Exploring the genetic contribution to idiopathic Parkinson disease

Abstract

Background: Parkinson disease (PD) is a major cause of death and disability and has a devastating global socioeconomic impact. It affects 1-2% of the population above the age of 65 and its prevalence increases as the population ages. Several biological processes have been implicated in Parkinson disease, including mitochondrial dysfunction, aberrant protein clearance, and neuroinflammation. To which degree these processes are cause, effect or bystander to disease initiation and progression, remains however largely unknown. Having limited understanding of the mechanisms underlying the pathogenesis and pathophysiology of Parkinson disease, we are unable to develop disease-modifying therapies and patients face a future of progressive disability and premature death. There is a clear hereditary component to idiopathic PD, established through both twin studies and genome-wide association studies. However, only a minor fraction of the total estimated heritability can be explained by known associated genetic variability. It has been hypothesized that the cumulative effects of rare, low-impact mutations spread across genes and biological pathways could explain some of this “missing heritability”.

Aims: The aim of this work was to explore the genetic contribution to idiopathic PD, focusing on the cumulative effects of rare mutations.

Materials and methods: The main study population utilized in all four papers was the ParkWest cohort, a Norwegian population-based cohort of incident PD. In paper I, ParkWest provided both cases and controls, including clinical longitudinal data up to and including 7 years after baseline. All ParkWest cases were whole-exome sequenced and combined with previously sequenced control samples to form the genetic cohort utilized in papers II-IV. Additionally, a whole-exome sequencing cohort from the Parkinson Progression Markers Initiative was used in papers II-IV. Finally, a publicly available chip-genotyped dataset (NeuroX) from the International Parkinson’s Disease Genomics Consortium was used as a replication cohort in paper IV. In paper I, we characterized the familial aggregation of Parkinson disease in the ParkWest cohort and explored the effect of family history on disease progression. Subsequently, we used genetic data from multiple cohorts to assess the impact of rare, protein-altering mutations in mitochondrial biological pathways (paper III) and in genes previously linked to PD (paper II and IV).

Results and conclusions: We show that, while familial aggregation is present in our Norwegian cohort, this has a slightly lower effect size compared to previous studies. Through regression analysis we also show that having a family history of PD among first degree relatives is associated with a slightly milder phenotype, which may be due to genetic variability. In paper II, we attempted to replicate the results of a recently published study reporting an association between genetic variation in the TRAP1 gene and Parkinson disease. Our analyses did not replicate this association in our Norwegian cohort. Moreover, using stricter quality control parameters abolished the association in the same dataset used in the original study. Our results do not support the proposed role of TRAP1 in idiopathic PD. In paper III, we sought to investigate the role of rare, amino acid changing variation in molecular pathways related to mitochondrial function. Using the sequence kernel association (SKAT) test, we detected a statistically significant enrichment in the pathway of mitochondrial DNA maintenance. Impaired mitochondrial DNA homeostasis has previously been shown to be present in PD neurons, and our results indicate that this dysfunction could be partly mediated by inherited genetic mutations. In paper IV, we performed a targeted single gene and gene-set association study on genes that had previously been implicated in PD through genome-wide association studies. We identified 303 genes of interest, but did not find statistically significant associations, either in the single gene or gene-set analyses. Our results do not therefore support a major role for rare variant enrichment in genes tagged by GWAS, but cannot rule out effects with small effect sizes.