Molecular mechanisms of amyloid formation

Molecular mechanisms of amyloid formation


184KB Sizes 0 Downloads 33 Views





Presenilin-l (PS 1) is an integral membrane protein mvolved in the development of familial Alzheimer disease (FAD). Cadherin-based cell-cell mteractions control critical events in cell-cell adhesion and recognition. PSI accumulates at cell-cell contact sites where it co-localizes with components of the cadherin-based adherens junctions. At these sites, PSI is linked to the cortical cytosksleton and forms detergent-stable complexes with E-cadherin, p-catenin, and wcatenin, all components of adherens junctions. PSI overexpression in human kidney cells enhances cell-cell adhesion. Together, our data show that PSI incorporates into the cadherin/catenin adhesion system and modulates cell-cell adhesion. In the brain, PSI forms complexes with cadherins, concentrates at rynaptlc contacts where it exhibits partial colocalization with synaptophysin and may incorporate into synaptic junctions. Incorporation of PSI into the synaptic cadherin/catenin complex maker that complex a potential target for PSI FAD mutations. PSI forms complexes with the endothelial adherensJunctions that regulate endothelial permeability. Endothelium constitutes the main barner to the passage ot macromolecules and cxculating cells from blood IO tibbues including brain. Incorporation of defective (I..% FAD mutant) PSI in the endothelial adherens junctions could change the permeability of the blood-bram barrier and this change may adversely affect neuronal survival.








Alzhelmer’s dlseahe (AD) can be defined as a syndrome that acne\ from a chrome imbalance between Approduction and Apclearance, leading to the gradual accumulation of the peptide in brain regions subserving memory and cognition. All genetic alterations linked to familial forms of AD have been shown to mcrea\e cerebral Aplevcls. For example, mutations in the presenilins (PS), the most common cause of autosomal dominant AD, selectwely enhance the geneation of Ap42by y-secretaae. We have shown that the immediate substrates of y-secretase, C99 and C83 of APP, co-immunoprecipitate with PS heterodimers in the subcellular loci in which [email protected] generated. Moreover, we observed and mutated two conserved intramembranous aspartatea in TM6 and TM7 of PSI and PS2 and found that each is absolutely required for y-secretase cleavage of C99 and CU. Because our studies with transition-state peptidomimetic inhibitors strongly suggested that y-secretax is an aspartyl protease which cIcuvcs it5 subwatcc withm the membrane, we hypothesized that these two aspartates are the active Gte of an unprecendented intramcmbranous aspartyl protease (i.e., PS is y-secretase). Inhibitors that bind the PS active sate should thus be useful to prevent and treat AD. We have also performed an unbiased screen for proteares that degrade APand found that insubn-degrading enLyme, a 110 kDa metalloprotease, is the principal A&cleaving protemase in neuronal and microglial cultures. Dysfunction of IDE could be a riak factor for developing AD. Finally, we have detected the initial dimeriution of [email protected] in cultured neurons, suggesting that these dimers could be released and serve as a nidus for the oligomeriration of extracellular Ab. These varmus fmdings prowde mechanistically distinct targets for the therapeutic lowering of cerebral Apin order to prevent It\ oligomeri7atmn and wbsequent cytotoxKxy.




AP is derived from APP by the sequential action7 of two proteases termed gamma and beta secretase. Cleavage of APP can occur in a variety of subcellular organelle\. AP42 can be produced in the ER/IC. whde both AP40 and 42 are generated in the Golgi or late in the secretory pathway. A type I membrane protein (BACE) has recently been identified a\ the beta \ecretaae while presenilin I (PSI) i& proposed to serve a, a gamma secretax. We have studied the intracellular locations at which APP is processed in neuronal and non-neuronal cells as well a\ the processing and localization of BACE. By redirecting the intracellular transport of APP, PSI, and BACE we hope to characterize the subcellular organelles where APP cleavage occur? as well as factors that may modify this process. BACE is efficiently glycosylated and transported to the Golgi before accumulating in endosomes. exhibiting a relatively long half-life. Our rtudie\ on the internalization and targeting of BACE will be prewnted Retention of APP in the ERllC rewlt\ in continued production of Ab42,

Amyloid Precursor Protein cmd Af3 Amyloidgenesis

but not AP40. This intracellular AP42 accumulates in the cell largely in an insoluble form. Expression of ER-retained APP in neurons derived from PSl knockout mice still results m the production of AP42, though secreted forms of AP42 as well as AP40 are largely eliminated. Studies using chimeric APP conrtmcts and pharmacological methods to retain APP in the ER or accumulate it in the TGN of PSI-negative neurons affected the production of AP in the TGN, but not in the ER. These results suggest that PSl has organelle-specific effects on AP production and that although PSI is largely found in the ER. it modulates gamma-secretase activity at more distal sites in the biosynthetic pathway of neurons. Thus, we conclude that PSI either plays a modulatory role in gamma-secretase enzymatic activity, or. if PSI is indeed itself a gamma-secretax. that other gamma-secretases must also exist to account for the full production of intracellular A!342 as well a\ the limited production of other AP species in the absence of PSl.





There are at least two distinct pathways for the production and accumulatmn of AP m senile plaque amyloid deposit? associated with Alzheimer’a disease. One ia the well-known and now classical pathway that gives rise to soluble secreted AP, involving the sequential cleavage of APP by p- and y-secretaaea. Although the production of AP in this pathway is well characterized, it is still not clear how AP becamea insoluble and accumulates in focal lesions or why amyloid is pathogenic. A less-known pathway involves the aggregation of misfolded APP molecules and amyloidogenic carboxyl-terminal fragments of APP inside the cell as a heterogeneous mixture and its ultimate conversion to the protease-resistant Apn-42 core by non-specific proteolysis. This pathway doe? not exist in normal cells, but it can be seeded by the endocytosis of AP42 from the extracellular media. It may also be capable of initiatmg spontaneously from mi\folded APP and amyloidogenic APP fragments. Once initiated. the pathway is self-propagating. In this model. the explanation for the accumulation of insoluble AD is obvious, since aggregation to form an insoluble residue is the initial step in the pathway. The model also provides a facile explanation for the focal nature of amyloid lesions. rince the pathway may initiate in single cells that establishes the nidus for amyloid deposition. The accumulation of intracellular, insoluble misfolded APP and its amyloidogemc CTFs may represent the pathophysiology of dystrophic neurites, where the insoluble residue is released from the neurite and convened to the protease-resistant AD fibril after phagocytosia by micro&ha and released back Into the extracellular space as senile plaque amyloid. The essential features of the solid-phase pathway will be presented along with new data employing conformation-dependent antibodies against AP and cell-cpecific posttranslational modification of AD that cupport key prediction\ of the pathway. An interesting aspect of the wlid-phase pathway for amyloid accumulation is that both sides of the long-standing and often acrimoniouc debate about whethel wule plaques only represent tombrtone markers and whether amyloid is central to Ad pathogenesi$ may both be correct.






Many neurodegenerative diseases are characterized by the observatmn. at autopsy, of characteri& fibrillar protein deposits in the affected region\ of the brain. These include Alzheimer’s disease (extracellular amyloid plaques in the cortex) and Parkmron’\ disease (cytoplasmic Lewy bodies in the wbutantia nigra), the wbjects of this talk, as well as the triplet repeat diseases and frontotetnporal dementia. In all of the case\, mutations or abnormalitw in the gene encoding the fibrillar protein are linked to disease. When in vitro biophysical studies have been done, the mutations \eem to promote fibrillogenesis. This convergence of pathology, genetics, and bmphysics provides strong circumstantial evidence that the process of fibrillization is crItwd. but does not address the isaue of which species along the pathway from mormal protein (AP in the case of AD. wsynuclein in the case of PD) to dlrease-asrociated fibril. if any, is pathogenic. This is a critical question for the development of diagnostic and therapeutic strategies. This talk will review in vitro biochemical studies done in our labratory which attempt to dissect the complex pathways that terminate in fibril formation by both AP and u-aynuclein. Special attention will be paid to the characterization of transient intermediates (spherical, linear, and annular protofibrila) by atomic force microscopy. The implications of our work with respect to the dwgn of ammal models of AD and PD and the discovery of therapeutx agent\ will be discwsed.