ABSTRACT
Objective: The objective of the present study was to formulate, optimize and characterize dual drug loaded nanoliposomal system containing metformin hydrochloride (MTN) and glimepiride (GLD).
Material and methods: Nanoliposomal formulations were optimized by Box-Behnken design using ethanol injection method. A 3-factor, 3-levels Box-Behnken design was used to derive a second order polynomial equation and construct contour plots to predict responses. The independent variables selected were molar ratio phospholipid:cholesterol (X1), aqueous phase concentration (X2) and lipid phase concentration (X3). The responses analyzed were particle size and entrapment efficiency.
Results: The final optimized formulation (LP-F) had the particle size, polydispersity index (PDI), %EEMTN, %EEGLD and %yield of 180.3 ± 2.69nm, 0.13 ± 0.07, 88.6 ± 1.99%, 99.3 ± 0.13% and 95.6 ± 1.23% respectively. TEM revealed Vesicles were dispersed and non-aggregated form with the average size of 185 ± 4.76 nm. SEM revealed 3D- surface view of the liposome which exposed spherical uniformly distributed particles.
Conclusion: From in-vitro release profiles of LP-F, it was observed that both the drugs have nearly simultaneous release profile in sustained manner.
Keywords: Nanoliposome, Box-Behnken dsign, Ethanol injection method, Antidiabetic, Metformin hydrochloride, Glimepride.

INTRODUCTION

Diabetes is a chronic disease caused either by insufficient production of insulin by pancreas (Type I) or inability of body’s cells to effectively utilize glucose (Type II). In both cases there is raised blood glucose level (hyperglycaemia), and might progress to serious damage to nerves and blood vessels, if uncontrolled. In 2014, it is estimated that 9% of adults above 18 years had diabetes and was the cause of 1.5million deaths during 2012 out of which >80% of diabetic mortality occurred in low- and middle-income countries (Guariguata et al., 2014). Type 1 or insulin-dependent diabetes is characterized by limited insulin production and necessitates daily administration of insulin. However, Type 2 or non-insulin dependent diabetes is more prevalent form and constitutes 90% of the diabetic population globally. For the effective management of type 2 diabetes mellitus various mono and combination therapies have been applied which include biguanides, sulfonylureas, meglitinide derivatives, alpha-glucosidase inhibitors, thiazolidinediones, glucagonlike peptide–1 (GLP-1) agonists, dipeptidyl peptidase IV (DPP-4) inhibitors, selective sodium-glucose transporter-2 (SGLT-2) inhibitors, insulins, amylinomimetics, bile acid sequestrants and dopamine agonists (Freeland and Farber, 2015).
Metformin hydrochloride (MTN) (N, N-dimethylimidocarbonimidicdiamide hydrochloride) belongs to biguanides class of anti-diabetic agent and has been used as first line treatment of type 2 diabetes. It acts by lowering hepatic gluconeogenesis, decreases intestinal absorption of glucose and increases insulin sensitivity by promoting peripheral glucose uptake and utilization. MTN is classified as class III in biopharmaceutical classification system (BCS) because of high aqueous solubility and poor intestinal permeability (Li et al., 2015) with absolute oral bioavailability of 50-60% (Dunn et al., 1995). It has been observed that there is an inverse relationship between the dose (ranging from 0.5 to 1.5g) and the relative absorption which is indicative of active absorption process.
Glimepiride (GLD) (1-[[p-[2-(3-ethyl-4-methyl-2-oxo-3-pyrroline-1-carboxamido) ethyl] phenyl] sulfonyl]-3-(trans-4-methylcyclohexyl) urea) is another antidiabetic agent of sulfonylurea class of anti-diabetic agents. It stimulates insulin secretion from the β-cells of pancreatic islets of Langerhans and is also known to increase peripheral insulin sensitivity thereby decreasing insulin resistance. After oral administration, GLD is completely absorbed from the gastrointestinal tract and Cmax is reached 2.4 to 3.75 h after dosing. GLD is completely metabolized in liver by CYP2C9 to the active M1 (hydroxy) metabolite, which is consequently dehydrogenation to form inactive M2 (carboxy) metabolite. All-phase half-life of GLD is 1.2 to 1.5 hours(Langtry and Balfour,1998).GLD has frequently been used in combination with metformin, thiazolidinediones, alpha-glucosidase inhibitors and insulin for improved diabetic control (Yamanouchi et al., 2005; Charpentier et al., 2001). Currently, many tablet dosage forms are available in market as combination of MTN and GLD. However, since the recommended dose of MTN is quite high (0.5-1.5g), it creates patient incompliance especially in elderly patients. Therefore, in this research we aim to develop dispersible liposomal formulation with simultaneous loading of MTN and GLD which is expected to increase the bioavailability, reduce the dose and increase the patient compliance. We have used ethanol injection method to prepare the liposome. Box-Behnken statistical design was applied to optimize the formulation and finally in vivo pharmacokinetics study was performed and compared with a marketed formulation Glycomet-GP2 (US Vitamins Limited, Mumbai, India) containing 500mg of MTN and 2mg of GLD.

MATERIALS AND METHODS

Materials
Metformin (MTN) and Glimepiride (GLD) were gift samples from Dr. Reddy’s Laboratories Ltd, Hyderabad, India. Glycomet-GP2®tablets containing 500mg of MTN and 2mg of GLD (manufactured by US Vitamins Limited, Mumbai, India) were procured from a local pharmacy. Phospholipid, Phospholipon®90H was a gift sample from Lipoid, GmbH, Germany. All other chemicals and reagents were of AR-grade and used without further purification.
High performance liquid chromatography
For the simultaneous quantification of MTN and GLD in liposomes and plasma, HPLC method used was adopted from previous study conducted by (Ramesh and Habibuddin, 2014). Shimadzu HPLC system consisting of quaternary pump, SPD-10AVP column oven, a variable-wavelength UV–visible detector. Supelco reversed phase C-18 column with internal diameter of 25 cm × 4.6 mm, 5-μm particle, was used. Mobile phase used was acetonitrile: phosphate buffer pH 3.0 (60: 40% v/v). For the calibration curve, different concentration having both MTN and GLD (MTN: 100, 200, 400, 600, 800, 1000µg/mL; GLD: 0.4, 0.8, 1, 1.2, 1.6, 2, 4µg/mL) was prepared in mobile phase and injected in HPLC system with injection volume of 25µL. The eluate was monitored at 235 nm. More information about drug quantification in liposomes, drug release medium and plasma is described in the respective sections.
Design of experiment
Box-Behnken statistical design with 3 factors, 3 levels and 17 runs was selected for the optimization study. Theoretically, experimental design consists of a set of points lying at the midpoint of each edge and the replicated center point of the multidimensional cube. The independent and dependent variables for liposome are listed in Table 1. Using the software Design Expert®9.0, a polynomial equation (quadratic model) was generated for experimental design for liposomal formulation.
Yi = b0 + b1X1 + b2X2 + b3X3 + b12X1X2 + b13X1X3 + b23X2X3 + b11X12 + b22X22 + b33X32
Where, Yi is the dependent variable; b0 is the intercept; b1 to b333 are the regression coefficients; X1, X2 and X3 are the independent variables (factors) that were selected on the basis of pilot experiments. The dependent variable or responses analysed were liposomal size, entrapment efficiencies of both MTN (%EEMTN) and GLD (%EEGLD). After the analysis of Box-Behnken system, Design Expert®9.0 software was used to get the formula for optimized liposome with minimum particle size and maximum entrapment efficiency. The formula obtained was prepared in the to get the final formulation (LP-F).
Preparation of nanoliposomes
17runs of liposomal formulation as per Box-Behnken design (Table 2) were performed by ethanol injection methodas mentioned by Pons et al (Pons et al.,1993). Briefly, suitable quantities of phospholipid, cholesterol and GLD (Table 3) was dissolved in99.9% ethanol (2.5mL)at 45°C to form lipid phase. Aqueous phase was prepared by dissolving required quantity of MTN in milli Q water (15mL). The aqueous phase was kept on a magnetic stirrer at 500rpm and at the temperature of 45±2°C. Total lipid phase was taken into 5mL syringe and rapidly injected into the aqueous phase and kept on stirring for 1h to get a liposomal dispersion. To get uniformity of vesicular size all batches of liposomal dispersion were subjected to probe sonication using probe sonicator [Vibra-Cell™ VC 750; Sonics, USA] for 1min at amplitude 35% and pulse 2sec: 5sec (on: off). Dispersion was then centrifuged at 10,000 rpm for 45min to separate liposomes from unetrapped drugs. The supernatant was removed and the sediment was re-dispersed in 10ml of water, frozen at -80°C and freeze-dried in a lyophilizer (Labconco Corp., MI, USA) using 10%w/w of sucrose with respect to the total solid weight.

Characterization of nanoliposomes
Particle size
Vesicle size was determined by dynamic light scattering technique using Zetasizer (Nano-ZS, Malvern Instruments, UK). 10mg of liposomal powder was dispersed in 20 mL of water, mixed thoroughly and analyzed at 25°C (Ahmad et al., 2014).
Entrapment efficiency
10mg of liposome was mixed with 10mL of HPLC mobile phase containing 0.1%v/v Triton X-100. The mixture was bath sonicated for 10min for complete lysis of vesicles and solubilization of drugs into mobile phase. The solution was then centrifuged at 10,000 rpm for 10min, 1mL of supernatant was filtered through 0.2µm syringe filter and 25µL was injected into HPLC system. Concentration (µg/mL) of MTN (CMTN) and GLD (CGLD) were determined using calibration curve and entrapment efficiency was calculated as follows:
Entrapment efficiency of MTN (%EEMTN) = QMTN/ (CMTN x T) x 100
Entrapment efficiency of GLD (%EEGLD) = QGLD/ (CGLD x T) x 100
QMTN=Quantity of MTN added (µg), QGLD=Quantity of GLD added (µg), T=Total weight of all ingredients in the run (mg)
Transmission electron microscopy
Morphology of optimized liposomal formulation (LP-F) was confirmed by transmission electron microscopy (TEM) (Morgagni 268D SEI, USA). Combination of bright field imaging at increasing magnification and diffraction modes was used to observe the form and size of the liposomes. In order to do the TEM observations, 10mg ofliposome was dispersed in 50ml water and a drop wasdeposited on the holey film grid and observed after drying.
Scanning Electron Microscopy (SEM)
Surface morphology of the optimized liposome was studied by SEM. The lyophilized formulation was coated with 5nm gold by sputter coating technology and observed through Zeiss Gemini 5 1530 FEG Scanning electron microscope (Carl Zeiss Microscopy GmbH, Germany).
Differential scanning calorimetry
Comparative differential scanning calorimetry of the optimized liposomal formulation (LP-F) was performed to obtain the phase transition temperature by DSC apparatus (Perkien Elmer, USA). 20mg of freeze-dried liposome (LP-F) and the amount equivalent to loaded MTN, GLD and pure phospholipid (Phospholipon®90H) was sealed in a 40µL aluminum crucible and empty aluminium crucible is taken as reference. The temperature of the pans was raised from 100-300°C, at a rate of 5°C/min. The heat flow calibration was performed with indium. The reproducibility of the thermograms was determined by repeating the temperature cycle three times for each sample.
In vitro drug release in simulated gastric and intestinal fluid
The drug release from LP-F was performed in simulated gastric and intestinal fluid using the dialysis bag method ( Singh et al., 2013). The dialysis bag with molecular weight cut off of 12KD was pre-soaked in double-distilled water for 12 h. 801.73mg of optimized liposome (LP-F) containing 500mg MTN and 2.25mg of GLD (calculated from entrapment efficiency and percentage yield data) was dispersed in 2 mL of milli Q water and put in the dialysis bag and a beaker containing 900 mL of dissolution medium. The temperature and rpm was adjusted to 37°C and 50 rpm respectively. 3 mL of sample was withdrawn at each time interval of 0, 0.25, 0.5, 0.75, 1, 2, 3, 4, 5, 6, 8, 12, 18 and 24h. The same volume of fresh dissolution medium was replaced with fresh medium each time the sample was withdrawn. All the samples were then vacuum dried at 50°C and 1mL of mobile phase (Section 2.2) was added, shaken, filtered through 0.2µ syringe filter and 25µL was injected into HPLC system for analysis. Marketed formulation GP-2 (containing 500mg MTN and 2mg GLD) tablet was triturated and dispersed in 2mL water and put into dialysis bag and release study was performed as for LP-F. All the operations were carried out in triplicate.

RESULTS

High performance liquid chromatography
HPLC method was performed for the quantification of MTN and GLD for the determination of entrapment efficiency, drug release study as well as stability studies. Method was cross validated with respect to the study presented by Ramesh and Habibuddin (2014) with some modifications and was found to be in concordance with the ICH guidelines. The method was found to be linear (R2 = 0.9992), sensitive, accurate, precise, robust and reproducible.
Box-Behnken Design and formulation optimization
Box-Behnken design was the chosen experimental design of the experiment. It is type of response surface designs, specifically made to require only 3 levels, coded as -1, 0, and +1. Box-Behnken designs can be applied for 3 to 21 factors or independent variables. They are basically a combination of two level factorial design and incomplete block design, which are able to provide desirable statistical properties in lesser number of runs as compared to full factorial design. In this experiment we used 17runs, with varying levels of independent factors. The responses i.e. liposomal size and encapsulation efficiencies of MTN and GLD were analyzed individually. In order to draw the mathematical relationship between the factors and responses various models were studied.

It was observed that liposomal size and entrapment efficiency of MTN followed quadratic model while entrapment of GLD does not show any significant variability with respect to changing levels of factors. The entrapment efficiency was more the 98% in each case. This effect might be due to very high ratio of phospholipid with respect to GLD (varied from molar ratio of 78.6 to 1179; Table 3) and nearly all the GLD molecule was entrapped within the liposomal bilayer. Quadratic equations which governs the liposomal size and entrapment efficiency of MTN (%EEMTN) are given below:
Liposomal Size (nm) = 228.54 -21.925 X1 -1.3 X2 + 61.875 X3+ 15.45 X1X2 + 9.3 X1X3 -2.75 X2X3 + 5.18X12 + 0.03 X22+ 16.33 X32
%EEMTN =79.18 -4.2375 X1 + 4.5 X2 + 3.438 X3-1.5 X1X2-5.825 X1X3 -1.7 X2X3 + -1.878 X12 + 5.798 X22 + 1.973X32
Characterization of optimized nanoliposomes
Percentage yield, Particle size, Polydispersity index and Entrapment efficiency
The formula given by the software were reproduced in lab in triplicate and final optimized formulation, LP-F, has the particle size, polydispersity index (PDI), %EEMTN, %EEGLD and %yield of 180.3 ± 2.69nm, 0.13 ± 0.07, 88.6 ± 1.99%, 99.3 ± 0.13% and 95.6 ± 1.23% respectively. The actual quantities of ingredients are also shown in the table 3. The obtained results were very close to predicted values which is practical validation of the experimental design. PDI value equal to zero is indicative of entirely mono-disperse vesicular size distribution while 1 indicates a completely poly-disperse system. In our case,PDI of 0.13 ± 0.07 was obtained which means more than 80% of the total particles was uniform in size. The size distribution pattern of LP-F is given in Figure 2.
Transmission electron microscopy and scanning electron microscopy
The transmission (TEM) and scanning (SEM) electron microphotograph of LP-F is shown in Figure 3A and 3B respectively. TEM revealed clear outline and the core of the well identified vesicles displaying the vesicular structure. Similar structural feature was reported earlier. Vesicles were presented in dispersed and non-aggregated form with the average vesicular size of 185 ± 4.76 nm which was in support of the size obtained with zetasizer.

SEM revealed 3D-surface view of the liposome which revealed spherical uniformly distributed particles.
Differential scanning calorimetry
DSC thermograms of liposome (LP-F), MTN, GLD and pure phospholipid (Phospholipon®90H) are illustrated in Figure 4A, 4B, 4C and 4D respectively. MTN, GLD and Phospholipid showed an endothermic peak at 207°C, 224°C and 124°Crespectively.
In vitro drug release in simulated gastric and intestinal fluid
The comparative in-vitro release profiles of LP-F and GP-2 tablet in simulated gastric (SGF) and intestinal fluid (SIF) is shown in Figure 5. It was observed that form GP-2, more than 90% of MTN release occurred within 3h in SGF and 4h in SIFdue very high water solubility while GLD release occurred relatively slowly and only about 50% of GLD was able to be released in 24h in both dissolution media.

Table 1. Independent variables (factors) and their levels. Click here to view full image

Table 2. 17 runs of the Box-Behnken design and their experimental responses. Click here to view full image

Table 3. Calculated values of ingredients in each run in mg and no. of moles. Click here to view full image

Table 4. Predicted formula given by Design Expert® 9.0 software and obtained results (n=3) for minimum particle size and maximum entrapment efficiency. Click here to view full image

Figure 1. 3D-response surfaces showing the variability of response with respect to simultaneous variation in levels of two simultaneous factors. Click here to view full image

Figure 2. It shows particle of uniform size with Polydispersity index (PDI). Click here to view full image

Figure 3. (TEM) and scanning (SEM) image of LP-F is shown in Figure 3A and 3B respectively. Click here to view full image

Figure 4. DSC thermograms of liposome (LP-F), MTN, GLD and pure phospholipid (Phospholipon®90H) are illustrated in Figure 4A, 4B, 4C and 4D respectively. Click here to view full image

Figure 5. The comparative in-vitro release profiles of LP-F and GP-2 tablet in simulated gastric (SGF) and intestinal fluid (SIF) are shown in the respective figure. Click here to view full image

DISCUSSION

Analysis of variance (ANOVA) on each equation showed that for particle size equation terms X1, X3, X1X2 and X32 significantly effecting the response while for % EEMTN all the terms i.e. X1, X2, X3, X1X2, X1X3, X2X3, X12, X22 and X32 are significant (p<0.05). For particle size, increased Ph:Ch ratio has negative effect because of lower entrapment of cholesterol molecules in the bilayer leading to enhanced flexibility and closer phospholipid arrangement. Concentration of aqueous phase had insignificant effect on the particle size of the liposome but concentration of lipid phase has positive effect which is due to more availability of phospholipid molecule to form the bilayer. Similar results were also observed in previous study conducted by Pons et al. (1993). Although, aqueous phase concentration (X2) itself has insignificant effect on the particle size, but it is affecting the response by the factor, X1 as observed with the interaction term X1X2 which is negating the size lowering effect of factor X1. This could be because at higher aqueous phase concentration, there might be slower interaction between the hydrophobic tails of phospholipid to form the liposome. For %EEMTN the quadratic equation revealed that all the terms including higher order and interaction terms are governing the response. Ph:Ch ratio has the negative effect on the entrapment of MTN because low cholesterol ratio is known to cause loss of bilayer rigidity and leaching of hydrophilic drug from the liposomal core. Higher aqueous phase concentration lead to higher encapsulation of MTN. The magnitude of this effect was nearly similar to the factor X1 but in inverse direction. Concentration of lipid phase has positive effect on the entrapment efficiency this might be because of relatively larger vesicle size as well as higher number of vesicles. All the three interaction terms X1X2, X1X3 and X2X3have negative effect on the entrapment efficiency. From the equation it can be concluded that factor X1 that is the Ph:Ch is dominant over both X2 and X3 as it is inversing the effect produced by the later. In case of X2X3, the factors individually have positive effect on the %EEMTN but they are minimizing the effects of each other. The 3D-response surfaces taking two variables at a time show the correlation between the factors and response are presented in Figure 1. This response surface helps to analyze the variability of response at different levels of factors, where it is easier to determine if there is interaction between the two factors. From the data of particle size, %EEMTN and %EEGLD, optimization was performed to derive an optimal formula which could provide lowest particle size, highest %EEMTN and %EEGLD within range of 98.8 to 99.9%. The predicted optimized formula in terms of coded factors, their actual values are given in Table 4. DSC results revealed that drug loaded liposome (LP-F) showed disappearance of all characteristic melting endotherms of ingredients except that of phospholipid which suggests that there is a complete encapsulation of drug in the formulations. Similar results were also observed in previous studies as well with liposomal and niosomal preparations (Chen et al., 2012; Hathout et al., 2007; Akhter et al., 2012). The remarkable difference in the release profile could have direct impact on the therapeutic efficacy since both would have different gastrointestinal absorption and the achievement of combination tablet is not up to the mark. However, with LP-F, it was observed the both the drugs have nearly simultaneous release profile in sustained manner. The liposomal encapsulation of GLD, a lipophilic drug would have been in the form of a lipid bilayer where the drug is molecularly embedded and losses its crystallinity and thereby assuring its easy release in the dissolution medium. Simultaneously, MTN was released in a sustained manner from the liposomal core. In both media, the initially, there was higher rate of GLD release from LP-F, but later on the rate of MTN release was higher. The possible reason would be the behavior of lipid bilayer as a rate controlling membrane MTN. This sort of simultaneous release of both drug could be expected to get increased therapeutic effectiveness. The drug release through the liposomal system have added benefit of sustained release and hence longer duration of action can be expected at relatively lower dose.

CONCLUSION

Nanoliposome containing metformin hydrochloride (MTN) and glimepiride (GLD) formulations were successfully optimized by employing statistical tool. The higher entrapment efficiency, smaller particle size and prolonged release were achieved. The effect of different constituents on dependent variables was successfully investigated and good quality correlation was established by using statistical design. The drug release through the liposomal system has added benefit of sustained release and hence longer duration of action can be expected at relatively lower dose.

CONFLICT OF INTEREST

None declared.

REFERENCES
Ahmad I, Akhter S, Ahmad MZ, Shamim M, Rizvi MA, Khar RK, et al. Collagen loaded nano-sized surfactant based dispersion for topical application: formulation development, characterization and safety study. Pharmaceutical development and technology. 2014;19(4):460-7.
Akhter S, Kushwaha S, Warsi MH, Anwar M, Ahmad MZ, Ahmad I, et al. Development and evaluation of nanosized niosomal dispersion for oral delivery of Ganciclovir. Drug development and industrial pharmacy. 2012;38(1):84-92.
Charpentier G, Fleury F, Kabir M, Vaur L, Halimi S. Improved glycaemic control by addition of glimepiride to metformin monotherapy in type 2 diabetic patients. Diabetic Medicine. 2001;18(10):828-34.
Chen Y, Wu Q, Zhang Z, Yuan L, Liu X, Zhou L. Preparation of curcumin-loaded liposomes and evaluation of their skin permeation and pharmacodynamics. Molecules. 2012;17(5):5972-87.
Dunn CJ, Peters DH. Metformin. Drugs. 1995;49(5):721-49.
Freeland B, Farber MS. Type 2 diabetes drugs: a review. Home healthcare now. 2015;33(6):304-10; quiz 11-2.
Guariguata L, Whiting DR, Hambleton I, Beagley J, Linnenkamp U, Shaw JE. Global estimates of diabetes prevalence for 2013 and projections for 2035. Diabetes research and clinical practice. 2014;103(2):137-49.
Hathout RM, Mansour S, Mortada ND, Guinedi AS. Liposomes as an ocular delivery system for acetazolamide: in vitro and in vivo studies. AAPS PharmSciTech. 2007;8(1):1.
Langtry HD, Balfour JA. Glimepiride. Drugs. 1998;55(4):563-84.
Li Y, Song J, Tian N, Cai J, Huang M, Xing Q, et al. Improving oral bioavailability of metformin hydrochloride using water-in-oil microemulsions and analysis of phase behavior after dilution. International journal of pharmaceutics. 2014;473(1-2):316-25.
Morilla MJ, Benavidez P, Lopez MO, Bakas L, Romero EL. Development and in vitro characterisation of a benznidazole liposomal formulation. International journal of pharmaceutics. 2002;249(1-2):89-99.
Pons M, Foradada M, Estelrich J. Liposomes obtained by the ethanol injection method. International journal of pharmaceutics. 1993;95(1):51-6.
Ramesh D, Habibuddin M. Stability Indicating RP-HPLC Method for the Simultaneous Determination of Atorvastatin Calcium, Metformin Hydrochloride, and Glimepiride in Bulk and Combined Tablet Dosage Form. International Scholarly Research Notices. 2014;2014.
Singh A, Ahmad I, Akhter S, Jain GK, Iqbal Z, Talegaonkar S, et al. Nanocarrier based formulation of Thymoquinone improves oral delivery: stability assessment, in vitro and in vivo studies. Colloids and surfaces B, Biointerfaces. 2013;102:822-32.
Yamanouchi T, Sakai T, Igarashi K, Ichiyanagi K, Watanabe H, Kawasaki T. Comparison of metabolic effects of pioglitazone, metformin, and glimepiride over 1 year in Japanese patients with newly diagnosed Type 2 diabetes. Diabetic medicine. 2005;22(8):980-5.
Zeisig R, Eue I, Kosch M, Fichtner I, Arndt D. Preparation and properties of sterically stabilized hexadecylphosphocholine (miltefosine)-liposomes and influence of this modification on macrophage activation. Biochimica et biophysica acta. 1996;1283(2):177-84.