This study is part of our effort to develop a vaccine cream that will be applied to the skin in the same way as cosmetic formulations. Skin may be a potent immunological induction site since it acts as an immune barrier by its immunocompetent cells. For long time it was thought that the skin barrier was impenetrable to large molecules, but several macromolecules have been successfully transported through the skin; for instance, application of the antigens cholera toxin (CT) and heat-labile enterotoxin (LT) onto hydrated skin using a patch has induced strong systemic and mucosal immune responses [Glenn, 2000]. Thus, transcutaneous immunization has the potential to make vaccines administration easier and cheaper, while maintaining efficiency and safety, which would facilitate the implementation of worldwide mass vaccination campaigns and provide the means for a fast response to terrorist bio-attacks.

Water-in-oil-in-water (W₁/O/W₂) double emulsions have a compartmentalized structure that provides high capacity of entrapment, protection of fragile substances, combination of incompatible substances in one product and controlled release. These multiphase systems have been regarded as suitable carriers for vaccines due to their harmless emulsification procedure and adjuvant properties as a vehicle [Bozkir, 2004]. Unfortunately, the fact that double emulsions are thermodynamically unstable systems has greatly limited their practical applicability and when long-time stability is indeed achieved, it is hard to break the globules to release the encapsulated material. While small molecules can transport through the oil phase, the delivery of proteins requires an external stimulus to readily destabilize the emulsion. We have demonstrated how temperature changes can induce globules bursting of otherwise stable double emulsions. Using fluorescence capillary videomicroscopy with temperature-variation features, we induced instability in individual W₁/O/W₂ double-emulsion globules by a temperature increase (oil thawing) that led to instant and efficient release of the model protein FITC-BSA [Rojas, 2007]. We used the model oil n-hexadecane, which melts at about 18^oC and therefore greatly facilitates study of this system by allowing the oil phase to freeze while both aqueous phases remain liquid.

In addition, the proposed temperature-sensitive double emulsion was prepared in bulk and studied under more realistic conditions. Keeping the oil phase frozen successfully preserved emulsion stability and hence banned release during storage; on the other hand, key factors that affect the amount of protein released after oil thawing were identified and subsequently optimized. Macroscopically, this globule-breakage mechanism translates into sharp phase separation that leaves an oil layer on top of an aqueous layer that puts the protein in direct contact with the skin. We suggest that these layers could prevent sample loss while occluding and hydrating the skin to improve penetration, thus eliminating the need for a patch. Other components could be incorporated to the system to facilitate antigen penetration and immunogenicity. A biocompatible oil phase that melts above room temperature will eventually replace n-hexadecane so that it remains frozen during storage and readily delivers the protein upon skin administration.