Functional domains in the Alzheimer's disease-associated presenilin 1 protein

Abstract

Alzheimer's disease (AD), the most common cause of dementia, is neuropathologically characterized by the presence of neurofibrillary tangles and amyloid plaques in the brain. Amyloid plaques are extracellular deposits primarily composed of the amyloid beta-peptide (Abeta). Abeta is derived from the amyloid beta-precursor protein (APP) by sequential cleavages at the betasecretase and gamma-secretase sites. gamma-Secretase cleavage is performed by a high molecular weight protein complex containing presenilin (PS), nicastrin (Net), Aph-1 and Pen-2. The gamma-secretase complex is an unusual transmembrane aspartyl protease that cleaves APP within the transmembrane domain. Two aspartate residues within transmembrane regions of the multispanning PS protein (Asp-257 and Asp-385) are thought to form the catalytic site of the enzyme complex.

The PS protein is endoproteolytically cleaved to generate an N-terminal fragment (NTF) and a C-terminal fragment (CTF) that forms an NTF/CTF heterodimer in the active gamma-secretase complex. Recently, a large number of single membrane-spanning proteins have been shown to be cleaved within their transmembrane domains by the gamma-secretase complex, in a process referred to as regulated intramembrane proteolysis (RIP). One of the gammasecretasedependent RIP substrates recognized is Notch, a protein involved in cell differentiation during embryonic development. Hence, recent identification and experimental analysis of the gamma-secrctase complex has lead to the understanding that the Abeta-peptide is produced by an enzyme complex that is also involved in processing of a variety of other substrates by an apparently novel catalytic mechanism. PS is a key component of the gamma-secretase complex. Mutations in PS are associated with familial AD and PS was the first component in the complex shown to associate with gamma-secretase activity. Thus, the biochemical understanding of the gamma-secretase complex and the role of PS is important from a basic biological point of view and could help in the development of small molecules that modulate gamma-secretase processing in an APP-specific manner.

The work presented in this thesis has focused on identifying and characterizing domains in PS1 that are critical for mediating gamma-secretase activity. To further understand the role of these functional domains, the study was extended to include determination of the topology of PS1. In paper I we showed that co-expression of PS1 NTF and CTF can functionally substitute for expression of the full-length molecule in PS-deficient mammalian cells. This finding enabled studies of domains involved in regulating PS1 function, under conditions bypassing endoproteolysis (papers I-III). In paper I, co-expression of NTF and CTF, with the catalytic aspartates substituted for alanines, demonstrated that Asp-257 and Asp-385 affect gamma-secretase activity independently of their effect on PS1 endoproteolysis. However, the Asp-385 residue was also shown to have a direct effect on endoproteolysis, since co-expression of wild type CTF, but not CTF with Asp-385 substituted for an alanine, and a C-terminally truncated endoproteolysisdefective PS1 molecule rescued endoproteolysis of the latter molecule. In paper II, cointroduction of PS1 NTF and CTF allowed the search for functional domains in the NTF, which lead to the identification of an amino acid residue (Tyr-288) that controls PS1 endoproteolysis and -y-secretase activity. Importantly, bypassing endoproteolysis by co-expressing mutant PS1 NTF and wild type CTF did not rescue gamma-secretase activity for non-functional Tyr-288 mutants. Moreover, non-functional Tyr-288 mutants interacted with the other gamma-secretase components and were distributed to the same subcellular compartments as wild type PS1. In paper III the C terminus of PS1 was shown to be essential for gamma-secretase complex integration, gamma-secretase activity, and PS1 endoproteolysis.

Co-expression of NTF and C-terminally truncated CTFs, did not rescue gammasecretase complex formation or activity. Moreover, stably expressed PS1 CTF interacted with Net and Aph-1 in the absence of full-length PS1 or PS1 NTF. The topology study in paper IV proposed a nine transmembrane domain topology for PS1 with the C terminus in the lumenal/extracellular space. Hence, our topological model for PS1 suggests that the functional domain in the C terminus identified in paper III faces the lumen. Taken together these results contribute to the basic knowledge of how the gamma-secretase complex assembles and performs its enzymatic action.