Andrew Singleton, PhD - US grants

We continue to investigate families without known mutations including rare families with parkinsonism, dystonia and ataxia, by genome wide SNP mapping and most recently by exome sequencing. In the last year we have identified and published mutations that cause Brown Vialetto van Laere (BVVL) syndrome and expanded this work to find an additional cause of an atypical form of BVVL (in preparation). In addition we have identified a novel genetic lesion that causes a rare form of spinocerebellar ataxia (nominated as SCA32) and rapid identification of a known SCA mutation causing SCA14. We have in addition, identified the cause of recessive forms of ataxia in 8 in bred families, and this work has revealed 2 novel genetic loci and mutations for this disease (in preparation). Our current efforts have also lead to the identification of significant phenotypic heterogeneity associated with known disease causing mutations;identifying, for example, NOTCH3 mutations as a cause of Alzheimer's disease. In Parkinson's disease our efforts have focussed on exome sequencing across families with an apparent recessive form of disease;this analysis is currently ongoing.

During the past year we have published several studies identifying the extent and nature of LRRK2, PRKN and PINK mutations in Parkinson's disease. This work aims to further our understanding of the phenotypic spectrum associated with mutation in these genes. We have extended our work in the spinocerebellar ataxias by screening of TTBK2 and ITPR1 in ataxia probands (mutations in these genes were identified earlier by us to be causes of ataxia). Notably we also performed a large candidate association analysis within multiple system atrophy (MSA). This work showed for the first time that common variability in the gene SNCA is an unequivocal risk factor for MSA with an odds ratio of approximately 8. This provides critical etiologic insight into this disorder, for which no genetic risk factor had previously been identified. Our work has also included testing of the idea that OMI and GIGYF2 mutations may lead to Parkinson's disease, and analysis of DYT16, ATP13A2, PLA2G6, FBXO7 and spatacsin mutations in various young onset movement disorders.

The genetic contribution to a number of neurological disorders is thought to be complex in nature, disease risk being driven by a combination of risk alleles commonly present in the human genome. Recently the completion of stages I and II of the international Haplotype Map project and the availability of high-plex SNP assays has made genome wide assay of common genetic variability a realistic endeavor. We are applying genome wide association analysis using 500,000SNPs to a Parkinsons disease cohort from the NINDS funded neurogenetics repository. We have developed and implemented the necessary hardware and software infrastructure to store and manipulate the 2 billion datapoints associated with these experiments. The data the experiments analysing Parkinsons disease are complete and have recently been published; we are now nearing completion of the second stage of this work, analysing 1000 Parkinsons disease cases and 1000 neurologically normal controls, these data will be posted at dbGAP for mining an augmentation by interested researchers.

We continue to be a contributor to the reanalysis of large tracts of genome wide association data and much of this work, which identifies several new loci, and a converging molecular pathway, has been published. Most recently we continue our efforts using second generation sequencing (which previously lead to the identification of TREM2 and TREML2, and PLD3 variants as a risk factor for Alzheimer's disease). We have followed up on that work with continued characterization of known AD loci in relatively small series of patients. We are contributing members to consortia that continue to investigate various aspects of AD genetics. We are analyzing longitudinal high dimension data from AD patients using a series of machine learning and artificial intelligence methods

During this year we have found several families with progranulin mutations and now have papers published and in press on this. This work has shown that progranulin mutations produce a fairly typical phenotype and are not associated with overlapping disorders such as amyotrophic lateral sclerosis. We have used dense SNP based linkage mapping to identify novel loci associated with familial frontotemporal dementia. This work shows linkage overlapping an extant locus on chromosome 9 and we are now involved in a positional cloning project to determine the underlying genetic mutation in this family; this project involves sequencing of approximately 80 genes within the critical interval and is currently 50% complete.[unreadable] Our ongoing work on dementia with Lewy bodies and Parkinson's disease with dementia primarily focuses on analysis of loci identified in our genome wide association studies in Alzheimer's disease and Parkinson's disease to see if we can tease out the genetic basis of dementia in these disorders which neuropathologically sit between Alzheimer's and Parkinson's disease. This work complements our analysis of quantitative neuropathology in a series of brains from a longitudinal aging study from Finland; in particular the analysis of these neuropathological measures in terms of genetic risk - we have performed quantitative trait locus mapping in approximately 250 such brains and identified several loci positively associated with amyloid, tau and synuclein pathologies.

In this year, we have completed our labwork for stage I of a whole genome association analysis of stroke. The initial phase of this work is now published and we are embarking on a second phase, using additional samples from the Coriell repository. The initial work suggests that there are no major, common genetic loci that predisopse to ischemic stroke; however they suggest that common variability on chromosome 9 associated with heart disease and diabetes may also be a minor risk factor for ischemic stroke.

Using more than 500 samples from 31 distinct worldwide human populations we performed very dense genome wide single nucleotide polymorphism (SNP) genotyping at 550,000 loci. We analyzed these data and the distribution of genotypes, haplotypes and copy number variants across populations. We showed that these data were able to assign individuals to populations and that the resulting predictions supported fine-scale inferences about population structure. Increasing linkage disequilibrium was observed with increasing geographic distance from Africa, as expected under a serial founder effect for the out-of-Africa spread of human populations. We have extended upon this work to use the data from these populations to determine whether imputation of unknown genotypes is feasible and the best approach to this prediction. To understand the effects of genetic variability on DNA methylation we have complete genome wide genotyping and epigenome wide DNA methylation typing in 1000 brain samples. These data show a striking effect of genetic variation on DNA methylation levels and show clearly that such variation is likely to be physically close the the DNA methylation site under influence. Further we show that these effects are generally consistent across tissues, although there are some notable exceptions to this noted as tissue specific methylation Quantitative Trait Loci. Further, analysis of these data have revealed age related DNA methylation changes that occur across various tissues, including brain.

In the last year, we have completed three projects designed to elucidate the genetic pathogenesis of ALS: In the first project, we sought to identify causative variants for ALS by conducting a genome-wide association study in a cohort of 432 Irish individuals. The relatively homogeneous genetic background of this island population enhances its power to detect relevant loci. Following replication in our previously published North American dataset, the strongest association was a variant in the gene encoding DPP6, a component of type A neuronal transmembrane potassium channels.[unreadable] In the second project, we screened the TARDBP gene for mutations in a cohort of 279 sporadic ALS cases and 806 neurologically normal control individuals. Mutations in the TARDBP gene have been recently described in families with ALS, but the importance of the gene in the pathogenesis of the commoner sporadic form of the disease was unknown. No mutations or pathogenic structural variants were found in the ALS cases suggesting that this particular gene is not a common cause of sporadic motor neuron degeneration. [unreadable] In the third project, we undertook a two-stage genome-wide association study to identify the genes involved in ALS: we followed our initial genome-wide association study of 545,066 SNPs in 553 individuals with ALS and 2,338 controls of European descent by testing the 7,600 most associated SNPs from the first stage in three independent cohorts consisting of 2,160 cases and 3,008 controls. Two adjacent SNPs on chromosome 7p13.3, rs2708909 and rs2708851, were significantly associated with disease in the combined joint analysis (P-value = 5.47x10-7 and 7.22x10-7). The most associated SNP, rs2708909, is located in intron 3 of SUNC1, which encodes a nuclear envelope protein known to interact with microtubules and dynein. Our findings suggest that SUNC1 variants contribute to susceptibility to sporadic ALS. [unreadable] In summary, the current year has been successful in identifying genetic variants important in the pathogenesis of ALS using both candidate gene approaches and genome-wide association methods. Each of the three studies employed large cohorts of research subjects, and utilized the Illumina genotyping platform and the sequencing facilities available within the Laboratory of Neurogenetics, NIA.

The genetic contribution to a number of neurological disorders is thought to be complex in nature, disease risk being driven by a combination of risk alleles commonly present in the human genome. Recently the completion of stages I and II of the international Haplotype Map project and the availability of high-plex SNP assays has made genome wide assay of common genetic variability a realistic endeavor. We have applied genome wide association analysis using 500,000SNPs to a Parkinsons disease cohort from the NIH funded neurogenetics repository. We have combined the data from this work with collaborators from Germany, the US, the UK and Japan. This work has lead to the identification of common variability in SNCA and MAPT as unequivocal risk factors for Parkinson's disease. Further this work has identified a novel locus on chromosome 1 as a risk factor for PD. We have now extended this work and combined results with other large genotyping projects in PD and related diseases to identify additional risk loci for this disease, identifying an additional 7-10 risk loci for Parkinson's disease. We plan to follow this with replication genotyping and resequencing to nominate common and rare risk variants in these diseases.

Our ongoing work on dementia with Lewy bodies and Parkinson's disease with dementia primarily focuses on analysis of loci identified in our genome wide association studies in Alzheimer's disease and Parkinson's disease to see if we can tease out the genetic basis of dementia in these disorders which neuropathologically sit between Alzheimer's and Parkinson's disease. We have performed and published a genome wide association study in frontotemporal dementia, a focused association study in dementia with Lewy bodies, an analysis of the heritability of dementia with Lewy Bodies, and an assessment of known loci in DLB. We have ongoing a large whole exome sequencing project in both FTD and DLB and have extended this work with a new Investigator, Dr. Scholz, to include whole genome sequencing. This work has now expanded to include whole genome sequencing in 3000 DLB, 3000 FTD cases, the data for which is being made publicly available.