Using genetic variability in whole transcriptome expression in cells and tissues to understand the pathogenesis of Parkinson's and Alzheimer's disease

Parkinson's and Alzheimer's disease (PD & AD) are common diseases that affect the brain, causing certain cells to die. When these cells are lost people experience, amongst other problems, difficulty remembering recent events in the case of AD and difficulties moving in the case of PD. At the moment these diseases affect over 1 million people in the United Kingdom with a devastating impact on patients and their families. Since both these diseases become more common with age, the number of sufferers is likely to increase as the number of elderly people rises in the UK. Unfortunately, at the moment we know relatively little about why certain people suffer from these diseases and until we do it is hard for researchers and doctors to develop effective treatments.
In the past 3 years scientists have been able to show that some individuals have common changes in their DNA (the instruction manual for all the building blocks needed to make the human body) that put them at higher risk of developing these conditions. Although this has been an amazing step forward, it has not been easy for scientists to understand how these changes in the DNA, termed genetic risk factors, act to help cause disease. Many people believe that one way in which these genetic risk factors might cause disease is by changing how much or the way in which particular genes (the basic building blocks) are expressed in brain cells. In order to check this, researchers need to make measurements about the genetic variation an individual carries and link this information to the genes they express in their cells.
Although this type of investigation is particularly challenging for diseases that affect the brain, it is possible and meeting those challenges is the aim of this project. One of the difficulties is accurately measuring the amount of the gene expressed in the right cells in the brain. The brain is extremely complex and even within a single region there are many different types of cells. We will address this problem by making use of a technique called laser capture microdissection to obtain pure collections of cells from brain regions relevant to PD and AD. In this technique, using very thin slices of the human brain tissue, we shoot out just the cells we are interested in collecting with a high power laser and then extract the RNA (the gene copies) from the pellets we shoot out. Another advance is the advent of new technology for measuring all the genes expressed in a single cell type or tissue sample. This technology is called "RNA-seq" or "whole transcriptome sequencing" and it allows us to measure all kinds of gene product from any gene. It also allows us to measure the relative quantities of alternative versions of the same gene, and to measure these quantities in such a way that the influence of genetic risk factors can be more sensitively detected by directly comparing the relative product quantities in individuals who happen to have both a "good" and a "bad" copy of a given genetic variant.
We believe that we are in an ideal position to complete this project and so add important insights to the understanding of PD and AD. We already have many of the RNA samples needed to perform this study and while some new RNA samples will be created during the project, through our close collaboration with the MRC Sudden Death Brain and Tissue Bank in Edinburgh we have the brain samples needed to do this (all of which have been donated for research). We have also already carried out extensive analysis of DNA from these samples, which means that the new information we generate during this grant can be quickly and cost-effectively added to existing sources of data in order to progress this research. As a final benefit, all the data we will generate on the human brain will be made publically available so that other scientists interested in understanding how the human brain works in health and disease can use this information.

The aim of this study is to determine the effect of genetic risk factors on whole transcriptome expression in brain regions and cell types most vulnerable to Parkinson's and Alzheimer's disease (PD and AD) in order to understand the pathogenic processes underlying these conditions. This project is inspired by the recent discovery of 15 new genetic risk variants for PD and 13 new loci for AD. However, knowing the genomic position of risk variants is not equivalent to knowing how they act in terms of the underlying molecular processes or where they act in terms of the regional or cellular location. We aim to address both these issues. Using control post-mortem human brain tissue originating from 150 individuals, we will use expression quantitative trait locus (eQTL) mapping and allele-specific expression (ASE) analysis to identify loci that both change disease risk and regulate the expression of specific gene transcripts within specific tissues/cells. We will conduct these studies within the tissues and cells most relevant to PD and AD, namely the substantia nigra, hippocampus, temporal cortex, dopaminergic neurons, CA1 neurons, pyramidal neurons and astrocytes. Existing exome sequencing and SNP array data will be combined with whole transcriptome RNA sequencing data generated on the Illumina HiSeq2000 platform for downstream analysis. These two types of information - genetic variation and transcript expression - will be used for eQTL mapping in tissues and ASE analysis in cells. In both cases the eQTLs identified will be annotated for disease-relevance in the first instance. All the data generated by this project will be publicly released as both raw data files for re-analysis and processed information suitable for non-expert users through NCBI's Gene Expression Omnibus repository and Phenotype-Genotype Integrator sites. Thus, we will generate novel disease-relevant findings and provide the neuroscience community with a world-class resource.

The goal of this work is to develop comprehensive databases which allow the interpretation of the genome wide association studies (GWAS) for the two most common neurodegenerative diseases, Parkinson's and Alzheimer's disease (PD and AD). In this way, we will improve our understanding of the pathogenic processes underlying these diseases and in the long term improve the diagnosis and treatment of patients struggling to cope with these debilitating diseases. Reaching this goal will require many intermediate steps and we describe who we envisage using our research findings and how this data will help them.

Researchers interested in the follow up of GWAS data: This study was designed to improve our understanding of the molecular mode of action and cellular location in which genetic risk variants for PD and AD act. Therefore, the most obvious users of our data are the original AD and PD consortia. However, the results of this study will also be useful for the interpretation of GWAS for other neurological and psychiatric disorders, making it relevant to researchers interested in the results of GWAS in frontotemporal dementia, epilepsy, depression and schizophrenia to name a few. Since we are publicly releasing all findings through the NCBI's searchable database, Phenotype-Genotype Integrator, we envisage other less specialist researchers also accessing and using the data. These individuals include clinicians, epidemiologists and drug companies interested in following up GWAS hits for clinical trial design or biomarker development.

Researchers interested in the creation and use of in vitro model systems: The cell-specific whole transcriptome sequencing data produced in this project will be a baseline for the detailed molecular assessment of stem cell-based disease models. We will work collaboratively with researchers in the University of Edinburgh (Dr. Kunath & Dr. Chandran) and Cambridge University (Dr. Livesey) to identify the strengths and weaknesses of existing cell models and work to improve the accuracy of these systems. Thus, we will contribute to the development of new research tools for use by the neuroscience research community within universities and in the private setting.

Basic neurobiologists: The large gene expression data sets we are producing will provide insights into other aspects of basic neurobiology. For example, it will be of value to those researchers interested in analysing sex differences in transcript expression. Using this data it will be possible to improve our understanding of the basis of well recognised differences in the incidence, presentation and response to treatment of many neurological disorders. This could in turn contribute to the drive towards individualised patient assessment and care.

Bioinformaticians & statistical geneticists: The raw genotyping, exome and RNA sequencing data will be used by statistical genetics and bioinformatics wishing to test and develop new statistical and analytical methods. In particular, the existence of paired exome sequencing and SNP array data will allow researchers to objectively test and improve methods for the imputation of genetic data. Therefore, we will directly contribute to one of the MRC's stated aims, the training of more researchers and clinicians in bioinformatics.

Charities & patient groups: The discovery of genetic risk variants for PD and AD represent one of the most significant steps forward in the understanding of these diseases in the last 10 years. Therefore, they are hugely important to patients and carers who desperately want to see progress in the diagnosis and treatment of their diseases. From our own experience of hosting visits for Parkinson's Disease UK, Alzheimer's Research UK and the Michael J Fox Foundation, we recognise the importance of engaging with patients and communicating the findings generated from our research in order to maintain hope, create novel resources and improve the uptake of clinical trials.