A feed-forward pathway drives LRRK2 kinase membrane recruitment and activation

Curation statements for this article:
  • Curated by eLife

    eLife logo

    Evaluation Summary:

    This paper, which is of interest to membrane biologists and colleagues in signal transduction, examines the interesting question of whether LRRK2 recruitment to membranes may regulate its activity. Membrane association involves binding to membrane-tethered Rab GTPases via LRRK2's armadillo domain, and the authors propose an elegant feedforward mechanism to describe how recruitment could lead to Rab phosphorylation, but not all features of the feed-forward model are directly supported by data.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)

This article has been Reviewed by the following groups

Read the full article See related articles

Abstract

Activating mutations in the leucine-rich repeat kinase 2 (LRRK2) cause Parkinson’s disease, and previously we showed that activated LRRK2 phosphorylates a subset of Rab GTPases (Steger et al., 2017). Moreover, Golgi-associated Rab29 can recruit LRRK2 to the surface of the Golgi and activate it there for both auto- and Rab substrate phosphorylation. Here, we define the precise Rab29 binding region of the LRRK2 Armadillo domain between residues 360–450 and show that this domain, termed ‘site #1,’ can also bind additional LRRK2 substrates, Rab8A and Rab10. Moreover, we identify a distinct, N-terminal, higher-affinity interaction interface between LRRK2 phosphorylated Rab8 and Rab10 termed ‘site #2’ that can retain LRRK2 on membranes in cells to catalyze multiple, subsequent phosphorylation events. Kinase inhibitor washout experiments demonstrate that rapid recovery of kinase activity in cells depends on the ability of LRRK2 to associate with phosphorylated Rab proteins, and phosphorylated Rab8A stimulates LRRK2 phosphorylation of Rab10 in vitro. Reconstitution of purified LRRK2 recruitment onto planar lipid bilayers decorated with Rab10 protein demonstrates cooperative association of only active LRRK2 with phospho-Rab10-containing membrane surfaces. These experiments reveal a feed-forward pathway that provides spatial control and membrane activation of LRRK2 kinase activity.

Article activity feed

  1. Author Response

    Reviewer #1 (Public Review):

    In this manuscript, Vides et al. performed a functional analysis of the Parkinson's disease-associated leucine-rich repeat kinase 2 (LRRK2). In particular, the authors sought to address how membrane recruitment of LRRK2 leads to an increase in its kinase activity. Briefly, the authors showed that LRRK2 utilizes two distinct binding sites (350-550 #1, 17/18 #2) for Rab GTPases within its N-terminal Armadillo domain to achieve membrane association. Intriguingly, these two sites differ substantially in their preference for binding phosphorylated (Rab8a, Rab10) and non-phosphorylated (Rab8a, Rab10, Rab29, Rab32, Rab39) substrates. In cells, a LRRK2 site #2 mutant showed a significantly reduced colocalization with phosphorylated Rab10. Using LRRK2 inhibitor washout experiments, the authors demonstrate that disrupting site #2 led to slower re-phosphorylation kinetics. Lastly, the authors employed an elegant in vitro system to demonstrate that LRRK2 membrane association and Rab phosphorylation are coupled in a feed-forward reaction. Overall, the work of Vides and colleagues provide compelling mechanistic insights into the spatial regulation of LRRK2.

    Nevertheless, a few critical points remain.

    Major points:

    1. Since LRRK2 is reported to form dimers and multimers, the authors should perform their colocalization studies (Figure 6) in cells lacking endogenous LRRK2.

    Co-localization with wild type LRRK2 is not seen with the mutant in question, so dimerization/oligomerization with endogenous protein appears not to be an issue for this construct.

    1. To what extent does modification of K17 and/or K18 (e.g., acetylation or ubiquitylation) play a role in regulating LRRK2 pRab binding?

    Phosphosite indicates LRRK2 ubiquitylation at K1118, K1129, K1833, K1963, K2091, with none in the ARM domain. We have not looked at either acetylation or ubiquitylation directly but now mention that this could regulate interaction with pRabs.

    1. In their lipid bilayer-based in vitro assay, the authors should also examine the effect of an LRRK2 variant that lacks site #1.

    We have included the opposite mutant with similar impact on the model: we show that lack of pRab binding site at the N-terminus removes the cooperativity of the otherwise wild type protein.

    Reviewer #2 (Public Review):

    Vides and colleagues describe a novel feed-forward mechanism of LRRK2-mediated phosphorylation of Rab8a and Rab10. The work underlies the importance of the N-terminal armadillo domain in the binding of different Rabs. They further characterized the Rab29 binding epitope, which is involved in the membrane targeting of LRRK2 mediated by Rab29 (site #1). Beyond previous work, the authors could demonstrate that one point mutation (K499E) is sufficient to abolish Rab29 binding. Furthermore, they could show that this binding site also binds the substrate Rabs Rab8a and Rab10. In addition to this binding site (#1), the authors identified one additional site (site #2) particularly involved in the specific binding of Rab8a and Rab10 but not of Rab29 nor the non-LRRK2 substrate Rab7, providing an explanation for the LRRK2 substrate specificity observed in vivo. While the Rab29 binding site bind nonphosphorylated Rabs, the newly identified site around the N-terminal Lysine 18 shows increased binding to phosphorylated Rab and provides support for a feed-forward mechanism in the substrate phosphorylation.

    The authors provide a sound biochemical characterization of critical steps of LRRK2 activation, which is of broad interest to the field. Beyond scientific interest, a well- characterized activation mechanism might guide future drug development strategies.

    We thank the reviewer for noting that we should document the bound nucleotide identity. Rab8 and Rab10 are not the easiest to work with–much harder than other Rabs to retain full nucleotide exchange capacity–preps show at best, 50% active molecules in terms of ability to exchange nucleotide. We maintain Mg-GTP throughout all purification steps and assays and use Q mutants in vitro to stabilize GTP binding. Even so, we now monitored the nucleotide state of purified Rabs by mass spec and found that our routine preps of Rab8A-Q and Rab10-Q each show a 50:50 ratio of bound GTP to GDP. We have noted this caveat in the text –our work will underestimate affinities since GTP-bound forms likely predominate in these interactions.

    Major concerns:

    • The nucleotide states of the different Rabs (after nucleotide exchange), need to be experimentally confirmed, i.e. by HPLC.
    • It is not always clear, which Rab variants (i.e. WT or Q63L) have been used for a particular experiment (information provided in the main text vs material and methods). While irrelevant for in vitro experiments, for studies in cells it should be considered that the use of Rab Q63L constructs (Q60L in Ras), does not necessarily imply that the GAP catalyzed GTP hydrolysis is completely abolished. In contrast to Ras GAPs, some RAB GAPs can provide the water-coordinating glutamine residue, critical for hydrolysis (see: Müller and Goody, 2018; PMID: 28055292).

    All studies within cells were done with endogenous Rab GTPases (WT). We have also clarified the text throughout as to which Rab form is used.

    Reviewer #3 (Public Review):

    Vide et al. present new insights into the interactions between LRRK2 and Rab GTPases. They identified two distinct Rab-binding sites in the N-terminal Armadillo (ARM) domain of LRRK2, which they named Site #1 and Site #2. One of the main findings is the striking effect of Rab GTPase phosphorylation on LRRK2's recruitment to and activation on membranes; both unmodified and phosphorylated Rabs (pRab) bind to the N-terminus of LRRK2, but to different regions. Site #1, located closer to the C-terminus of the ARM domain, binds unmodified Rab8A, Rab10, and Rab29, with Rab29 showing the highest affinity. Site #2, located at the extreme N-terminus of LRRK2, binds to the modified pRab8A and pRab10. Combining structure prediction and conservation analysis they identified the potential interaction interfaces of Site #1 and Site #2, including two conserved lysine residues (K17 and K18) in Site #2 that are critical for pRab binding. The authors propose a model where initial membrane association is mediated by binding unphosphorylated Rab8A, 10, or 29 to the lower-affinity Site #1. Membrane-associated LRRK2 then phosphorylates one of its substrates, which can now engage the higher-affinity Site #2, starting a cascade of phosphorylation events (the feed-forward mechanism).

    Overall, the authors present clear and convincing data showing the interaction between LRRK2's Nterminal ARM domain and Rab/pRab, and supporting their feed-forward mechanism. The main shortcoming in the manuscript is the absence of data directly addressing two important features of their feed-forward model: (1) The proposal that the increased activity of LRRK2 upon recruitment to membranes is only the result of its increased local concentration (without any contributions from a potential Rab-dependent activation); and (2) The ability of LRRK2 to simultaneously bind Rab and pRab. Despite this shortcoming, this manuscript presents an important contribution to our understanding of LRRK2 function, providing an elegant model for LRRK2's recruitment to and activation on membranes. This paper will be of much interest to a broad readership.

    We have fully addressed the “shortcoming”: we now demonstrate that phosphoRab10 can bind LRRK2 Armadillo domain simultaneously with Rab8 and also that pRab8 can activate kinase activity on Rab10. We thank the reviewer for these terrific suggestions.

  2. Evaluation Summary:

    This paper, which is of interest to membrane biologists and colleagues in signal transduction, examines the interesting question of whether LRRK2 recruitment to membranes may regulate its activity. Membrane association involves binding to membrane-tethered Rab GTPases via LRRK2's armadillo domain, and the authors propose an elegant feedforward mechanism to describe how recruitment could lead to Rab phosphorylation, but not all features of the feed-forward model are directly supported by data.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)

  3. Reviewer #1 (Public Review):

    In this manuscript, Vides et al. performed a functional analysis of the Parkinson's disease-associated leucine-rich repeat kinase 2 (LRRK2). In particular, the authors sought to address how membrane recruitment of LRRK2 leads to an increase in its kinase activity. Briefly, the authors showed that LRRK2 utilizes two distinct binding sites (350-550 #1, 17/18 #2) for Rab GTPases within its N-terminal Armadillo domain to achieve membrane association. Intriguingly, these two sites differ substantially in their preference for binding phosphorylated (Rab8a, Rab10) and non-phosphorylated (Rab8a, Rab10, Rab29, Rab32, Rab39) substrates. In cells, a LRRK2 site #2 mutant showed a significantly reduced colocalization with phosphorylated Rab10. Using LRRK2 inhibitor washout experiments, the authors demonstrate that disrupting site #2 led to slower re-phosphorylation kinetics. Lastly, the authors employed an elegant in vitro system to demonstrate that LRRK2 membrane association and Rab phosphorylation are coupled in a feed-forward reaction. Overall, the work of Vides and colleagues provide compelling mechanistic insights into the spatial regulation of LRRK2. Nevertheless, a few critical points remain.

    Major points:

    1. Since LRRK2 is reported to form dimers and multimers, the authors should perform their colocalization studies (Figure 6) in cells lacking endogenous LRRK2.

    2. To what extent does modification of K17 and/or K18 (e.g., acetylation or ubiquitylation) play a role in regulating LRRK2 pRab binding?

    3. In their lipid bilayer-based in vitro assay, the authors should also examine the effect of an LRRK2 variant that lacks site #1.

  4. Reviewer #2 (Public Review):

    Vides and colleagues describe a novel feed-forward mechanism of LRRK2-mediated phosphorylation of Rab8a and Rab10. The work underlies the importance of the N-terminal armadillo domain in the binding of different Rabs. They further characterized the Rab29 binding epitope, which is involved in the membrane targeting of LRRK2 mediated by Rab29 (site #1). Beyond previous work, the authors could demonstrate that one point mutation (K499E) is sufficient to abolish Rab29 binding. Furthermore, they could show that this binding site also binds the substrate Rabs Rab8a and Rab10. In addition to this binding site (#1), the authors identified one additional site (site #2) particularly involved in the specific binding of Rab8a and Rab10 but not of Rab29 nor the non-LRRK2 substrate Rab7, providing an explanation for the LRRK2 substrate specificity observed in vivo. While the Rab29 binding site bind non-phosphorylated Rabs, the newly identified site around the N-terminal Lysine 18 shows increased binding to phosphorylated Rab and provides support for a feed-forward mechanism in the substrate phosphorylation.

    The authors provide a sound biochemical characterization of critical steps of LRRK2 activation, which is of broad interest to the field. Beyond scientific interest, a well- characterized activation mechanism might guide future drug development strategies.

    Major concerns:

    - The nucleotide states of the different Rabs (after nucleotide exchange), need to be experimentally confirmed, i.e. by HPLC.

    - It is not always clear, which Rab variants (i.e. WT or Q63L) have been used for a particular experiment (information provided in the main text vs material and methods). While irrelevant for in vitro experiments, for studies in cells it should be considered that the use of Rab Q63L constructs (Q60L in Ras), does not necessarily imply that the GAP catalyzed GTP hydrolysis is completely abolished. In contrast to Ras GAPs, some RAB GAPs can provide the water-coordinating glutamine residue, critical for hydrolysis (see: Müller and Goody, 2018; PMID: 28055292).

  5. Reviewer #3 (Public Review):

    Vide et al. present new insights into the interactions between LRRK2 and Rab GTPases. They identified two distinct Rab-binding sites in the N-terminal Armadillo (ARM) domain of LRRK2, which they named Site #1 and Site #2. One of the main findings is the striking effect of Rab GTPase phosphorylation on LRRK2's recruitment to and activation on membranes; both unmodified and phosphorylated Rabs (pRab) bind to the N-terminus of LRRK2, but to different regions. Site #1, located closer to the C-terminus of the ARM domain, binds unmodified Rab8A, Rab10, and Rab29, with Rab29 showing the highest affinity. Site #2, located at the extreme N-terminus of LRRK2, binds to the modified pRab8A and pRab10. Combining structure prediction and conservation analysis they identified the potential interaction interfaces of Site #1 and Site #2, including two conserved lysine residues (K17 and K18) in Site #2 that are critical for pRab binding. The authors propose a model where initial membrane association is mediated by binding unphosphorylated Rab8A, 10, or 29 to the lower-affinity Site #1. Membrane-associated LRRK2 then phosphorylates one of its substrates, which can now engage the higher-affinity Site #2, starting a cascade of phosphorylation events (the feed-forward mechanism).

    Overall, the authors present clear and convincing data showing the interaction between LRRK2's N-terminal ARM domain and Rab/pRab, and supporting their feed-forward mechanism. The main shortcoming in the manuscript is the absence of data directly addressing two important features of their feed-forward model: (1) The proposal that the increased activity of LRRK2 upon recruitment to membranes is only the result of its increased local concentration (without any contributions from a potential Rab-dependent activation); and (2) The ability of LRRK2 to simultaneously bind Rab and pRab. Despite this shortcoming, this manuscript presents an important contribution to our understanding of LRRK2 function, providing an elegant model for LRRK2's recruitment to and activation on membranes. This paper will be of much interest to a broad readership.