Genetic Epidemiology and Cognitive Endophenotyping in Amyotrophic Lateral Sclerosis

Abstract

Amyotrophic Lateral Sclerosis (ALS) is a progressive debilitating and ultimately fatal neurodegenerative disorder affecting both upper (UMN) and lower (LMN) motor neurons. While it has been long known to exist on a spectrum of motor and cognitive dysfunction, recent genetic and epidemiological studies have also found a significant overlap between ALS and other neuropsychiatric disorders. While the exploration of this spectrum of liability is still in its infancy, delineating the nature of these phenotype-genotype correlations may provide us with a novel means for categorising these disorders as profiles of multi-endophenotypic traits and their associated genetic determinants. This is hoped will provide insight into the pathogenesis of both ALS and associated disorders, ultimately assisting in the identification of novel targets for drug development. In this project, a “clan genomics” model was applied to deconstruct this shared genetic liability into population level and family specific genetic risk. The influence of population-specific genetic signatures on ALS risk and phenotype was explored both within the genetically homogenous Irish population and across Irish and Latin-American clinic-based cohorts, adding to our knowledge of ALS epidemiology in these understudied regions. In the largest such study conducted to date, ALS heritability estimates showed that genetic and environmental factors contribute approximately equally to variation in ALS risk. For the first time, it was shown that genetic factors play a greater role in ALS risk among women. A significantly younger age of onset was observed among Cuban ALS patients, not accounted for by known ALS genes or demographic factors. This suggests that larger scale genome-phenotype correlation consortium efforts may be highly informative in identifying rare large-effect variants in Cuban ALS kindreds and in exploring epigenetic interactions in these regions. How ALS risk manifests within kindreds was explored in three studies. Analysis of Irish ALS register data showed that, in Ireland, at least 20% of ALS is familial and could be much higher if extended phenotypes associated with ALS were also considered. However, while the importance of extended phenotypes within ALS kindreds is increasingly recognised, there remains a lack of clarity as to how best to define significant familial clustering of these disorders. The largest ALS family aggregation study performed to date confirmed that relatives of those with ALS are at risk of developing many neuropsychiatric disorders, particularly FTD and schizophrenia. These data were used in conjunction with novel probability modelling approaches to propose new criteria for redefining familial ALS to encompass significant clustering of FTD or schizophrenia. Similar to more traditionally defined familial ALS, probands from these ALS and neuropsychiatric familial clusters also developed ALS at a younger age, evidencing the clinical relevance of the high shared genetic risk burden within these families. That this increased genetic risk was not entirely attributable to the pathogenic C9orf72 repeat expansion, suggests that other rare, pleiotropic gene variants are yet to be discovered in the Irish population. This is further supported by the findings of extensive cognitive and neuropsychiatric changes among relatives of people with ALS. Both asymptomatic carriers of the pathogenic C9orf72 repeat expansion and their non-carrier kindred, showed changes mirroring patterns seen in ALS patients. In particular, clustering of both verbal fluency deficits and initiation apathy in C9orf72 negative relatives from C9orf72 kindreds infers the existence of selective neural networks vulnerabilities in this cohort. While this may reflect additional unidentified genetic risk in this cohorts, aligning with the oligogenic model of ALS, how epigenetic factors contribute to this clustering will need to be explored in future studies. Nonetheless, this current work raises concerns as to the common practice of using asymptomatic non-carriers as the only controls in studies of pre-symptomatic mutation carriers. Finally, much of the work of this project has direct applications for ALS genetic counselling. Furthermore, this work has contributed significantly to our understanding of ALS genetic epidemiology and provides strong support in favour of the hypothesis that ALS is a disorder of aberrant neural networks. It has become increasingly clear that a simple neurodevelopmental or neurodegenerative model does not best fit ALS. Instead, both genetic and environmental risk can be considered under a ‘Development Risk Factor Model of ALS’, where genetic or time sensitive environmental exposures result in the development of vulnerable neural networks, whose phenotypic manifestations may interact with environmental factors over a lifetime, ultimately resulting in decompensation and the development of overt symptomology. As such, high risk family units may play an essential role in future research teasing out the risk architecture of ALS, by allowing for the exploration of how epigenetic interactions with kindreds may alter phenotypic disease manifestations. Developing biomarker profiles indicating dysregulation in specific neural networks will facilitate the implementation of timely and personalised interventions which will hopefully lead to substantial improvements in outcomes.