Jan Holmgren

By a double-diffusion precipitation-in-gel technique, isolated cholera toxin as well as its natural toxoid were shown to be fixed and precipitated by the ganglioside G(M1) but not by any of the related glycolipids G(M3), G(M2), G(M1)-GlcNAc, G(D1a), G(D1b), G(T1), globoside, G(A1), and tetrahexoside-GlcNAc. Twenty-five nanograms of G(M1) was enough to give a precipitation line with 1.2 mug of toxin, whereas about 50 ng was required with this amount of toxoid. G(M1) also inactivated the toxin in the ileal loop as well as in the intradermal models in rabbits. A 1: 1 molar ratio of ganglioside to toxin was found limiting, e.g., 100 pg of G(M1) could inactivate 5 ng (about 50 blueing doses) of isolated toxin. G(M1) inactivated crude toxin (culture fil rate) with the same efficiency as isolated toxin, and the inactivating capacity of G(M1) was unaffected by mixing with other gangliosides, indicating the specificity in the reaction between G(M1) and toxin. The other glycolipids tested did not inactivate toxin except G(D1a) and G(A1) which did so with approximately 1,000 times less efficiency than G(M1). This identified the portion Gal --> GalNAc [Formula: see text] as the critical region in G(M1) for toxin fixation, and it is postulated that this may be the tissue receptor structure for the cholera toxin.

There is a great need for substances that can act as adjuvants on local mucosal immune responses to perorally (p.o.) administered immunogens and which could be included in future oral vaccines. In this study we show that in mice cholera toxin (CT) is a potent adjuvant on enteric mucosal immune responses to related (cholera B subunit) as well as unrelated (KLH) antigens presented by the p.o. route. The adjuvant action of CT was dose-dependent and was achieved only when CT was given p.o. and together with the antigen. Both priming (memory induction) and boosting of the gut mucosal immune system by the oral route were greatly potentiated by CT. High numbers of specific antibody-producing cells as well as substantial mucosal memory in the lamina propria were stimulated by p.o. priming immunizations if CT adjuvant was included. Anamnestic responses could be elicited by a single p.o. booster immunization for at least 10 weeks and probably much longer. The adjuvant action of CT is suggested to involve activation of adenylate cyclase and cyclic AMP-mediated signals with differential effects on B and regulatory T intestinal lymphocytes. The adjuvant-active dose of CT, 100-500 ng, was lower than the immunogenic dose (2 micrograms) and much below the p.o. dose needed for detectable net fluid secretion in mouse intestine (5-10 micrograms). Cholera B subunit (10 micrograms) administered p.o. together with 500 ng of CT was 50 times more effective in stimulating gut mucosal anti-toxin responses compared with B subunit vaccine alone. Our results suggest that CT or substances that use similar adjuvant mechanisms may substantially increase the mucosal immunogenicity and efficacy of non-replicating oral vaccines.