This video and manuscript describe an emulsion-based method to encapsulate mammalian cells in 0.5% to 10% alginate beads which can be produced in large batches using a simple stirred vessel. The encapsulated cells can be cultured in vitro or transplanted for cellular therapy applications.
Cell encapsulation in alginate beads has been used for immobilized cell culture in vitro as well as for immunoisolation in vivo. Pancreatic islet encapsulation has been studied extensively as a means to increase islet survival in allogeneic or xenogeneic transplants. Alginate encapsulation is commonly achieved by nozzle extrusion and external gelation. Using this method, cell-containing alginate droplets formed at the tip of nozzles fall into a solution containing divalent cations that cause ionotropic alginate gelation as they diffuse into the droplets. The requirement for droplet formation at the nozzle tip limits the volumetric throughput and alginate concentration that can be achieved. This video describes a scalable emulsification method to encapsulate mammalian cells in 0.5% to 10% alginate with 70% to 90% cell survival. By this alternative method, alginate droplets containing cells and calcium carbonate are emulsified in mineral oil, followed by a decrease in pH leading to internal calcium release and ionotropic alginate gelation. The current method allows the production of alginate beads within 20 min of emulsification. The equipment required for the encapsulation step consists in simple stirred vessels available to most laboratories.
Mammalian cell encapsulation has been broadly studied as a means to protect transplanted cells from immune rejection1 or to provide a three-dimensional support for immobilized cell culture2,3,4. Pancreatic islet encapsulation in alginate beads has been used to reverse diabetes in allogeneic5,6 or xenogeneic7,8,9,10,11,12 rodents. Preclinical and clinical trials of encapsulated pancreatic islet transplantation to treat type 1 diabetes are ongoing13,14,15. For transplantation applications or larger-scale in vitro immobilized cell production, nozzle-based bead generators are generally used. Typically, a mixture of alginate and cells is pumped through a nozzle to form droplets that fall into an agitated solution containing divalent cations, resulting in the external gelation of the droplets. Coaxial gas flow16,17, nozzle vibration18, electrostatic repulsion19 or rotating wires20 facilitate droplet formation at the nozzle tip.
The main drawbacks of conventional bead generators are their limited throughput and the limited range of solution viscosities that will result in adequate bead formation21. At high flow rates, the fluid exiting the nozzle breaks up into droplets smaller than the nozzle diameter, decreasing size control. Multi-nozzle bead generators can be used to increase the throughput, but the uniform distribution of flow among the nozzles and the use of solutions >0.2 Pas is problematic22. Lastly, all of the nozzle-based devices are expected to impart some damage to islets, since the diameter of the nozzles used is between 100 µm and 500 µm, while ~15% of human islets can be larger than 200 µm23.
In this video, we describe an alternative method to encapsulate mammalian cells by forming droplets in a single emulsification step instead of drop-by-drop. Since bead production is performed in a simple stirred vessel, the method is suitable for small (~1 mL) to large-scale (103 L range) bead production with low equipment costs24. This method allows the production of beads with high sphericity using a broad range of alginate viscosities with short (e.g. 20 min) bead generation times. This method was originally developed by Poncelet et al.25,26 and used to immobilize DNA27, proteins28 including insulin29, and bacteria30. We have recently adapted these methods to the encapsulation of mammalian cells using pancreatic beta cell lines31,32 and primary pancreatic tissue32.
The principle of the method is to generate a water-in-oil emulsion consisting of alginate droplets in mineral oil, followed by internal gelation of the alginate droplets (Figure 1). First the encapsulant (e.g., cells) is dispersed in an alginate solution containing a fine grain calcium salt with low solubility at the initial process pH. The alginate mixture is then added to an agitated organic phase to create an emulsion, usually in the presence of a surfactant. In the case of mammalian cell encapsulation, components present in serum can act as surfactants. Next, the pH is reduced in order to solubilize the calcium salt by adding an oil-soluble acid that partitions into the aqueous phase. Acetic acid, with a mineral oil/water partition coefficient <0.00533, should be pre-dissolved in oil, then added to the emulsion where it is mixed in the oil phase and rapidly partitions into the aqueous phase34. Figure 2 illustrates the chemical reactions and diffusion that take place during the acidification and internal gelation step. Finally, the encapsulated cells are recovered by phase inversion, phase separation accelerated by centrifugation, repeated washing steps and filtration. These steps can then be followed by bead and cell sampling for quality control analyses, in vitro cell culture and/or transplantation of the encapsulated cells.
Figure 1: Schematic of the emulsification-based process to encapsulate mammalian cells. Beads are first produced by emulsifying an alginate, cell and CaCO3 mixture in mineral oil (steps 1 and 2 in the schematic), triggering internal gelation by adding acetic acid (step 3). The gelled beads are then separated from the oil by adding an aqueous buffer to trigger phase inversion (step 4), followed by centrifugation and oil aspiration (step 5), and then filtration (step 6). Finally, the beads collected on the filter are transferred into cell culture medium for in vitro culture or for transplantation. Please click here to view a larger version of this figure.
Figure 2: Reactions and diffusion steps occurring during internal gelation. (1) Acetic acid is added to the organic phase and is transported to the alginate droplets by convection. (2) The acetic acid partitions into the aqueous phase. (3) In the presence of water, the acid dissociates and diffuses to reach the CaCO3 grains depicted in dark blue. (4) The H+ ions are exchanged with the Ca2+ ions in CaCO3, releasing Ca2+ ions. (5) The calcium ions diffuse until they encounter unreacted alginate, leading to the ionotropic cross-linking of the alginate chains. Please click here to view a larger version of this figure.
Contrary to conventional nozzle-based cell encapsulators, a broad bead size distribution is expected from this process due to the mechanism of droplet formation in stirred emulsification. For a subset of applications, this bead size distribution may be problematic. For example, a larger fraction of cells may be exposed at the bead surface in smaller beads. If nutrient (e.g. oxygen) limitations are a concern, these limitations may be exacerbated in larger beads. An advantage of the stirred emulsification method is that the average bead size can readily be adjusted by changing the agitation rate during the emulsification step. The broad bead size distribution can also be exploited to study the effect of bead size on encapsulated cell performance.
Mammalian cell encapsulation by emulsification and internal gelation is an interesting alternative for laboratories that are not equipped with a bead generator. Furthermore, this method gives users the option of reducing the processing time, or generating beads at very low or very high alginate concentrations.
The protocol outlined below describes how to encapsulate cells in 10.5 mL of 5% alginate solution prepared in 10 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) buffer. The alginate consists of a 50:50 mixture of transplantation-grade LVM (low viscosity high mannuronic acid content) and MVG (medium viscosity high guluronic acid content) alginate. Calcium carbonate at a final concentration of 24 mM is used as the physical cross-linking agent. Light mineral oil constitutes the organic phase, while acetic acid is used to acidify the emulsion and trigger internal gelation. However, the alginate type and composition, as well as the process buffer selected depend on the desired application32. A variety of alginate types (see Table of materials) have been used to produce beads with this protocol.
1. Prepare the Alginate Solution, the CaCO3 Suspension and the Acidified Oil
2. Alginate Bead Generation by Emulsification and Internal Gelation
3. Bead Recovery
4. Quality Control and Applications
NOTE: In order to ensure encapsulated cell and bead quality, the bead size distribution and cell survival after the process should be quantified. Reversing the gel to recover the cells from within the beads for further analysis is commonly performed.
At the end of the emulsification and internal gelation process, a bead volume similar to the initial alginate and cell mixture volume should be recovered. The beads should be highly spherical with few defects (Figure 3). The beads should be sufficiently strong to withstand pipetting through large-bore pipettes. At high alginate concentrations, encapsulated oil or air droplets may be observed in the beads. This likely occurs due to the entrapment of oil in alginate droplets during the emulsification step (oil/water/oil double emulsion). In our hands, these beads represented a small fraction (<5%) of the bead volume and most of these beads were removed while aspirating the oil during the bead recovery step, due to their lower density.
Figure 3: Representative 5% alginate beads containing encapsulated βTC3 cells obtained by the emulsification and internal gelation process. Phase contrast image of 5% alginate beads containing 5 x 106 βTC3 cells/mL beads. Please click here to view a larger version of this figure.
The size of the beads obtained will not be uniform – this is an intrinsic feature of stirred emulsions (Figure 4A). During emulsification in turbulent flow, pressure fluctuations act to disrupt droplets37,38, counteracting the surface tension and the viscous forces in the dispersed phase39. Viscous forces are non-negligible if the viscosity of the dispersed phase is much higher than that of the continuous phase. Assuming that the maximum local energy dissipation rate is proportional to the average energy dissipation rate in the vessel, the maximum stable droplet diameter dmax (if >> than the Kolmogorov scale) can be calculated as follows39,40,41,42
where ρc is the density of the continuous phase, ε is the average energy dissipation rate in the vessel, σ is the interfacial tension between the two fluids, µd is the viscosity of the dispersed phase and C is a proportionality constant. When surface tension forces predominate over viscous forces, the right-hand side simplifies to C, whereas if the viscous forces predominate it simplifies to:
In stirred vessels, the energy input under turbulent flow scales is proportional to the power input per unit volume, i.e. para Ni3Di2, where Ni is the agitation rate and Di is the impeller diameter. Therefore, combined with Equation 1 and given a constant Di, if surface tension forces predominate the dmax is predicted to be proportional to Ni-1.2 (Figure 4B), while it is proportional to Ni-0.75 if viscous forces within the beads predominate as the main retentive force. Figures 4B and 4C suggest that interfacial effects determine the droplet size for low alginate concentrations, while viscous forces become more dominant at higher alginate concentration (e.g. 5% for alginate #3 described in the Table of materials). Since it can be impractical to determine dmax experimentally, high quantile values (e.g. the 95th quantile) of the diameter or the surface area moment mean diameter (d32) have been used instead, as they generally are proportional to dmax43.
Before encapsulating cells, it is recommended that a curve of d32 as a function of Ni be generated (Figure 4B). This curve will help select the adequate Ni to obtain a desired average bead size. A new standard curve will also likely be required for different batches of oil or alginate, or if the alginate concentration is modified. Depending on the vessel geometry and on the alginate concentrations, a relatively wide range of average bead diameters should be achievable. For example, d32 values between 200 µm and > 1 mm can be obtained using 2% (Figure 2C in Hoesli et al.32) to 5% alginate (Figure 4) using the protocol described above.
Figure 4: Bead size distribution. (A) Toluidine blue-O stained 5% alginate beads. (B) d32 as a function of the agitation rate (Ni) for 1.5% and 5% alginate beads using a 50:50 mixture of alginates #1 and #2 (see Table of materials). A single point is shown for the 5% alginate beads. (C) d32 as a function of Ni for 2% and 5% alginate #3 (see Table of materials). The error bars represent the standard deviation of bead sizes within a sample (N >150 beads). Please click here to view a larger version of this figure.
The cell survival from the emulsification and internal gelation process mainly depends on the extent and duration of the pH drop experienced by the cells. Using the process described in this video, 76 ± 2% βTC3 cell survival was measured31. When encapsulating dispersed cells at 105 cells/mL alginate to 107 cells/mL alginate seeding densities, all beads should contain cells. When encapsulating cell clusters such as pancreatic cells, a subset of beads can be expected to be empty. Figure 5 shows typical live cell/ dead cell staining results obtained using this process. Higher cell survival can be achieved by increasing the process buffer capacity and by decreasing the duration of the acidification step. For example, 90 ± 2% MIN6 cell survival was obtained by using 60 mM MOPS (3-(N-morpholino)propanesulfonic acid) as the process buffer and by decreasing the overall process time to 4 min32. However, the beads generated were significantly stronger with the 10 mM HEPES process buffer than with the 60 mM MOPS process buffer due to the increased amounts of Ca2+ released at lower pH31. Both the short MOPS process and the longer HEPES process should generate beads that are sufficiently stable for >2 weeks in vitro culture or transplantation. However, beads may swell or degel entirely in solutions with low calcium or with high chelator concentrations (for example, calcium-free phosphate-buffered saline solution).
For encapsulated MIN6 cells grown in vitro, visible cell aggregates form after ~5-7 days of in vitro immobilized cell expansion and ~150 µm diameter spheroids can be harvested after 2 weeks. The growth rate of alginate-immobilized cells is expected to be lower than in non-immobilized cultures, particularly for gels with higher alginate concentration or higher guluronic acid content.
Figure 5: Encapsulated cells immediately after the process. (A) Live cell (green, calcein AM) and dead cell (red, ethidium homodimer) staining of encapsulated cells. (B) Phase contrast image of the same bead. The MIN6 cells were encapsulated in 2% alginate beads using the 12 min emulsification, 8 min acidification and a 10 mM HEPES buffer as described in this protocol. Please click here to view a larger version of this figure.
Various steps (depicted in Figure 2) during the internal gelation reaction can limit the overall kinetics. For calcium carbonate grains larger than ~2.5 µm, the rate of carbonate dissolution has been shown to be rate-limiting26,44. The acidification step that leads to internal calcium release has also been shown to be the critical process variable affecting cell survival32. The conditions that lead to internal gelation are therefore crucial both for bead quality as well as cell survival. Depending on the final application for the beads, different optimal process conditions may be chosen. For example, a higher pH drop (e.g. using 10 mM HEPES to buffer the alginate solution) may be desirable for transplantation applications where the long-term stability of the beads is crucial. On the other hand, a lower pH drop (e.g. using 60 mM MOPS to buffer the alginate solution) may be preferable for in vitro applications.
In addition to adjusting the level and duration of acidification, other process modifications can be envisioned. For example, other stirred vessel configurations, alginate concentrations, process buffers, organic phases and acidifying agents can be used. As discussed previously, the average bead size can easily be adjusted by changing the agitation rate during the emulsification step. In addition, longer emulsification times can lead to smaller bead sizes, but can cause decreased cell viability. The publications by Poncelet et al.25,26,45 and Hoesli et al.31,32 have tested and discussed some of these options.
Slight changes in the protocol described above can have a profound impact on bead generation, bead quality and cell survival. Table 1 provides a troubleshooting guide to understand and solve problems that may occur.
Problem | Typical cause | Suggested solutions |
The impeller stops rotating during alginate or acid oil addition | The magnetic field is insufficient for adequate mixing or the impeller shaft is misaligned | Ensure that the spinner flask is well-centered on the magnetic plate and taped onto the plate in an optimal position prior to adding the alginate mixture. Test very high agitation rates and then decreased to the desired agitation rate prior to adding the alginate mixture. |
Ensure that the impeller shaft is well-aligned and stable. For the Bellco spinner flasks, ensure that the impeller is firmly held in place by bolts on both sides of the spinner flask cap. Ensure that the impeller movement is limited to the radial direction (the impeller should not wobble on the shaft). | ||
No beads obtained | The alginate was improperly emulsified | Ensure that alginate microdroplets are visible during the emulsification. If not, verify that the oil phase density and the vessel geometry23 are similar to those described in this protocol. A dense oil phase or vessel geometry that does not provide sufficient local energy dissipation may not allow proper alginate emulsification. |
No beads obtained | The internal gelation did not occur | A common error is incomplete dissolution of the acetic acid in the oil phase. Ensure that the acid oil mixture is sufficiently vortexed and no acetic acid is observed at the bottom of the tube. |
If the impeller does not reach the oil, more oil can be added prior to adding the alginate. However, the amount of acetic acid should be adjusted according to the following equation:![]() Where ![]() ![]() |
||
No beads obtained, or beads are mechanically unstable | The alginate solution zero-shear viscosity is below 0.004 Pa·s or above 112 Pa·s | Due to the high batch-to-batch variability of alginate lots, it is recommended to measure the viscosity of different concentrations of alginate solutions. The combination of a high-viscosity batch with a low-viscosity batch may be required to reach a target viscosity. For the 5% LVM and MVG mixture used here, the zero-shear viscosity was 3.3 Pa·s. It should be noted that alginates with too short guluronic acid block lengths or short overall chain lengths may not allow gel formation. |
Beads are too large or too small | The agitation rate during the emulsification is too low or too high for the alginate concentration or viscosity used | Increasing turbulence by increasing the agitation rate or adding baffles will decrease the bead size. The alginate concentration, the temperature, the vessel and impeller geometry will all impact the bead size obtained. |
Many broken beads are observed | Harsh bead handling, damage during the internal gelation step or insufficient bead strength. | Mechanical damage to the beads can lead to bead rupture once internal gelation is initiated. Reduce the agitation rate immediately prior to the acidification step and ensure that the beads are handled gently by using 25 mL pipettes or 1000 µL pipettes with cut tips. If damaged beads are observed even with careful handling and reduced agitation during acidification, the bead strength may be inadequate. This could be due to insufficient guluronic acid content or incomplete gelling (e.g. insufficient acidification). |
Low cell viability | Initial process pH too high or final pH too low | Mammalian cell process survival can be increased by limiting the pH drop and the acidification time23. However, these modifications may reduce the bead mechanical strength. Another potential issue may be the use of process solutions (e.g. alginate solution) with inadequately adjusted osmolalities. |
Low cell viability in the smaller beads | Insufficient alginate solution buffering capacity | A greater pH drop is expected in the smaller beads than in the larger beads due to the greater surface/volume ratio of these beads. Since the acetic acid added to the emulsion is limited, higher acetic acid concentrations are predicted in smaller beads for the same CaCO3 concentration. Increasing the alginate solution buffering capacity (e.g. using 60 mM MOPS instead of 10 mM HEPES) can increase cell survival in the smaller beads32. Small beads with high cell losses could be selectively removed by filtration or sedimentation without incurring large volumetric losses. |
Incomplete CaCO3 dissolution | Insufficient acidification or acidification time. Note that this issue may not be problematic if bead stability is sufficient. | Incomplete CaCO3 dissolution has been observed in the larger beads following this protocol32. The extent of CaCO3 dissolution can vary depending on the CaCO3 grain size, the pH drop during the process, the duration of the pH drop, as well as the alginate concentration used. This issue is not considered problematic as long as the bead mechanical stability is sufficient for the desired application. To increase bead cross-linking, ensure that the CaCO3 grain is sufficiently small (~2.5 µm). Sonicating the CaCO3 suspension can help disrupt aggregated grains26. At very low or high alginate concentrations, adjustments in the CaCO3 concentration may be required. Lastly, chelators or solutions with low levels of divalent ions will lead to the gradual loss of Ca2+ in the gel. If gradual loss of bead mechanical stability is observed, verify that the bead storage or culture buffer contains >2 mM Ca2+. |
Table 1: Troubleshooting guide. List of problems, probable causes and potential solutions.
Potential limitations of the emulsification and internal gelation process for in vitro culture of mammalian cells and transplantation applications include: (1) the necessity for transient exposure to low pH, (2) the need to remove residual oil, (3) the polydisperse bead size distribution, (4) the potential shear stress during alginate droplet formation and (5) the higher porosity of internally gelled beads compared to externally gelled beads47.
High cell survival and function of MIN6 and βTC3 have been obtained after process optimization32. However, this process may not be suitable for more pH-sensitive cells. The complete elimination of oil from the beads may also be challenging, and Food and Drug Administration (FDA)-approved oil types are required for eventual clinical applications. The polydisperse bead size distribution may also be somewhat problematic for immunoisolation applications, where complete cell encapsulation and long-term cell survival are desirable. Whereas small bead sizes can lead to cell protrusion48, large beads can hinder nutrient diffusion and/or cell function35,49. On the other hand, the emulsification process offers an interesting opportunity to investigate the effect of bead size on cell survival after transplantation by generating both small and large beads in a single process. The average bead size can readily be adjusted by modifying the agitation rate during the emulsification step (Figure 4). Lastly, the higher porosity of internally gelled beads may also be detrimental for immunoisolation, where molecules such as antibodies and cytokines should be excluded from the beads. However, the emulsification process also allows bead generation from very concentrated alginate solutions, thereby reducing the alginate bead pore size.
The emulsification and internal gelation process is an interesting alternative to nozzle-based encapsulators for laboratories wishing to scale up cell encapsulation or lacking access to an encapsulator. This process is a robust and simple method to immobilize cells in alginate beads using very dilute or very concentrated alginate solutions. Some of the subsequent experiments that can be performed with the immobilized cells are to determine the effect of the alginate matrix on cell fate decisions, or to determine to what extent the beads provide a barrier between transplanted cells and the host immune system. The emulsification and internal gelation method provides a new approach to encapsulate primary islets in large batches. The method also allows cell encapsulation in highly concentrated alginate beads with reduced porosity and hence reduced antibody access to the encapsulated cells. The emulsification and internal gelation method therefore provides a new means to generate clinically relevant quantities of encapsulated therapeutic cells.
The authors have nothing to disclose.
We thank Jill Osborne for her ground-laying work on the emulsification process and Lauren Wilkinson for technical support. We thank Dr. Igor Laçik, Dr. Timothy J. Kieffer and Dr. James D. Johnson for their input and collaboration. We thank Diabète Québec, JDRF, ThéCell, the Centre québécois sur les matériaux fonctionnels (CQMF), the Natural Sciences and Engineering Research Council (NSERC), the Centre for Human Islet Transplantation and Beta-cell Regeneration, the Canadian Stem Cell Network, the Michael Smith Foundation for Health Research, le Fonds québécois de la recherche sur la nature et les technologies and COST 865 for financial support.
Reagents and consumables | |||
LVM alginate (transplantation-grade) | Novamatrix | Non-applicable | Referred to as "alginate #1" in the results. |
MVG alginate (transplantation-grade) | Novamatrix | Non-applicable | Referred to as "alginate #2" in the results. |
Alginate (cell culture-grade) | Sigma | A0682 (low viscosity) or A2033 (medium viscosity) | A2033 is referred to as "alginate #3" in the results. |
DMEM | Life Technologies | 11995-065 | |
Fetal bovine serum, characterized, Canadian origin | Thermo Fisher Scientific | SH3039603 | |
Glutamine | Life Technologies | 25030 | |
Penicillin and streptomycin | Sigma | P4333-100ML | |
HEPES, cell culture tested | Sigma | H4034-100G | |
NaCl | Thermo Fisher Scientific | S271-1 | |
Fine-grain CaCO3 | Avantor Materials | 1301-01 | After preparing the CaCO3 suspension, sonicate and use within one month. |
Light mineral oil | Thermo Fisher Scientific | O121-4 | Sterile filter through a 0.22 μm pore size membrane prior to use. |
Glacial acetic acid | Thermo Fisher Scientific | A38-500 | Handle with caution: refer to MSDS. |
Sterile spatulas | Sigma | CLS3004-100EA | |
Sterile nylon cell strainers, 40 µm | Thermo Fisher Scientific | 08-771-1 | |
Serological pipettes (2 mL, 5 mL, 10 mL, 25 mL) | Sarstedt | 86.1252.001, 86.1253.001, 86.1254.001 and 86.1685.001 | |
Pasteur pipettes | VWR | 14673-043 | |
Toluidine Blue-O | Sigma | T3260 | |
Equipment | |||
100 mL microcarrier spinner flasks | Bellco | 1965-00100 | The impeller configuration with recent models may not be suitable for adequate emulsification. A blade able to sweep the oil down to 0.5 cm from the bottom of the flask can be custom-made from a Teflon sheet. |
Magnetic stir plate with adjustable speed | Bellco | 7760-06005 | The rotation speed should be calibrated (e.g. using a tachometer) prior to use. |
Cell counter | Innovatis | Cedex AS20 | This system is now sold by Roche. This automated cell counter can also be replaced by manual cell enumeration after Trypan blue staining using a hemocytometer. |
LED light box | Artograph | LightPad® PRO | This item can be replaced by other types of illuminators. |
Handheld camera | Canon | PowerShot A590 IS | A variety of handheld cameras can be used to capture toluidine blue-o stained bead images. A ruler should be placed next to the Petri dish containing the beads prior to acquiring images. |
Fluorescence microscope with phase contrast and adequate fluorescence filters | Olympus | IX81 | Several microscopy systems were used to image the beads. The results shown here were obtained with an IX81 microscope equipped with GFP and TRITC fluorescence filters. To capture entire beads, 4X to 20X objectives were used depending on the agitation rate. Live/dead staining images were typically captured with 20X to 40X objectives. |
Image aquisition software | Molecular Devices | Metamorph | A variety of image acquisition software can be used to acquire phase contrast and fluorescence images. |
Image analysis freeware | CellProfiler | Non-applicable | A variety of image analysis software can be used to identify beads as objects and analyze bead size (e.g. ImageJ). |