, 2012) has presented with a similar microhemorrhage liability (Ostrowitzki et al., 2012), whereas several other antibodies that bind strongly to soluble Aβ appear to lack the adverse event (Adolfsson et al., HDAC inhibitor 2012; Farlow et al., 2012; La Porte et al., 2011). Our results have significant implications for the clinical development of N-terminal nonplaque-selective

antibodies. The proposed mechanism of action for the N-terminal antibodies was based upon preclinical studies in transgenic mice and, as such, their respective clinical implementation has been modeled upon the phagocytosis mechanism, maximal effector function, and studies performed in patients with extensive deposition. A small neuroimaging study performed with the PET ligand PIB did report a significant amyloid reduction (∼25%) in patients receiving the monoclonal antibody bapineuzumab as compared to placebo controls (Rinne et al., 2010). It is important to highlight that ∼17% of the perceived amyloid reduction Linsitinib was due to an atypical rise in the PIB signature in the small (n = 7) placebo cohort (Ossenkoppele et al., 2012). We demonstrate that N-terminal antibodies that bind both soluble and insoluble Aβ fail to lower existing Aβ deposition in our preclinical PDAPP transgenic model in line with previous reports in the literature. Many studies

have demonstrated that N-terminal Aβ antibodies prevent plaque deposition (Bard et al., 2000; Schenk et al., 1999; Schroeter et al., check 2008) including studies in FcRγ knockout mice (Das et al., 2003). These, along with our current results, suggest that it is unlikely that plaque prevention is due to phagocytosis. A more plausible mechanism for N-terminal Aβ antibodies would be their direct binding to soluble Aβ, either monomer or oligomeric complexes within the CNS, to

facilitate the prevention or elimination of amyloid-promoting Aβ seeds. The anti-seeding and spreading mechanism would not be dependent upon effector function and thus if correct should enable the generation of a safer therapeutic antibody. This strategy is already being clinically implemented with MABT5102A, also known as Roche’s Crenezumab, a humanized N-terminal Aβ antibody engineered as a minimal effector function IgG4 (Adolfsson et al., 2012). Indeed, testing multiple humanized anti-Aβ antibodies that target different mechanisms of action in the clinic (mono or combination therapy) will be important for the field, especially as clinicians begin investigating the presymptomatic populations. In summary, these studies have demonstrated that the development of plaque-specific Aβp3-x antibodies that lack binding to soluble Aβ peptides leads to significant engagement of Aβ deposits (i.e., plaque binding) and to the subsequent removal of existing plaque without a microhemorrhage liability.