The Alzheimer’s disease related gamma-secretase complex : localization and novel interacting proteins

Abstract

Alzheimer’s disease (AD) is caused by synaptic and neuronal loss in the brain that eventually results in cognitive decline. Characteristic hallmarks of AD are senile plaques containing the amyloid beta-peptide (Abeta) and neurofibrillary tangles containing hyperphosphorylated tau protein. Abeta is produced from the amyloid precursor protein (APP) by sequential proteolytic cleavages by beta-secretase and gamma-secretase, and the polymerization of Abeta into amyloid plaques is thought to be the main pathogenic event in AD. Since gamma-secretase mediates the final cleavage that liberates Abeta from APP, gamma-secretase has been widely studied as a potential drug target for the treatment of AD.

gamma-Secretase is a transmembrane protease complex containing presenilin, nicastrin, Aph-1, and Pen-2, which are sufficient for gamma-secretase activity. gamma-Secretase has more than 60 substrates including APP and Notch. Inhibitors of gamma-secretase caused side effects in clinical trials, probably due to the involvement of gamma-secretase in Notch signaling. Therefore, more specific regulation or modulation of gamma-secretase activity is needed. In the last years, gamma-secretase associated proteins (GSAPs) such as CD147, TMP21, and proteins in the tetraspanin web family have been reported to regulate Abeta production. In this thesis, we have characterized gamma-secretase in different subcellular localizations, and identified and characterized novel GSAPs in mammalian brain.

Studies in cell lines have shown that g-secretase is partially localized to lipid rafts, microdomainsenriched in cholesterol and sphingolipids, which can be prepared biochemically as detergent resistant membranes (DRMs). In Paper I, we show that DRMs from human brain were enriched in gamma-secretase. gamma-Secretase activity was also high, as accessed by Abeta and AICD production levels. The DRM fraction was subjected to size exclusion chromatography, and all of the gamma-secretase components and a lipid raft marker were found in the void volume (> 2000 kDa). The size of the DRMs indicates that they contain other proteins and lipids. In Paper II, the distribution of active g-secretase in different subcellular compartments in brain was investigated. Previously, active gamma-secretase has been localized to the Golgi apparatus, endosomes, and plasma membranes in cell studies. Here, we showed that highly active gamma-secretase was present in endosomes, synaptic vesicles, and synaptic/plasma membranes in rat brain.

The localization of active gamma-secretase in synapses and endosomes was also confirmed by fluorescent labeling with an active site inhibitor using confocal microscopy. In Paper III, we developed an efficient way to purify g-secretase and GSAPs. Microsomal membranes from brain were incubated with a gamma-secretase inhibitor with cleavable biotin group (GCB). After affinity purification, bound proteins were subjected to LCMS/ MS analysis. All of the known gamma-secretase components were identified, and TMP21 and the PS associated protein syntaxin1 were identified as GSAPs. In total, over 90 potential GSAPs were identified. In Paper IV, we used GCB affinity purification and identified novel GSAPs in DRMs. By siRNA mediated gene knockdown that a subset of the GSAPs affected Abeta production; voltagedependent anion channel 1, syntaxin12, and cytochrome C oxidase subunit IV isoform 1.

In summary, we conclude that the active gamma-secretase complex is localized to lipid rafts, endosomes, plasma membranes, and at synapses. We identified novel GSAPs in human brain and in DRMs purified from rat brain. We suggest that the interactions between these proteins and g- secretase could be potential drug targets to modulate Abeta production in AD.