Adhesion molecule L1 binds to amyloid beta and reduces Alzheimer's disease pathology in mice

Alzheimer's disease (AD) is a devastating neurodegenerative disorder and the most common cause of elderly dementia. In an effort to contribute to the potential of molecular approaches to reduce degenerative processes we have tested the possibility that the neural adhesion molecule L1 ameliorates some characteristic cellular and molecular parameters associated with the disease in a mouse model of AD. Three-month-old mice overexpressing mutated forms of amyloid precursor protein and presenilin-1 under the control of a neuron-specific promoter received an injection of adeno-associated virus encoding the neuronal isoform of full-length L1 (AAV-L1) or, as negative control, green fluorescent protein (AAV-GFP) into the hippocampus and occipital cortex. Four months after virus injection, the mice were analyzed for histological and biochemical parameters of AD. AAV-L1 injection decreased the A beta plaque load, levels of A beta 42, A beta 42/40 ratio and astrogliosis compared with AAV-GFP controls. AAV-L1 injected mice also had increased densities of inhibitory synaptic terminals on pyramidal cells in the hippocampus when compared with AAV-GFP controls. Numbers of microglial cells/macrophages were similar in both groups, but numbers of microglial cells/macrophages per plaque were increased in AAV-L1 injected mice. To probe for a molecular mechanism that may underlie these effects, we analyzed whether L1 would directly and specifically interact with A beta. In a label-free binding assay, concentration dependent binding of the extracellular domain of L1, but not of the close homolog of L1 to A beta 40 and A beta 42 was seen, with the fibronectin type III homologous repeats 1-3 of L1 mediating this effect. Aggregation of A beta 42 in vitro was reduced in the presence of the extracellular domain of L1. The combined observations indicate that L1, when overexpressed in neurons and glia, reduces several histopathological hallmarks of AD in mice, possibly by reduction of A beta aggregation. L1 thus appears to be a candidate molecule to ameliorate the pathology of AD, when applied in therapeutically viable treatment schemes. (c) 2013 Elsevier Inc. All rights reserved.