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