The Role of the Alzheimer's Disease Risk Gene PICALM in Blood Brain Barrier Transport, Alzheimer's Disease and Amyloid Angiopathy

Recent genetic studies by this and other groups has revealed that PICALM is the third strongest strongest genetic risk factor (after APOE and BIN1) for late onset Alzheimer disease (AD). PICALM's cellular function is to facilitate the formation and intracellular trafficking of a specific subtype of vesicles involved in transporting material from the cell surface to intracellular compartments (termed clathrin-coated vesicles). There are several controversies in the field about how PICALM variants increase risk for AD. Over the last 3 years, we have generated extensive preliminary data which resolve several of these controversies. We have shown that genetic variants in PICALM increase risk both for AD and for pathological deposition of Abeta in blood vessel walls (Cerebral Amyloid Angiopathy - CAA). We have shown that the AD/CAA risk allele increases the production of PICALM protein in cells lining blood vessel walls in the brain (Cerebral Microvascular Endothelial Cells - CMECs). These cells are an important route for transporting proteins such as Abeta into/from the brain. CMECs express high levels of PICALM. We (and others) have shown that in most cells, PICALM has small effects on intracellular trafficking of the amyloid precursor protein (APP) and the presenilin complex, and on the subsequent cleavage of APP by the presenilin complex to generate neurotoxic Abeta peptide. Others have shown that PICALM also plays a role in the formation of autophagosomes, which is involved in in clearance of intracellular protein aggregates. However, the effects of PICALM on these processes are modest. Perturbations in these processes per se are therefore unlikely to account for PICALM-induced increased risk for AD/CAA. Intriguingly, we have recently shown that, modest overexpression of PICALM in cultured human CMECs (similar in degree to that seen in PICALM risk-allele carriers) can: i) dramatically increase the internalisation of Abeta; and ii) increase the subsequent re-externalisation of that Abeta as aggregation-prone seeds which then promote further Abeta aggregation in blood vessel walls. We hypothesize that in CMECs overexpressing PICALM, the increased Abeta uptake and increased subsequent re-secretion of Abeta as fibrillogenic seeds would: i) impair the Abeta across blood vessel walls; and ii) increase the tendency for Abeta aggregation in blood vessel walls. In the present proposal, we will test this hypothesis by employing well-characterised cellular models of the blood brain barrier. These models possess several features not used in prior studies that render our models more realistic models of Abeta transport than those used previously by other groups. We will use these models to work out how PICALM (and its interacting protein partners) affect Abeta trafficking in human CMECs. The experiments proposed here will help identify how PICALM affects amyloid clearance, and may provide targets for novel therapies directed at improving Abeta clearance. Finally, we will seek to acquire data from recent clinical trials of Abeta vaccines to determine whether PICALM genotypes can be used to predict the occurrence of amyloid-associated vascular events (e.g. microhaemorrhages) following treatments directed at Abeta. This proposal will: clarify the role of PICALM in modulating AD/CAA risk; may identify useful targets to manipulate transvascular Abeta clearance; and may reveal whether PICALM genotype predicts risk of complications of anti-Abeta therapies. It will lay the foundation for future work (to be separately funded) in which we will provide in vivo validation of these results.

We have developed exciting data suggesting that: a) in human brain, PICALM is expressed at highest levels in Cerebral Microvascular Endothelial Cells (CMECs); b) non-coding variants in PICALM are associated with increased risk for AD and Cerebral Amyloid Angiopathy (CAA); c) the AD/CAA risk allele promotes increased PICALM expression in CMECs from human brain; d) equivalent overexpression of PICALM in cultured CMECs modulates several aspects of Abeta biology, but the strongest effect is to enhance uptake and subsequent re-secretion of Abeta as amyloid seeds. To explore the underlying biology we will: 1) Deploy well-characterised models (immortalised human CMECs and polarised, multicellular trans-well culture systems) that possess several advantages over other models; 2) Use super resolution microscopy to characterize the effect of changes in PICALM expression on the quantity and direction of secretion of Abeta seeds (brainvessel/vessel-brain); 3) define the subcellular localization of PICALM in models expressing normal, reduced or increased PICALM; 4) map changes in intracellular distribution of endocytosed Abeta in models expressing normal, reduced or increased PICALM; 5) investigate how changing PICALM expression alters the protein interactome of PICALM; 6) validate which of the changes in the PICALM interactome are necessary and sufficient to alter Abeta uptake and secretion. Our proposal builds upon our expertise in the biology of the BBB (Abbott), bioinformatics (Oliver), intracellular vesicular trafficking (CIMR collaborators), advanced imaging (Kaminski), and human functional genomics (St George-Hyslop). This work will clarify the role of PICALM and may identify potential therapeutic targets to ameliorate vascular clearance of Abeta in AD.

Impact through innovation/novelty:
There are several potential beneficiaries of this research program. In the immediate timeframe of 1-5 years during the grant, the principal beneficiaries will be academic researchers working on the pathogenesis, diagnosis and treatment of AD. By understanding the role of PICALM, they will be able to: 1) better model this disease; 2) design improved diagnostics (including potential diagnostics for complications for certain types of therapies); and 3) improve therapeutics based on this knowledge. A second group of academics will be those working on blood brain barrier. If our preliminary data indicating that genetic variants in PICALM predict the presence of cerebral amyloid angiopathy is confirmed in clinical trial data, this knowledge will facilitate the design of therapeutic trials with anti-Abeta therapies (all of which are thought to increase the risk of amyloid related microvascular complications). This could lead to cheaper and improved clinical trials directed at the most appropriate genotypically defined subgroup. If these new therapies were successful, they would then bring potential benefits to the well-defined group with low risk of complications. The methods developed by the applicants represent some of the first validated assays permitting measurements of the aggregation of neurotoxic proteins at the blood brain barrier. The Kaminski group in particular has made significant advances in the development of advanced multi-parametric imaging techniques for the study of protein aggregation. With dSTORM we have shown that we obtain structural information of oligomers at a resolution down to 10 nm, in situ, in the environment where amyloids naturally occur..

Impact through collaboration:
Staff working on the project will develop skills in generating comprehensive cellular models of blood brain barrier, a skill which is widely transferable to many aspects of CNS biology including CNS pharmacology. They will also gain skills in cell biology, molecular biology and biochemistry of amyloid and amyloid angiopathy. The applicants have participated in the Neurodegenerative Disease Initiative funded by the Wellcome Trust and the MRC and in the BBB group at Kings. These collaborative work groups will increase the impact of our research. Indeed, by engaging in regular interest group meetings, we have the opportunity to communicate our findings to a multi-disciplinary group of researchers focussing on neurodegenerative diseases.

Impact for Cambridge University:
The expertise acquired through the proposed research project, of these techniques and the optimisation of protocols for sample preparation, will benefit Cambridge University. Indeed, a large number of researchers from the University may take advantage of the presence of platforms developed here in imaging and in blood brain barrier science. All technologies by this group will be made available to all researchers in the University.

Impact for Society:
By understanding the role of PICALM, they will be able to: 1) better model this disease; 2) design improved diagnostics (including potential diagnostics for complications for certain types of therapies); and 3) improve therapeutics based on this knowledge. Over 820,000 people were affected by neurodegenerative diseases in the UK in 2010. This number is forecast to increase rapidly. The annual cost of these age related diseases was over £23 billion in 2009.