| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Riddet Centre, Massey University, Private Bag 11 222, Palmerston North, New Zealand
1 Corresponding author: h.singh{at}massey.ac.nz
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Key Words: milk fat globule membrane • liposome • encapsulation • Microfluidizer
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Phospholipids have been shown to have a number of health benefits, including liver protection (Koopman et al., 1985) and memory improvement (Crook et al., 1991, 1992). Sphingolipids are required for cellular signaling, and have been shown to be involved in the control of cell proliferation, apoptosis, inflammation, and cancer (Huwiler et al., 2000). Sphingomyelin also inhibits intestinal absorption of cholesterol and fat in rats, with milk SM being more effective than egg SM (Peel, 1999). Sphingolipids are traditionally extracted from bovine brain, and are not only very expensive but also unsuitable for use in foods for vegetarians. Issues relating to bovine spongiform encephalopathy and Creutzfeldt-Jakob disease make it undesirable to use material extracted from bovine brain in food systems. The relatively high concentrations of sphingolipids in dairy phospholipid could avoid many of the negative issues surrounding many of the other sources of phospholipids and sphingolipids.
The identification of these biological functions of phospholipids, in particular sphingolipids, has led to increasing interest in techniques for isolating phospholipid fractions from waste dairy streams, such as buttermilk. These techniques range from traditional methods using solvent extraction to emerging technologies such as microfiltration and supercritical fluid extraction (Astaire et al., 2003; Corredig et al., 2003). Currently, the Fonterra Cooperative Group Ltd. (New Zealand) is the only company known to extract and purify MFGM phospholipid fractions from buttermilk commercially. In addition to their biological functions, MFGM phospholipids have been to shown to have good emulsification properties (Corredig and Dalgleish, 1998; Roesch et al., 2004), and have been used for the production of emulsions for drug delivery (Sato et al., 1994; Yuasa et al., 1994).
In the pharmaceutical and cosmetic industries, highly purified phospholipids extracted from soy oil or egg yolk are used to produce liposomes. Liposomes are spherical structures consisting of one or more phospholipid bilayers enclosing an aqueous core (Zeisig and Cämmerer, 2001). They may be used for the entrapment and controlled release of drugs or nutraceuticals, as model membranes or cells, and for specialist applications such as gene delivery (Lasic, 1998).
There are many potential applications for liposomes in the food industry, ranging from the protection of sensitive ingredients to increasing the efficacy of food additives. However, the high cost of the purified soy and egg phospholipids, combined with problems in finding a production method suitable for use in the food industry, has limited the use of liposomes in foods. To our knowledge, MFGM-derived phospholipids have never been used in making liposomes. The high levels of sphingolipids in MFGM phospholipids may provide nutritional benefit for the consumer, as well as improved liposome functionality.
There are several methods that may be used to produce liposomes, and a number of excellent reviews have been published that provide preparation details of the more common production techniques (Watwe and Bellare, 1995; Betageri and Kulkarni, 1999; Frezard, 1999). The standard preparation procedure is via the rotary evaporation of a chloroform solution of phospholipid, cholesterol, and other hydrophobic compounds to produce a thin phospholipid film. Addition of water and hydrophilic compounds causes bilayer sheets of the lipid to separate from the bulk and form liposomes (Picon et al., 1994). Jackson and Lee (1991) stated that the large-scale production of liposomes was limited by poor encapsulation efficiencies, the lack of a continuous production process, and the use of organic solvents. They concluded that the solution to this problem might be the use of a microfluidization technique. The Microfluidizer is a high-pressure homogenizer that can rapidly produce a large volume of liposomes in a continuous and reproducible manner (Chen et al., 2001), without use of sonication, detergents, solvents, or alcohols. The liposome population produced appears to be relatively stable, without rapid aggregation or fusion (Kim and Baianu, 1991). In microfluidization, the phospholipid and the material to be entrapped are dispersed in a liquid phase. This may be water, an aqueous buffer solution, or a solvent, depending on the solubilities of the components. The solution is pressurized in continuous flow, and split into 2 streams that are then forced together at high velocity (> 500 m/s). The resulting release of kinetic energy provides the required activation energy to break up the large phospholipid bilayer sheets into smaller fragments (Kim and Baianu, 1991). To minimize surface energy, the ends wrap around, forming bilayer vesicles known as liposomes.
The objective of the research presented in this paper was to study the formation of liposomes using a microfluidization technique. A phospholipid-rich fraction isolated on an industrial scale from MFGM was used to prepare liposomes.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Fatty Acid Profile of Phospholipid Fraction
The lipid was extracted from the phospholipid fraction using chloroform and methanol in a ratio of 1:2 (by volume). Fatty acids were methylated by acid-catalyzed transesterification at 80°C for 12 h in a sealed tube. The fatty acid methyl esters were separated using a BPX-70 capillary column, 100 m x 0.22 mm i.d., 0.25-µm film (SGE, Melbourne, Australia). The gas chromatographic system consisted of a model 6890 gas chromatograph equipped with an autosampler (HP7673) and Chem Station integration (all Hewlett Packard, Avondale, PA). The column oven was held at an initial temperature of 165°C for 52 min, and then increased at a rate of 5°C/min to a final temperature of 210°C for 59 min (total run time: 120 min). Both the injector port and the flameionization detector port were at 250°C. The carrier gas flow (helium) was maintained at 1.0 mL/min (linear gas velocity: 20 cm/s) throughout the temperature program with an inlet split ratio of 30:1. Fatty acid peaks were identified by retention time matching with authentic standards, including a composite standard made from commercially available methyl esters (NuChek Prep, Elysian, MN; Sigma, St. Louis, MO).
Analysis of Phospholipid Head Group
31Phosphorus-nuclear magnetic resonance (31P-NMR) analysis was conducted on the MFGM phospholipid material in its original powder state. The analysis was performed by Spectral Service (Köln, Germany) using a Bruker AC-P 300-MHz NMR spectrometer (Diehl, 2001, 2002).
Preparation of Liposome Dispersions
A 10% lipid dispersion was made in imidazole buffer (20 mM imidazole, 50 mM sodium chloride, and 0.02% sodium azide in Milli-Q water, adjusted to pH 7, with 1 M HCl) and thoroughly mixed using an Ultra-Turrax blender (JKA, Staufen, Germany). The phospholipid dispersion was then processed using an M-110Y Microfluidizer (Microfluidics International Corp., Newton, MA) with a 75-µm F12Y-type interaction chamber (Talsma et al., 1989; Larivière et al., 1991).
Determination of Size and Size Distribution of Liposomes
The average hydrodynamic diameter of the liposome dispersions was measured on a Zetasizer 4 (Malvern Instruments Ltd., Worcestershire, UK) using photon correlation spectrometry. Samples of the liposome dispersions were diluted in imidazole buffer to the required turbidity (< 250 kilocounts/s). Preliminary experiments had found no effect of concentration on the measured diameter at turbidities of up to 500 kcps. Each sample was analyzed 3 times at 25°C with a sampling time of 99 s and a scattering angle of 90°. A medium viscosity of 1.054 cP and a refractive index of 1.34 were used for the aqueous phase, with a typical liposome refractive index of 1.45 (Blessing et al., 1998; Ardhammer et al., 2002).
Asymmetrical flow field-flow fractionation was used to provide more detailed information regarding the size distribution of the liposome dispersions. This was performed using a Postnova Avalanche AF4 AFFF (Post-nova, Munich, Germany) equipped with a Postnova refractive index detector (PN 3140) and a PD Expert multiangle dynamic and static light-scattering system (Precision Detectors, Bellingham, MA). A channel spacer of 0.25 mm and a field programming method using power field decay were used. The field was initially held constant at 70% for 4.8 min, and then decayed at the rate of –p x 4.8 min, where P = 2 (to obtain a constant fractionation power). This method was designed using the software provided with the system to give the optimum separation (in terms of maximum resolution) in 60 min for particles between 7.5 and 750 nm in diameter. The crossflow and the expected particle size of eluted material as a function of time are shown in Figure 1
. The channel outlet flow was held constant at 0.3 mL/min and the total run time was 60 min.
|
Determination of 
-Potential
Liposome dispersions were prepared using 0.1 M NaCl, and the pH was adjusted using 1 M HCl and 1 M NaOH. The samples were diluted as required using 0.1 M NaCl solution that had been adjusted to the specific pH. The -potential was measured using the Zeta-sizer 4 (Malvern Instruments Ltd.) with an AZ104 cell. Five measurements of 25 s duration at 100 mV were used to measure the -potential at the stationary layer 14.63% of the capillary diameter in from the wall.
Negative-Staining Transmission Electron Microscopy
The liposome dispersions were diluted approximately 1:10 with distilled water. One drop of the diluted sample was combined with a drop of 2% ammonium molybdate and left for 3 min. The solution was then placed on a copper mesh for 5 min before the excess liquid was drawn off with filter paper. The mesh was examined using a Philips 201C transmission electron microscope (Eindhoven, The Netherlands).
Thin-Section Transmission Electron Microscopy
The liposome dispersion was mixed with low-temperature gelling agarose. The agarose-embedded samples were cut into ~1 mm3 cubes, placed in a bijoux bottle containing 3% glutaraldehyde in 0.2 M sodium cacodylate buffer, and kept at 5°C for 24 h. The glutaraldehyde was removed by rinsing twice with 0.2 M sodium cacodylate buffer for 2 h, and then the samples were left in 1% osmium tetroxide overnight at room temperature. The samples were washed twice with distilled water, placed in 1% uranyl acetate for 30 min, and then washed twice more with distilled water.
The embedded samples were dehydrated at 5°C using 25% acetone for 15 min, and then 50, 70, and 90% acetone for 30 min each, followed by 100% acetone. The acetone was replaced with Procure 812 embedding resin, and the samples were put on rollers for 24 h. A cube of each sample was placed into an embedding capsule, which was then cured at 60°C for 48 h.
The samples were then sectioned to a thickness of 90 nm using a Reichert Ultracut microtome. These sections were mounted on 3-mm copper grids and stained with lead citrate before examination in a Philips 201C transmission electron microscope at an accelerating voltage of 60 kV.
Cryo-Field Emission Scanning Electron Microscopy
Samples of liposome dispersions were plunge-frozen using liquid propane, before being frozen with slushy liquid nitrogen at –140°C. They were then fractured with a knife on an Alto 2500 cryo stage (Gatan, Abingdon, UK), and freeze-etched by raising the temperature from –140 to –90°C before returning to –140°C. The exposed surface was then coated with Au and Pd for 120 s, before being examined under a JSM-6700F field emission scanning electron microscope (JEOL, Tokyo, Japan). An accelerating voltage of 10 kV was used.
Determination of Phospholipid Oxidation
The peroxide value and the level of conjugated dienes were measured before and after microfluidization to provide an indication of the effect of the processing on phospholipid oxidation. The peroxide value was obtained using a technique based on ISO 3960:2001. Briefly, 1 mL of the liposome dispersion was dissolved in 6 mL of acetic acid:chloroform (3:2), and 0.5 mL of saturated potassium iodide solution was added. The sample was mixed for 1 min, with further addition of 6 mL of distilled water. The solution was titrated against 0.1 M sodium thiosulfate solution until the yellow color had almost disappeared. Then, 0.5 mL of a 1.0% starch solution was added, and titration was continued until the blue color disappeared. The peroxide value was calculated as milliequivalents of peroxide per 1,000 g of sample.
The levels of conjugated dienes and trienes formed during oxidation were determined by a method based on IUPAC method no. 2.505 (IUPAC, 1987) and Lethuaut et al. (2002). A 25-µL aliquot of the liposome dispersion was dissolved in 10 mL of isopropanol, mixed for 4 s, and centrifuged for 5 min at 2,500 x g (CentraMP4R centrifuge, International Equipment Company, Needham Heights, MA). The absorbance was read against a blank containing 10 mL of isopropanol and 25 µL of Milli-Q water at 232 nm (linoleic hydroperoxides and conjugated dienes) and 268 nm (conjugated trienes and secondary products).
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
). This is in general agreement with previous reports on the phospholipid composition of MFGM (Boyd et al., 1999; Keenan and Mather, 2002; Astaire et al., 2003). This composition is considerably different from that of phospholipids derived from other sources. For example, the phospolipid products manufactured by Avanti Polar Lipids (Alabaster, AL) derived from egg yolk are composed almost entirely of PC, with small amounts of PE and PI, and a very small amount of SM. Soy phospholipid material may contain PC, PE, and PI, but does not contain any SM.
|
The fatty acid profile of the phospholipid fraction is shown in Table 2. As expected, this fraction contained relatively high proportions of saturated and monounsaturated fatty acids, with palmitic acid (16:0), stearic acid (18:0), and oleic acid (18:1) being the most common. These values are in general agreement with those reported by Fauquant et al. (2005) for the fatty acid profile of MFGM; however, there was a significantly higher proportion of C22:5 fatty acid chains than reported in the literature. This may be a consequence of the selective removal of buttermilk components during fractionation. In contrast, the phospholipids of soy or egg origin tend to have a higher proportion of mono- and polyunsaturated fatty acids, and are predominantly composed of palmitic acid (16:0), oleic acid (18:1), and linoleic acid (18:2) (Weiner, 1995; Pheko et al., 1998).
|
Particle Size and Size Distribution
The first pass through the Microfluidizer had the largest effect on liposome size, resulting in a decrease in average hydrodynamic diameter from over 500 nm to between 100 and 150 nm, with a polydispersity of between 0.4 and 0.5, as measured using photon correlation spectrometry. Successive passes continued to reduce the average size, but the changes were quite small. Increasing the number of passes had no significant effect on the polydispersity.
Figure 2 shows the effect on the average liposome diameter of passing the phospholipid dispersions [10% (wt/wt)] through the Microfluidizer a number of times at different operating pressures. The trends in the average diameter at the 3 operating pressures were approximately in parallel. Increasing either the operating pressure or the number of passes decreased the hydrodynamic diameter of the liposomes. It was evident that the effects of the pressure and the number of passes were additive. There was an approximately 40% decrease in the diameter between the largest (produced at 1 pass at 73 MPa) and the smallest (produced at 10 passes at 103 MPa) average liposome diameter measured.
|
. The dispersions containing 5 and 10% phospholipid showed a slightly larger reduction in average diameter after one pass through the Microfluidizer, but additional passes caused smaller size reductions in the 10% dispersion than in either the 1 or 5% phospholipid dispersions. This suggests that the 10% phospholipid dispersion was less sensitive to the shear and turbulence produced by the microfluidization process. This would be expected as the higher viscosity in more concentrated dispersions could resist deformation and breakup of liposome particles.
|
. This provided an indication of the relative proportion of liposomes present for each diameter. However, care must be taken when interpreting the results as increasing the particle size results in increased scattering for the same number of particles.
|
The trends observed for the effects of pressure, number of passes, and phospholipid concentration on the average liposome size agree with those reported by Barnadas-Rodríguez and Sabés (2001). They used a smaller laboratory-scale Microfluidizer (model 110S) to investigate the effect of a number of production variables on liposome size. They found that the mean liposome diameter decreased with increasing pressure and number of passes. Bachmann et al. (1993) used a high-pressure homogenizer to produce liposomes and found that, although higher homogenization pressures and repeated recirculation led to reductions in vesicle diameter and heterogeneity, size reduction was less effective at phospholipid concentrations above 10%. This supports the observations shown in Figure 3
, with the 10% phospholipid dispersion having a larger average diameter than lower phospholipid concentrations after 5 or more passes.
Koide and Karel (1987) were among the first to use a Microfluidizer (model M-110) to produce liposomes, reporting an average liposome diameter of 196 nm after 10 passes at 14 MPa (140 bar). Peel (1999) used a high-pressure homogenizer to process preformed multilamellar vesicles, and found that, after 5 passes at 13 MPa (130 bar), the average diameter stabilized at approximately 125 nm. The diameters measured by these researchers were generally larger than those obtained in our experiments, presumably because of the lower pressures used. However, Brandl et al. (1998) used a high-pressure homogenizer at pressures of up to 140 MPa (20,000 psi or 1,400 bar), and obtained liposome dispersions with median diameters of < 40 nm, which were much smaller than those found in our study.
The high polydispersity of liposome dispersions indicates a broad particle-size distribution, and is similar to the results of Zeisig and Cämmerer (2001), who reported that liposomes produced by microfluidization had an average diameter of 100 to 200 nm with a polydispersity of between 0.2 and 0.6.
-Potential
The -potential values for liposome dispersions at various pH values are shown in Figure 5. The liposome dispersions exhibited negative potentials at pH values above 3, with a flat region at approximately –65 mV between pH 5 and pH 10. Below pH 5, the -potential became rapidly less negative, with a potential of zero at pH 2.6 and rising to a positive potential of 15 mV at pH 2.0. Very few -potential values for liposomes have been reported in the literature, and the significant effect of the composition of the buffer salts used on the potential renders any comparison between different systems difficult. However, a -potential of ~–65 mV is much larger than those of between 0 and –43 mV reported for other liposome dispersions (Talsma et al., 1989; Ruel-Gariepy et al., 2002). The negative -potential presumably arose because of the presence of negatively charged phospholipids, particularly PI, with the positive -potential below pH 3 reflecting the positive charge of PC and PE molecules at these pH values.
|
Microstructure
In addition to simply confirming the presence of liposomes, electron microscopy can provide some valuable information regarding the size, shape, and structure of the liposomes within a dispersion. Negative-staining and thin-section transmission electron microscopy allows determination of liposome size at the very low end of the size spectrum, and can allow the lamellarity of a liposome dispersion to be estimated in a qualitative or semiquantitative manner (Perkins, 1993). Weiner (1995) stated that this was the method of choice below 5 µm. Cryo electron microscopy techniques are reported to be more likely to preserve the original structure of the liposomes (Weiner, 1995) and may provide three-dimensional confirmation of the liposome structure.
Negative-Staining Transmission Electron Microscopy.
The negative-staining technique was very quick and easy compared with the other 2 methods. A typical micrograph of the liposome dispersions is shown in Figure 6a. The liposomes could easily be identified as discrete particles that were predominantly spherical or rod-like in shape. The outer membrane surrounding the internal aqueous space could be clearly seen in many of the liposomes, and internal membranous structures could be identified. The micrographs contained a large number of very small particles (
40 nm) interspersed with much larger particles (100 to 200 nm).
|
). Although multilamellar liposomes are usually thought of as being composed of neatly stacked lamellae at regular spacing, according to Perkins et al. (1993), most large heterogeneous liposome systems typically contain liposomes with irregularly spaced bilayers or "liposome within liposome" structures. These are perhaps more correctly referred to as multivesicular liposomes (Talsma et al., 1987; Brandl et al., 1998). Freeze-fracture transmission electron micrographs in 
kalko et al. (1998) of liposomes produced via microfluidization also showed irregular-shaped vesicles trapped inside other vesicles, as did liposomes formed from phospholipid pastes produced by high-pressure homogenization (Brandl et al., 1998).
Thin-Section Transmission Electron Microscopy.
A typical micrograph of the liposome dispersions produced using the standard thin section technique is shown in Figure 7a. Visual analysis of the images indicated that the liposomes were primarily between 80 and 100 nm in diameter. They appeared to be clumped together, with groups of densely packed liposomes surrounded by empty space. Many of the vesicles in the clumps seemed to be nonspherical, although the liposomes in the less-crowded areas had retained their spherical shape. This suggested that some of the non-spherical nature of the liposomes could be due to membrane deformation caused by the crowding within the liposome clumps. However, some liposomes appeared to be shrunken and collapsed, which did not seem to be because of tight packing. It seems likely that the dehydration steps used during preparation for this technique were responsible for the clumping of the liposomes, and may have contributed to the nonspherical nature of the vesicles.
|
. There was little or no clumping in any of the micrographs. However, many of the liposomes still appeared to be dehydrated and collapsed, although a number of intact spherical vesicles were present in each micrograph. The liposome diameters appeared to range from 50 to 150 nm. The apparent multilamellar nature of many of the liposomes was obvious, with others appearing to have only a single membrane. It was difficult to know whether these were indeed unilamellar liposomes, or whether the sample had been sectioned between the inner and outer membranes of those liposomes without cutting through the inner bilayers.
Cryo-Field Emission Scanning Electron Microscopy.
The micrographs produced by this technique showed three-dimensional confirmation of smooth spheres consistent with that of liposomes (Figure 8). As expected, the approximately spherical vesicles had a wide particle size distribution, with both large and small liposomes evident in all dispersions. It was possible to see a number of large liposomes with diameters around 200 nm as well as some very small liposomes of approximately 40 nm in diameter.
|
Overall, this work showed that it is possible to produce liposomes from an MFGM phospholipid material using microfluidization. Electron microscopy revealed that a number of liposome structures were formed, including unilamellar, multilamellar, and multivesicular liposomes. The liposome dispersion had a broad particle size distribution, with an average hydrodynamic diameter of ~95 nm. Because of the origin of the phospholipids, the fatty acid profile of the MFGM phospholipids was more highly saturated than that of nonhydrogenated soy or egg phospholipid, which is likely to result in a higher phase transition temperature for the membrane and an increased resistance to oxidation. The high level of SM present in the phospholipid fraction may increase the stability of the liposome dispersion and reduce the membrane permeability. Initial experiments showed that the surface of these liposomes has a strong negative charge, which may further enhance their stability. Future studies will be undertaken to continue the characterization of the MFGM phospholipid liposomes, assessing permeability and phase transition temperature and comparing the results with liposome dispersions produced from soy phospholipids.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Received for publication August 30, 2005. Accepted for publication October 4, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
kalko, N., J. Bouwstra, F. Spies, M. Stuart, P. Frederik, G. Gregoriadis. 1998. Morphological observations on liposomes bearing covalently bound protein: Studies with freeze-fracture and cryo electron microscopy and small angle X-ray scattering techniques. Biochim. Biophys. Acta 1370:151–160.[Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
), 90 MPa (
), and 73 MPa (•). Each data point is the average of 3 or more replicates.
