Article pubs.acs.org/Langmuir
Encapsulation of Yeast Cells in Colloidosomes Polly H. R. Keen, Nigel K. H. Slater, and Alexander F. Routh* Department of Chemical Engineering and Biotechnology, BP Institute, University of Cambridge, Madingley Road, Cambridge CB3 0EZ, United Kingdom
ABSTRACT: Polymeric colloidosomes encapsulating viable Baker’s yeast cells were prepared. To make the capsules, an aqueous suspension of 153 nm poly(methyl methacrylate-co-butyl acrylate) latex particles plus yeast cells is emulsified in a continuous phase of sunflower oil. By adding a small amount of ethanol to the oil phase, the latex particles at the surface of the emulsion droplets aggregate, forming the colloidosome shells. The microcapsules have been examined using optical, confocal, and scanning electron microscopies. The viability of the yeast cells was tested using fluorescent molecular probes. The encapsulated Baker’s yeast cells were able to metabolize glucose from solution, although at a slower rate compared to nonencapsulated yeast. This demonstrates diffusion limitation through the colloidosome shell. The diffusive resistance could be increased by manufacturing colloidosomes with a double latex shell.
■
INTRODUCTION The encapsulation of microorganisms has many promising applications in a broad range of fields from bioreactors to medicine. Advantages of microencapsulated cells include increased cell density, ease of separation, protection from harsh external conditions, and reduced susceptibility to contamination by foreign organisms.1,2 Various microencapsulation techniques have been developed including using sacrificial cores,3 interfacial polymerization,4 or solvent evaporation.5 However, these methods usually require chemicals which are toxic to biological organisms, such as harsh organic solvents or monomers in the case of interfacial polymerization. Another encapsulation technique is layer-by-layer polyelectrolyte deposition. The toxicity of the polyelectrolytes (polycations especially) or the use of chemicals to dissolve the core or cross-link the polymers is problematic when encapsulating biological materials. Another disadvantage of this method is the loading of the microcapsule: this can be done either before the layer-by-layer deposition, in which case biologically friendly dissolution of the core can be difficult, or after the capsule is created, in which case there is low encapsulation efficiency.6 However, yeast cells have been encapsulated by this method with preserved metabolic activity.7 Continued biological activity of an enzyme in multilayer polymeric microcapsules has also been shown.8 Although this © 2011 American Chemical Society
method allows control of shell thickness and can avoid the use of harsh solvents, it is time-consuming with multiple deposition and washing steps. There is also the problem of polyelectrolyte induced particle aggregation; hence, low particle concentrations must be used, resulting in a low yield. Numerous methods have been specifically developed for cell encapsulation. One promising technique involves encapsulating microbial cells in a polymeric support composed of a hydrogel such as calcium alginate, poly(vinyl alcohol), or chitosan.1 The disadvantages of these microcapsules, however, include disruption and cell leakage during long-term cultivation.9 Microcapsules can also be manufactured using the selfassembly of colloidal particles. The self-assembly of colloidal particles at an oil−water interface was first reported by Ramsden10 and Pickering,11 and the emulsions stabilized by this phenomenon are called Pickering emulsions. Velev et al. exploited Pickering emulsions to create microcapsules.12−14 The colloidal particles creating the shell of the microcapsule were locked into place by the deposition of polyelectrolytes. The term “colloidosomes” was coined by Dinsmore, who fixed the shell of the microcapsule by heating the system above the Received: October 25, 2011 Revised: November 25, 2011 Published: December 7, 2011 1169
dx.doi.org/10.1021/la204183u | Langmuir 2012, 28, 1169−1174
Langmuir
Article
in a glass vial using the vortex for 60 s then left to settle for 5 min until all bubbles disappeared. A 30 wt % dispersion of latex particles in water was prepared. Equal volumes of the latex and the yeast suspensions were mixed using the vortex in a microtube for 20 s. Then 0.08 mL of the yeast and latex mixture was added to the oil mixture and immediately homogenized with the vortex for 1 min. The mixture was then left to stand for 1 min. When creating colloidosomes with a double latex shell, 0.04 mL of 50 wt % latex dispersion was added to the sample, and again, the sample was vortexed for 1 min before being left to stand for 1 min. This step was omitted to create colloidosomes with a single latex shell. To kill any unencapsulated yeast, 4 mL of 0.25% Virkon solution was added, and the mixture was gently shaken by hand for a further 30 s. To transfer the colloidosomes to an aqueous phase, the mixture was centrifuged at 1000 rpm for 10 min. This compacts the colloidosomes to the boundary between the oil and aqueous phases. The oil and aqueous phases could then be removed using a Pasteur pipet. This left colloidosomes encapsulating viable yeast (with a few dead yeast cells outside the microcapsules). The activity of the colloidosome encapsulated yeast cells (with both a single and a double latex shell) was investigated by adding 6 vials worth of the aforementioned colloidosome sample to 10 mL of a 6 mM glucose solution. A small volume (20 μL) was extracted from the sample with a pipette at various time intervals and fed to the test strip of the glucose monitor. This gave measurements of the glucose concentration in the sample at increasing times. Cell Viability Tests. Fluorescent molecular probes were used to assess the microbial cells’ viabilities. LIVE/DEAD Yeast Viability Kit was also purchased from Molecular Probes (Invitrogen). The FUN 1 stain shows cells which are both metabolically active and have plasma membrane integrity marked clearly with orange-red fluorescent intravacuolar structures. Dead cells, however, exhibit extremely bright, diffuse green-yellow fluorescence. The experimental procedure for staining the samples involved adding 1 μL FUN 1 solution to 1 mL of the colloidosome and yeast sample. The sample was then thoroughly mixed before incubating in the dark for 30 min at 30 °C prior to observation with the confocal microscope. Observation. To observe the microcapsules, a Leica DME transmitted light optical microscope equipped with an XLICap color digital camera and capture software (XLI Imaging Ltd., Version 12.0) was used. Observation was also carried out using a Leica TCS SP5 confocal laser scanning microscope with a 1.4 numerical aperture 63× oil immersion objective. To achieve fluorescence with the fluorescent molecular probe SYTO 9, an excitation wavelength of 488 nm was used and emission at 510−540 nm was observed. Fluorescence of propidium iodide was achieved via excitation at 488 nm with emission at 620−650 nm. When using the molecular probe FUN 1, an excitation wavelength of 488 nm was used and emission at 580−610 nm was observed. For higher resolution observation, a JEOL-6340F scanning electron microscope at 5.0 kV was used. The microcapsule samples, in the aqueous phase, were rinsed with 50 vol % aqueous ethanol solution to remove any excess oil from their surface and then were air-dried on a stainless steel SEM stub. The colloidosome samples were then platinum-coated using an Emitech K575 sputter coater (argon environment, 1 × 10−3 mbar, 40 mA, 1 min).
glass transition temperature of the latex fusing the latex particles together.15 In this work the latex particles forming the shells of the colloidosomes are locked by transitioning the self-assembled colloidal latex from a stable to an unstable state, reducing the electrostatic potential between the particles, by the addition of a small amount of ethanol to the continuous phase.16 The method involves emulsifying an aqueous phase containing latex particles and yeast cells in a continuous oil phase. Unlike many microcapsule fabrication techniques, this method avoids the use of high temperatures, harsh solvents, or chemical reactions and hence can be adopted to encapsulate viable biological material such as microbial cells. These authors are not aware of any reports that colloidosomes have previously been used for cellular encapsulation.
■
EXPERIMENTAL SECTION
Materials, Microorganisms, and Methods. Poly(methyl methacrylate-co-butyl acrylate) latex particles were synthesized via emulsion polymerization. The diameter of the latex particles was determined by dynamic light scattering using a Brookhaven ZetaPALS instrument. The measured diameter was 156 nm. The instrument was also used to determine the zeta potential of the particles and of the Baker’s yeast cells. Figure 1 shows the zeta potentials of the colloidal particles as a
Figure 1. Zeta potentials as a function of pH measured for latex and Sainsbury’s Baker’s yeast. The pH was varied by the addition of aqueous HCl. function of pH demonstrating the colloidal stability of the system. The glass transition temperature of the latex was found, using differential scanning calorimetry, to be 35 °C. The water used in all experiments was deionized water of resistivity 18.2 MΩ·cm (Milli-Q, Millipore) produced by a Pure Lab Ultra apparatus. Sunflower oil from the supermarket Sainsbury’s was used as the continuous phase without purification. Ethanol (Fisher Scientific, 99.5%) was used without purification. Baker’s yeast cells (Saccharomyces cerevisiae) were purchased from the supermarket (Sainsbury’s Fast Action Dried Bread Yeast). The dried yeast was mixed with ultrapure water at a concentration of 10 mg/mL and was washed by five cycles of centrifugation (1500 rpm, 2 min) and resuspension in water before use. The vortex mixer used was a TopMix FB15024 (Fisher Scientific). A 0.25% solution of the disinfectant Virkon (Antec, DuPont) was made by dissolving the powder in deionized water. Glucose solution was made up to a concentration of ∼6 mM using D-glucose (Sigma) and distilled water. Glucose concentration measurements were taken using an Accu-chek Aviva blood glucose monitor with Accu-chek Aviva glucose test strips. Preparation of Colloidosomes. A mixture of 4 vol % ethanol in sunflower oil (7.6 mL of sunflower oil: 0.32 mL of ethanol) was mixed
■
RESULTS AND DISCUSSION Figures 2 and 3 show that water-core colloidosomes encapsulating yeast cells can be manufactured. Shortly after the microcapsules have been made, some colloidosomes are empty of cells, some contain a single yeast cell, a few contain more than one yeast cell, and yeast cells are also present outside the colloidosomes. Most colloidosomes are fairly spherical and range in size from 10 to 30 μm. Some are more deformed, as can be seen in the top right-hand side of Figure 2. These deformations are the result of centrifugation: all colloidosomes 1170
dx.doi.org/10.1021/la204183u | Langmuir 2012, 28, 1169−1174
Langmuir
Article
Figure 2. Optical microscope image of colloidosomes encapsulating yeast cells plus nonencapsulated cells.
Figure 5. Optical microscope image of colloidosomes encapsulating yeast cells (the bright spots are out of focus, unencapsulated yeast cells).
was then left in the same position and examined at various times after the sample had been made. At 10 and 85 min after manufacture, the images show one yeast cell encapsulated. However, at 150 min, two yeast cells are visible, and at 240 min, three cells can be seen. Figure 7 demonstrates that the encapsulated Baker’s yeast cells remain viable. The red fluorescence of the yeast cells shows their metabolic activity, and their ability to divide within the colloidosomes could also be viewed. The size of the colloidosomes was measured from optical microscope images; the average diameter (approximating all microcapsules as spherical) was 12.4 μm, the standard deviation was 5.4 μm, and a histogram of the sizes measured is shown in Figure 8. The Baker’s yeast cells were typically seen to have a diameter of between 1 and 5 μm. Figure 9 is an SEM photograph of the original latex particles used in the method, while Figure 10 is an SEM image of a manufactured colloidosome. The ethanol has fused the latex particles together to form the shell of a fairly spherical microcapsule. The imperfect sphericity derives from the use of centrifugation in order to transfer the colloidosomes to an aqueous phase and the rinsing of the colloidosomes with ethanol solution during SEM sample preparation. Total Encapsulation of Yeast Cells. As shown in Figure 2, some yeast cells were always observed outside the colloidosomes, while a significant number of cells were also encapsulated in the water cores. Therefore, the encapsulation process is not 100% efficient if sufficient numbers of encapsulated cells are needed for further experimentation. In an attempt to explain this inefficiency, the electrokinetic properties of the yeast cells were investigated. At pH 7 (the pH at which the method is carried out)and all pH’s for that matterthe zeta potential of the yeast cells is always less negative than that of the latex particles (see Figure 1). This suggests that the latex particles are more colloidally stable than the yeast cells. Therefore, the yeast cells preferentially move to the oil−water interface over the latex particles, forming part of the colloidosome shell. This allows the yeast cells to replicate outside of the colloidosomes. Another explanation for the
Figure 3. Optical microscope image of colloidosome encapsulating two yeast cells plus empty water-core colloidosomes.
are spherical before this process. No colloidosomes are formed without the presence of ethanol in the oil phase; the latex particles do not aggregate so fall apart on transfer to an aqueous phase. Figure 4 shows a colloidosome full of yeast cells 1 day after the microcapsules have been made. This verifies that the yeast
Figure 4. Optical microscope image of a colloidosome full of yeast cells.
cells are able to stay alive within the colloidosomes and can reproduce to the capacity of the microcapsule. Figure 5 again shows the success of the method at manufacturing mainly spherical colloidosomes encapsulating yeast cells. It also shows that while some yeast cells replicate within their colloidosome, as seen in the top right-hand corner of the image, others stay as single cells for a longer period of time, as seen in the colloidosomes at the bottom of the image. The ability of the Baker’s yeast cells to divide within the colloidosomes is shown in Figure 6. This series of images was achieved by focusing the optical microscope on a single colloidosome encapsulating a yeast cell. The microscope slide 1171
dx.doi.org/10.1021/la204183u | Langmuir 2012, 28, 1169−1174
Langmuir
Article
Figure 6. Time series of optical microscope images showing yeast cells moving and dividing within a colloidosome.
Figure 7. Fluorescence and transmission confocal microscopy images of a colloidosome encapsulating viable Baker’s yeast cells stained with FUN 1.
Figure 8. Histogram of measured colloidosome sizes.
unencapsulated yeast is that some yeast cells get stuck in the extra latex agglomerates or to the outside of the colloidosomes; therefore when the cells divide, the number of cells outside the colloidosome grows. Because the doubling time of Saccharomyces cerevisiae cells is only ∼75 min,17 the number of external microbial cells rapidly increases with time. It was therefore necessary to either separate these free yeast cells from the encapsulated ones or kill the free cells and maintain viability of those encapsulated. Otherwise, any measurements on the sample (such as measuring a decreasing glucose concentration with time) would relate to the free cells as well as those encapsulated, and this was obviously undesirable. Using a higher ratio of latex particles to yeast cells in an attempt to encapsulate all cells did not solve the problem. Neither did the use of more vigorous stirring in the method or
Figure 9. SEM image of original latex particles.
adding the yeast cells and then the latex particles separately to the oil/ethanol mixture. This was because the preferential movement of yeast to the oil−water interface meant that there would always be at least one yeast cell outside the microcapsules which could replicate and hence increase in number. Trialed methods of separating out the nonencapsulated yeast included filtering, centrifuging, and settling by gravity. No success was had with any of these methods. Again, at least one 1172
dx.doi.org/10.1021/la204183u | Langmuir 2012, 28, 1169−1174
Langmuir
Article
yeast. The glucose is metabolized by the yeast cells mainly for cell maintenance and fermentation. For comparison, glucose
Figure 10. SEM image of a typical colloidosome.
yeast cell could always be seen outside the colloidosomes in the sample. In order to ensure all external yeast cells would not contribute to the investigations of the encapsulated cells, it was decided to kill the unencapsulated microbes. The most common method of killing yeast is heating to temperatures above 50 °C; however, this would overly fuse the latex particles comprising the colloidosomes forming latex agglomerates as well as killing the encapsulated yeast cells. Therefore, killing the yeast cells outside the colloidosomes was attempted using various chemicals: detergents such as SDS, acids (acetic and hydrochloric), and the laboratory disinfectant Virkon. These chemicals were applied to the colloidosome samples at various points in the method, both before and after centrifuging the samples. It was found that the Baker’s yeast was a robust species, and it was difficult to cause cell death. The strong acids, high temperatures, or harsh fungicides generally required to kill the cells are undesirable to apply to the colloidosomes. The use of strong acids (acetic acid and hydrochloric acid at pH 3 for 200 min, as recommended by Ludovico et al.18) overly aggregated the latex particles, causing agglomerates and disrupted the colloidosome structure. It was found that the Baker’s yeast cells were also resistant to high concentrations of all the detergents experimented with. Success was had, however, with Virkon. Virkon contains oxone (potassium peroxymonosulphate), sodium dodecylbenzenesulfonate, sulfamic acid, and inorganic buffers. It causes cell death by oxidation of proteins and other components of the cell protoplasm, resulting in inhibition of enzyme systems and loss of cell wall integrity.19 Because of its combination of ingredients and hence high effectiveness as a disinfectant, it kills yeast at a low concentration. 0.25% Virkon was found not to aggregate or disrupt the colloidosomes yet metabolically inactivate the unencapsulated yeast. Further experiments allowed the deduction of a method to maintain viability of encapsulated yeast cells but kill those outside the microcapsules using 0.25% Virkon. Because of the presence of the detergent SDS in Virkon, more of the oil phase was also able to be removed than previously after centrifugation of the sample, which was an added bonus. Diffusion into Colloidosomes. Figure 11 shows the consumption of glucose by the colloidosome-encapsulated
Figure 11. Glucose consumption by colloidosome-encapsulated yeast.
concentration measurements were also taken for an unencapsulated yeast sample to investigate the effect of encapsulation on the rate of glucose consumption. For control purposes, the same measurements were recorded for colloidosomes without yeast inside and for yeast exposed to 0.25% Virkon for 10 min (to ensure that the method kills all unencapsulated yeast). Putting the yeast cells in the colloidosomes delays and slows down the reduction in glucose concentration over time. This is partly due to the inefficient encapsulation of the yeast cells by the colloidosome manufacturing method. However, there is still a large difference in the trend between the encapsulated yeast and half the number of unencapsulated yeast (i.e., assuming 50% encapsulation efficiency). Therefore, there must be a considerable diffusion limitation of glucose molecules through the colloidosomes’ shells. This is understandable: the ethanol fuses the latex particles together, resulting in only a few small pores in the shell and glucose molecules are fairly large (∼1 nm). There is also some extra diffusive resistance from the presence of residual oil from the colloidosome manufacturing method on the surface of the microcapsule shells. In addition, it is possible that the use of Virkon to kill the external yeast may have affected the yeast cells inside the colloidosomes, slowing their metabolic activity. The delay in the glucose concentration reduction is increased slightly when the colloidosomes are manufactured with a double latex shell. This is because the microcapsule walls are thickened by the addition of extra latex particles and any pores in the shell get filled by the additional particles. Therefore, the diffusive resistance against glucose into the aqueous yeastcontaining core increases. The double latex shell adds structural rigidity to the colloidosomes; a larger number retain more of a spherical shape after centrifugation. Also, it was microscopically 1173
dx.doi.org/10.1021/la204183u | Langmuir 2012, 28, 1169−1174
Langmuir
Article
(12) Velev, O. D.; Furusawa, K.; Nagayama, K. Langmuir 1996, 12, 2374−2384. (13) Velev, O. D.; Furusawa, K.; Nagayama, K. Langmuir 1996, 12, 2385−2391. (14) Velev, O. D.; Nagayama, K. Langmuir 1997, 13, 1856−1859. (15) Dinsmore, A. D.; Hsu, M. F.; Nikolaides, M. G.; Marquez, M.; Bausch, A. R.; Weitz, D. A. Science 2002, 298, 1006−1009. (16) Nomura, T.; Routh, A. F. Langmuir 2010, 26, 18676−80. (17) Robert, V. Yeasts in Food: Beneficial and Detrimental Aspects; Behr’s Verlag DE: Hamburg, 2003; p 488. (18) Ludovico, P.; Sousa, M. J.; Silva, M. T.; Leão, C.; Côrte-Real, M. Microbiology (Reading, England) 2001, 147, 2409−15. (19) Antec International Limited, 2005.
observed that the yield of colloidosomes, and the efficiency of the encapsulation of yeast cells slightly increased by the inclusion of the second latex addition step in the method; more colloidosomes were formed and a higher proportion contained at least one yeast cell. No decrease in glucose concentration was seen over time for the samples of colloidosomes without yeast inside or for the unencapsulated yeast exposed to 0.25% Virkon for 10 min. This verifies that the glucose consumption seen in the samples of encapsulated yeast is due to the encapsulated yeast only. No decrease in glucose concentration was seen over time for unencapsulated yeast exposed to Virkon at concentrations down to 0.05% for 10 min. This suggests that the concentration of Virkon which diffuses into the colloidosomes is less than 0.05%. It also suggests that the delay in the metabolisation of glucose by the encapsulated yeast is not due to the effect of more than 0.05% Virkon on the metabolic ability cells.
■
CONCLUSIONS A method has been developed allowing the encapsulation of viable Baker’s yeast cells in water-core colloidosomes of between 5 and 30 μm in diameter. Poly(methyl methacrylateco-butyl acrylate) latex particles were fused to form the colloidosome shell by the addition of a small amount of ethanol to the continuous oil phase. This biologically friendly method does not involve high temperatures or the use of harsh chemicals. The encapsulated Baker’s yeast cells were able to metabolize glucose from solution. There was a large diffusive resistance through the colloidosome shell for the glucose molecules. This diffusive resistance could be increased by manufacturing colloidosomes with a double latex shell.
■ ■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected].
ACKNOWLEDGMENTS The authors thank Kamran Yunus for his assistance with use of the confocal microscope and Damien Dupin (University of Sheffield) for his assistance with the manufacture of latex particles. Polly H. R. Keen is grateful to the EPSRC for funding.
■
REFERENCES
(1) Park, J. K.; Chang, H. N. Biotechnol. Adv. 2000, 18, 303−319. (2) Yoshida, M.; Mardriyati, E.; Tenokuchi, D.; Uemura, Y.; Kawano, Y.; Hatate, Y. J. Appl. Polym. Sci. 2003, 89, 1966−1975. (3) Yang, H. G.; Zeng, H. C. Angew. Chem., Int. Ed. 2004, 43, 5206− 9. (4) Frere, W.; Danicher, L.; Gramain, P. Eur. Polym. J. 1998, 34, 193−199. (5) Lorenceau, E.; Utada, A. S.; Link, D. R.; Cristobal, G.; Joanicot, M.; Weitz, D. A. Langmuir 2005, 21, 9183−6. (6) Cock, L. J.; De; Koker, S.; De; Geest, B. G.; De; Grooten, J.; Vervaet, C.; Remon, J. P.; Sukhorukov, G. B.; Antipina, M. N. Angew. Chem., Int Ed. 2010, 49, 6954−73. (7) Diaspro, A.; Silvano, D.; Krol, S.; Cavalleri, O.; Gliozzi, A. Langmuir 2002, 18, 5047−5050. (8) Caruso, F.; Trau, D.; Möhwald, H.; Renneberg, R. Langmuir 2000, 16, 1485−1488. (9) Hertzberg, S.; Moen, E.; Vogelsang, C.; Østgaard, K. Appl. Microbiol. Biotechnol. 1995, 43, 10−17. (10) Ramsden, W. Proc. R. Soc. London 1903, 72, 156−164. (11) Pickering, S. U. J. Chem. Soc., Dalton Trans. 1907, 91, 2001. 1174
dx.doi.org/10.1021/la204183u | Langmuir 2012, 28, 1169−1174
