Rethinking LRRK2 Toxicity: It’s How Much Is Present, Not What It Does
Attempts to discern how mutations in the LRRK2 gene contribute to Parkinson’s disease have focused largely on the protein’s role as a kinase. In the January 8 Journal of Neuroscience, researchers led by Steven Finkbeiner at the Gladstone Institute of Neurological Disease, San Francisco, California, challenge this view. The authors report that in cultured neurons, cell death correlated poorly with kinase activity, but strongly with the amount of soluble LRRK2 in each cell. While it remains unclear how soluble LRRK2 poisons cells, the authors found that toxicity depended on the presence of α-synuclein, strengthening previous evidence for an interaction between these proteins.
Other researchers called the data a significant advance. “How increased LRRK2 results in cell death remains to be elucidated, but the results now make it clear that non-kinase-dependent effects of soluble LRRK2 are relevant for cellular physiology,” Patrik Verstreken at K.U. Leuven, Belgium, wrote to Alzforum (see full comment below). Likewise, Derya Shimshek at Novartis Pharma AG, Basel, Switzerland, and Martin Herzig, previously with that company, wrote to Alzforum, “The exciting findings provide key insights into the cellular role of LRRK2 in neurodegeneration underlying Parkinson’s disease.” (See full comment below.)
Previous studies of LRRK, short for leucine-rich repeat kinase 2, implicated the protein in a variety of functions, including apoptosis, synaptic vesicle recycling, and inflammation (see Jan 2009 news story; Oct 2012 news story; Oct 2012 conference story; and Mar 2013 conference story). Researchers do not know, however, which of these most contributes to disease. One clue comes from the most common pathogenic mutation, G2019S, which ramps up LRRK2’s kinase activity. This, in turn, has been linked to neuronal toxicity (see Greggio et al., 2006; Lee et al., 2010; Ramsden et al., 2011). Pharmaceutical companies have a wealth of experience designing kinase inhibitors and are highly interested in LRRK2 as a therapeutic target.
Finkbeiner and colleagues wanted to examine the relationship between kinase activity, LRRK2 levels, and cell death. First author Gaia Skibinski used an automated robotic microscope to follow the fates of individual cells in rat primary cortical and midbrain neuronal cultures. The neurons expressed wild-type LRRK2, a G2019S mutant, or a Y1699C mutant, which decreases GTP turnover and serves as a control. Transgenes were expressed at about fivefold endogenous levels and were fluorescently tagged, allowing the authors to quantify the amount of LRRK2 protein in each cell. By comparing protein levels and other aspects of cell physiology to survival, the authors were able to estimate the toxic effects of each feature on a cell-by-cell basis. Skibinski and colleagues found that for each neuron, the likelihood of dying over seven days correlated with the level of soluble LRRK2 in the cell.
To test the role of kinase activity in toxicity, the authors blocked it with inhibitors, or by using kinase-dead transgenes. Either method improved cell survival. However, blocking kinase activity had previously been shown to reduce LRRK2 stability and lower its amount (see Herzig et al., 2011). When the authors controlled for protein level, they found that kinase activity had no bearing on cell survival.
After adjusting for protein quantity, the two mutant forms of LRRK2 conferred a 50 percent higher risk of death than did the wild-type protein. Why might this be? Mutant LRRK2 forms inclusion bodies in neurons, but these aggregates did not affect cell survival, the authors found. Previous studies have shown that mutant LRRK2 exacerbates α-synuclein pathology in PD model mice (see Dec 2009 news story). In agreement with this, the authors found elevated levels of α-synuclein in cells carrying mutant LRRK2 transgenes, but not in those with wild-type protein. Knocking down α-synuclein improved survival in cells with mutant LRRK2. Skibinski and colleagues then transfected LRRK2 transgenes into neurons from synuclein knockout mice. Neither mutant nor wild-type LRRK2 were toxic, with cell death equivalent to that of untransfected cells. This implied to the authors that α-synuclein mediates LRRK2 toxicity.
To test the results in human cells, the authors made neuronal cultures from induced pluripotent stem cells generated from patients with the G2019S mutation. About half the neurons were dopaminergic. These cells accumulated more α-synuclein than neurons derived from normal controls, and they died more quickly. Skibinski noted that the ability to track survival and protein levels in single cells made this model more sensitive than typical cell culture experiments. “This was one of the first models for LRRK2 patients where we were able to see a difference in cell death without the addition of exogenous stress,” she told Alzforum. As in the rodent cultures, knockdown of α-synuclein protected the human neurons.
How might α-synuclein conspire with LRRK2 to damage cells? By raising the amount of LRRK2, the authors found. Knocking down α-synuclein in the rodent cultures lowered transgenic LRRK2, suggesting some sort of feedback loop or interaction. However, Skibinski thinks there is more to α-synuclein’s toxicity than just its effect on LRRK2 levels. “Several papers link both proteins to proteostasis pathways such as autophagy. Maybe α-synuclein and LRRK2 converge on that pathway, and it’s the balance between the two that causes degeneration,” she speculated. In future work, Skibinski will investigate the interaction between α-synuclein and LRRK2 to identify therapeutic targets. Suppressing LRRK2 itself might not be a viable strategy for Parkinson’s, since lack of the protein can cause problems in kidney and other peripheral organs (see, e.g., Baptista et al., 2013). Another intriguing question regards whether the interaction between LRRK2 and α-synuclein would hold up in Parkinson’s patients with wild-type LRRK2. The authors plan to make induced pluripotent stem cells from sporadic patients to test this.
Commentators agreed that these data help explain the toxicity of those LRRK2 mutations that have no effect on kinase activity. However, Ted Dawson at Johns Hopkins University, Baltimore, cautioned against concluding the kinase is irrelevant. Work from his group shows that even when levels of kinase-dead LRRK2 are equivalent to active forms, they are less harmful to cells (see Smith et al., 2006; West et al., 2007; Lee et al., 2010). “I think Skibinski et al.’s data suggests that LRRK2 mutants induce both kinase-dependent and independent forms of cell death,” Dawson wrote to Alzforum. Kinase-independent toxicity may arise from LRRK2’s GTPase domain, he noted (see Xiong et al., 2010; Xiong et al., 2012).—Madolyn Bowman Rogers
- LRRK2 Pathway Offers Up New Targets in Parkinson’s
- New Substrate for Parkinson’s Protein Is Picky About Phosphate
- The Many Faces of LRRK2
- LRRK Watchers’ Eyes Turn to Inflammation, Autophagy, Kinase
- α-Synuclein Conspires With LRRK2 to Corrupt Neurons
- Greggio E, Jain S, Kingsbury A, Bandopadhyay R, Lewis P, Kaganovich A, van der Brug MP, Beilina A, Blackinton J, Thomas KJ, Ahmad R, Miller DW, Kesavapany S, Singleton A, Lees A, Harvey RJ, Harvey K, Cookson MR. Kinase activity is required for the toxic effects of mutant LRRK2/dardarin. Neurobiol Dis. 2006 Aug;23(2):329-41. PubMed.
- Lee BD, Shin JH, VanKampen J, Petrucelli L, West AB, Ko HS, Lee YI, Maguire-Zeiss KA, Bowers WJ, Federoff HJ, Dawson VL, Dawson TM. Inhibitors of leucine-rich repeat kinase-2 protect against models of Parkinson's disease. Nat Med. 2010 Sep;16(9):998-1000. PubMed.
- Ramsden N, Perrin J, Ren Z, Lee BD, Zinn N, Dawson VL, Tam D, Bova M, Lang M, Drewes G, Bantscheff M, Bard F, Dawson TM, Hopf C. Chemoproteomics-based design of potent LRRK2-selective lead compounds that attenuate Parkinson's disease-related toxicity in human neurons. ACS Chem Biol. 2011 Oct 21;6(10):1021-8. PubMed.
- Herzig MC, Kolly C, Persohn E, Theil D, Schweizer T, Hafner T, Stemmelen C, Troxler TJ, Schmid P, Danner S, Schnell CR, Mueller M, Kinzel B, Grevot A, Bolognani F, Stirn M, Kuhn RR, Kaupmann K, van der Putten PH, Rovelli G, Shimshek DR. LRRK2 protein levels are determined by kinase function and are crucial for kidney and lung homeostasis in mice. Hum Mol Genet. 2011 Nov 1;20(21):4209-23. PubMed.
- Baptista MA, Dave KD, Frasier MA, Sherer TB, Greeley M, Beck MJ, Varsho JS, Parker GA, Moore C, Churchill MJ, Meshul CK, Fiske BK. Loss of leucine-rich repeat kinase 2 (LRRK2) in rats leads to progressive abnormal phenotypes in peripheral organs. PLoS One. 2013;8(11):e80705. Epub 2013 Nov 14 PubMed.
- Smith WW, Pei Z, Jiang H, Dawson VL, Dawson TM, Ross CA. Kinase activity of mutant LRRK2 mediates neuronal toxicity. Nat Neurosci. 2006 Oct;9(10):1231-3. PubMed.
- West AB, Moore DJ, Choi C, Andrabi SA, Li X, Dikeman D, Biskup S, Zhang Z, Lim KL, Dawson VL, Dawson TM. Parkinson's disease-associated mutations in LRRK2 link enhanced GTP-binding and kinase activities to neuronal toxicity. Hum Mol Genet. 2007 Jan 15;16(2):223-32. PubMed.
- Xiong Y, Coombes CE, Kilaru A, Li X, Gitler AD, Bowers WJ, Dawson VL, Dawson TM, Moore DJ. GTPase activity plays a key role in the pathobiology of LRRK2. PLoS Genet. 2010 Apr;6(4):e1000902. PubMed.
- Xiong Y, Yuan C, Chen R, Dawson TM, Dawson VL. ArfGAP1 is a GTPase activating protein for LRRK2: reciprocal regulation of ArfGAP1 by LRRK2. J Neurosci. 2012 Mar 14;32(11):3877-86. PubMed.
- Not Just a Family Affair: Dardarin Mutations Predict Sporadic PD
- SfN: Latest Intel on LRRK2, Parkin
- PARK9 Is Unveiled—Mutations Compromise Orphan Lysosomal ATPase
- Come and Get ’em—Foundation Develops Antibodies for PD Research
- Research Brief: A Better Mouse Model of Parkinson Disease?
- More Than a LRRK: PD Field Thinks Big With Genetic Cohort
- Skibinski G, Nakamura K, Cookson MR, Finkbeiner S. Mutant LRRK2 toxicity in neurons depends on LRRK2 levels and synuclein but not kinase activity or inclusion bodies. J Neurosci. 2014 Jan 8;34(2):418-33. PubMed.
To make an annotation you must Login or Register.
In this paper, Skibinski et al. use cellular models to assess LRRK2 toxicity. They use automated imaging to assess the time of death of large cohorts of cells and come to the interesting observation that “soluble” LRRK2 levels positively correlate with earlier cell death. Neither kinase activity nor the presence of inclusion bodies contribute to this phenotype. How increased LRRK2 results in cell death remains to be elucidated, but the results now make it clear that non-kinase-dependent effects of soluble LRRK2 are relevant for cellular physiology. The results also place in proper context the fact that numerous pathogenic mutations in LRRK2 exist that do not result in increased (or decreased) kinase activity. The existence of such mutations is indeed consistent with the hypothesis that altered kinase activity per se is not the sole driving force that causes Parkinson’s disease.
Is the LRRK2 kinase activity then irrelevant? Previous work has implicated LRRK2 kinase function in important neuronal and cellular processes, including synaptic transmission and vesicle trafficking. Furthermore, this domain is conserved over several million years of evolution, suggesting it harbors functional relevance. However, whether these actions of the kinase domain are relevant to the disease is indeed unclear. Similarly, additional work will be needed to determine to what extent the phenotype assessed in the work by Skibinski et al., “time of cell death in vitro,” is relevant to Parkinson’s disease in vivo. Nonetheless, the current work aids in shifting focus toward addressing the non-kinase-dependent effects of LRRK2, and this is an important contribution.View all comments by Patrick Verstreken
Novartis Institutes for BioMedical Research
Institute of Pathology, University Hospital Basel
The exciting findings presented here by Skibinski et al. provide key insights into the cellular role of LRRK2 in neurodegeneration underlying Parkinson’s disease (PD). The authors used a longitudinal, real-time imaging platform to individually track primary rodent neurons ectopically expressing tagged mutant or wild-type LRRK2, as well as tracking LRRK2 patient-derived human neurons. They elegantly demonstrated that toxicity of ectopically expressed LRRK2 depended on the levels of diffuse LRRK2 and α-synuclein, but not LRRK2 kinase activity, per se. Pharmacological or genetic inhibition of LRRK2’s kinase function lowered the protein’s levels, thereby causing less neurodegeneration.
These findings are very appealing as they could explain why pathogenic mutations are found in functional domains other than the kinase domain of LRRK2. They could potentially all lead to the generation of diffuse toxic LRRK2 entities. Together with the synergistic or independent/indirect event of α-synuclein build-up, and interference with cellular homeostasis during aging, this could indeed shed some light on the pathophysiological mechanisms of PD.
In subsequent studies, it would be important to determine whether endogenously expressed LRRK2 with pathogenic mutations (e.g., in LRRK2 G2019S patient-derived neurons) produces more diffuse toxic LRRK2 protein in comparison to LRRK2 wild-type, or if such diffuse toxic LRRK2 protein is only found when ectopically overexpressed. If the former turns out to be true, therapeutics that help promote more efficient sequestration of toxic diffuse LRRK2 may be a viable therapeutic strategy against LRRK2-mediated toxicity, as the authors suggest. Though difficult to achieve, this would leave normal LRRK2 unaffected, which would be highly desirable as the findings from LRRK2 knockout studies in rodents demonstrated an important role for LRRK2 in kidney and lung function.
Of note, it remains puzzling why overexpression of G2019S-mutated human LRRK2 in non-dopaminergic mouse neurons in vivo resulted in diffuse LRRK2 staining in the brainstem and cortical neurons, but did not lead to overt degeneration of these neurons or to accelerated α-synuclein-driven neurodegeneration in the brainstem (see Herzig et al., 2012).
The LRRK2 field has suffered from a lack of validation experiments. We would embrace efforts to strengthen the interesting hypothesis of Skibinski and co-workers with experiments from different labs that confirm these new insights. We rely on solid, reproducible data to establish preclinical models to develop promising medication for those in need.View all comments by Martin Herzig
Medical Research Council Protein Phophorylation and Ubiquitylation Unit
This is an intriguing study. LRRK2 belongs to a class of enzymes termed “protein kinases.” LRRK2 is frequently mutated in Parkinson’s. Between 1-2 percent of Parkinson’s cases may be caused by mutations in the gene that encodes for the LRRK2 enzyme. To my knowledge, the evidence that LRRK2 is implicated in Alzheimer’s is weak. The assumption has always been that the protein kinase enzymatic function of LRRK2 was the critical entity for understanding how it operates and how it is linked to Parkinson’s disease.
Indeed, there is some evidence to support this. For example, the most frequent LRRK2 mutation (G2019S) lies within the heart of the kinase domain and this mutation significantly increases kinase catalytic activity. However, in contrast, this study by Skibinski et al. provides some evidence that the kinase activity of LRRK2 may be less critical than assumed by most researchers. The authors show, in a cell culture overexpression toxicity model, that LRRK2 toxicity is independent of its kinase activity. Instead, their data reveal that the absolute levels of LRRK2 expressed in cells is the most important factor. The data also suggest that expression of another protein, α-synuclein, which is strongly linked to Parkinson’s as well as Alzheimer’s, is also a critical determinant for LRRK2-mediated toxicity.
I feel that this paper makes a useful contribution to the literature as it highlights the importance of studying absolute LRRK2 protein levels, and studying whether variation of these levels is a key factor in Parkinson’s. It is possible that some of the LRRK2 Parkinson’s mutations could influence the expression levels of the kinase and this could be the mechanism by which they drive toxicity. LRRK2 also contains a GTPase catalytic domain, a leucine-rich repeat, and a highly conserved C-terminal tail, which are all likely to play critical roles. The importance of some of these other domains has been previously overlooked while everyone has focused on finding substrates for LRRK2. The Skibinski study suggests that more analysis of the other domains of LRRK2 is warranted, and that we should also be studying the importance of absolute LRRK2 expression as a factor mediating LRRK2 toxicity.
I am not an expert in the analysis of neuronal toxicity, but I notice that many of the studies in the Skibinski paper are undertaken by overexpressing wild-type and mutant forms of LRRK2 in primary neuronal cells. Previous work has shown that there is always the danger in overexpression studies that this approach could lead to the observed phenotype through a non-physiological perturbation of the cell system. Nevertheless this is an interesting paper that will stimulate further research into LRRK2. It will also be interesting to learn more about the molecular mechanism by which LRRK2 interacts with α-synuclein, and whether this has a role to play in Alzheimer's.View all comments by Dario Alessi
Make a Comment
To make a comment you must login or register.