Impact of HPMC on inhibiting crystallization and improving permeability of curcumin amorphous solid dispersions

Carbohydrate Polymers 181 (2018) 543–550

Contents lists available at ScienceDirect

Carbohydrate Polymers journal homepage: www.elsevier.com/locate/carbpol

Impact of HPMC on inhibiting crystallization and improving permeability of curcumin amorphous solid dispersions ⁎

T

⁎

Na Fan, Zhonggui He , Pingping Ma, Xin Wang, Chang Li, Jin Sun, Yinghua Sun, Jing Li Wuya College of Innovation, Shenyang Pharmaceutical University, Shenyang, China

A R T I C L E I N F O

A B S T R A C T

Keywords: Curcumin amorphous solid dispersions Inhibit crystallization Improve permeability Hydrogen bonding Ionic interactions

The purpose of this paper was to elucidate the impacts of hydroxypropylmethyl cellulose E5 as assistant excipient on inhibiting crystallization and improving membrane permeability in curcumin amorphous solid dispersions that formulated by Eudragit E100. Intermolecular interactions formed between curcumin and polymers were probed using in situ Raman imaging and infrared spectroscopy. The abilities of hydroxypropylmethyl cellulose E5 in inhibiting crystallization and improving membrane permeability were confirmed by fluorescence spectroscopy, dynamic light scattering analysis and in vitro permeability experiment. The results demonstrated hydroxypropylmethyl cellulose E5 was significant in maintaining the amorphous drug concentration owing to the hydrogen bond interactions formed with curcumin, rending its ability to inhibit crystallization by reducing drug droplet size. Furthermore, the addition of hydroxypropylmethyl cellulose E5 in curcumin amorphous solid dispersions promoted drug membrane permeability through lowering the order level of phospholipid bilayer layer.

1. Introduction The poor solubility of water-insoluble drugs is a limiting factor for achieving good oral bioavailability. Many methods have been used to overcome this issue. For example, Stella introduced a recent “surprising” prodrug bortezomib (marketed as Velcade). Bortezomib was significantly more soluble due to the in situ formation of boronic acid esters by reaction with diol groups of mannitol (Stella & Nti-Addae, 2007). Chaudhari studied the effect of hydroxypropylmethyl cellulose (HPMC) and polyvinylpyrrolidone (PVP) on the model drug fenofibrate. It confirmed that the polymers governed the dissolution of amorphous solid dispersions (ASDs) (Chaudhari & Dave, 2015). Md. Akhlaquer Rahman systematically clarified the oral lipid based formulations improved the bioavailability by increasing the solubility, facilitating gastrointestinal absorption of poorly water-insoluble, lipophilic drug (Rahman, Harwansh, Mirza, Hussain, & Hussain, 2011). Among these methods, ASDs can be considered as the most promising technique (Serajuddin, 1999). ASDs are defined as drug molecules dispersed in hydrophilic carriers and the carriers with different properties can modulate drug release profile. At present, a wide variety of polymers are used in solid dispersions to achieve specific goals. For example, cellulose derivative matrices like carboxymethylcellulose acetate butyrate (CMCAB), hydroxypropylmethyl cellulose acetate succinate (HPMCAS), or cellulose acetate adipate propionate (CAAdP) were

⁎

applied for quercetin ASDs (Li & Konecke, 2013) and ellagic acid ASDs (Li, Harich, Wegiel, Taylor, & Edgar, 2013). It was also reported that poly(acrylic acid) was applied for increasing the dissolution of the drug, and cellulose derivatives were added to inhibit drug crystallization (Li, Konecke, Wegiel, Taylor, & Edgar, 2013). Therefore, it can be accepted that cellulose derivatives have huge values in assisting the design of ASDs. Curcumin (Cur, Supplementary data, Fig. S1) is a hydrophobic polyphenol derived from the rhizome of Curcuma longa. Cur exhibits keto–enol tautomerism and its di-keto form and keto–enol form are shown in Fig. S1. The diverse pharmacological and biological properties include anti-oxidant (Sharma, 1976), anti-inflammatory (Motterlini, Foresti, Bassi, & Green, 2000), anti-tumor (Sadzuka, Nagamine, Toyooka, Ibuki, & Sonobe, 2012), anti-cancer (Aggarwal, Kumar, & Bharti, 2003), cardioprotective effects (Imbaby, Ewais, Essawy, & Farag, 2014) and so on. Cur has been investigated in a wide range because of its wide application and good safety (Li, Konecke, Wegiel, Taylor, & Edgar, 2013). However, Cur with characters of moderately hydrophobic (logP 2.5) and low aqueous solubility (11 ng/mL at pH 5.0) (Tønnesen, Másso, & Loftssonb, 2002) degrades very quickly at neutral or alkaline pH (Wang et al., 1997). These disadvantages limit the clinical development of Cur. In order to overcome the challenges, different strategies have been applied (Chen, Ormes, Higgins, & Taylor, 2015). Cur ASDs is one of the most promising strategies due to the high

Corresponding authors at: Wenhua road 103, Shenyang Pharmaceutical University, Shenyang, 110016, China. E-mail addresses: [email protected] (Z. He), [email protected] (J. Li).

https://doi.org/10.1016/j.carbpol.2017.12.004 Received 6 September 2017; Received in revised form 22 October 2017; Accepted 4 December 2017 Available online 06 December 2017 0144-8617/ © 2017 Elsevier Ltd. All rights reserved.

Carbohydrate Polymers 181 (2018) 543–550

N. Fan et al.

region 500–4000 cm−1. Samples were prepared by grounding with KBr gently and respectively.

melting point (180 °C) of Cur (Li & Wegiel, 2013). In the current stage, Cur ASDs can be formulated by different polymers, including Eudragit E100 (E100), PVP, CMCAB, HPMC and HPMCAS (Wegiel, Zhao, Mauer, Edgar, & Taylor, 2014). As cellulose derivatives have potential value as assistant excipient for ASDs, the present paper innovatively introduced HPMC into the design of Cur ASDs formulated by E100 to systemically elucidate the advantages of HPMC that mainly covered its ability in inhibiting Cur crystalline and enhancing membrane permeability in Cur ASDs. E100 is a polybase and proton acceptor, and the average pKa of the basic monomer is 8.4 (Menjoge & Kulkarni, 2007). Cur has pKa values of 10.51, 9.88 and 8.38 (Bernabé-Pineda, Ramı́rez-Silva, Romero-Romo, González-Vergara, & Rojas-Hernández, 2004). Therefore, ionic interactions were formed between E100 and Cur, resulting in the enhancement of drug dissolution (Wegiel et al., 2014). Herein, the existence of tautomeric Cur in ASDs was vividly showed by Raman imaging plus spectroscopy and the intermolecular interactions were determined by IR spectroscopy. Furthermore, the impact of HPMC on inhibiting crystallization in Cur ASDs was mainly studied by fluorescence spectroscopy and dynamic light scattering analysis. As for membrane permeability experiment, a facile in vitro permeability test was conducted for as-synthesized Cur ASDs. It is believed that impact of HPMC on inhibiting drug crystallization and improving permeability in Cur ASDs that elucidated in this paper can provide valuable instruction for the study of Cur ASDs in future.

2.3.3. Effect of polymer on the equilibrium solubility of cur The equilibrium solubility of Cur was determined by adding an excess amount of Cur to 1 mL pH enzyme-free simulated gastric fluid (pH 1.0 hydrochloric acid) with the presence of 500 μg/mL pre-dissolved polymer(s) in Eppendorf tubes (EP tubes). The EP tubes were equilibrated at 37 °Cfor 48 h in an agitating water bath. Samples were then ultracentrifuged to separate excess crystalline Cur particles from the supernatant. Ultracentrifugation was performed at 12000 rpm for 10 min. High performance liquid chromatography (HPLC) analysis was performed using an HITACH HPLC (Tokyo, Japan). The chromatographic separation was performed with a C18 Column. A water (22%), methanol (77%) and glacial acetic acid (1%) mixture was used as mobile phase, and the flow rate was 1 mL/min. The ultraviolet detection wavelength was 428 nm. All measurements were performed in triplicate at room temperature. 2.4. The impact of HPMC on inhibiting crystallization 2.4.1. Fluorescence spectroscopy Fluorescence spectroscopy was used to reflect the nature of phase separation phenomena occurring in solution. Herein, 50 μL Cur ethanol solution with the concentration of 10, 20, 30, 40, 50, 60, 70, 80, 90 μg/ mL were added to 50 μL pH 1.0 hydrochloric acid with the presence of 500 μg/mL pre-dissolved polymer(s) in 96-well plates. Fluorescence measurements were made using a Microplate reader (ThermoFisher scientific, USA). The excitation wavelengths used were settled in the range of 200 ∼ 500 nm, the emission wavelength was 520 nm.

2. Materials and methods 2.1. Materials Cur with purity of more than 99.8% was purchased from meilun bio Co., Ltd. (Dalian, China). Eudragit E100 (E100) was kindly provided by Evonik Co., Ltd. (Germany). Hydroxypro-pylmethyl cellulose E5 (HPMC) was supplied by Anhui Shanhe Pharmaceutical Excipients Co., Ltd. (Huainan, China). Other chemical agents were obtained from Tianjin Bodi Chemical Holding Co., Ltd. (Tianjin, China).

2.4.2. The effect of polymer(s) on amorphous drugs concentration The effect of polymer(s) on amorphous Cur concentration was determined by adding 100 μL Cur supersaturated ethanol solution (Cur supersaturated ethanol solution was obtained from the supernatant when dissolving an excess amount of Cur in ethanol solution until the existence of Cur crystals at the bottom) to 400 μL pH 1.0 hydrochloric acid with the presence of 500 μg/mL pre-dissolved polymer(s) in EP tubes. A batch of samples was immediately filtered with an oil microporous membrane (0.45 μm) and measured by HPLC, while another batch of samples was filtered after standing for 2 h and then measured by HPLC. All samples were performed in triplicate. The polymer solutions in the test included E100, HPMC, E100/HPMC 1:1, E100/HPMC 3:1, E100/HPMC 6:1 and E100/HPMC 9:1, and the working conditions of HPLC were the same as described in Section 2.5.

2.2. Preparation of bulk cur ASDs Cur and the polymer(s) were dissolved in ethanol. Solvent removal was achieved by rotary evaporation. The ASDs with various compositions (Supplementary data, Table S1) were subsequently dried in vacuum oven overnight to remove any residual solvent. The ASDs were ground using a mortar and pestle and then sieved (60 mesh) to obtain uniform particles. The short names for as-synthesized Cur ASDs were listed in Table S1, which were Cur-E100 1:1, Cur-E100 1:2, Cur-E100 1:4, Cur-E100 1:6 or Cur-E100, Cur-E100 1:8, Cur-E100/HPMC1:1, CurE100/HPMC 3:1, Cur-E100/HPMC 4:1, Cur-E100/HPMC 6:1, CurE100/HPMC 9:1.

2.4.3. Dynamic light scattering (DLS) The particle size of each sample was determined by DLS (Nano ZS90, Malvern, Worcestershire, UK). It can not only be used for nanosuspension, but also for evaluating particle size with nanometer level. The following procedure was applied for sample preparation. Briefly, 0.5 mL saturated ethanol solution of Cur was pipetted into 2 mL polymer(s) aqueous solution (pH 1.0 hydrochloric acid) with 500 μg/ mL pre-dissolved polymer(s). The parameters included: (1) Equilibrium time 10s; (2) 13 of number of runs; (3) 5 of run duration (s); (4) 3 of number of measurements; (5) 1 of delay between measurements (s).

2.3. Intermolecular interactions between cur and polymers 2.3.1. Raman imaging and spectroscopy The existence form of tautomeric Cur in different ASDs was assessed by Raman imaging and spectroscopy. Samples (Cur-E100 1:6 (CurE100), Cur-E100/HPMC 1:1, Cur-E100/HPMC 3:1, Cur-E100/HPMC 6:1 and Cur-E100/HPMC 9:1) were analyzed in situ through a quartz sight window via a Raman spectrometer (inVia Laser Micro Raman Spectroscopy, Renishaw PLC) equipped with a thermoelectrically cooled CCD detector and a fiber optic probe. Then, the samples were measured at room temperature using a 500 mW laser source with a wavelength of 785 nm.

2.4.4. In vitro dissolution Cur and Cur ASDs (containing 4 mg Cur) were dissolved in 250 mL pH 1.0 hydrochloric acid by using small cup method with a RC806D dissolution tester (Tianjin, China). The experiment was carried out for 2 h with working conditions of 37 °C and 50 rpm. Aliquots (5 mL) were withdrawn at appropriate time intervals and replaced with 5 mL of fresh dissolution medium after each sampling to maintain constant volume. For convenience, the sample medium was analyzed using UV1120 (Shimadzu, Japan) at the wavelength of 425 nm after going

2.3.2. IR spectroscopy Infrared spectroscopy (IR, Spectrum 1000, PerkinElmer, USA) spectra of drug, polymers, and Cur ASDs obtained from the spectral 544

Carbohydrate Polymers 181 (2018) 543–550

N. Fan et al.

through 0.45 μm microporous membrane. The standard curve for UV analysis was A = 0.1555C-0.0024 (R2 = 0.9995), where A and C stood for absorbance and concentration respectively. All samples were performed in triplicate.

significantly stronger than any polymer used, the Raman spectrum reflected the existing state of Cur in ASDs. As can be seen, the Raman images consisted of red and green colors and the intensity of red was much stronger than green based on analysis result. Similarly as the literature (Li et al., 2017), these two kinds of Raman spectra can be ascribed to the spectra of Cur with keto form and Cur with enol form. Since according to the in situ Raman images, the green Raman spectra belonged to Cur of enol form owing to the co-shift of −OH group (around 1250 cm−1 to lower wavenumber) and carbonyl group (around 1625 cm−1 to higher wavenumber) while the red raman spectra that similar as spectrum of Cur belonged to Cur of keto form (Li et al., 2017). In another word, red color represented Cur with keto form, green color stood for Cur with enol form and Cur with keto form has stronger intensity than Cur with enol form in Cur ASDs. Obviously, there was more Cur with keto form that existed in Cur-E100, Cur-E100/HPMC 6:1 and Cur-E100/HPMC 9:1. The existing Cur form in Cur ASDs was possibly related with the main molecular interactions formed between drug and polymer, which will be discussed in IR analysis.

2.5. The impact of HPMC on improving membrane permeability In the course of drug screening, Kansy, Senner, & Gubernator, 1998 established parallel artificial membrane permeability assay (PAMPA), which can quickly detect membrane permeability of drug through passive diffusion. Vertical diffusion pool (Supplementary data, Fig. S2) was an effective tool for carrying out in vitro permeability experiment. The initial step was to establish PAMPA. Briefly, soybean lecithin solution that dissolved in ethanol was spread on a non-smooth surface of a 0.22 μm polyester fiber filter and was dried at room temperature. Afterwards, 1 mL and 7 mL pH 1.0 hydrochloric acid were added to the supplying pool and receiving pool of vertical diffusion pool respectively, and the artificial membrane was displayed in the middle of the two pools. After that, the sample was put into supplying pool. Aliquots (20 μL) were withdrawn at appropriate time intervals and the samples were analyzed by HPLC (The working conditions of HPLC were the same as described above). All samples were performed in triplicate.

3.1.2. IR spectroscopy Drug–polymer interactions in solid dispersions can be analyzed by IR spectroscopy (Wegiel et al., 2013). The IR spectra of Cur, polymer(s) and Cur ASDs are presented in Fig. 2 and specific figures and data are shown in Supplementary data (Fig. S3, Table S2 and Table S3). Owing to the hydrogen bonds or other forces between Cur and polymer(s), peak shifts can be seen in Cur ASDs. It is well known that Cur is special for its enol and keto two isomers. In order to better explore the interaction between Cur and the polymer, the intramolecular forces of CurCur will be initially explored. The spectrum of Cur did not exhibit the strong carbonyl stretching band that belonged to the di-keto form of Cur at around 1700 cm−1. Rather, there was a prominent stretching band at 1628.8 cm−1 that was consistent with formation of a keto–enol

3. Results and discussions 3.1. Intermolecular interactions between cur and polymers 3.1.1. Raman imaging and spectroscopy Molecular interactions formed between drug and polymer(s) from nonpolar bond aspect can be analyzed by Raman spectroscopy (Haichen Nie, Marks, Taylor, Byrn, & Marsac, 2016). The Raman results of Cur ASDs are shown in Fig. 1. Since the Raman signal of Cur was

Fig. 1. Raman images and spectroscopy of Cur and Cur ASDs. (1) and (2) stand for two forms of Cur in ASDs. (1) represents for Cur with enol form, and (2) represents Cur with keto form. The formulations include A, Cur-E100; B, Cur-E100/HPMC 1:1; C, Cur-E100/HPMC 3:1; D, Cur-E100/HPMC 6:1; E, Cur-E100/HPMC 9:1.

545

Carbohydrate Polymers 181 (2018) 543–550

N. Fan et al.

Fig. 2. IR spectra of Cur, Cur ASDs and corresponding excipients.

E100/HPMC 6:1 or Cur-E100/HPMC 9:1, the shift of OH group moved to higher wavenumber. The result illustrated that Cur-E100/HPMC 1: 1 with the most amount of HPMC can form better hydrogen bond with Cur owing to the increased OH groups. As for Cur-E100, E100 did not have OH group, therefore the OH peak in its Cur ASDs derived from Cur. The OH peak of Cur in ASDs shifted by 18.2 cm−1 compared to Cur, indicating that the OH of Cur and the C]O of E100 formed hydrogen bond interaction. The C]O stretching vibrations assigned to the carbonyls for the relevant polymers are listed in Supplementary data (Table S3), along with the solid dispersion shifts observed in the Cur ASDs. The carbonyl group of the E100 can potentially form hydrogen bond with the phenolic OH groups in Cur. The result showed that the carbonyl group of the E100 shifted to lower wavenumbers, giving hints for the formation of hydrogen bond (Tang, Pikal, & Taylor, 2002). In the E100 spectrum, there were two peaks at 2820 cm−1 and 2770 cm−1 that corresponded to the non-ionized dimethylamino groups. It had reported that these dimethylamino groups of E100 can also potentially interact with Cur through ionic interactions (Wegiel et al., 2014).

Cur stabilized by a strong intramolecular hydrogen bond. Strong intramolecular hydrogen bonding made the keto–enol group less available for intermolecular hydrogen bonding between Cur with E100 or E100/HPMC mixture. Since two phenolic hydroxyl groups of Cur contributed to intermolecular hydrogen bonding, it might be anticipated that the hydrogen bonding forces formed between Cur and excipient will be weak due to limited possibilities for intermolecular hydrogen bonding. As for intermolecular interaction between Cur and polymer(s), the first consideration was the peak in the range of 3300 cm−1–3500 cm−1. The stretch band of OH in Cur ASDs between 3300 cm−1 and 3500 cm−1 belonged to polymer or Cur can be determined by overall peak morphology and intensity (Li et al., 2017). The stretch peak of OH in Cur IR spectrum was at 3421.6 cm−1. If the overall peak morphology and intensity of Cur ASDs were similar to the IR spectrum of polymer, the peak shift should compare Cur ASDs with polymer, not Cur ASDs and Cur. It was obvious that the peak morphology and intensity of Cur ASDs (except Cur-E100) was similar to the peak of excipient. Therefore, the peaks of Cur ASDs located in the range of 3300 cm−1–3500 cm−1 were compared to the peaks of corresponding excipient, which demonstrated that OH group of HPMC formed hydrogen bond with Cur. At the same time, the stretch of C]O from Cur shifted in Cur ASDs, indicating that the C]O in the Cur and the OH of the polymer formed hydrogen bond. As can be seen from Supplementary data (Table S2), when comparing Cur-E100/HPMC1:1 with Cur-E100/HPMC 3:1,Cur-

3.1.3. Effect of polymer on the equilibrium solubility of cur The measured equilibrium solubility of Cur and Cur ASDs are shown in Fig. 3. The equilibrium solubility data confirmed the poor water solubility of Cur and the addition of E100 or E100/HPMC mixture improved the Cur equilibrium solubility. The first way to improve the 546

Carbohydrate Polymers 181 (2018) 543–550

N. Fan et al.

Fig. 3. Equilibrium solubility of Cur with or without different excipients.

equilibrium solubility of Cur was by breaking intermolecular hydrogen bonds between Cur and Cur. Since the ionic interaction between Cur and E100 was stronger than the intermolecular hydrogen bond between Cur and Cur, E100 greatly increased the Cur equilibrium solubility. The Cur equilibrium solubility in most of E100/HPMC mixture solutions (E100/HPMC1:1, E100/HPMC 6:1, E100/HPMC 9:1) was lower than E100 solution possibly because the ionic interactions formed between Cur and E100 was weakened with the increase of HPMC. Additionally, the water-soluble polymers used in Cur ASDs contributed to the higher Cur equilibrium solubility than Cur. The reason was that the addition of polymers weakened the interaction between Cur molecules due to the interaction forces formed between polymers with Cur, thus leading to the enhancement of dissolving Cur molecules. According to the result of equilibrium solubility, HPMC had very little impact on the crystal solubility of Cur, which was consistent with the description of published literature (Raina, Alonzo, Zhang, Gao, & Taylor, 2014). The P-value derived from statistical analysis (Cur and Cur + E100, Cur and Cur + E100/HPMC 1:1, Cur and Cur + E100/HPMC 3:1, Cur and Cur + E100/HPMC 6:1, Cur and Cur + E100/HPMC 9:1 were 0.04, 0.0008, 0.004, 0.003 and 0.0001) confirmed the significant enhancement of Cur equilibrium solubility in E100 solution or E100/HPMC mixture solutions. 3.2. The impact of HPMC on inhibiting crystallization 3.2.1. Fluorescence spectroscopy It has reported that ASDs can undergo liquid-liquid phase separation (LLPS) in these highly supersaturated solutions as a precursor to crystallization (Jackson, Kestur, Hussain, & Taylor, 2016a). The initial concentration of LLPS has been defined to be the amorphous solubility. A colloidal drug-rich droplet phase originated from exceeded drug that has a metastable equilibrium with a water-rich phase containing amorphous drug (Indulkar, Gao, Raina, Zhang, & Taylor, 2016). It has reported that the occurrence of LLPS is beneficial in vivo (Xie, Gao, & Taylor, 2017), therefore, fluorescence spectroscopy was used to investigate the nature of phase separation phenomena occurring in solution (phase separation to a disordered liquid-like phase or crystallization) (Almeida e Sousa, Reutzel-Edens, Stephenson, & Taylor, 2015). The fluorescence spectroscopy images of Cur + E100 or Cur + E100/HPMC 6:1 are shown in Fig. 4, and others that include noise, Cur + HCl, Cur + HPMC, Cur + E100/HPMC 1:1, Cur + E100/HPMC 3:1 and Cur + E100/HPMC 9:1 are shown in Supplementary data (Fig. S4). The noise of fluorescence spectroscopy can be neglected since it did not influence the fluorescence signal of samples. There were two kinds of wavenumbers should be seriously considered in Fig. 4. The

Fig. 4. Fluorescence spectra of different concentrations of Cur in E100 solution or E100/ HPMC 6:1 solution.

wavenumber around 410 nm represented Cur with keto form, and the wavenumber around 440 nm stood for Cur with enol form. The main form of Cur in the acid medium was keto form, therefore, the wavelength of 410 nm was mainly observed. When amorphous Cur

547

Carbohydrate Polymers 181 (2018) 543–550

N. Fan et al.

concentration reached the highest level and the system was prone to crystallize during LLPS, the peak of 410 nm would display blue shift or disappear. Therefore, the concentrations of Cur ethanol solution were marked in Fig. 4 when the peak of 410 nm displayed blue shift or disappeared. As seen from result, when the peak of 410 nm displayed blue shift or disappeared, the amorphous Cur concentration in HCl, E100, HPMC and E100/HPMC3:1 were 20 μg/mL，and the amorphous Cur concentration in E100/HPMC 1:1, E100/HPMC 6:1 and E100/HPMC 9:1 were 40 μg/mL. It proved that the combination of E100 and HPMC was better than the single E100 in inhibiting crystallization. When E100 occupied a large proportion in E100/HPMC mixture, such as E100/ HPMC 6:1 and E100/HPMC 9:1, a small amount of HPMC can exert outstanding advantage in inhibiting crystallization.

Table 1 Particle size data of the different suspensions evaluated. Sample Cur + E100 Cur + E100/HPMC Cur + E100/HPMC Cur + E100/HPMC Cur + E100/HPMC

1:1 3:1 6:1 9:1

Mean particle size (nm)

SD

RSD(%)

76.05 204.7 205.7 108.0 159.5

7.251 8.054 9.199 15.49 2.491

9.53 3.93 4.47 14.3 1.56

3.2.2. The effect of polymer(s) on amorphous drugs concentration When the Cur supersaturated ethanol solution was added to different polymer aqueous solution, the LLPS occurred (Jackson, Kestur, Hussain, & Taylor, 2016b). The limited activity of the drug in the drugrich phase resulted in low solute concentration where LLPS occurred (Jackson et al., 2014). The addition of polymers to aqueous solution containing the drug-rich droplets can inhibit crystallization. As can be seen in Fig. 5, amorphous Cur concentrations in all excipient solutions were high at the beginning. It showed that the addition of HPMC in E100 solution to initially achieve high amorphous drugs concentration was not necessary. After 2 h, the amorphous Cur concentration significantly decreased, demonstrating that crystal Cur was generated in the system. The trend of amorphous Cur concentration at 2 h was consistent with that at 0 h. The amorphous Cur concentration in E100/ HPMC 1:1 mixture was the highest and the amorphous Cur concentration decreased as lowering the proportion of HPMC in E100/ HPMC mixture. It confirmed that HPMC was significant in maintaining the amorphous Cur concentration owing to the hydrogen bond interactions formed with Cur. 3.2.3. Particle size evaluation DLS analysis was performed to analyze Cur nano droplets generated during LLPS. The mean particle size, as determined from DLS analysis, suggesting that the droplet size was the smallest (around 76 nm) in E100 solution (Table 1). It was because the ionic interactions formed between Cur and E100 rendered the good compatibility of Cur and E100, resulting in small particle size of drug droplets. The result also can be confirmed by effect of E100 on amorphous drugs concentration during LLPS in the above study (3.2.2 The effect of polymer(s) on amorphous drugs concentration). The amorphous Cur concentration in E100 solution at 2 h was close to that at 0 h, which confirmed the good compatibility of Cur and E100. The drug droplet size in E100/HPMC 9:1 aqueous solution was bigger than E100/HPMC 6:1. This result was consistent with fluorescence spectroscopy analysis, demonstrating that when E100 occupied a large proportion in E100/HPMC mixture, a small amount of HPMC can exert obvious function in inhibiting

Fig. 6. In vitro release profiles of (a) Cur and a series of Cur ASDs formulated by E100; (2) Cur, Cur-E100, and a series of Cur ASDs formulated by E100 and HPMC.

crystallization (Wegiel, Mosquera-Giraldo, Mauer, Edgar, & Taylor, 2015). 3.2.4. In vitro dissolution First of all, in order to determine the optimal drug-the main polymer ratio, E100 was used as the main polymer. Cur ASDs were prepared with different weight proportions of Cur and E100. These Cur ASDs were named as Cur-E100 1:1, Cur-E100 1:2, Cur-E100 1:4, Cur-E100 1:6, Cur-E100 1:8 and the result was shown in Fig. 6(a). All the Cur ASDs improved drug dissolution compared with pure Cur, which was

Fig. 5. Amorphous Cur concentration in different excipient solutions.

548

Carbohydrate Polymers 181 (2018) 543–550

N. Fan et al.

Fig. 7. (a) The schematic illustration of membrane permeability enhancement caused by HPMC; (b) in vitro permeability profiles of Cur, Cur-E100, and a series of Cur ASDs formulated by E100 and HPMC. ** P < 0.05, ***P < 0.001.

evidenced by IR analysis resulted in crystallization inhibition effect of HPMC; (2) The high viscosity of HPMC improved the viscosity of the whole system and reduced Cur molecular mobility (Li et al., 2017).

ascribed to amorphous phase Cur distributed in the matrix of ASDs. For in vitro release profiles of Cur-E100 1:1, Cur-E100 1:2, Cur-E100 1:4 and Cur-E100 1:6, the Cur dissolution improved as increasing the E100 amount in Cur ASDs. It should be noted that in vitro release profile of Cur-E100 1:8 can not accord with this increasing trend because drug completely released in Cur-E100 1:6 and Cur-E1001:8 (The cumulative drug release were both close to 100%). After Cur completely released, amorphous Cur from Cur-E100 1:8 exhibited higher tendency to crystallize than Cur-E100 1: 6, resulting in its higher reduction of drug dissolution compared with Cur-E100 1: 6. In this case, Cur-E100 1: 6 not only showed the highest release amount at 30 min, but also had the lowest reduction of cumulative drug release in the following time period, confirming that the proper weight ratio of Cur-E100 should be settled at 1: 6 due to the significant good drug dissolution with low drug crystalline in solution. There are various mechanisms to inhibit crystallization during processing and storage of ASDs like reducing the drug molecular mobility and forming hydrogen bonds with the drug (Li & Konecke, 2013) (Li & Harich, 2013). Therefore, HPMC is a good choice as auxiliary polymer because it has high viscosity, form hydrogen bonds and promote gastrointestinal motility (Li et al., 2017). Thus, Cur-E100/ HPMC1:1, Cur-E100/HPMC 3:1, Cur-E100/HPMC 4:1, Cur-E100/ HPMC 6:1, Cur-E100/HPMC 9:1 were prepared and the results were shown in Fig. 6(b). During dissolution within 120 min, the reduced cumulative drug release of Cur-E100 was significantly higher than that of Cur-E100/HPMC 3:1, Cur-E100/HPMC 4:1, Cur-E100/HPMC 6:1 and Cur-E100/HPMC 9:1, except for Cur-E100/HPMC1:1. This result confirmed two points. (1) The hydrogen bond between Cur and HPMC that

3.3. The impact of HPMC on improving membrane permeability of cur ASDs PAMPA was used to detect membrane permeability of drug through passive diffusion. The result was shown in Fig. 7(b). The final penetration quality depended on two factors, including the amount of Cur dissolved in the medium and the amount of Cur permeated through the phospholipid bilayer in Cur ASDs. Obviously, the drug amount that penetrated from membrane for Cur-E100 did not present gradually increasing trend. On the contrary, the addition of HPMC in the formation of Cur ASDs showed increasing drug amount that penetrated through membrane along with time, confirming the impact of HPMC on improving drug membrane permeability. Furthermore, there were three key points that needed to be considered. Firstly, the drug membrane permeability amount of Cur-E100/HPMC 1:1 was higher than Cur-E100 after 6 h, though in vitro dissolution of Cur-E100/HPMC 1:1 was much lower than Cur-E100. As for the drug membrane permeability amount at 12 h, p-value derived from statistical analysis (compared Cur-E100 with Cur-E100/HPMC 1:3, Cur-E100 with Cur-E100/HPMC 1:6) showed that both Cur-E100/HPMC 1:3 and Cur-E100/HPMC 1:6 were significantly higher than Cur-E100. Thirdly, Cur-E100/HPMC 1:9 displayed lowest drug membrane permeability amount, possibly demonstrating that solution crystallization phenomena that occurred to CurE100/HPMC 1:9 (see Fig. 6(b)) hindered the overall drug membrane permeability quality. The above statements strongly convinced the 549

Carbohydrate Polymers 181 (2018) 543–550

N. Fan et al.

ability of HPMC in enhancing drug membrane permeability. The function mechanism of HPMC for improving drug membrane permeability was shown in Fig. 7(a). Firstly, under acidic conditions, H+ reacted with O of OH group on the chain of HPMC, so that the molecular chain was charged. Secondly, the molecular chain was attached to the methyl group of the hydrophobic part of the phospholipid bilayer. Finally, membrane permeability increased due to the disorder of phospholipid bilayer layer. Therefore, HPMC exerted function in promoting Cur membrane permeability in Cur ASDs formulated by E100.

Experimental Toxicology, 33(8), 800–813. Indulkar, A. S., Gao, Y., Raina, S. A., Zhang, G. G., & Taylor, L. S. (2016). Exploiting the phenomenon of liquid-liquid phase separation for enhanced and sustained membrane transport of a poorly water-soluble drug. Molecular Pharmaceutics, 13(6), 2059–2069. Jackson, M. J., Toth, S. J., Kestur, U. S., Huang, J., Qian, F., Hussain, M. A., et al. (2014). Impact of polymers on the precipitation behavior of highly supersaturated aqueous danazol solutions. Molecular Pharmaceutics, 11(9), 3027–3038. Jackson, M. J., Kestur, U. S., Hussain, M. A., & Taylor, L. S. (2016a). Characterization of supersaturated danazol solutions – Impact of polymers on solution properties and phase transitions. Pharmaceutical Research, 33(5), 1276–1288. Jackson, M. J., Kestur, U. S., Hussain, M. A., & Taylor, L. S. (2016b). Dissolution of danazol amorphous solid dispersions: Supersaturation and phase behavior as a function of drug loading and polymer type. Molecular Pharmaceutics, 13(1), 223–231. Kansy, M., Senner, F., & Gubernator, K. (1998). Physicochemical high throughput screening: Parallel artificial membrane permeation assay in the description of passive absorption processes. Journal of Medicinal Chemistry, 41, 7. Li, B., Harich, K., Wegiel, L., Taylor, L. S., & Edgar, K. J. (2013). Stability and solubility enhancement of ellagic acid in cellulose ester solid dispersions. Carbohydrate Polymers, 92(2), 1443–1450. Li, B., Konecke, S., Harich, K., Wegiel, L., Taylor, L. S., & Edgar, K. J. (2013). Solid dispersion of quercetin in cellulose derivative matrices influences both solubility and stability. Carbohydrate Polymers, 92(2), 2033–2040. Li, B., Konecke, S., Wegiel, L. A., Taylor, L. S., & Edgar, K. J. (2013). Both solubility and chemical stability of curcumin are enhanced by solid dispersion in cellulose derivative matrices. Carbohydrate Polymers, 98(1), 1108–1116. Li, J., Chang Li, X. W., Fan, N., Wang, J., He, Z., & Sun, J. (2017). Viewing molecular and interface interactions of curcumin amorphous solid dispersions for comprehending dissolution mechanisms. Molecular Pharmaceutics, 14(8), 2781–2792. Menjoge, A. R., & Kulkarni, M. G. (2007). Mechanistic investigations of phase behavior in Eudragit E blends. International Journal of Pharmaceutics, 343(1-2), 106–121. Motterlini, R., Foresti, R., Bassi, R., & Green, C. J. (2000). Curcumin, an antioxidant and anti-inflammatory agent, induces heme oxygenase-1 and protects endothelial cells against oxidative stress. Free Radical Biology & Medicine, 28(8), 1303. Rahman, M. A., Harwansh, R., Mirza, M. A., Hussain, S., & Hussain, A. (2011). Oral lipid based drug delivery system (LBDDS): Formulation, characterization and application: A review. Current Drug Delivery, 8(4), 330–345. Raina, S. A., Alonzo, D. E., Zhang, G. G., Gao, Y., & Taylor, L. S. (2014). Impact of polymers on the crystallization and phase transition kinetics of amorphous nifedipine during dissolution in aqueous media. Molecular Pharmaceutics, 11(10), 3565–3576. Sadzuka, Y., Nagamine, M., Toyooka, T., Ibuki, Y., & Sonobe, T. (2012). Beneficial effects of curcumin on antitumor activity and adverse reactions of doxorubicin. International Journal of Pharmaceutics, 432, 42–49. Serajuddin, A. (1999). Solid dispersion of poorly water-Soluble drugs: Early promises, subsequent problems, and recent Breakthroughs. Journal of Pharmaceutical Sciences, 88(10), 1058–1066. Sharma, O. (1976). Antioxidant activity of curcumin and related compounds. Biochemical Pharmacology, 25(15), 1811–1812. Stella, V. J., & Nti-Addae, K. W. (2007). Prodrug strategies to overcome poor water solubility. Advanced Drug Delivery Reviews, 59(7), 677–694. Tønnesen, H. H., Másson, M., & Loftssonb, T. (2002). Studies of curcumin and curcuminoids. XXVII. Cyclodextrin complexation: Solubility, chemical and photochemical stability. International Journal of Pharmaceutics, 244, 127–135. Tang, X. C., Pikal, M. J., & Taylor, L. S. (2002). A spectroscopic investigation of hydrogen bond patterns in crystalline and amorphous phases in dihydropyridine calcium channel blockers. Pharmaceutical Research, 19(4), 477–483. Wang, Y. J., Pan, M. H., Cheng, A. L., Lin, L. I., Ho, Y. S., Hsieh, C. Y., et al. (1997). Stability of curcumin in buffer solutions and characterization of its degradation products. Journal of Pharmaceutical and Biomedical Analysis, 15, 1867–1876. Wegiel, L. A., Mauer, L. J., Edgar, K. J., & Taylor, L. S. (2013). Crystallization of amorphous solid dispersions of resveratrol during preparation and storage-Impact of different polymers. Journal of Pharmaceutical Sciences, 102(1), 171–184. Wegiel, L. A., Zhao, Y., Mauer, L. J., Edgar, K. J., & Taylor, L. S. (2014). Curcumin amorphous solid dispersions: The influence of intra and intermolecular bonding on physical stability. Pharmaceutical Development and Technology, 19(8), 976–986. Wegiel, L. A., Mosquera-Giraldo, L. I., Mauer, L. J., Edgar, K. J., & Taylor, L. S. (2015). Phase behavior of resveratrol solid dispersions upon addition to aqueous media. Pharmaceutical Research, 32(10), 3324–3337. Xie, T., Gao, W., & Taylor, L. S. (2017). Impact of Eudragit EPO and hydroxypropyl methylcellulose on drug release rate, supersaturation, precipitation outcome and redissolution rate of indomethacin amorphous solid dispersions. International Journal of Pharmaceutics, 531(1), 313–323.

4. Conclusion The present paper reported the impacts of HPMC as assistant excipient on inhibiting crystallization and improving membrane permeability in Cur ASDs that formulated by E100. The results confirmed that HPMC was significant in maintaining the amorphous drug concentration owing to the hydrogen bond interactions formed with curcumin, rending its ability to inhibit crystallization by reducing drug droplet size. Furthermore, the addition of HPMC in Cur ASDs that formulated by E100 promoted drug membrane permeability through lowering the order level of phospholipid bilayer layer. It is believed that impact of HPMC on inhibiting drug crystallization and improving permeability in Cur ASDs that elucidated in this paper can provide valuable instruction for the study of Cur ASDs in future. Acknowledgement This work was supported by Innovation Team from Liaoning Education Department of China (No. L2014022). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.carbpol.2017.12.004. References Aggarwal, B. B., Kumar, A., & Bharti, A. C. (2003). Anticancer potential of curcumin: Preclinical and clinical studies. Anticancer Research, 23, 363–398. Almeida e Sousa, L., Reutzel-Edens, S. M., Stephenson, G. A., & Taylor, L. S. (2015). Assessment of the amorphous solubility of a group of diverse drugs using new experimental and theoretical approaches. Molecular Pharmaceutics, 12(2), 484–495. Bernabé-Pineda, M., Ramı́rez-Silva, M. T., Romero-Romo, M., González-Vergara, E., & Rojas-Hernández, A. (2004). Determination of acidity constants of curcumin in aqueous solution and apparent rate constant of its decomposition. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 60(5), 1091–1097. Chaudhari, S. P., & Dave, R. H. (2015). Evaluating the effects of different molecular weights of polymers in stabilizing supersaturated drug solutions and formulations using various methodologies of the model drug: Fenofibrate. Journal of Pharmaceutical Sciences and Pharmacology, 2(3), 259–276. Chen, J., Ormes, J. D., Higgins, J. D., & Taylor, L. S. (2015). Impact of surfactants on the crystallization of aqueous suspensions of celecoxib amorphous solid dispersion spray dried particles. Molecular Pharmaceutics, 12, 533–541. Haichen Nie, Z. L., Marks, B. C., Taylor, L. S., Byrn, S. R., & Marsac, P. J. (2016). Analytical approaches to investigate salt disproportionation in tablet matrices by Raman spectroscopy and Raman mapping. Journal of Pharmaceutical and Biomedical Analysis, 118, 328–337. Imbaby, S., Ewais, M., Essawy, S., & Farag, N. (2014). Cardioprotective effects of curcumin and nebivolol against doxorubicin-induced cardiac toxicity in rats. Human &

550