Parkinson's disease and Gaucher disease : focus on the GBA1 link

Abstract

Parkinson’s disease (PD) is the 2nd most common neurodegenerative disorder in the world, characterized by both motor as well as many non-motor features. There are many symptomatic treatments available, that may shift significantly during the course of the disease. The biggest medical need in PD management however is treatments that actively targets disease pathology, thus being able to slow or stop the disease progression. When such therapies become available, another obstacle to administer them as early as possible in the disease progress is the fact that PD diagnosis is based on motor symptoms, that does not appear until late in the neurodegenerative process. Thus, the notion of prodromal PD has gained increased research attention in the search for earlier signs and biomarkers specific for PD.

The most important genetic risk marker for PD development is mutations in the GBA1 (Glucocerebrosidase 1) gene. The gene codes for a lysosomal enzyme called Glucocerebrosidase, that normally degrades Glucocerebroside - a glucolipid mainly found in cellular membranes. When the GBA1 gene is altered, the enzyme becomes defective, and Glucocerebroside is pathologically accumulated within cells. In its biallelic state, GBA1 mutations gives rise to Gaucher disease (GD), a systemic disease mainly characterized by visceral, hematological and skeletal manifestations, but in rare cases also shows severe neurological signs and symptoms. There is a particular high endemic incidence of neuronopathic GD in the northern parts of Sweden with a distinct phenotype called “the Norrbottnian GD variant”. The GBA1-PD connection has therapeutic implications, since disease modifying treatments targeting the lysosomal pathway may be a way to find diseasemodifying treatments for PD. Since GBA1 mutations, and subsequent lysosomal dysfunction, most likely are linked to PD pathology, an overall aim of this thesis is to investigate different aspects of this gene in relation to the Swedish PD population.

In Study I and II we screened the most common variations in the GBA1 gene in the largest Swedish PD cohort reported. Comparing 1625 PD patients and 2025 healthy controls from 5 different sites throughout Sweden, we found the most significant association with highest effect size with the severe mutation L444P. There was also a markedly skewed geographical distribution of this mutation, with 58% of the mutation carriers coming from northern Sweden, consistent with the high GD prevalence in Norrbotten. The other GBA1 variants N370S and E326K also showed association to PD risk, but not as markedly as the L444P mutation. The relevance for the GBA1-variant T369M on PD risk has been debated, and thus we also analyzed this in our cohort. However, despite a slight overrepresentation of T369Mcarriers in the PD cohort, this difference did not prove significant in our population. In conclusion, alterations in the GBA1 gene results in a significantly greater risk of developing PD in Sweden. However, there are marked differences between which specific genetic alteration is investigated – the mutation matters. There is also a marked geographical accumulation of the severe GBA1 mutation L444P in northern Sweden.

Study III assessed polyneuropathy (PNP) in a small number of PD patients and healthy controls. The increased frequency of polyneuropathy we saw in Parkinson’s disease patients can be both a clinical feature of the disease, but also aggravated by the L-dopa treatment and subsequent disturbances in the B-vitamin metabolism, especially Cobalamin. Genetic analyses also revealed an increased PNP risk for PD patients with mutations in the COMT gene, coding for an enzyme implicated in Cobalamin metabolism. However, other investigated mutations in another gene related to B-vitamin metabolism (MTHFR), or to GBA1, did not correlate to PNP risk in PD patients.

Study IV followed 12 Norrbottnian GD patients for 10 years, reporting on clinical outcomes. Neurological GD affection was measured with the modified Severity Scoring Tool (mSST), showing slow progressive deterioration over time. The neurological symptoms that mostly contributed to the decline was Ataxia of gait and Cerebellar tremor. Almost all patients had horizontal gaze palsy and 50% had epilepsy at end-of-follow-up. One patient deteriorated rapidly both cognitively and functionally, being diagnosed with Dementia with Lewy bodies. However, for the rest cognition remained mostly stable, although at a slightly impaired level. The main conclusion is that there is an overall slow worsening of neurological signs and symptoms among the Norrbottnian GD patients over time, but many live long and the clinical heterogeneity, despite almost identical genotype, is large.

Also, plasma Neurofilament Light (NfL) was assessed in selected GD patients and compared to healthy controls, investigating its possible function as a biomarker for neurological affection in GD. There were no significant differences in NfL levels between GD patients and controls, but a trend for association between NfL levels and neurological affection. This suggests that following NfL over time in larger neurologically affected GD cohorts might reveal interesting associations. However, the clear lack of correlation also raises a warning about being too liberal in using the unspecific marker NfL as a proxy for neurological affection in neuronopathic GD.