Background: New pre-clinical trials in AD mouse models may help to develop novel immunogen-adjuvant configurations with the potential to avoid the adverse responses that occurred during the clinical trials with AN-1792 vaccine formulation. Recently, we have pursued an alternative immunization strategy that replaces QS21 the Th1 type adjuvant used in the AN-1792 clinical trial with a molecular adjuvant, mannan that can promote a Th2-polarized immune response through interactions with mannose-binding and CD35/CD21 receptors of the innate immune system. Previously we established that immunization of wild-type mice with mannan-A beta(28) conjugate promoted Th2-mediated humoral and cellular immune responses. In the current study, we tested the efficacy of this vaccine configuration in amyloid precursor protein (APP) transgenic mice (Tg2576). Methods: Mannan was purified, activated and chemically conjugated to A beta(28) peptide. Humoral immune responses induced by the immunization of mice with mannan-A beta(28) conjugate were analyzed using a standard ELISA. A beta(42) and A beta(40) amyloid burden, cerebral amyloid angiopathy (CAA), astrocytosis, and microgliosis in the brain of immunized and control mice were detected using immunohistochemistry. Additionally, cored plaques and cerebral vascular microhemorrhages in the brains of vaccinated mice were detected by standard histochemistry. Results: Immunizations with low doses of mannan-A beta(28) induced potent and long-lasting anti-A beta humoral responses in Tg2576 mice. Even 11 months after the last injection, the immunized mice were still producing low levels of anti-A beta antibodies, predominantly of the IgG1 isotype, indicative of a Th2 immune response. Vaccination with mannan-A beta(28) prevented A beta plaque deposition, but unexpectedly increased the level of microhemorrhages in the brains of aged immunized mice compared to two groups of control animals of the same age either injected with molecular adjuvant fused with an irrelevant antigen, BSA (mannan-BSA) or non-immunized mice. Of note, mice immunized with mannan-A beta(28) showed a trend toward elevated levels of CAA in the neocortex and in the leptomeninges compared to that in mice of both control groups. Conclusion: Mannan conjugated to A beta(28) provided sufficient adjuvant activity to induce potent anti-A beta antibodies in APP transgenic mice, which have been shown to be hyporesponsive to immunization with A beta self-antigen. However, in old Tg2576 mice there were increased levels of cerebral microhemorrhages in mannan-A beta(28) immunized mice. This effect was likely unrelated to the anti-mannan antibodies induced by the immunoconjugate, because control mice immunized with mannan-BSA also induced antibodies specific to mannan, but did not have increased levels of cerebral microhemorrhages compared with non-immunized mice. Whether these anti-mannan antibodies increased the permeability of the blood brain barrier thus allowing elevated levels of anti-A beta antibodies entry into cerebral perivascular or brain parenchymal spaces and contributed to the increased incidence of microhemorrhages remains to be investigated in the future studies.