Rab29-dependent asymmetrical activation of leucine-rich repeat kinase 2

Abstract

Gain-of-function mutations in LRRK2, which encodes the leucine-rich repeat kinase 2 (LRRK2), are the most common genetic cause of late-onset Parkinson's disease. LRRK2 is recruited to membrane organelles and activated by Rab29, a Rab guanosine triphosphatase encoded in the PARK16 locus. We present cryo-electron microscopy structures of Rab29-LRRK2 complexes in three oligomeric states, providing key snapshots during LRRK2 recruitment and activation. Rab29 induces an unexpected tetrameric assembly of LRRK2, formed by two kinase-active central protomers and two kinase-inactive peripheral protomers. The central protomers resemble the active-like state trapped by the type I kinase inhibitor DNL201, a compound that underwent a phase 1 clinical trial. Our work reveals the structural mechanism of LRRK2 spatial regulation and provides insights into LRRK2 inhibitor design for Parkinson's disease treatment.

Fig. 1.. Structural determination of the Rab29–LRRK2 complex.

(A) Schematic diagram showing Rab29-mediated LRRK2 membrane recruitment and activation. (B) Cryo-EM structures of the Rab29–LRRK2 complex in three oligomerization states. (C) Top view of the cryo-EM structure of the Rab29–LRRK2 tetramer.

Fig. 2.. Molecular basis of Rab29-dependent LRRK2 recruitment.

(A) Rab29–LRRK2 interface in the LRRK2 monomer state. LRRK2 and Rab29 are colored gray and hot pink, respectively. Side chains of interface residues are shown as sticks. (B and C) Impact of Rab29 or LRRK2 mutations on LRRK2 localization in HEK293 cells. Quantification of a portion of LRRK2 overlapping with Rab29 according to Mander’s coefficient for confocal analysis is shown in fig. S3, A and B. Each empty circle represents colocalization coefficient (Mander’s coefficient) measured in one cell. Error bars represent SEM. Significance was determined by the Kruskal-Wallis one-way analysis of variance (ANOVA) test. **** P < 0.0001; ns (not significant). (D and E) Quantification of the immunoblotting data shown in fig. S3, C and D. Data are presented as ratios of pRab10-Thr⁷³/total Rab10, pLRRK2-Ser¹²⁹²/total LRRK2, and pRab29-Thr⁷¹/total Rab29, normalized to the average of LRRK2 WT values. The data shown are the mean ± SD of three determinations. Single-letter abbreviations for the amino acid residues are as follows: A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; and Y, Tyr.

Fig. 3.. Structure of the Rab29–LRRK2 tetramer.

(A) Cryo-EM structure of the Rab29–LRRK2 tetramer with two different views. Peripheral Rab29 (Rab29^peri) and LRRK2 (LRRK2^peri) are colored in hot pink and gray, respectively; central LRRK2 (LRRK2^cent) copies are colored in blue and orange. (B and C) Interactions between (B) the WD40 domain of LRRK2^cent and ARM-ANK domains of LRRK2^peri and (C) the ROC domain of LRRK2^cen and the ARM domain of LRRK2^peri. (D) Interactions between two LRRK2^cent copies. (E) Conformational changes in the C-terminal halves of LRRK2 upon activation. A dashed circle indicates the central cavity between the KIN and COR domains. Color codes for different parts of LRRK2 are as follows: ROC, green; COR-A, light orange; COR-B, bright orange; N-lobe of KIN, cyan; C-lobe of KIN, marine; WD40, pink.

Fig. 4.. An active conformation of LRRK2.

(A) Superposition of kinase domains of LRRK2^cent and LRRK2^peri. N- and C-lobes of the LRRK2^cent kinase domain are colored in cyan and marine, respectively; the LRRK2^peri KIN domain is colored in gray. (Inset) Image shows the Cryo-EM density of the ATP molecule. (B and C) Key catalytic residues in LRRK2^peri (B) and LRRK2^cent (C) KIN domain with side chains shown as ball-and-stick models. The distances between the side chain of D2017 and the phosphate group of ATP are indicated with dashed lines. (D) R-spine of the LRRK2^peri (left) and LRRK2^cent (right) KIN domains. The four residues forming the R-spine (L1935, L1924, Y2018, and Y1992) are shown as green surfaces. (E) Docking of C-terminal catalytic halves of LRRK2^cent into the cryo-ET map of microtubule-bound LRRK2. (F) Movement of the KIN domain relative to the COR domain upon activation. (Inset) Interactions between the KIN and COR domains in the active conformation; side chains of the interface residues are shown as sticks. Dk, docking helix; APE, conserved APE motif; AL, activation loop. (G) Quantitative immunoblotting analysis of the cellular kinase activity of LRRK2-bearing mutations in the interface between the KIN and COR domains in fig. S6C. Data are presented as ratios of pRab10-Thr⁷³/total Rab10, pLRRK2-Ser¹²⁹²/total LRRK2, and pRab29-Thr⁷¹/total Rab29, normalized to the average of LRRK2 WT values. The data shown are the mean ± SD of three experiments.

Fig. 5.. Structure of LRRK2^RCKW with DNL201.

(A) Cryo-EM map of LRRK2^RCKW in complex with type I inhibitor DNL201. (Insets) Cryo-EM densities of the DNL201 inhibitor (top) and the surrounding residues (bottom) are shown. (B) DNL201 binding site [magnified from bottom inset of (A)]. Side chains of DNL201-interacting residues are shown as sticks. (C) Structural comparison of DNL201-bound LRRK2^RCKW and LRRK2^cent structures. (D) Comparison of KIN-ROC interface between LRRK2^RCKW-DNL201 (gray) and LRRK2^cent (blue and orange). Key structural elements from KIN domain involved in the interaction are labeled.

Fig. 6.. Rab29-LRRK2 tetramer and kinase activation.

(A and B) Quantitative immunoblotting analysis of the cellular kinase activity of LRRK2 in the presence of Rab29 or Rab32. HEK293 cells were transiently cotransfected with WT LRRK2 and hemagglutinin (HA)–tagged empty vector (“–”), HA-tagged Rab29, or HA-tagged Rab32 (WT or Q85L mutant). Data are presented as ratios of pLRRK2-Ser¹²⁹²/total LRRK2, pRab10-Thr⁷³/total Rab10, and pLRRK2-Ser⁹³⁵/total LRRK2, normalized to the average of LRRK2 WT values. The data shown are the mean ± SD of three determinations. (C) Cryo-EM maps of the Rab32–LRRK2 complex. (D) Summary of LRRK2 kinase activity and LRRK2 states observed in the cryo-EM study of LRRK2 alone, in the presence of Rab32, or in the presence of Rab29.