OBa of curcumin. These strategies improved the OBa of curcumin, but in all cases the improvement was demonstrated only in vitro or in vivo. Additionally, most previous studies did not identify controllable factors influencing the OBa of curcumin. Controlled digestion of emulsions in the gastro- intestinal tract (GIT) and solubilization of the encapsulated molecules have been reported (Devraj et al., 2013). Unfortunately, these effects were dependent on the quantities of bile acids (BAs) and calcium ions, which are not controllable factors. Moreover, most studies of con- trollable digestion overlooked absorption of the bioactives loaded in the carriers through the gut epithelium (Zhang et al., 2016). Thus, further studies of the digestion/absorption of loaded curcumin, and of the factors related to enhancement of its OBa, are needed.
The biocompatible solid lipid nanoparticles (SLNs) can increase the OBa of curcumin (Baek & Cho, 2017). The solid lipid matrix of SLNs immobilizes core materials and protects them against physical/bio- chemical stressors, such as free radicals, pH, high-ionic-strength solu- tions, and metabolizing enzymes, during passage through the GIT (Ban, Park, Lim, Choi, & Choi, 2015). Orally ingested bioactive material- loaded SLNs are digested by lipases and form micelles with BAs, phospholipids, and the digestion products, and the bioactives solubi- lized in the micelles can be absorbed into the lymph through en- terocytes (McClements & Xiao, 2012). Additionally, the SLNs enable sustainable release of the core material. Indeed, the OBa of curcumin was enhanced by polyethylene glycol (PEG)ylated SLNs, likely due to sustained release and absorption through the GIT (Ji et al., 2016). However, orally administered SLNs must pass through the digestive tract, which could alter their size, surface charge, and curcumin loading efficiency. Hence, further studies of the digestion/absorption mechan- isms are needed to identify controllable factors that allow the limita- tions of curcumin to be overcome.
We previously suggested that gastrointestinal digestion of SLNs could be controlled using PEGylated emulsifiers, i.e., by altering the length and concentration thereof (Ban, Jo, Lim, & Choi, 2018). Here, we examined if the OBa of curcumin could be increased by PEGylated SLNs and conducted digestion/absorption studies using in vitro/in vivo models of the human GIT, to assess the underlying mechanisms. In the in vitro digestion study, changes in the size, surface charge, and lipolysis pattern of the SLNs were recorded after treatment with digestion fluid, and the amount of solubilized curcumin in the mixed micellar fraction was determined. In the absorption study, gut permeation of curcumin loaded in SLNs was assessed using mucus-covered Caco-2 cell mono- layers, and a pharmacokinetics study was performed in a rat model.
2.Materials and methods
2.1.Chemicals
Tristearin (TS), polyoxyethylene (10) stearyl ether (PEG10SE; Brij® S10), polyoxyethylene (1 0 0) stearyl ether (PEG100SE; Brij® S100), bovine serum albumin (BSA), polyacrylic acid, cholesterol, polysorbate 80, mucin, and phosphatidylcholine were purchased from Sigma Aldrich Co. (St. Louis, MO, USA). Liquid canola oil (LCO) was obtained from CJ Cheiljedang Co. (Seoul, South Korea). Curcumin and linoleic acid were supplied by Acros Organics (Pittsburg, PA, USA) and Tokyo Chemical Industry (Tokyo, Japan), respectively. Eagle’s minimum es- sential medium (EMEM) and fetal bovine serum (FBS) were purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA) and GE Healthcare (Chicago, IL, USA), respectively. Nonessential amino acids (NEAA), antibiotic-antimycotic, trypsin-ethylenediaminete- traactic acid (EDTA), and Hank’s balanced salt solution (HBSS) were obtained from Thermo Fisher Scientific (Waltham, MA, USA). All other chemicals were of analytical-reagent grade.
2.2.Fabrication of solid lipid nanoparticles and emulsion
The curcumin-loaded SLNs/emulsion were prepared using an oil-in- water emulsion technique, as reported previously (Ban, Lim, Chang, & Choi, 2014) but with slight modifications. First, the lipid (5 wt%) and aqueous (95 wt%) phases were heated to 85 °C and mixed using a high- speed blender (Ultra-Turrax T25D; Ika Werke GmbH & Co., Staufen, Germany) at 8,000 rpm for 1 min followed by at 11,000 rpm for 1 min. The lipid phase of the curcumin-loaded SLNs was prepared by blending TS (95 wt% of the lipid phase) and curcumin (5 wt% of the lipid phase) in ethanol (25 mg mL−1) and evaporating the ethanol with stirring for 30 min (85 °C). The aqueous phase was fabricated by dissolving PEG10SE or PEG100SE (5.3–46.9 mM) in double-deionized water (DDW) with stirring for 1 h. After preparing the coarse emulsion, the droplet size was reduced by sonication (VCX 750; Sonics & Materials Inc., Newtown, CT, USA) for 4 min at 60% amplitude, with a tem- perature of 85 °C and a duty cycle of 1 s. Next, sonication was per- formed for 6 min was applied to the emulsions during cooling to 45 °C, and the samples were immediately stored at 4 °C. The curcumin-loaded emulsion was prepared in the same manner as the curcumin-loaded SLN stabilized by 46.9 mM PEG10SE, except that LCO was used in place of TS.
2.3.Characterization of solid lipid nanoparticles and emulsion
Solid lipid nanoparticles diluted 10-fold with DDW were passed through a 1 μm pore-size filter (GF/B; Whatman Ltd., Loughborough, UK). The aggregated SLNs (> 1 μm) remaining on the filter were weighed after drying in an oven at 60 °C. The difference in filter weight before and after this procedure, i.e., the weight of the aggregated SLNs, was recorded and the yield was calculated as the percentage of the filter weight difference per the SLN weight. To measure the mean particle size (PS, z-average) and ζ-potential (ZP), using a Zetasizer with a 173° angle helium–neon laser (λ = 633 nm; Nano ZS; Malvern Instruments Ltd., Worcestershire, UK), the SLNs and emulsion were passed through filters to eliminate aggregates. The ZP measurement was based on the Smoluchowski equation at 25 °C with an electric field strength of 20 V cm−1.
Emulsifier surface load (Γs) was calculated as s Ca D/6Φ, where Ca is the concentration of the emulsifier adsorbed to the surface of the SLNs/emulsion, D is the PS, and Φ is the lipid phase volume fraction (i.e., 0.05). The method for determining Ca was introduced previously by our group (Ban et al., 2018). Briefly, emulsifier not coating the lipid particles/droplets was collected using a Sephadex G-25 column (GE Healthcare, Chalfont St. Giles, UK) and quantified using ammonium cobalt thiocyanate (ACTC) solution and a spectrophotometer (Phar- maspec UV-1700; Shimadzu Corp., Kyoto, Japan). The Ca for the SLNs/ emulsion was calculated by subtracting the non-coating emulsifier concentration from the total concentration of emulsifiers.
To solubilize unentrapped curcumin and precipitate the curcumin- loaded particles/droplets, 0.2 mL of the SLN/emulsion dispersion was added to a 1.8 mL mixture of methanol and acetonitrile (1:1), vortexed for 10 s, and centrifuged (15,000 rcf, 10 min; 5427R; Eppendorf AG, Hamburg, Germany). The supernatant was filtered through a 0.22-μm polyvinylidene difluoride (PVDF) membrane (Millex-GV 33MM syringe filter; Merck Millipore, Bedford, MA, USA) and curcumin was quanti- fied by high–performance liquid chromatography (HPLC; 3.125–100 μg mL−1, R2 = 1.0000; λ = 426.9 nm). The entrapment ef- ficiency (EE) of the curcumin-loaded SLNs/emulsion was determined using the equation: EE(%) Wt Ws 100Wt where Wt is the total weight of curcumin and Ws is the weight of cur- cumin in the supernatant (i.e., the weight of unentrapped curcumin).
2.4.High-performance liquid chromatography
A HPLC system equipped with a Waters 2695 Separations module (Waters, Milford, MA, USA) and analytical C18 column (Venusil XBP C18, 5 μm, 100 Å, 4.6 × 250 mm; Bonna-Agela Technology, Newark, DE, USA) was utilized for curcumin quantification. Detection was conducted at 25 °C using a Waters 996 Photodiode Array Detector. Curcumin in the samples (20 μL) was isocratically eluted at a flow rate of 0.8 mL min−1 using a mixture (35:55:10) of methanol, acetonitrile,
and 5 vol% acetic acid as the mobile phase. This HPLC system was used for quantification of curcumin, the soluble fraction in the mixed mi- celles, permeation via the simulated gut membrane, and pharmacoki- netic analysis.
2.5.Measurement of the size and ζ-Potential of the solid lipid nanoparticles and emulsion after simulated gastrointestinal digestion
Prior to determination of the colloidal stability of SLNs/emulsion under high ionic strength/acidic conditions, the SLNs/emulsion were diluted 10-fold and aggregates were eliminated by filtration (1 μm). To set high ionic strength conditions, 5 mL of the diluted/filtered SLNs/ emulsion was blended with 3.8 mL of the mixture of all media and juices, except HCl solution, proteins, bile, and enzymes (Table S1). The mixture was adjusted to pH 7 using 1 M NaOH or 1 M HCl solution and a pH meter (Professional Meter PP-15; Sartorius AG, Göttingen, Ger- many). For acidic conditions, the diluted/filtered dispersions were ad- justed to pH 3 using 50 mM HCl and incubated in a shaking water bath (BS-31; JEIO Tech., Seoul, South Korea) (2 h, 37 °C, 100 rpm). Next, 2 mL of the samples were centrifuged (10 min, 25,000 rcf; Eppendorf 5427R) to eliminate creamed SLN aggregates. The PS and ZP of SLNs/ emulsion in the supernatant were measured using the Zetasizer. The relative centrifugal force (25,000 rcf) was determined in preliminary experiments and did not induce creaming or sedimentation of freshly prepared SLNs/emulsion.
To determine the effects of pancreatic-lipase and BA on changes in the PS and ZP of SLNs/emulsion, 5 mL of diluted (10-fold) and filtered (1 μm) SLNs/emulsion was treated with pancreatic lipase (3.965 mg mL−1), bile extract (41.32 mg mL−1), and a mixture of lipase and bile. After incubation for 2 h at 37 °C with shaking at 100 rpm, 2 mL samples were centrifuged for 10 min at 25,000 rcf to eliminate creamed SLN aggregates. The PS and ZP of SLNs/emulsion in the supernatant were measured using the Zetasizer.
