Apolipoprotein (apo) E and its polymorphism are linked to the pathogenesis of late-onset and sporadic Alzheimer's disease (AD). ApoE facilitates the deposition and fibrillogenesis of beta-amyloid (Abeta), and may participate in Abeta clearance. We recently found that apo(E-AII) complex binds to Abeta much more strongly than does monomeric apoE. Here, we investigated the effect of apoAII on the interaction between apoE and Abeta. Addition of apoAII to apoE monomers increased the binding of apoE2 and apoE3 to Abeta(1-42), presumably following the formation of apo(E3-AII), apo(E2-AII), and apo(AII-E2-AII) complexes. This increased binding was not seen in the case of apoE4. When neuroblastoma cells were cultured in media containing Abeta(1-42) and a mixture of apoE3 and apoAII, intracellular Abeta was significantly reduced and cell viability was maintained at a higher level than in cells cultured without apoAII. ApoE2 itself seemed to act as an inhibitor of the endocytosis of Abeta, and we did not observe a significant effect of apoAII on the movement of Abeta in apoE2-containing medium. However, cell viability could be maintained at a higher level (as with apoE3) by adding apoAII to apoE2, despite the reduced viability of cells incubated without apoAII. In medium containing apoE4, both the amount of Abeta accumulated into cells and the cell viability were unchanged by the presence of apoAII in the medium. In addition, apoE4 itself was toxic, as previously suggested. These findings demonstrate that the type of apo(E-AII) complex present could underlie the isoform-specific role of apoE in the pathogenesis of AD.