Mutations in LRRK2 (leucine-rich repeat kinase 2) are the most common cause of familial PD (Parkinson’s disease). Mutations that cause PD are found in either the GTPase or kinase domains of LRRK2 or an intervening sequence called the COR [C-terminus of ROC (Ras of complex proteins)] domain. As well as the two catalytic domains, LRRK2 possesses several protein–protein interaction domains, but their function and the proteins with which they interact are poorly understood. In this issue of the Biochemical Journal, Nichols et al. study the interaction of the N-terminal region of LRRK2 with 14-3-3 proteins, regulatory proteins that often bind to phosphorylated regions of components of cell signalling pathways. Using a combination of techniques, Nichols et al. have identified two residues (Ser910 and Ser935) that are critically responsible for 14-3-3 binding. The interaction of LRRK2 with 14-3-3 proteins can prevent dephosphorylation of Ser910/Ser935 and stabilize LRRK2 structure, perhaps by influencing the dimerization of LRRK2. The ability to interact with 14-3-3 correlates with the pattern of intracellular LRRK2 distribution. Collectively, these new results identify a potentially important regulatory mechanism of this complex protein and might provide ways to think about therapeutic opportunities for PD.
- 14-3-3 protein
- leucine-rich repeat kinase 2 (LRRK2)
- Parkinson's disease
- protein–protein interaction
PD (Parkinson's disease) has, in recent years, been shown to have a significant genetic component. The most common single cause of PD is mutation of the LRRK2 (leucine-rich repeat kinase 2) gene. Several mutations have been found worldwide in different families [1–3].
LRRK2 is a large multidomain protein of 2527 amino acids (286 kDa) belonging to the Roco protein family. All members of this family include a ROC (Ras of complex proteins)–COR (C-terminus of ROC) bidomain . The ROC and COR domains interact and may regulate the kinase activity of LRRK2, although this has been challenged . PD-associated mutations are found in this central region and also in the kinase domain for which LRRK2 is named. As reviewed previously , mutations in the kinase domain increase kinase activity. Although several kinase substrates have been proposed, none has been confirmed. The remainder of the protein contains domains likely to be involved in protein–protein interactions.
14-3-3 proteins are ubiquitously expressed small (28–33 kDa) acidic polypeptides that self assemble into homo- and hetero-dimers. Binding motifs of 14-3-3 proteins have high affinity for phospho-serine/phospho-threonine residues of many proteins involved in transcription, biosynthesis, apoptotic signalling and cytoskeletal dynamics. 14-3-3 proteins change signalling either by protecting important regulatory sites of proteins from dephosphorylation or by influencing the formation of dimers or multiprotein complexes .
The study of Nichols et al.  in this issue of the Biochemical Journal show that several members of the 14-3-3 protein family bind to LRRK2. They identified sites for this binding in the N-terminus of LRRK2 (Ser910 and Ser935). The authors designed streptavidin–agarose beads with a conjugated di-phosphorylated peptide encompassing Ser910 and Ser935 of LRRK2 that were able to bind 14-3-3. The dephosphorylation of Ser910 and Ser935 by λ phosphatase completely abolished the binding. This experiment strongly supports the idea that 14-3-3 binds directly to phosphorylated Ser910/Ser935.
A key finding is that three PD-associated mutants (R1441G, Y1699C and I2020T) and a truncated form of LRRK2 (E1874stop) had decreased ability to bind 14-3-3. The interaction with 14-3-3 was also substantially weaker in six other LRRK2 mutants (M712V, R1441H, R1441C, A1442P, L1795F and G2385R). For the majority of these mutants, the authors also found decreased Ser910/Ser935 phosphorylation. Taken together, these data suggest that the 14-3-3 interaction may be affected by a number of different mutations and thus might be a common pathogenic output. However, G2019S, the most common LRRK2 mutation causing autosomal-dominant parkinsonism, as well as some other potentially pathogenic mutations (I2012T, I1371V and T2356I) did not cause a decrease in 14-3-3 binding. This implies that 14-3-3 binding is not the only mechanism influencing progression of PD associated with mutations in LRRK2.
In their paper, the authors speculate that the interaction with 14-3-3 proteins could be important for the stabilization of LRRK2 structure in vivo by protecting Ser910 and Ser935 from dephosphorylation . An alternative hypothesis is that perhaps 14-3-3 proteins could influence LRRK2 dimer formation. Several lines of evidence suggest that a major intracellular form of LRRK2 is a dimer  and that the kinase activity of LRRK2 depends on its dimerization . It is therefore possible that 14-3-3 proteins might influence LRRK2 dimerization and thus modulate its kinase activity.
LRRK2 forms not only a dimer, but also a series of apparently higher-molecular-mass species on native PAGE [7,8], although some recent reports interpret these species as monomers . The range of species that LRRK2 exists as is a pivotal question to resolve as identification of protein–protein binding partners in dimeric compared with high-molecular-mass complexes might help us understand the signalling roles of LRRK2. 14-3-3 proteins may influence the relative amounts of dimer compared with other forms of LRRK2, either by mediating an association of LRRK2 with regulatory proteins or by keeping the LRRK2 structure compact that, in turn, could be important for its functional activity.
In another set of assays, Nichols et al.  show that LRRK2 mutants with a lower ability to bind 14-3-3 proteins form inclusions in cytosol. The interpretation of these data is unclear. The formation of inclusions could be due to the inability of 14-3-3 to maintain the conformational structure of LRRK2 as a monomer, and LRRK2 might then form unusual multiprotein complexes. Alternatively, if 14-3-3 proteins stabilize higher-molecular-mass complexes of LRRK2, then the absence of 14-3-3 might abolish normal multiprotein complexes and hence facilitate the formation of aberrant, perhaps misfolded, protein inclusions.
A kinase-inactive form of LRRK2 showed lower affinity for 14-3-3, but did not abolish the interaction completely. Interestingly, the kinase-activated forms of LRRK2 such as G2019S did not increase association of LRRK2 with 14-3-3 to any measureable extent. This finding leads to the question of whether Ser910/Ser935 are autophosphorylation sites or whether they are instead phosphorylated by other kinases. The finding that λ phosphatase can dephosphorylate both of them suggests that they are located on the surface of LRRK2 and thus could be available for many intracellular kinases. Interestingly, another recent study mapping phosphorylated sites in LRRK2 found evidence for a constitutive phosphorylation of peptides containing the same two residues . Because the phosphorylation of Ser910/Ser935 is not completely dependent on autophosphorylation, this supports the idea that external kinase(s) may regulate phosphorylation and hence 14-3-3 binding and LRRK2 function.
This leads to some important questions to follow up on the work of Nichols et al. . Clearly, identification of the kinase(s) responsible for the regulation of Ser910/Ser935 phosphorylation is a critical next step. Further characterization of the mode of 14-3-3 binding will also be important. Does 14-3-3 bind LRRK2 in cis or in trans? As 14-3-3 proteins are often dimeric, it would be important to understand whether dimers of 14-3-3 bind dimers of LRRK2 or whether the stoichiometry is different. As relative partitioning of LRRK2 into dimers or other complexes is thought to regulate its function , this may represent a regulatory step for LRRK2.
All of this has implications for therapeutics in PD, something that is sorely lacking for halting neurodegeneration and not just merely treating symptoms. One could at least imagine that disrupting the LRRK2–14-3-3 interaction could modify LRRK2 participation in signalling cascades and therefore limit the toxic effects of the protein, assuming that the inclusions are not themselves detrimental to cells. Whether this would be more or less specific than small molecules targeted to the ATP-binding pocket of the kinase domain is uncertain, but studies like the one from Nichols et al.  show how basic research can stimulate translational efforts.
This work was supported by the Intramural Research Program of the National Institute on Aging of the National Institutes of Health.
Abbreviations: LRRK2, leucine-rich repeat kinase 2; PD, Parkinson's disease; ROC, Ras of complex proteins; COR, C-terminus of ROC
- © The Authors Journal compilation © 2010 Biochemical Society