Multiple Emulsions Oral Vaccine Vehicles for Inducing Immunity or Tolerance zy THOMAS L. HEARN, MARGARET OLSEN, AND ROBERT L. HUNTER zyxw Department of Pathology and Laboratory Medicine Woodruff Memorial Building, zyxwv Room 762 Emoly University 1639 Pierce Drive Atlanta, Georgia 30322 It has been known since the 1960s that water-in-oil-in-water or multiple emulsions are effective adjuvants, but instability of available preparations precluded widespread use.’ z Our research on the adjuvant activity of nonionic block copolymers suggested that larger hydrophobic molecules would be more effective.’ When such agents were synthesized, we found that they were not only excellent adjuvants but that they also effectively stabilized multiple emulsions.2 Being insoluble in both oil and water, these amphipathic copolymers preferentially localize on and stabilize oil-water interfaces. These multiple emulsions are formed of particles ranging from 5 to 50 microns in diameter and consisting of 80% saline and 20% oil phase (squa1ene:span zyxwv XO at 9:1), with the copolymer in the aqueous phase. The antigen is also in the saline, which is dispersed in droplets that are smaller than the resolution of light microscopes. These particles are suspended in a second aqueous phase of arbitrary volume. Multiple emulsions can be prepared with practically any antigen with equal ease. When fed to animals, multiple emulsions pass through the upper gastrointestinal tract to the colon intact. Emulsion particles can be seen over Peyer’s patches in mice from 20 minutes, and to 8 hours after oral infusion. Most of the particles break in the colon as the feces dehydrate. Using titanium dioxide particles as a marker, it was shown that the emulsions promote particle uptake by the domes of Peyer’s patches. Very few particles were taken up by Peyer’s patches without copolymer in emulsion even though they are not digestible. Multiple emulsions can be effective vehicles for inducing both serum TgC and secretory IgA antibody responses.2 For example, ICR outbred mice were immunized orally one, two, or three times at two-week intervals with various doses of TNP,,-ovalbumin and were followed for a year before boosting with an identical oral immunization. The proportion of responders and titers increased with increasing dose and number of immunizations so that 100% of animals produced both intestinal secretory IgA and circulating TgG antibody responses to three infusions of 100 zyxwv pg of antigen. The IgG titers persisted for one year and were boosted by a further oral immunization. Oral immunization using multiple emulsions also uniformly primed animals for an increased response following parenteral immunization (FIG. 1). We found no evidence of tolerance (defined as a reduction in response to parenteral immunization) in this model. Recently, M. Elson reported that oral immunization with egg albumin in the same delivery system induced tolerance. We were aware of these results and tried to reproduce them. We were unable to induce tolerance with TNPlo-ovalbumin, even though tolerance was regularly induced with ovalbumin. These data suggest that the outcome of immuniza- 388