Luc Buee, Neil Cashman, Fabrizion Chiti, Blas Frangione, John Fryer, Jorge Ghiso (organiser), Charles Glabe, John Harding, Mariusz Jaskolski, David Lomas, Ronald Melki, Harry Orr, Sheena Radford, Tamas Revesz (organiser), Jean-Michel Verdier, Nicholas Wood
17 – 22 October 2003
Jorge Ghiso, Tamas Revesz
Biological systems have evolved specific procedures to ensure that, once synthesized, proteins fold correctly or, if they do not, they are rapidly degraded before to avoid any serious harm. Despite these tightly regulated physiologic controls, several human diseases (e.g. Alzheimer, Prion, Parkinson, Huntington, etc.) are associated with protein misfolding resulting in protein aggregation and cellular malfunction. Recently, attention has been focused on a group of diseases where either intact proteins or their proteolytic fragments convert from their normally soluble forms to insoluble inclusion bodies and amyloid fibrils, accumulating in several organs and leading to cell death. Despite the range of proteins involved, all of which have a unique and characteristic native fold, the aggregation and/or fibrillogenesis process found in the disease states are remarkably similar. The ability of protein sequences to fold efficiently under the guidance of molecular chaperones is a very highly developed evolutionary adaptation to suppress amyloid formation or aggregation in vivo. Compromising such protective mechanism may allow a natural sequence to revert to its altered conformation. Recent studies in yeast prions and fungi raise the question whether protein misfolding is not simply linked to disease but also involved in the mediation of normal cellular functions.
Un système de contrôle de la qualité a été mis en place dans le monde vivant pour vérifier le repliement des protéines néo-synthétisées. Ce système permet la dégradation rapide de protéines mal repliées afin d’éviter les sérieux dégâts qu’elles pourraient occasionner dans la cellule. Malgré la sensibilité et l’efficacité de ce système physiologique de contrôle de la qualité, un certain nombre de protéines mal repliées ne sont pas dégradées. Elles s’agrégent, s’accumulent dans les cellules et sont à l’origine de dysfonctionnements et de maladies chez l’homme comme les maladies d’Alzheimer, de Parkinson, de Huntington, les maladies à prions etc. Dans ce groupe de maladies, des protéines entières ou leurs produits de dégradation normalement solubles deviennent insolubles. Ils forment alors des corps d’inclusions et des fibrilles amyloïdes qui s’accumulent dans divers organes et entraînent la mort cellulaire. Si les protéines impliquées dans ces maladies diffèrent significativement les unes des autres par leur nature et leurs structures tridimensionnelles, leur mécanisme d’agrégation et/ou d’assemblage en fibres, qui sont associés aux diverses pathogenèses, sont remarquablement similaires.
Les chaperons moléculaires facilitent le repliement des protéines in vivo. Ils constituent de ce fait un outil hautement efficace, retenu et développé par l’évolution, pour empêcher où défavoriser l’agrégation des protéines et/ou leur assemblage en fibrilles amyloïdes à la suite de leur dépliement ou à leur mauvais repliement. L’inactivation des chaperons moléculaires enlève toute chance à une protéine mal repliée de revenir à sa conformation native et active. Des études récentes réalisées grâce à des prions de levures et de champignons filamenteux suggèrent que le mauvais repliement de certaines protéines pourrait être non seulement à l’origine de pathogenèses mais aussi de l’acquisition ou du maintien de fonctions biologiques dans la cellule.
The concept of Disorders of Protein Folding from the biochemical, genetic and neuropathological points of view was introduced and summarized in the first two presentations by the Colloquium co-chairs.
Jorge Ghiso from New York University (USA) reviewed Amyloidosis as a Protein Folding disease, discussing the Lessons from systemic and localized amyloid disorders.
Protein misfolding and aggregation are involved in a variety of human disorders. Among them, particular interest has been devoted to a large group of chronic and progressive neurodegenerative conditions characterized by the selective loss of neurons associated either with cognitive, motor or sensory systems. Although the sporadic cases constitute the vast majority of these disorders, in many instances genetic abnormalities can be linked to a particular clinical and/or pathological hallmark. Amyloidosis, a subset of disorders of protein folding, comprises a wide spectrum of diseases of different etiology characterized by the deposition of insoluble fibrillar proteins in different organs. From the clinical standpoint, amyloid diseases can be systemic (when several organs are affected by amyloid deposits) or localized (when amyloid lesions are restricted to a single organ or tissue) and in both instances, hereditary conditions have been identified. Amyloid fibrils are composed of self-assembled, low molecular-weight peptides usually representing fragments of larger precursor molecules normally present in body fluids. In humans, 25 different proteins are known to self-assemble and form fibrillar amyloid structures. Their precursor molecules are totally unrelated proteins exhibiting a wide variety of biological functions (i.e., immune system-related molecules, apolipoproteins, transport and regulatory components, coagulation factors, enzymes, protease inhibitors, hormones, cytoskeletal proteins, cell adhesion molecules, infectious agents and molecules of still unknown biological function). Despite these differences, all amyloid proteins share a number of biochemical and structural properties. All amyloid peptides are molecules in the mass range of 4 to 30 kDa, and many of them are heterogeneous at the amino- and/or carboxyl-terminal ends. In general, amyloids are rich in β-pleated sheet secondary structure, a conformation largely responsible for their high tendency to aggregate and polymerize and for their tinctoreal characteristics with Congo red or thioflavin S. The self-assemble process results into long, unbranched, 8-nm-wide twisted fibrils that are highly insoluble and bear poor antigenic properties, features that preclude their effective physiologic removal by macrophages. The transition of a soluble molecule to an insoluble fibril appears to be modulated by multiple factors, among them concentration, pH, post-translational modifications, amino acid substitutions, certain metal ions as well as by the presence of various unrelated amyloid associated proteins.
Tamas Revesz from the Queen Square Brain Bank and the Institute of Neurology, London (UK) discussed Neuropathology of Protein Folding Disorders.
Neurodegenerative diseases such as Alzheimer’s disease (AD), Parkinson’s disease (PD), Huntington’s disease (HD), prion diseases, frontotemporal dementias and motor neuron disease share the common feature that deposition of abnormal protein(s) gives rise to characteristic morphological changes in the central nervous system. Insoluble aggregates of disease-related proteins can deposit in intracellular inclusions and/or form extracellular protein deposits (plaques). These morphological changes can be characteristic enough for a specific neuropathological diagnosis to be made, although more often similar inclusions (e.g. neurofibrillary tangles) occur in a number of neurodegenerative conditions and are not specific for a certain disease. An example of the intraneuronal inclusions is the Lewy body, which is the hallmark lesion of idiopathic PD and dementia with Lewy bodies. In AD, which is the most common neurodegenerative disease, two different proteins form pathological lesions: Aβ, which deposits in extracellular plaques and often in blood vessel walls and the microtubule‑associated protein tau, which composes the intracellular neurofibrillary tangles. The latter are also found in a large group of neurodegenerative diseases, also known as tauopathies. In several diseases (e.g. in some tauopathies) not only neurons, but also glial cells are affected by inclusion formation. A group of neurodegenerative diseases associated with expanded polyglutamine tracts (for example HD and spinocerebellar ataxias) are also characterized by intranuclear inclusions composed of abnormal proteins.
David Lomas from the University of Cambridge (UK) discussed the issue of Serpin polymerisation: antitrypsin deficiency, thrombosis and dementia.
It is known that most Northern Europeans are homozygous for the M variant of the proteinase inhibitor α1-antitrypsin but some 4% carry the Z allele (342Glu-Lys) which in the homozygote results in a profound plasma deficiency, liver disease and early onset panlobular emphysema. The Z α1-antitrypsin accumulates in the liver of homozygotes by loop-sheet polymerization in which the reactive centre loop of one molecule is inserted into a β-pleated sheet of a second. In addition to the Z variant, only two other mutations of α1-antitrypsin are similarly associated with plasma deficiency and the formation of hepatic inclusions: Siiyama (53Ser-Phe) and Mmalton (52Phe deleted). The Siiyama variant isolated from the plasma of a homozygote was found to be in the form of long chains of loop-sheet polymers identical to those of Z α1-antitrypsin. Similarly α1-antitrypsin Mmalton forms loop-sheet polymers in vivo and in vitro showing that this is a general mechanism underlying α1-antitrypsin deficiency related liver disease. This process of polymerization also accounts for the mild plasma deficiency of the S (264Glu-Lys) and I (39Arg-Cys) variants of α1-antitrypsin and it has been described in mutants of other members of the serine proteinase inhibitor or serpin superfamily: C1-inhibitor, antithrombin and α1-antichymotrypsin in association with angio-oedema, thrombosis and emphysema, respectively. Recent studies have demonstrated that polymerization of a neurone-specific serpin, neuroserpin, underlies a novel inclusion body dementia. This dementia is called Familial Encephalopathy with Neuroserpin Inclusion Bodies (FENIB).
Harry T. Orr, from the University of Minnesota (USA) discussed Phosphorylation of Serine 776 of ataxin-1 is critical for polyglutamine-induced disease in SCA1 transgenic mice and mediates an interaction with protein 14-3-3.
Polyglutamine-induced neurodegeneration in transgenic mice carrying the spinocerebellar ataxia type 1 (SCA1) gene is modulated by subcellular distribution of ataxin-1 and by components of the protein folding/degradation machinery. Since phosphorylation is a prominent mechanism by which these processes are regulated, the phosphorylation of ataxin-1 was examined and found that phosphorylation of serine 776 (S776) of mutant ataxin-1 was critical for polyglutamine-induced pathogenesis in transgenic mice. The protein 14-3-3 was identified as an essential mediator of ataxin-1’s neurotoxicity. 14-3-3 binds to and stabilizes ataxin-1, leading to its accumulation. The ataxin-1/14-3-3 association is regulated by Akt phosphorylation at serine 776. The finding that phosphatidylinositol 3-kinase/Akt signaling and 14-3-3 cooperate to modulate ataxin-1’s neurotoxicity provides new insight into SCA1 pathogenesis and identifies potential targets for therapeutic interventions in this disease.
Nicholas W. Wood from the National Hospital for Neurology and Neurosurgery (UK) discussed the Genetic approaches to understanding Parkinson’s Disease and the implications for the ubiquitin protezone complex.
Parkinson’s Disease is a common, progressive, incurable, neurodegenerative condition affecting the basal ganglia. Its pathological characteristic is an intra-neuronal inclusion, the Lewy body, and is felt to represent the gold standard for diagnosing PD. The aetiology of PD has remained obscure until recently. The major progress in the last 5 or 6 years comes from the study of genetics and an increasing number of genes and loci have now been implicated in PD. At the present time the system most reliably identified is one of protein handling. The first gene identified was α-synuclein. This is a natively unfolded protein and when mutated is more likely to form fibrils and deposits itself within the Lewy body. It has also been shown that α-synuclein is found in Lewy bodies in patients with sporadic PD, i.e. there must be other pathways in determining this α-synucleinopathy. The second gene to be cloned, Parkin, is an ubiquitin ligase. This is an autosomal recessive condition and is numerically the most important. The mutations in the gene reduce ubiquitin ligase activity and the search is now on for the major targets for this ligase activity in the hope that we can identify further downstream events. It is noteworthy that a glycosylated form of α-synuclein has been shown to be a target for Parkin. A third gene, UCHL1, has been implicated in one very small family but the exact role of of this gene is not clear. There are two other currently identified genes, NURR and DJ1, which have also been implicated in rare forms of familial PD.
Neil R. Cashman form the Centre for Research in Neurodegenerative Diseases (Canada) discussed The Side-Chain Accessibility Hypothesis and Misfolded Prion Protein
The immune system apparently does not recognize infectious prions, although cellular and perhaps humoral mechanisms participate in prion propagation. Even the normal cellular isoform of the prion protein (PrPC) is poorly immunogenic. Several laboratories have now reported that antibodies preferentially reactive against PrPC can interfere with prion propagation in vitro and in vivo, although immune recognition of this essentially ubiquitous cell surface protein could prove deleterious. Conformational changes of proteins in disease must be accompanied by molecular surface exposure of previously sequestered amino acid side chains. Low-pH induction of beta sheet structure in recombinant PrP is associated with increased solvent accessibility of tyrosine. Antibodies directed against the prion protein repeat motif Tyr-Tyr-Arg recognize PrPSc, but not PrPC by immunoprecipitation, plate capture immunoassay, and flow cytometry. Antibody binding to the pathological epitope is saturable, specific, and can be recapitulated in vitro by low pH treatment of PrP from normal brain. Conformation-selective exposure of the Tyr-Tyr-Arg epitope provides a probe for the distribution and structure of pathologically misfolded prion protein, and may aid the search for novel diagnostics and therapeutics for prion diseases.
Ronald Melki from the CNRS-LEBS (France) discussed the Assembly of the yeast prion Ure2p into protein fibrils.
The [URE3] phenotype in the yeast Saccharomyces cerevisiae propagates by a prion mechanism. It is thought to result from conformational changes in the normally soluble, dimeric and highly helical protein Ure2 leading to its assembly into long, straight, insoluble fibrils that are similar to amyloids in that they bind Congo red and show green-yellow birefringence and have an increased resistance to proteolysis. The assembly process of Ure2p was recently dissected and showed that dissociation of the normally dimeric protein to its constituent monomers is a prerequisite for assembly into fibrils that are formed by the polymerisation of native-like helical subunits. Using specific ligand binding, FTIR spectroscopy and X-ray fiber diffraction Ure2p fibrils assembled under physiologically relevant conditions are devoid of a cross-beta core. A model for fibril formation, based on assembly of native-like monomers, driven by interactions between the N-terminal glutamine and asparagine-rich region and the C-terminal functional domain was presented and discussed. By analyzing the effect of heat-treatment on Ure2p fibrils evidences were brought for a large conformational change that occurs within the fibrils with the loss of the ligand binding capacity, decrease of the alpha helicity and the formation of a cross-beta core that characterizes amyloids. The extent of the conformational change suggests that it is not limited to the N-terminal part of Ure2p polypeptide chain. Heat-treated fibrils are unable to propagate their structural characteristic while native-like fibrils are. This indicates that heat-treated cross-beta rich fibrils are inert end products irrelevant to the prion concept in contrast to native-like fibrils. The potential evolution of native-like fibrils into amyloid fibrils was discussed.
Charles Glabe form the University of California-Irvine (USA) discussed Common Structure and Mechanism of Amyloid Pathogenesis in Degenerative Diseases.
The Alzheimer amyloid peptide Aβ forms high molecular weight soluble aggregates (micelles and protofibrils) that represent intermediates on the pathway to fibril formation. Recent evidence indicates that these soluble oligomers are common to most amyloids and may represent the primary toxic species of amyloids, like Aβ. Based on information about the organization of the Aβ within micelles, a molecular mimic of micelles was synthesized. The polyclonal antibody produced by vaccination of rabbits with the molecular mimics is remarkably specific for the soluble oligomeric intermediates and does not recognize soluble low MW species or Aβ fibrils. Surprisingly, the anti-oligomer antibody also specifically recognizes soluble oligomers from some other types of amyloidogenic proteins and peptides with no sequence homology alpha-synuclein, IAPP , poyglutamine KKQ40KK, prion 106-126, lysozyme, human insulin and transthyretin, indicating that they have a common structure. All of the oligomers listed above display in vitro neurotoxicity that is rescued by anti-oligomer antibody, suggesting that this structure mediates toxicity by a common mechanism. The effect of soluble oligomers on membrane conductance was examined using synthetic lipid bilayers and cell plasma membranes. These results indicate that oligomeric or protofibrillar oligomers from different amyloids have a common structure and suggest that they share a common mechanism of pathogenesis.
Fabrizio Chiti from the Università di Firenze (Italy) discussed Protein aggregation: an apparently complex problem governed by relatively simple rules.
A biophysical characterization of the aggregation of a protein of choice was presented with the aim of identifying principles of general validity that govern the apparently complex process of protein aggregation. Over 60 mutations of this sample protein have been produced. When aggregation is studied under conditions in which the native globular state is populated, as described previously by other investigators, mutations were found to favor aggregation to an extent proportional to the destabilization of the native state caused by mutation. When aggregation is studied under conditions in which an ensemble of partially denatured conformations is populated, mutations were found to accelerate the aggregation process when they induce either an increase of hydrophobicity and/or propensity to convert from a-helical to b-sheet structure within key regions or a reduction of the overall net charge of the protein. These experiments, conducted in vitro under controlled conditions, have been extended within cells to observe and analyze protein aggregation in vivo. This analysis basically confirms that aggregation in vivo obeys to similar rules. Finally, an equation has been edited for the prediction of the effect of any mutation on the aggregation of an unstructured polypeptide chain. The equation is able to predict with a confidence of 85-90% whether a mutation favors or inhibit aggregation, regardless of the protein sequence considered.
Jean-Michel Verdier from the University of Montpellier II (France) discussed Lithostathine : a fascinating protein that teaches us how fibrils can form.
Lithostathine is a small glycoprotein of 144 amino acids related to animal C-type lectins. It belongs to the « Reg » gene family containing 18 members. It was originally described in pancreas, but its expression is ubiquitous, including brain. The formation of stones in pancreatic juice leads to the development of a chronic calcifying pancreatitis (CCP). The first pathological sign of CCP is the formation in pancreatic ducts of precipitated fibrillary proteins in the center of the stones. The main protein component of these stones is a truncated form of lithostathine where the N-terminal undecapeptide is absent. We demonstrated that these fibrils resulted from an autolytic mechanism due to a specific cleavage of the Arg11-Ile12 peptide bond. This cleavage separated the N-terminal undecapeptide from the rest of the molecule that became insoluble and precipitated under the form of fibrils. By resolving the three-dimensional structure of lithostathine by X-ray crystallography, it was observed that it was shared into two parts: one flexible, unstructured N-terminal undecapeptide, and a globular C-terminal, rich in β-sheet, with a folding similar to C-type lectins. Lithostathine was present in both pathognomonic lesions of AD: senile plaques and neurofibrillary tangles. In addition, by monitoring lithostathine in the cerebrospinal fluid of patients with AD, we observed that it was overexpressed since the very first (pre-symptomatic) stages of the disease. Finally, it was recently observed that lithostathine is also present as plaques in the brain of patients with Creutzfeldt-Jakob or Gertsmann-Sträussler-Scheinker diseases. Interestingly, these plaques were resistant to strong proteinase K treatment. Throughout these peculiar properties, lithostathine may prove to be a very useful protein to understand the formation of fibrils, and especially to test drugs able to disassemble fibrillar bundles.
John. J. Harding form the University of Oxford (UK) discussed Cataracts, a conformational disease caused by post-translational modification of proteins
Lens proteins are particularly interesting because in central lens they are not renewed and so can accumulate slow chemical change over decades. Many chemical changes detected in human cataracts cause unfolding, yellowing, increased fluorescence, aggregation and cross-linking of proteins in vitro as seen in the ageing and cataractous human lens. Glycation, the non-enzymic reaction of sugars with proteins, has been widely studied and implicated in cataractogenesis. Glycation is increased not only in cataract, especially in diabetes, but also in Alzheimer’s Disease. The crystallins may not be the primary targets of glycation and other post-translational modifications of proteins, and recently it was studied the inactivation of enzymes by these processes as well as the protection against this inactivation mostly by alpha-crystallin, a major lens protein and molecular chaperone.
Mariuz Jaskolski from the A. Mickiewicz University (Poland) discussed the 3D Domain swapping of full-length and N-truncated human cystatin C
Human cystatin C (hCC) is the first disease-causing amyloidogenic protein for which oligomerization via 3D domain swapping has been observed. The aggregates arise in the crystallization buffer and have the form of two-fold symmetric dimers in which a long a-helix of one molecule, flanked by two adjacent b-strands, has replaced an identical domain of the other molecule, and vice versa. One of the exchanged b-chains is the very N-terminal strand, which is necessary for the protein’s potency as inhibitor of cysteine proteases. It can be speculated that a similar mechanism of 3D domain swapping operates when the molecules of hCC polymerize in an open-ended fashion characteristic of amyloid fibrils. This view is supported by the observation of hCC dimers in blood plasma of patients with severe amyloidosis caused by a variant of the protein with one amino acid exchange. The picture of hCC aggregation in vivo is complicated by the fact that the 120-residue protein undergoes slow proteolytic degradation resulting in an N-truncated variant lacking the first 11 amino acids. While the truncated protein is severely crippled in its inhibitory capacity, it has been suggested to be the dominating building material in the amyloidogenic process.
Sheena E. Radford from the University of Leeds (UK) presented Structural insights into the mechanism(s) of assembly of β2-microglobulin into amyloid-like fibrils in vitro
β2-microglobulin (β2m) is involved in the human disorder, dialysis-related amyloidosis (DRA). Native b2m is 99-residues in length and has a seven stranded β-sandwich fold typical of the immunoglobulin superfamily. The protein contains two β-sheets held together by a single disulphide bond. In vivo, b2m is present as the nonpolymorphic light chain of the class I major histocompatibility complex (MHC-I). As part of its normal catabolic cycle, b2m dissociates from the MHC-I complex and is transported in the serum to the kidneys where the majority (95 %) is degraded. Renal failure disrupts the clearance of β2m from the serum, resulting in an increase in β2m concentration therein. In patients suffering from renal failure b2m then self-associates into amyloid fibrils that typically accumulate in the musculoskeletal system. Analysis of ex-vivo material has shown that the majority of amyloid fibrils in patients with DRA consist of full-length wild-type β2m. The most recent studies on the mechanism(s) β2m fibrillogenesis have focused on the conformational transitions that occur at different pH values and the structural features of the ensembles of species that form under key fibril-forming conditions. Using a combination of approaches including X-ray crystallography, CD, AFM, NMR, mutagenesis, and mass spectrometry, together with seeding and cross-seeding experiments, it was proposed an outline for the early stages in the mechanism of β2m fibrillogenesis and identified key properties of these amyloidogenic precursors.
Luc Buée from the INSERM U422 at the Université de Lille 2 (France) discussed Tau aggregation in neurodegenerative disorders: a common feature but an unknown mechanism.
Tau proteins are microtubule-associated proteins. They are mostly found in neurons where they modulate microtubule assembly. There are six Tau isoforms in the human brain that are generated by alternative splicing. The main posttranslational modification is phosphorylation. Aggregation of microtubule-associated Tau proteins into filaments is a common feature encountered in AD and other neurodegenerative disorders referred to as tauopathies. Aggregation may be enhanced by polyanions and/or posttranslational modifications (glycosylation, oxidation, phosphorylation, etc). However, abnormal phosphorylation is the major modification of these proteins aggregated into intracellular filamentous inclusions. Since most of these conformational phosphorylation-dependent epitopes are detected in animal and cell models before Tau aggregation, abnormal Tau phosphorylation is likely to be a pre-requisite to aggregation. Because many of the phosphorylation sites are proline directed, one possible hypothesis is that the proline conformation would change upon phosphorylation and/or aggregation. Indirect evidence for a role of the proline conformation comes from the recent finding that Pin1, a prolyl cis/trans isomerase essential for the cell cycle, interacts with Tau. Pin1 is a recently characterized human peptidyl-prolyl cis/trans isomerase that modulates the assembly, folding, activity and transport of cellular proteins. Because Pin1 was found to be a key regulator of the mitotic transition, the earlier finding that Alzheimer’s disease might be related to a reactivation of the cell cycle is in agreement with a functional role for Pin1 in AD. In conclusion, molecular mechanisms leading to changes in Tau conformation include activation of kinases such as cdk5 or other regulators such as Pin1. Such molecules could be important drug targets as they are possibly involved in early stages of neurodegeneration.
Tamas Revesz from the Queen Square Brain Bank and the Institute of Neurology (UK) discussed Cerebral Amyloid Angiopathies: A Pathologic, Biochemical and Genetic View.
The most important clinicopathological manifestations of CAA include cerebral hemorrhage, ischemic lesions and dementia. One of the classifications of CAA takes into consideration the nature of the amyloid protein deposited in the vascular lesions. In the most common sporadic CAA, and in CAA related to sporadic AD deposition of Aβ is characteristic. CAA can also be severe in several variants of familial AD caused by mutations of the amyloid‑β precursor protein or presenilin-1 genes, in which Aβ variants and/or wild-type Aβ deposition takes place. Further amyloid proteins involved in familial CAAs include a.) the mutant cystatin C (ACys) in hereditary cerebral hemorrhage with amyloidosis of Icelandic type, b.) variant transthyretins (ATTR) in meningo-vascular amyloidosis, c.) mutated gelsolin (AGel) in familial amyloidosis of Finnish type, d.) disease-associated prion protein (PrPSc) in a variant of the Gerstmann-Sträussler-Scheinker syndrome, and e.) ABri and ADan in CAAs observed in the recently discovered BRI2 gene related dementias, familial British dementia and familial Danish dementia, respectively.
John Fryer from Washington University (USA) discussed The role of apolipoprotein E and J in mouse models of Alzheimer’s disease and cerebral amyloid angiopathy
In 1993, the ε4 allele of apolipoprotein E (apoE) was identified as a risk factor for sporadic AD but the specific molecular mechanisms underlying this effect are still not completely understood. ApoE has been shown to be an Aβ binding molecule, localizes to Aβ plaques, and effects the fibrillogenesis of Aβ in vitro. To further understand the effects of apoE on AD pathology in vivo, our lab has examined the PDAPP and APPsw models on a murine Apoe-/- background. Lack of murine apoE results in a reduction of Aβ plaque load and an almost complete absence of fibrillar plaques in brain parenchyma until very old ages. To further explore the effects of apoE on Aβ deposition, we crossed PDAPP and APPsw mice to transgenic mice expressing human apoE2, E3, or E4 on a murine Apoe-/- background. These mice show a marked delay in the development of plaques, suggesting a role for apoE not only in fibrillogenesis but also in the clearance of Aβ. However, once Aβ deposited, there was a clear genotype effect on Aβ deposition that follows the findings in human AD cases (E4>E3>E2). The effect of apoE on cerebral amyloid angiopathy (CAA) is even more pronounced, such that in the absence of apoE no animal tested to date has developed either diffuse or fibrillar plaques in cerebral vessels even up to 24 months of age. Additionally, one consequence of CAA, hemorrhage, was completely blocked in the absence of apoE. Furthermore, the ratio of Aβ40:Aβ42 was increased in isolated cerebral vessels from aged APPsw and PDAPP mice in relation to brain parenchyma. In the absence of apoE this ratio remained low, suggesting that perhaps a high Aβ40:Aβ42 ratio combined with apoE facilitates CAA. Another abundant CNS lipoprotein, apoJ (clusterin), has previously been shown to bind to Aβ and localize to amyloid deposits. In the absence of apoJ, PDAPP mice have a reduction in amyloid deposition, but even more striking is the reduction in neuritic dystrophy normally associated with amyloid plaques.
Blas Frangione from New York University (USA) closed the Colloquium by reviewing the major issues discussed during the meeting. The discussion concentrated in the specificity of Congo Red staining, the relationship between aggregation and cell toxicity, the influence of amyloid-associated proteins (chaperones) in the mechanism of amyloidogenesis and common therapeutic avenues in the treatment of these Conformational Disorders.