Characterization of Parkinson's Disease-associated LRRK2 Kinase

Abstract

Parkinson’s disease (PD) is a progressive neurodegenerative disorder, which primarily results from the selective loss of dopaminergic neurons in the substantia nigra. Mutations in leucine-rich repeat kinase 2 (LRRK2) have been linked to autosomal familial and sporadic PD. It is now estimated that approximately 1% of Parkinson’s disease (PD) results from mutations in LRRK2. Previous studies have established that the most common mutation, which replaces G2019, located within the magnesium binding DFG motif in the kinase domain, with a serine residue enhances protein kinase activity between 2 to 4 folds. Recently our lab in collaboration with Matthias Mann’s lab discovered that LRRK2 phosphorylates a subset of Rab GTPases including Rab8A and Rab10. Moreover, it was shown that LRRK2 PD-associated mutations including G2019S, Y1669C and R1441C/G exhibit an enhanced phosphorylation of Rab10 and Rab8A in cells. Based on this data it was suggested that overactivation of LRRK2, which leads to an increased phosphorylation of indicated Rab GTPases, may lie behind the mechanism by which LRRK2 causes disease. To date, a number of compounds such as LRRK2-IN-1, GSK2578215A and MLI-2 have been described as fairly selective LRRK2 inhibitors.

It was reported that treatment of cells and mice with these inhibitors completely blocks Rab10 and Rab8A phosphorylation in cells. These results indicated that inhibition of the LRRK2 kinase activity could be a possible treatment for PD. In a previous work, our laboratory found that treatment of cells or mice with structurally diverse LRRK2 inhibitors results in rapid dephosphorylation of LRRK2 at two residues: S910 and S935. Our results suggest that these sites do not comprise LRRK2 autophosphorylation sites but are instead regulated by a distinct signalling mechanism that is controlled by LRRK2.

In the first part of my thesis I investigate the physiological function of LRRK2 S910 and S935 residues in cell using LRRK2 [S910A+S935A] knock-in mouse model. I found that in vivo, LRRK2 S910A+S935A mutation evidently reduced Rab10 and Rab8A phosphorylation in MEFs suggesting that this mutation is important for LRRK2 kinase activity. Moreover, I found that LRRK2 [S910A+S935A] mutation leads to decreased total levels of full length LRRK2 in kidneys, indicating that LRRK2 could be regulated differently in these tissues. In this chapter I also report the histopathological data for brains, kidneys and lungs that shows that the LRRK2 [S910A+S935A] mutation does not cause any major pathology in these tissues. This data indicates that the previously reported kidney and lung phenotype observed in LRRK2 KO and KD mice is not likely to be due to S910 and S935 dephosphorylation. However, behaviour phenotyping of LRRK2 [S910A+S935A] knock in mice revealed that they performed significantly worse in the rotarod test compared with their littermate wild type mice. In future, it would be interesting to perform additional behaviour tests with these mice to confirm these results.

In the second chapter, the initial aim was to address whether or not MYPT1 comprised a physiological substrate of LRRK2, and if so to investigate whether it was involved in the negative feedback regulation of LRRK2 on S910/S935. For this purpose, I used transcription activator-like effector nucleases (TALENs) as well as CRISPR/Cas9 gene editing techniques to attempt to knock out MYPT1 from HEK293 cells but I was unable to obtain homozygous knockout cell lines, which might be result of inviability of MYPT1 deficiency. In parallel experiments, in an attempt to confirm the interaction of MYPT1 and LRRK2, I immunoprecipitated endogenous LRRK2 and MYPT1 from wild type and homozygous LRRK2 knockout mouse embryonic fibroblasts (MEFs). This revealed that MYPT1 did not coimmunoprecipitate with endogenous LRRK2, casting doubt over the previously reported data suggesting these proteins interact. Finally, I used orbitrap-based phospho-site mass spectrometry to analyse changes in overexpressed and endogenous MYPT1 phosphorylation, and was unable to demonstrate that treatment of cells with LRRK2 inhibitors induced dephosphorylation of MYPT1 on residues that had been identified to be phosphorylated in vitro by LRRK2. In conclusion, results presented in this study cast doubt that MYPT1 is a direct LRRK2 substrate.

In the third part of the thesis, my aim was to widen to identify novel LRRK2 substrates or regulators. For this purpose I performed a comparative mass spectrometry analysis to analyse LRRK2 interactors that co-immunoprecipitate with endogenous LRRK2 immunoprecipitated from LRRK2 wild type, LRRK2 [S910A+S935A] knock-in and LRRK2 KO MEFs. I present the table comprising identified potential LRRK2 interactors and attempt to validate some of the proteins from the list. Unfortunately, my data showed that tested proteins including PALM protein do not seem to bind LRRK2. In this chapter, I also performed LRRK2 substrate specificity analysis, which suggested that LRRK2 could phosphorylate serines as well as threonines in vitro. Particularly, LRRK2 G2019S seems to phosphorylate serine residues much more efficiently than threonine. This data is supported by previous studies, for instance, LRRK2 G2019S was reported to have an enhanced autophosphorylation at S1292. Moreover, Rab12 is another LRRK2 substrate being validated at the moment and it was shown that it is phosphorylated by LRRK2 at S106. In future, it would be interesting to investigate how LRRK2 G2019S mutation affect phosphorylation at this site.

Date of Award	2016
Original language	English
Sponsors	Medical Research Council
Supervisor	Dario Alessi (Supervisor) & Carol MacKintosh (Supervisor)