Hydrogenated soybean phosphatidylcholine (HSPC, COATSOME^®NC-21) and dipalmitoylphosphatidyl-gly- cerol (DPPG, COATSOME^®MG-6060LS) were purchased from Nippon Oil and Fats Co., Ltd. (Tokyo, Japan). Nifedipine (JPXIV, NI) was provided by Nippon Fine Chemical Co., Ltd. (Osaka, Japan). Trehalose was pur- chased from Wako Pure Chemical Industries, Ltd. (Osaka, Japan). The membrane filters (pore size: 0.2 µm) were purchased from Toyo Roshi Kaisha Ltd. (Tokyo, Japan). All of the reagents used in the current study were purchased as the highest grades commercially available, and all solutions were prepared using de-ionized dis- tilled water.

NI-SLN suspensions and the NI-SLNs lyophilized with trehalose were prepared in accordance with the methods reported in our previous papers [9] [10] . Briefly, to prepare the NI-SLNs, a mixture of 40 mg of NI and 1000 mg of lipids (HSPC:DPPG = 5:1 molar ratio) was physically mixed in a mortar for 5 min, and the resulting mixture was co-ground with an R3-IR roll mill (Kodaira Seisakusho Co., Ltd., Tokyo, Japan) for 5 min. The resulting roll mixture was dispersed in 200 ml of de-ionized distilled water, and premixed using a Speed Stabilizer (Kinema- tica Co., Luzern, Switzerland) at 9000 rpm for 10 min followed by a high pressure homogenization (Microflui- dizer^®, M110-E/H; Microfluidics, Co., Newton, MA, USA) at 175 MPa with the pass cycle of 100. The resulting suspension was filtered through a 0.2 µm membrane filter and stored in a refrigerator at 4˚C for further use.

To prepare the NI-SLN-Tre material, 2 ml of the SLN suspension prepared above was diluted to 4 ml with distilled water and added to a vial containing 25 mg of trehalose and mixed well. The vial was then frozen at −40˚C for 3 h and then freeze-dried in a glass chamber for 24 h using a vacuum pump equipped with a vapor condenser (−20˚C, 0.0225 Torr). The lyophilized material was then subjected to a secondary drying step, where it was dried at 20˚C for 24 h.

The mean particle size and zeta potential properties of the NI-SLN and rehydrated NI-SLN-Tre suspensions were evaluated using an electrophoretic light scattering (ELS) photometer (Zetasizer Nano ZS, Malvern, Wor- cestershire, UK). The NI content of each SLN suspension was measured in accordance with the technique used in our previous study [9] using reversed phase HPLC (Shimadzu, Kyoto, Japan) with a reversed phase Cadenza 5CD-C18 column (150 × 4 mm, i.d.; Imtakt) at 40˚C. HPLC analyses were conducted in the binary mode (dual pump) with ultraviolet (UV) detection at a wavelength of 236 nm. The mobile phase consisted of a 3:2 (v/v) mixture of methanol and aqueous formic acid (0.2% v/v), which was delivered in an isocratic manner at a flow rate of 0.4 ml/min. Under these conditions, the retention time of NI was found to be 6.1 ± 0.1 min.

The hemocompatibilities of the NI-SLN and reconsti tuted NI-SLN-Tre suspensions were evaluated using a he-

molysis assay [15] . Erythrocytes were isolated from 2 ml of fresh rabbit blood, and diluted with normal saline to an erythrocyte dispersion of 2%. Different volumes of both samples containing 145 µg/ml of NI (0.1, 0.2, 0.3, 0.4 and 0.5 ml) were added to five tubes together with 2.5 ml of the 2% erythrocyte dispersion. Normal saline was then added to each tube to obtain a final volume of 5 ml. Positive and negative controls were prepared by the addition of distilled water (2.5 ml) and normal saline (2.5 ml) to 2.5 ml 2% erythrocyte dispersions. The samples were incubated at 37˚C for 2 h and then centrifuged at 1500 rpm for 10 min, and the amount of released hemoglobin was measured spectrophotometrically at a wavelength of 545 nm (UV 1750, LC, Japan). The per- centage hemolysis values were calculated as follows:

where A_s is the absorbance of the sample, A_n is the absorbance of the negative control and A_p is the absorbance of the positive control.

As previously reported [9] , a slight increase in particle size was observed in the reconstituted NI-SLN-Tre, al- though some of the other physicochemical properties, including the polydispersity index (PDI) and zeta potential of the NI-SLN-Tre, remained constant (Table 1), suggesting that the use of only 2.5% trehalose exerted an ac- ceptable lyoprotective effect on the NI-LNPs. This effect was attributed to the membrane-protecting effect of trehalose, which forms hydrogen bonding interactions with the polar head groups of the lipids, as well as dis- rupting the tetrahedral hydrogen bonding network of water, leading to a reduction in the amount of freezable water.

Figure 1 shows the time courses of NI in the rat plasma samples following the oral administration of the NI- SLN and reconstituted NI-SLN-Tre suspensions. The pharmacokinetic parameters of these materials are shown in Table 2. Compared with the NI-SLN suspension, the NI-SLN-Tre suspension was slowly and highly absorbed

^aThese values are taken from Barman et al. [9] .

These data are the average values obtained from five experiments ± S.E. AUC: area under the curve; C_max: peak plasma concentration; T_max: time of peak concentration; t_1/2: half-life; k_e: elimination rate constant; MRT: first- order moment mean residence time.

with T_max and C_max values of 100 min and 1673.3 ng∙ml⁻¹, whereas the NI-SLN suspension gave values of 80 min and 1429.4 ng∙ml⁻¹, respectively. A comparison of the AUC_0→∞ values for the NI-SLN-Tre and NI-SLN suspensions revealed values of 273.3 and 238.8 min μg∙ml⁻¹, respectively, indicating that although no significant differences were observed between the T_max, C_max and AUC_0→∞ values of these materials, NI tended to be ab- sorbed to a much greater extent from the lyophilized NI-SLN-Tre suspension than it was from the NI-SLN sus- pension. This difference in the absorption properties of the two suspensions was attributed to the enhanced sol- vation of the NI-SLN-Tre in the GI fluid, derived from the formation of hydrogen bonding interactions between the polar head groups of the lipid and the O-H groups of the trehalose. Given that the absorption of NI has been postulated to occur following its solvation, it is possible that the NI in the NI-SLN-Tre would require a compa- ratively longer period of time to go into solution, resulting in the delayed T_max and enhanced C_max values. In con- trast, the t_1/2 and k_e values of the NI-SLN-Tre were 76.5 min and 0.0118 min⁻¹, respectively, whereas those of the NI-SLN suspension were 96.9 min and 0.0076 min⁻¹, respectively, suggesting that the NI in the lyophilized formulation was being eliminated from the plasma much more rapidly than that of the intact suspensions. In fact, since the MRT values of the two materials were almost identical (148.7 and 155.6 min for NI-SLN-Tre and NI-SLN, respectively), it is possible that these two materials could be subjected to different elimination mecha- nisms following their absorption from the GI tract.

The results of our previous paper demonstrated that the rapid absorption of NI from NI-SLN suspensions was only observed from suspensions of the intact NI drug, in that the C_max, AUC_0→∞, T_max and t_1/2 of values of the NI-SLN were found to be significantly higher than those of the corresponding NI suspensions and there was also a 4-fold increase in the absorption of NI from the NI-SLN suspension compared with the NI suspension [10] . This report indicated that the use of an NI-SLN suspension would enhance the absorption process through a re- duction in the mean particle size of the NI and two phospholipids. Several mechanisms working either alone or in combination may have contributed to the observed increases in the bioavailability of drugs from SLN formu- lations, including 1) changes in the surface active properties of the phospholipids can augment the absorption of drug compounds through variations in the fluidity of the GI membrane or an increase in the affinity of the lipid particles for the GI membrane; 2) variations in the effective absorption of drugs from the SLN because of the inhibitory effect of the DPPG surfactant towards p-glycoproteins, which would lead to a reduction in the efflux of NI from the intestinal membrane; 3) promotion of the dissolution characteristics of these nanomaterials be- cause of their enormous effective surface areas, which would promote the dissolution characteristics of the drug, leading to a higher concentration gradient and an increased rate of absorption; 4) the small sizes of these SLN materials would allow them to adhere more strongly to the GI tract as well entering into the intervillar spaces, which would lead to an increase in the residence time for increased bioavailability; 5) the presence of the lipids would favor lymphatic transport; and 6) the influence of lipids on the preferential uptake of lipid particles by Peyer’s patches would also result in an increase in the bioavailability of NI by avoiding clearance through re- sulting from first pass metabolism. Unfortunately, although the mechanisms involved in the observed increases in the bioavailability of NI from the NI-SLN and NI-SLN-Tre suspensions remain unclear, it is envisaged on the basis of these results that the SLN offers many potential advantages as a suitable carrier for the improved oral bioavailability of NI even after they had been subjected to lyophilization.

In the present study, the hemocompatibilities of the NI-SLN and NI-SLN-Tre suspensions were evaluated using an in vitro hemolysis assay. The results (Figure 2) revealed the effects of different amounts of both samples on the erythrocyte or total blood hemoglobin levels. When sample volumes of 0.1, 0.2, 0.3, 0.4 and 0.5 ml were added to the assay, the percentage hemolysis values for the NI-SLN suspension were 0.39, 0.43, 0.49, 0.60, and 0.85%, respectively, indicating a high level of hemocompatibility. In addition, the percentage hemolysis values following the addition of the reconstituted NI-SLN-Tre suspension at the same lipid concentration were found to be less than 5% for each sample. Values of less than 5% have been reported to be safe for erythrocytes [15] , suggesting that the NI-SLN-Tre suspension also showed a moderate level of hemocompatibility with blood cells. Taken together, these results therefore suggested that both the NI-SLN and lyophilized NI-SLN-Tre suspensions would show good biocompatibility.