Conformation and Dynamics of the Kinase Domain Drive Subcellular Location and Activation of LRRK2.

Abstract Background: Leucine-Rich Repeat Kinase 2 (LRRK2) is a complex multi-domain protein where LRRK2-mutations are associated with Parkinson´s Disease (PD). To explore how pathogenic PD-mutations hijack the finely tuned activation process of LRRK2, we here used a multi-tiered approach. Methods: First, the spatial and temporal distribution of full-length LRRK2 was investigated by a real-time cell-based assay in the presence and absence of LRRK2-kinase inhibitors. In a 2nd layer we explored the consequences of PD mutations as well as removal of the N-terminal domains employing a construct containing the ROC, Cor, Kinase and WD40 domains (LRRK2RCKW). We focused on the biochemical characterization of LRRK2RCKW variants based on kinase assays using Rab8a or LRRKtide as substrates. Next, we used hydrogen-deuterium exchange mass spectrometry (HDX-MS) to map the solvent accessible regions of LRRK2RCKW in the presence and absence of the LRRK2 inhibitor MLi-2. Finally, Molecular Dynamics simulations on the kinase domain were applied to elucidate differences in breathing dynamics between wild type and mutants of the DYGψ motif. Results: Our cellular approaches revealed that the kinase inhibitors MLi-2 and rebastinib both freeze the kinase domain in a stable conformation, however, only MLi-2 resembles an active conformation and induces filament formation. LRRK2RCKW showed, regardless of the mutation it was combined with, filament formation, indicating a shielding function of the N-terminal domains. This shielding function is impaired for pathogenic mutations in full length LRRK2. LRRK2RCKW retained kinase activity similar to full-length LRRK2. HDX-MS provided a comprehensive allosteric portrait of the kinase domain and revealed how MLi-2 binding is sensed by the entire protein. Molecular Dynamics simulations suggest that, while all domains contribute to regulating kinase activity and spatial distribution, it is the highly dynamic kinase domain, driven by the DYGψ motif, that coordinates the overall domain crosstalk and serves as a regulatory hub for the intrinsic regulation of LRRK2.Conclusion: These studies confirm our hypothesis that the N-terminal scaffolding domains shield the catalytic domains in an inactive state. PD mutations, MLi-2, or Rab GTPases can all unleash the catalytic domains while the active kinase conformation, but not kinase activity, is essential for docking onto microtubules.