Homogeneous nanoparticles to enhance the efficiency of a hydrophobic drug, anti-hyperlipidemic Probucol, characterized by solid-state NMR

Abstract

A low absorption in the gastrointestinal tract of hydrophobic pharmaceutical compounds in use today considerably limits their bioavailability and therefore they are taken in large doses in order to reach the therapeutic plasma concentration, which inevitably results in undesired side effects. In this study, we demonstrate a new nanoparticle approach to overcome this problem and our experimental results show that this approach has a high efficiency of drug loading and is easily adaptable to industrial scale. Characterization of nanoparticles containing a cholesterol-lowering hydrophobic drug, probucol, using a variety of biophysical techniques revealed higher homogeneity of these particles compared to those prepared using other approaches. Intermolecular interactions of these nanoparticles are probed at a high-resolution by magic angle spinning solid-state NMR experiments.

Introduction

Poor solubility and therefore a low absorption in the gastrointestinal tract of >40% of the pharmaceutical compounds in use today considerably limits their bioavailability after being administered.^1,2 As a result, these hydrophobic drugs must be taken in large doses in order to reach the therapeutic plasma concentration, which inevitably results in undesired side effects.³ Therefore, it is essential to develop novel techniques to enhance the effectiveness of hydrophobic drugs. Recent studies have shown that nanoparticle formulation by wet-grinding^4-6 or cogrinding^7-10 is one of the promising methods to increase the solubility of hydrophobic drugs. However, the wet-grinding approach typically results in submicron size drug particles which aggregate when dried for the preparation of solid dosage forms.^11,12 While the cogrinding approach resulted in nanoparticles (<100 nm), they are unfortunately heterogeneous in size and therefore less effective. In this paper, we demonstrate a new approach by integrating cogrinding and spray drying to enable hydrophobic pharmaceutical compounds to maintain a fully soluble state and at the same time increase its rate of absorption to reach the therapeutic plasma concentration. Our results on the nanoparticle formation of a hydrophobic cholesterol-lowering drug, probucol (PBC), demonstrate that this one-step continuous process of nanoparticle formation approach has a high efficiency of drug loading and is easily adaptable to industrial scale.^13-15 This concept is schematically shown in Figure 1.

Figure 1.

A schematic representation of the preparation of pharmaceutical fine particles containing probucol nanoparticles.

Experimental Section

Materials

Probucol (MW 516.84, crystalline powder of PBC-I form), 4,4′-{(1-methylethylidene)bis(thio)}-bis{2,6-bis(1,1-dimethylethyl)}phenol, was supplied by Dai-ichi Sankyo Co., Ltd. (Tokyo, Japan). Methacryclic acid-methylmethacrylate copolymer (1:1) (MMC, MW ca 135000), was provided from Evonik Degussa Japan Co., Ltd. (Tokyo, Japan). Sodium dodecyl sulfate (SDS, MW 288.38) was purchased from Sankyo Co., Ltd. (Tokyo, Japan). Ammonium hydroxide, triethyl citrate, Tween 80, glycerol monostearate were purchased from Wako Pure Chemical Industries Ltd. All other chemicals used in this study were of reagent grades.

Preparation of coground mixture (GM) and PBC-II form

In the ternary system, blend of PBC, MML and SDS (weight ratio of 1:5:1) were physically mixed in a glass vial using a vortex mixer for 10 sec. The physical mixture of 2.5g was ground in a vibrational rod mill (TI-200, CMT Co., Ltd., Fukushima, Japan, see Supporting information, Figure S1) for 10 min under ambient condition. PBC was alone ground under the same conditions mentioned above to obtain the crystalline powder of PBC-II form and used for solid-state NMR experiments.

Preparation of spray dried (SD) particles

(Preparation of suspension) The formulation of spray drying suspension was summarized in Table S1. GM was dispersed in water, then 1.7% ammonium hydroxide was gradually added and equilibrated with slow agitation via magnetic stirrer for 60 min. Polysorbate 80, triethy citrate and glycerol monostearate were dispersed in hot water (80°C) and left until lowering ambient temperature with stirring. This glidant suspension was poured into GM suspension for a further agitation overnight. (Spray drying process) A desk top spray dryer (SD-1, Tokyo Rika Corp., Tokyo, Japan, Figure S2) was used to prepare the spray dried particles. The processing parameters were as follows: inlet air temperature 85-120°C; outlet air temperature, 75°C (SD75), 90°C (SD90) and 105°C (SD105); blow flow, 0.6 m³/min; spray rate, 2.0 g/min; and atomizing air pressure, 0.8 kg/cm³ (see Supporting information, Table S2).

Particle morphology

Spray dried particles were viewed by Scanning electron microscope (SEM) measurements, JSM-6360LV (JEOL, Tokyo, Japan) and samples were adhered to a sample stage using an adhesive tape. A sputter coater, JEC-1600 (JEOL Ltd., Tokyo, Japan) was used to coat the samples with gold/platinum at 20 mA for 30 s. E800 POL polarized microscope (Nikon, Tokyo, Japan) was used to observe the crystalline components in a polymeric film. Optical images of samples were captured using a digital sight camera system DS-5M-L1 (Nikon Co., Tokyo, Japan).

Caco-2 permeability experiments^16,17

(Cell culture) The Caco-2 cells (ATCC, Rockville, MD, USA) were seeded at 3×10⁵cells onto polycarbonate filter membranes (pore size 3.0 m, growth area 4.67 cm²) in clusters of 6 wells (Transwell®, Corning Costar Corp., Cambridge, MA, USA). The cells were grown in a medium consisting of DMEM (4.5 g/L glucose) supplemented with 10% FBS, 1% NEAA, 1% L-glutamine, penicillin (100 IU/ml), and streptomycin (100 μg/ml). The cultures were maintained at 37°C (CO₂ incubator, SANYO, Japan) in an atmosphere of 5% CO₂ and 95% air, at 95% relative humidity. The growth medium was changed three times a week until time of use. Cells from passage numbers 18–21 were used in the experiments at ages ranging from 14 to 21 days. (Permeability experiments) The permeability of pre-suspended PBC in SD particles was studied across Caco-2 cell monolayers in an apical-to-basolateral (AP-BL) direction at an apical pH of 6.8, and basolateral pH of 7.4. Before the permeability experiments, the cell monolayers were rinsed twice with HBSS, pH 7.4, and equilibrated in the transport buffers under experimental conditions for 15 min. Transepithelial electrical resistance (TEER) was measured using a Millicell® ERS Voltohmmeter (Millipore Corp., Bedford, MA, USA) and monolayers with TEER values below 250 Ωcm² were discarded. Particle samples (0.1 mg of PBC per monolayer) were suspended immediately in HBSS prior to experiments and then applied to apical buffer solution. Basolateral samples were withdrawn from the receiving chamber at 15, 30, 45, 60, 75, 90, 105 and 120 min and immediately replaced by an equivalent volume of fresh HBSS. All of the transport experiments were conducted at least in triplicate (n = 3).

Particle size analysis of PBC nanoparticles

The SD particles dispersed into pH6.8 HBSS and then sonicated for 10 min to make the suspension. The volumetric particle size distribution for each suspension was determined at 37°C by the dynamic light scattering method using NICOP 380ZLS® (NICOMO Co., Ltd., USA).

Solid-state NMR experiments

Solid-state NMR experiments were performed on a 400 MHz Varian/Chemagnetics spectrometer (9.4 T) at the University of Michigan and final spectra presented in this paper were acquired using a 900 MHz Bruker spectrometer (21.1 T) at biomolecular NMR facility, East Lancing, MI using a 4 mm triple-resonance MAS probe. All spectra were obtained under 18,000 ±5 Hz MAS at 27°C. Experimental parameters, sample conditions and other details are given in figure captions. All data were processed using the NMR Pipe software.

3D structural modelling

Materials Studio (Accelrys Software Inc., San Diego, CA, U.S.A.) was used for the construction of the 3D structural model based on solid-state NMR experimental constraints.

Results and Discussion

Scanning electron microscope (SEM) photographs of spray-dried powders prepared at 75°C (SD75), 90°C (SD90) and 105°C (SD105) are shown in Figure 2. These powder samples are spherical shaped particles with a smooth surface as seen in Figure 2 that are composed of nanoparticles. Although the mean diameters of these spherical particles were similar (~3μm, see Table S3) among SD75, SD90 and SD105 samples, smaller sized particles were found abundantly in SD75 and less in SD105 as seen in Figure 2 (See Supporting Information, Figure S3 and Table S3). This means suggest that inner water phase gradually evaporated at lower temperatures providing smaller particles, whereas a crust was formed on the outer surface of the spray droplets at higher temperatures, resulting in larger particle size¹⁸ (See Supporting Information, Figure S4). The dynamic-light scattering measurement of the size distribution of drug nanoparticles present in the spherical spray-dried powder revealed the presence of ~47 nm sized nanoparticles in SD75 and SD90 samples (Figure 3). On the other hand, the size of relatively large (~75 nm diameter) nanoparticles observed from SD105 sample increased with time as shown in Figure 3.

Figure 2.

SEM photomicrographs of powders spray-dried at SD75 (a), SD90 (b) and SD105 (c).

Figure 3.

Particle size distributions of PBC nanoparticles after SD in pH 6.8 HBSS solution. SD75 (a), SD90 (b) and SD105 (c).

The percentages of the probucol drug released from SD75, SD90 and SD105 powders at pH 1.2 and pH 6.8 buffers, mimicking the stomach and small intestine pH environments respectively, are given in Figure 4(a). The PBC releasing percentages from all three samples were below 2% at pH 1.2 even after 240 min. On the other hand, SD75, SD90 and SD105 respectively released 80, 60 and 50% of PBC immediately at pH 6.8. These data infer that SD75 had the best enteric property and released most of the drug nanoparticles within 15 min. These results further suggest that (a) the agglomeration of drug nanoparticles occurred during the spray drying process at a higher temperature as shown in Figure S4 and (b) these nanoparticles exhibit pH-dependent functional properties and therefore the hydrophobic drugs can be effective in the alkaline environment of small intestine.

Figure 4.

(a) Drug release profiles of PBC from SD75 (●), SD90 (■), and SD105 (×) nanoparticles at pH 1.2 (red) and pH 6.8 (blue). (b) Permeation of PBC drug across the living Caco-2 cell monolayer at 37°C and pH 6.8. Each error bar indicates the mean + S.D. (n=3).

The Caco-2 cell monolayer from human intestines was used to examine the cell membrane permeation of the nanoparticles. Experimentally measured permeation rate of PBC from nanoparticles are given in Figure 4(b). Our results showed the best permeation for SD75, that is ~8 μg of PBC was transported from apical to basal side of monolayer over 120 min. The 2-fold faster permeation speed of SD75 nanoparticles compared to that of nanoparticles prepared without the spray drying process⁷ suggests that the spray drying process can drastically improve the efficiency of hydrophobic drugs. This Caco-2 cell permeation test is consistent with the dissolution study (shown in Figure 4(a)) and particle size analysis (shown in Figure 3) of cogrinding-spray-drying processed drug nanoparticles. The solubility, fraction and size of nanoparticles correlate with the dissolution rate according to the Noyes-Whitney equation;¹⁹ therefore smaller-size nanoparticles would be efficient for the cell membrane permeation.

Our results suggest that the combination of cogrinding and spray drying process has significantly improved the solubility and bioavailability of the hydrophobic PBC drug. MAS solid-state NMR experiments were performed to characterize the structural integration of nanoparticles at a high-resolution. ¹H as well as ¹³C resonances of the coground mixture were assigned using 2D ¹H/¹H NOESY and 2D ¹H/¹³C chemical shift correlation experiments under MAS⁷ (Figure S7 in the Supporting Information). The two types of PBC crystal forms were distinguished based on ¹H and ¹³C spectra (Figures S8 and S9 in the Supporting Information). NMR spectra of nanoparticles prepared using the combined cogrinding and spray drying process revealed the existence of the drug in PBC-II form as shown in Figure 5 and Figure 6. ¹H spectral lines of nanoparticles prepared with this approach are significantly narrower than that of nanoparticles prepared without spray drying as shown in Figure 5. This observation suggests that the spray drying process not only increased the homogeneity of the physical size of nanoparticles but also their chemical composition. It also infers that the intermolecular interactions that enable the nanoparticle formation are effectively utilized in the spray drying process.

Figure 5.

¹H chemical shift spectra of a ground mixture of MMC, SDS and PBC prepared with (bottom) and without (top) spray drying. The 1D spectra were obtained using a solid-echo pulse sequence with a 1 ms refocusing time under 18 kHz MAS at 27°C in a Bruker 900 MHz NMR spectrometer and 16 scans.

Figure 6.

(a and b) 2D ¹H/¹H NOESY spectra of a ground mixture of MMC, SDS and PBC prepared with spray drying were obtained using a pulse sequence given in Figure S7 (in the Supporting Information) with t = 1 ms and a mixing time of 11.11 ms (a) and 55.55 ms (b) under 18 kHz MAS at 27°C in a Bruker 900 MHz NMR spectrometer; 16 scans and 62 t₁ increments were used for (b and c) with the synchronization of the MAS spinning speed and the mixing time. (c) A structural model based on NMR constraints depicts molecular interactions between MMC, SDS and PBC and gives an insight for the water-soluble mechanism of PBC, a hydrophobic molecule.

To further probe these intermolecular interactions, NOESY spectra of nanoparticles prepared with (Figures 6(a) and 6(b)) and without (Figure S10 in the Supporting Information) spray drying were recorded. The increase in the spectral resolution due to spray drying further confirmed the higher homogeneity of the intermolecular interactions in nanoparticles prepared with spray drying than that of the nanoparticles prepared without spray drying.⁷

The intensity of cross peaks in the 2D ¹H/¹H NOESY spectra can be used to determine the through space distance between protons and possibly the structural assembly of molecules that constitute the nanoparticle. The presence of cross peaks between MMC and SDS (indicated by a circle in Figures 6(a) and 6(b)) suggests that these two molecules are near in space within the nanoparticle. Specifically, CH₂ of MMC and CH₂ of SDS are within ~9 Å distance. Also, CH₂ of MMC and CH₃ of PBC are within ~3 Å distance and CH₂ of SDS and CH₃ of PBC are within ~3 Å distance. These results suggest that the methyl groups of the PBC drug are stabilized by hydrophobic interactions with CH₂ of MMC and SDS. Therefore, it is highly likely that the optimization of the hydrophobic interaction among drug, SDS and MMC could further enhance the homogeneity and the functional efficiency of nanoparticles. Our results also demonstrate that solid-state NMR experiments can be used to examine the quality of nanoparticles and to provide atomic-level resolution insights on the role of intermolecular interactions towards the design of nanoparticles. While these nanoparticles are not amenable for studies using most commonly used high-resolution physical techniques like X-ray crystallography and solution NMR spectroscopy, NMR data presented in this paper suggest that the use of a variety of MAS techniques and recently developed solid-state NMR approaches^20-25 on nanoparticles containing molecules labeled with specific isotopes such as ¹³C and ²H could provide additional high-resolution insights on the structure, intermolecular assembly, and dynamics of individual molecules. Such information may be utilized in the design of more effective nanoparticles.

Conclusions

In summary, our results suggest that the nanoparticles of PBC prepared with spray drying maintain the pharmaceutical property, exhibit enhanced oral bioavailability under the physiological alkaline condition (pH 6.8), and effectively absorbed through the mucous membranes in the human small intestine. We believe that this novel approach of cogrinding and spray drying could be of significant advantage for effective use of probucol to patients who require a higher and faster rate of mucosal absorption when taken orally. Our method could be generally applicable for other poorly soluble compounds on an industrial scale production. In addition this technique may find applications in material science, particularly in the design of multi-functional nanoparticles consisting of hydrophobic compounds.