Characterization of Antimicrobial-bearing Liposomes by ζ-Potential, Vesicle Size, and Encapsulation Efficiency
Liposome entrapment may improve activity of protein or polypeptide antimicrobials against a variety of microorganisms. In this study, ability of liposomes to withstand exposure to environmental and chemical stresses typically encountered in foods and food processing operations were tested. Liposomes consisting of distearoylphosphatidylcholine (PC) and distearoylphosphatidylglycerol (PG), with 0, 5, or 10 μg/ml of the antimicrobial peptide nisin entrapped, were exposed to elevated temperatures (25–75 °C) and a range of pH (5.5–11.0). Ability of liposomes to maintain integrity was assessed by measuring the encapsulation efficiency (EE), ζ-potential, and particle size distribution of liposomes. Distearoylphosphatidylcholine, PC/PG 8:2, and PC/PG 6:4 (mole fraction) liposomes retained between ~70–90% EE despite exposure to elevated temperature and alkaline or acidic pH. Particle size of liposomes averaged between 100 and 240 nm depending on liposome preparation. Liposomal surface charge depended primarily on phospholipid composition and changed little with inclusion of nisin. Surface charge was not affected by temperature for PC and PC/PG 8:2 but decreased for PC/PG 6:4 liposomes. Our results suggest that liposomes containing nisin may be suitable for use as antimicrobial-active ingredients in low- or high-pH foods subjected to moderate heat treatments.
KeywordsLiposomes Nisin Antimicrobials pH Temperature Stability Zeta-potential
Liposomes, spherical bilayer vesicles formed by dispersion of polar lipids in aqueous solvents, are widely studied for their ability to act as drug delivery vehicles by shielding reactive or sensitive compounds before release.1,2 Liposome entrapment may stabilize encapsulated bioactive materials against a range of environmental and chemical stresses, including presence of enzymes or reactive chemicals and exposure to extreme pH, temperature, and high ion concentrations.3, 4, 5 The functional properties of liposomes depend on their size, composition, and stability in food systems.6 Liposomes are typically spherical in shape and may consist of single or multiple bilayers composed of polar lipids.7 Their size can vary widely, from 40–50 nm to as much as ~1–3 μm, depending on the molecular properties of the lipid molecules, environmental conditions, and method of production.8 Of importance for application in foods is their ability to withstand various stresses that may be encountered during processing, packaging, or storage.9 Although traditionally used as model systems for biological membranes to study physiological processes in both eukaryotes and prokaryotes, liposomes are increasingly used in industrial applications, including encapsulation of pharmaceuticals and therapeutics, cosmetics, and anticancer and gene therapy drugs. In the food industry, liposomes were used to deliver food flavors and nutrients.10 More recently, liposomes were investigated for their ability to incorporate bioactive compounds to formulate novel functional foods and as carriers of food antimicrobials that might aid in the protection of food products against growth of spoilage and pathogenic microorganisms.11, 12, 13, 14 We previously demonstrated in our laboratories the successful production and high antimicrobial efficacies of liposomes containing two food antimicrobials, nisin, and lysozyme against strains of Listeria monocytogenes and Escherichia coli.13,14
Antimicrobial polypeptides such as nisin and lysozyme are known to be inhibitory toward the growth of gram-positive pathogens such as L. monocytogenes, Staphylococcus aureus, and Bacillus spp.15,16 Nisin is not inhibitory toward gram-negative pathogens if used without additives but its spectrum of activity can be broadened by combining the polypeptide with chelators such as EDTA or polyphosphates.17,18 Nevertheless, loss of antimicrobial activity was reported when nisin was directly added to foods. Jung et al.19 found significant loss of nisin activity in milk because of interactions with milk fat globules. Apparently, nisin preferentially adsorbed at the milk fat droplet surface to maximize interaction of the hydrophobic side chains with lipid phase molecules. Divalent cations associated with bacterial cell wall surfaces were shown to induce electrostatic repulsion preventing the cationic polypeptide nisin from interacting with the bacterial pathogens, thus reducing its activity.20 Furthermore, thermal processing of foods may lead to structural changes in polypeptide and protein antimicrobials and decrease their activity. For example, exposure to pasteurization temperatures can inactivate as much as 40% of nisin in a food product and lysozyme is generally completely inactivated above 75 °C.16 Storage above refrigeration temperatures may decrease nisin activity as well.20,21 Finally, nisin has shown to be less efficacious in solid food systems than in semisolid or liquid foods, which was attributed to the difficulty of maintaining a homogeneous distribution of the antimicrobial throughout the solid food matrix.21 Encapsulation of nisin may therefore offer a potential solution to overcome some of these stability issues resulting on antimicrobial formulations that are better suited for use in foods and can withstand a range of processing conditions.
The objectives of this study were to formulate liposomes bearing the antimicrobial polypeptide nisin and determine the integrity of liposomes exposed to environmental stresses (pH 5–11, temperature ≤75 °C). Integrity of liposomes was evaluated by determining the liposomal encapsulation efficiency (EE) using a fluorescent tracer dye (calcein). In addition, liposome surface charge and size changes as a function of nisin entrapment at two concentrations and temperature were evaluated as indicators of vesicle stability to the antimicrobial polypeptide.
Materials and methods
Phospholipids and antimicrobials
Distearoylphosphatidylcholine (PC) was purchased in chloroform from Avanti Polar Lipids (Alabaster, AL, USA). Distearoylphosphatidylglycerol (PG) was purchased in powdered form from Matreya Biochemicals LLC (Pleasant Gap, PA, USA). Distearoylphosphatidylglycerol was solubilized in a 4:1 chloroform/methanol mixture (Fisher, Fair Lawn, NJ, USA) before mixing with PC. All lipids were stored at −20 °C until ready for use. Nisin (CAS 1414-45-5; 2.5% w/w in sodium chloride and denatured milk solids) was obtained from Sigma-Aldrich (St. Louis, MO, USA) and stored in a desiccator at 5 °C until used.
Small unilamellar vesicles (SUVs) were prepared using the method of Pinnaduwage and Bruce22 with modification. Liposome samples were prepared from PC, PC/PG (8:2mol%), or PC/PG (6:4mol%). The lipids were first dispersed in chloroform or chloroform/methanol (4:1) and dried under N2 to form a lipid film on the wall of glass reaction tubes. Samples were desiccated overnight under vacuum to remove solvents. Lipid films were rehydrated with 0.1× phosphate buffered saline (0.1× PBS) (0.017 M of KH2PO4, 0.05 M of Na2HPO4, and 1.5 M of NaCl at pH = 7.4) (Biowhittaker, Rockland, ME, USA) to a lipid concentration of 10 mM. For nisin-containing liposomes, nisin was dissolved in 20 mM of HCl (Sigma) to obtain an acidic stock solution with a nisin concentration of 500 μg/ml. The nisin stock solution was immersed in boiling water for 5 min, passed through a 0.22-μm cellulose acetate (low-protein binding) membrane (Corning Inc., Corning, NY, USA), and diluted with 0.1× PBS to levels of 5 and 10 μg/ml for rehydration of lipid films.23 To allow for the determination of the EE, the fluorescent tracer calcein was solubilized in 0.1× PBS at a concentration of 50 mM and used to rehydrate lipid samples before SUV formation. After rehydration, vesicles were frozen in liquid nitrogen for ~10 s, gently thawed in water (25 °C) for ~10 s, and immersed in 70 °C water for 15 s. The thermal cycling was repeated four times. Liposomes were held at 60 °C for 1 h to ensure that phospholipids were above their respective gel–liquid crystalline phase transition temperature (TM) before extrusion to achieve size reduction of vesicles. Crude vesicle preparations were loaded into a Liposofast Basic handheld extruder (Avestin, Ottawa, Canada) and extruded sequentially through polycarbonate membranes (Avestin). Membranes were arranged in layers of three with liposomes being passed first 11 times through a 400/200/400-nm membrane stack followed by 11 passes through a 100/50/100-nm membrane stack. The extrusion was carried out at 60 °C to maintain vesicles above their respective TMs. Calcein-containing liposomes were subjected to size-exclusion chromatography to remove any unentrapped fluorophores from the liposomal preparation by passing them over a BioGel A0.5 M agarose (200–400 mesh) column (BioRad, Hercules, CA, USA) using a 0.1× PBS buffer as carrier. Liposomes were eluted in the void volume. All samples were stored at 5 °C until ready for use. Liposomes were assayed within 72 h of production.
Liposome vesicles were diluted to 100 μM in 2 ml of 0.1× PBS buffer, capped in film, and incubated in glass tubes at the desired temperatures in a Lauda RM-6 circulating water bath (Eppendorf, Westbury, NY, USA). Samples were incubated for 2 h before measurement. For pH-dependent entrapment experiments, 20 μl of liposomes was added to 10 μl of 1 M pH-adjusted 3-cyclohexylamino-propanesulfonic acid (CAS 1135-40-6) or 2-N-morpholino-ethanesulfonic acid (CAS 4432-31-9) buffers (Sigma) and brought to 2 ml total volume. Samples were incubated 1 h before measurement.
Fluorescence of incubated samples was scanned from 500 to 550 nm with an excitation of 495 nm and a maximum emission of ~515 nm in an LS-50B fluorescence spectrophotometer (PerkinElmer, Wellesley, MA, USA). Peaks were determined and reported as inherent fluorescence (F0). Twenty microliters of 10% Triton X-100 (Sigma) solution was added to the cuvette and allowed to stand for ~1 min to induce rupturing of liposomes and record the fluorescence of the released tracer (FT). FT is assumed to correspond to a 100% release of entrapped compounds. Detergent was added carefully so as to avoid the introduction of gas bubbles; samples were gently inverted eight to ten times to allow thorough mixing of detergent and liposomes.
The ζ-potential of liposomal preparations was determined by placing 1 ml of samples in a disposable cuvette. The cuvette was inserted into the measurement chamber of a particle electrophoresis instrument (Zetasizer Nanoseries ZS, Malvern Instruments, Worcestershire, UK), and equilibrated to 25 °C in the Peltier-controlled cuvette holder 5 min before. The ζ-potential was then determined by measuring the direction and velocity that the liposomes moved in the applied electric field. The Smoluchowsky mathematical model was used by the software to convert the electrophoretic mobility measurements into ζ-potential values. To ensure that the presence of the buffer did not influence the ζ-potential measurements, PC liposomes were initially dispersed in solutions containing a range of concentrations of buffer. Results indicated that ζ−potential was independent of buffer concentration unless the concentrations of buffer was larger than ten times the concentration used in this study (data not shown). Immediately after recording the ζ-potential at 25 °C, the instrument sample cell was heated to 75 °C and the sample again equilibrated for 5 min and the ζ-potential was recorded.
All experiments were carried out in triplicate with like means being averaged. Student’s t test was used to determine statistically significant differences with a confidence interval of 95%. Statistical analysis was carried out using Statistical Analysis Software (SAS) version 9.1 (SAS Institute, Cary, NC, USA).
Mean ζ-potential (mV) of liposomes composed of PC and PG at two temperatures (25 and 75 °C) containing 0, 5, and 10 μg/ml of nisin
Mean liposomal size (d32) and size distribution
On the other hand, particle size was influenced by phospholipid composition of the preparations (Figure 2). Distearoylphosphatidylcholine, PC/PG 8:2, and PC/PG 6:4 vesicles (empty) had d32 values at 25 °C of 103.1 ± 2.0, 181.7 ± 8.2, and 132.1 ± 2.8 nm, respectively. Upon inclusion of 5 μg/ml of nisin, particle sizes at 25 °C were 310.1 ± 104.9, 187.4 ± 13.3, and 122.7 ± 6.2 nm for PC, PC/PG 8:2, and PC/PG 6:4 vesicles, respectively (Figure 2). In the case of 10 μg/ml of nisin encapsulation, vesicle diameters for PC and PC/PG 6:4 liposomes, but not PC/PG 8:2 liposomes, were statistically different compared to nisin-free PC and PC/PG 6:4 liposomes (Figure 2). Interestingly, while PC/PG vesicles decreased in vesicle size with the inclusion of nisin, pure PC vesicles containing nisin increased in vesicle diameter (Figure 2a). Increases were not directly responsive to the level of nisin incorporated, as greater diameter increases were observed with 5 μg/ml of nisin vs 10 μg/ml of nisin in PC vesicles (Figure 2a).
Entrapment efficiency and ζ-potential of liposomes in the absence of nisin
Results of entrapment efficiency and ζ-potential measurements as a function of temperature correlate well with data reported elsewhere.24,25 The high entrapment efficiency of PG-containing liposomes with respect to temperature exposure was reported previously.26,27 In general, instability of liposomes was attributed to collisions and eventual merging of liposomal membranes of two or more liposomes.28 This process is thermodynamically driven because of the tendency of the system to decrease the energetically unfavorable curvature of the bilayer membrane in spherical liposomes. Collisions may be because of random (Brownian) movement of vesicles in solution or because of superimposed convection. Similar to emulsions, the probability of collisions of liposomes depends on the colloidal interactions between the particles.29 Increasing the repulsive interactions may reduce the frequency of collision whereas increasing the temperature can increase the probability for collisions. Thus, liposomes composed of charged polar lipids carrying higher electrical charges such as PC/PG 8:2 and PC/PG 6:4 liposomes with a ζ-potential of −52.28 and −72.60 mV, respectively, can be expected to be more stable than liposomes composed of neutral polar lipids. With respect to PC liposomes, a surface charge of ~−8.3 mV is in agreement with values reported for other PC vesicles.5,30,31
pH sensitivity of liposomes in the absence of nisin
Straubinger et al.32 demonstrated that liposomes composed of 100% PC were pH-insensitive and change of pH did not trigger release of encapsulated compounds. Lee et al.33 determined that incorporation efficiency of retinol in PC liposomes was 99.25% at pH 3.0 and 97.45% at pH 11.0, a statistically insignificant difference. Similarly, our data confirms that release of entrapped reporter did not vary with pH (Figure 1), i.e., negatively charged liposomes produced in this study did not exhibit significant differences in entrapment efficiencies as the pH was adjusted from 5.5 to 11.5. In contrast, inclusion of phosphatidylethanolamine has shown to yield liposomes that released constituents from their aqueous core under mild acidic conditions (pH < 6). In general, liposomes can be expected to retain the encapsulated material if the phospholipids used to formulate the liposomes maintain their charge regardless of the specific pH of systems in which they are applied. Consequently, knowledge of the pKa values of the involved phospholipids of which liposomes are composed is critical.
Encapsulation efficiency and ζ-potential of liposomes containing nisin
ζ-Potential measurements indicated that introduction of nisin in PC and PC/PG 6:4 liposomes significantly affected net surface charge of liposomes compared with nisin-free liposomes at 25 °C (Table 1). The addition of nisin was reported to perturb PC membranes, increasing fluidity and altering the overall bilayer structure.30,34 Model membrane studies have shown that nisin has a high specificity for anionic phospholipids, which is in fact part of the basis for its antimicrobial activity.34, 35, 36 Nisin is a member of a class of small antimicrobial polypeptides known as lantibiotics that are thought to self-assemble in the bacterial membrane to form pores that prevent bacteria from maintaining homeostasis.37 Whereas nisin rapidly creates pores in membranes that contain high levels of anionic lipids, these pores are transient in nature with nisin possibly translocating entirely across the membrane bilayer after porulation.34,38 In PC/PG 6:4 liposomes, more nisin may have been present in the liposomal membrane or might have translocated during incubation because of the higher content of the anionic phospholipids PG. Accordingly, ζ-potentials decrease at 25 °C as nisin is incorporated at higher levels (Table 1). Nisin-induced pores or leakage from vesicles was reported previously and given that 100% EE was not reported in this study, it is not unlikely that the cationic peptide neutralized charge of PG lipids on the outer surface of the membrane.14
An increase in temperature to >70 °C of liposomes containing nisin did not significantly affect the surface charge of liposomes. Similarly, Makino et al.5 reported that surface charge of PC liposomes remained unchanged as temperature was increased from 20 to 55 °C. However, at 75 °C, PC liposomes containing 10 μg/ml of nisin did display increased electronegativity; changes upon inclusion of 5 μg/ml of nisin were not significant (Table 1). It is possible that nisin/PC interactions and structural rearrangement of nisin within the liposome may have been altered as the lipids underwent their main gel–liquid crystalline phase transition, known to occur at ~54 °C.9 Heating of vesicles can increase the Brownian motion of lipids in vesicles and potentially induce rearrangement of lipid headgroups, increasingly exposing the phosphatidyl groups of liposomes to the environment. Alternatively, heating of nisin could have decreased its solubility and stability. Liu and Hansen39 reported almost complete insolubility of nisin at 25 °C at pH 7.0; others have reported that heating of nisin to temperatures approximating ultrahigh temperature processing can result in as much as 40% loss of nisin activity.20
Particle size distribution of liposomes containing nisin
Distearoylphosphatidylcholine vesicles, upon inclusion of 5 and 10 μg/ml of nisin, displayed increases in their size around the phase transition temperature of the polar lipid. The phase transition temperature of PC, the lipid used in this study, occurs around 54 °C. In the presence of nisin, the change from gel to liquid crystalline appears to have a destabilizing effect on the vesicles. The increases of liposome diameter according to the dynamic light scattering assay suggest changes in the liposome structure and morphology, a phenomenon recorded elsewhere.34 Breukink et al.40 demonstrated via calorimetric experiments that nisin has low penetration power for membranes formed from PC vs membranes formed from PC and PG, and that nisin/PC interactions are primarily hydrophobic in nature. El-Jastimi et al.34 reported the formation of large aggregates of nisin and liposomes after heating of liposomes past their phase transition temperatures. Wiener et al.41 demonstrated increases in vesicle diameter after loading with matrix protein of stomatitis virus. As previously mentioned, heating of nisin can also lead to decreases in solubility, which may also contribute to the observed size changes of PC liposomes.
Particle size distributions of PG-containing liposomes did not significantly change when samples were heated from 25 to 75 °C, even when nisin was entrapped (Figures 2 and 3). Overall, in the absence of nisin, PG-containing vesicles were of greater size then PC liposomes, a phenomenon reported elsewhere.14,42 This difference in size of PC liposomes vs those with PG incorporated is not thought to be a function of the zwitterion not being able to maintain stability because of charge repulsion at the surface, as the lipid has only a small electronegative surface charge at neutral pH (Table 1). With inclusion of nisin, the diameters of PG-containing liposomes decreased, whereas size of pure PC liposomes increased compared to their nisin-free counterparts (Figure 2). Laridi et al.43 reported substantial size reductions in liposomal preparations after nisin incorporation. Mean diameters of liposomes containing PC and other charged lipids decreased from ~1,741 to ~740 nm upon entrapment of 100 μg/ml of nisin.43 Other authors have speculated that the size decrease is because of nisin’s ability to reorganize the liposome via electrostatic and hydrophobic interactions34,44,45; it is possible that electrostatic interactions of the cationic polypeptide with the anionic lipid molecules may act to induce a rearrangement of vesicles into more tightly packed structures vs other nonlamellar phases.43
Nisin-containing PC/PG bilayers may be able to attain a more optimal packing, via charge neutralization of anionic lipids by the cationic peptide and a realignment of the phospholipids, including the formation of vesicles with decreased curvature stress, which could explain the reductions in diameters.34,46 The interaction between the polar lipids and nisin ultimately depends on the composition of the bilayer and thus the ratio of the two lipids in the liposomes and contribute to the observed differences in PC/PG 8:2 and PC/PG 6:4 vesicle size reductions.34,38,46, 47, 48 Nevertheless, a more in-depth elucidation of membrane composition and structural arrangement of phospholipids and nisin within the membrane using fluorescence or neutron scattering will be required to quantify these changes.34
Results presented in this study indicate that liposomes may be good candidates for the entrapment of bioactive compounds such as antimicrobials polypeptides. However, data presented in this study also show that formulation plays a critical role in terms of liposome diameter/size and surface charge of liposomes that in turn may impact the functionality of these particles in complex food systems. An in-depth understanding of the overall structure of liposomes will be needed to enable food processors to use these encapsulated compounds to control growth of foodborne pathogens in foods. We are encouraged by the fact that liposomes formulated with PC and PG appear to be relatively stable to pasteurization protocols although exposure to temperatures above 75 °C has yet to be tested. It should be noted though that we did not yet test whether nisin maintained its antimicrobial property after exposure to the elevated temperatures. The relatively high pH stability of liposomes demonstrates one of the main advantages of this encapsulation system, namely, that they could be formulated such that the internal pH is different from that of the surrounding continuous phase using appropriate polyelectrolytes.35,37 Thus, pH-sensitive compounds such as nisin may be particularly suited to be encapsulated in liposomal carrier systems. Finally, as liposomes are generally formed from lipids that are naturally occurring in various food staples, allergenicity and product labeling concerns can be expected to be minimal. In terms of encapsulation of food antimicrobials, data existing relating the bacteriostatic and bacteriolytic capability of liposomes with entrapped antimicrobials is still limited. Studies ongoing in our laboratory are currently focused on assessing the in vitro inhibitory effect of liposome encapsulation of antimicrobials and the inhibitory potential of liposome-entrapped antimicrobials within various food matrices.
This research was financially supported by a USDA NRI grant (USDA NRI 2004-35201-15358) and the Massachusetts and Tennessee Experiment Station (Hatch MAS 00911 and TEN 00263).
- 1.S.S. Chrai, R. Murari and I. Ahmad, Biopharm-Appl T Bio 15, 40 (2002).Google Scholar
- 6.R.R.C. New, In: Liposomes: A Practical Approach, edited by R.R.C. New (Oxford University Press, New York, NY 1990), p. 33.Google Scholar
- 8.J. Lasch, V. Weissig and M. Brandl, In: Liposomes: A Practical Approach, edited by V.P. Torchilin and V. Weissig (Oxford University Press, New York, NY 2003), p. 3.Google Scholar
- 10.B.F. Gibbs, S. Kermasha, I. Alli and C.N. Mulligan, Int J Food Sci Nutr 50, 213 (1999).Google Scholar
- 13.L.M. Were, B. Bruce, P.M. Davidson and J. Weiss, J Food Prot 67, 922 (2004).Google Scholar
- 15.J.M. Jay, M.J. Loessner and D.A. Golden, Modern Food Microbiology (Springer, New York, NY 2005), p. 301.Google Scholar
- 16.E.A. Johnson and A.E. Larson, In: Antimicrobials in Foods, edited by P.M. Davidson, J.N. Sofos and A.L. Branen (CRC Press, New York, NY 2005), p. 361.Google Scholar
- 17.J.S. Boland, P.M. Davidson, B. Bruce and J. Weiss, J Food Prot 67, 285 (2004).Google Scholar
- 20.L.V. Thomas and J. Delves-Broughton, In: Antimicrobials in Food, edited by P.M. Davidson, J.N. Sofos and A.L. Branen (CRC Press, New York, NY 2005), p. 237.Google Scholar
- 21.J. Delves-Broughton, Food Technol 44, 110 (1990).Google Scholar
- 26.Y.P. Zhang, R.N. Lewis and R.N. McElhaney, Biophys J 72, 779 (1997).Google Scholar
- 29.D.J. McClements, Food Emulsions: Principles, Practices, and Techniques (CRC Press, Boca Raton, FL 2005).Google Scholar
- 33.S.-C. Lee, H.-G. Yuk, D.-H. Lee, K.-E. Lee, Y.-I. Ludescher and R.D. Ludescher, J Biochem Mol Biol 35, 358 (2002).Google Scholar
- 34.R. El-Jastimi, K. Edwards and M. Lafleur, Biophys J 77, 842 (1999).Google Scholar
- 35.F.H. Gao, T. Abee and W.N. Konings, Appl Environ Microbiol 57, 2164 (1991).Google Scholar
- 39.W. Liu and J.N. Hansen, Appl Environ Microbiol 56, 2551 (1990).Google Scholar
- 45.D.B. Fenske and P.R. Cullis, Biophys J 64, 1482 (1993).Google Scholar