skibinski g 2014 j neuroscience subrayado

8/12/2019 Skibinski G 2014 J Neuroscience SUBRAYADO

1/16

Neurobiology of Disease

Mutant LRRK2 Toxicity in Neurons Depends on LRRK2Levels and Synuclein But Not Kinase Activity or InclusionBodies

Gaia Skibinski, 1 Ken Nakamura, 1,2,4 Mark R. Cookson, 5 and Steven Finkbeiner 1,2,3,41Gladstone Institute of Neurological Disease, the Taube-Koret Center for Neurodegenerative Disease Research, and the Hellman Family FoundationProgram in Alzheimers Disease Research, San Francisco, California 94158, 2Departments of Neurology, 3Physiology, and 4Graduate Programs inNeuroscience and Biomedical Sciences, University of California, San Francisco, California 94158, and 5Laboratory of Neurogenetics, National Institutes of Health, Bethesda, Maryland 20892

By combining experimental neuron models and mathematical tools, we developed a systems approach to deconvolve cellular mecha-nisms of neurodegeneration underlying the most common known cause of Parkinsons disease (PD), mutations in leucine-rich repeatkinase 2 (LRRK2). Neurons ectopically expressing mutant LRRK2 formed inclusion bodies (IBs), retracted neurites, accumulated sy-nuclein, and died prematurely, recapitulating key features of PD. Degeneration was predicted from the levels of diffuse mutant LRRK2that each neuron contained, but IB formation was neither necessary nor sufficient for death. Genetic or pharmacological blockade of itskinaseactivitydestabilizedLRRK2andlowered itslevels enoughto account forthe moderatereductionin LRRK2toxicity that ensued. By contrast, targetingsynuclein, includingneurons made fromPD patient-derivedinduced pluripotent cells, dramatically reduced LRRK2-dependent neurodegeneration and LRRK2 levels. These findings suggest that LRRK2 levels are more important than kinase activity perse in predicting toxicity and implicate synuclein as a major mediator of LRRK2-induced neurodegeneration.

Key words: LRRK2; mechanisms; Parkinsons disease; single cell; synuclein

IntroductionParkinsons disease (PD) is a progressive neurodegenerativedisorder resulting in motor and cognitive dysfunction. Nodisease-modifying treatment exists. Mutations in leucine-richrepeat kinase 2 (LRRK2), the most common genetic cause of PD (Paisan-Ruz et al. , 2004; Zimprich et al., 2004; Clark et al.,2006), lead to a dominantly inherited phenotype that is similarto idiopathic PD ( Haugarvoll et al., 2008). The mechanisms of mutant LRRK2-induced neurodegeneration are elusive. If identified, they might point to novel therapeutic targets.

Increased kinase activity is a prevailing hypothesis for mutantLRRK2-mediated toxicity. Of the known LRRK2 mutations thatcause PD, the G2019S mutation consistently enhances LRRK2kinaseactivity,butwhether this is a pathogenic mechanismcom-mon to all LRRK2 mutants is unclear (Greggio et al., 2006; Westet al., 2007). Alternatively, because the pathogenic mutations arenonconservative, their harmful effects may be mediated by mis-folded LRRK2. Misfoldedproteins tendto accumulateand aggre-gate within cells (Chiti and Dobson, 2006 ). In vitro, LRRK2aggregates and forms inclusion bodies (IBs) that correlate withmutant-LRRK2-induced toxicity and kinase activity (Greggio et

al., 2006; MacLeod et al., 2006). Although LRRK2 IBs are anindicatorof protein misfolding, their role and importance, if any,in disease progression are unknown.

Misfolded proteins may cause toxicity by adopting toxic confor-mations or by overloading the protein-homeostasis machinery (Gidalevitz et al., 2006). Overexpressing the C-terminus of Hsp70interactingproteinordownregulatingHsp90,both ofwhich interactwith LRRK2, reduce mutant LRRK2-induced toxicity in a kinase-independentmannerandareeachassociatedwith lower steady-statelevels ofmutantLRRK2( Wangetal.,2008 ;Koetal., 2009). Changesin LRRK2 levels correlate with changes in kinase activity, as geneticand pharmacological manipulations that inhibit kinase activity re-duceLRRK2stability (Lin etal., 2009;Herzig etal.,2011). IfLRRK2-

related toxicity dependson physiologicLRRK2levels, toxicity mighttherefore be better explained by changes in LRRK2 levels than by kinase activity itself.

Received June 26, 2013; revised Nov. 1, 2013; accepted Nov. 21, 2013.Author contributions: G.S., K.N., M.R.C., and S.F. designed research; G.S. performed research; K.N. and M.R.C.

contributed unpublished reagents/analytic tools; G.S. analyzed data; G.S. and S.F. wrote the paper.ThisworkwassupportedwithagiftfromtheDeClerqfamilyinmemoryofBettyBrown.TheGladstoneInstitutes

received support from a National Center for Research Resources Grant RR18928. G.S. is supported by the HillblomFoundation. K.N. is supported by the Burroughs-Wellcome Medical Scientist Fund Career Award, NINDS(1K08NS062954-01A1) and award no. P30NS069496 from the National Institute of Neurological Disorders andStroke. M.R.C. issupportedin partby theIntramural ResearchProgramof theNIH, NationalInstitute on Aging.S.F.issupportedby GrantsU24 NS078370,3R01NS039074,2R01NS04549,2P01AG0224074,CIRMRB4-06079researchGrant, the Koret/Taube Center, and the Hellman Family Foundation. We thank members of the Finkbeiner lab foruseful discussions, A.L. Lucido, S. Ordway, and G. Howard for editorial assistance, and K. Nelson for administrativeassistance. Dr. Dario Alessi (MRC-PPU, Dundee University, Dundee Scotland) who kindly provided LRRK2-IN1 andrabbit monoclonal anti-LRRK2 (100-500), anti-LRRK2 P-ser935, or anti-LRRK2 P-ser910.

The authors declare no competing financial interests.Correspondence should be addressed to Steven Finkbeiner, Gladstone Institute of Neurological Disease, 1650

Owens Street, Room 308, San Francisco, CA 94158. E-mail: [email protected]:10.1523/JNEUROSCI.2712-13.2014

Copyright 2014 the authors 0270-6474/14/340418-16$15.00/0

418 The Journal of Neuroscience, January 8, 2014 34(2):418433


2/16

Independent of kinase activity, the PD-associated protein-synuclein accumulates in cells with elevated LRRK2 levels (Lin

et al., 2009; Sanchez-Danes et al., 2012 ). PD patients with LRRK2mutations often display synuclein pathology ( Biskup and West,2009), and symptoms occur earlier if specific genetic variants inthe synuclein gene (SNCA; Botta-Orfila et al., 2012) are presentalong with LRRK2 mutations. However, whether -synuclein is

required for LRRK2-mediated degeneration is unknown. Givenits role in disrupting protein-degradation pathways ( Cuervo etal., 2004; Alegre-Abarrategui et al., 2009; Gomez-Suaga et al.,2011), -synuclein might mediate LRRK2 toxicity by disruptingprotein homeostasis and changing LRRK2 levels.

Typically, snap-shotcellularapproachesare used to identify disease-associated events. However, often a cellular event is mea-sured from populations of cells, masking the relationship be-tween theevent andcell death (e.g.,does it improve or exacerbatecell death). In contrast, following thousands of single cellsthroughout the course of neurodegeneration enables both theidentification of disease-associated events and the determinationof their prognostic value for cell death. Using an imaging plat-

form, we longitudinally tracked primary neurons expressingLRRK2 throughout their lives. LRRK2-associated parameters(e.g., levels, IB formation, and kinase activity) were capturedwithin each neuron and linked to when the neuron died. Usingmultivariate survival models, cell-to-cell variability was har-nessed to establish quantitative and temporal relationships be-tween LRRK2-related events and cell death. By combiningexperimental and statistical models ( Pastrana, 2012 ; Skibinskiand Finkbeiner, 2013 ), we established a basic cellular systemsmodelof LRRK2-mediated neurodegeneration that led to severalsignificant insights.

Materials and MethodsPlasmids. LRRK2 was derived from pAcGFP-LRRK2 (Greggio et al.,2006; fromMarkCookson, NIH) andsubcloned into pGW1-CMV (Brit-ish Biotechnologies). Y1699C, G2019S, and D1994A were generated us-ing Quikchange Site-Directed Mutagenesis Kit (Stratagene). Wild-typeandmutant LRRK2 were clonedinto pGW1-CMV ( Arrasate et al., 2004 ).All constructs were verified by sequencing. In cotransfection experi-ments with pGW1-Venus-LRRK2, we used pGW1-monomeric red flu-orescent protein (mRFP) or pGW1-mApple as a survival marker. Thehuntingtin (Htt) constructs (Htt586) with 17 or 136 polyglutamine re-peats were cloned into pGW1-GFP from full-length Htt constructsJM031 andJM032(gift from Ray TruantMcMasterUniversity, ON Can-ada). Knockdown of -synuclein was achieved using MISSION Sigma-Aldrich pLKO.1 constructs. To track Map2-expressing human cells,mApple was cloned into a plasmid that put its expression under thecontrol of the human Map2 promoter (System Biosciences).

Cell culture and transfections. For Western blot analysis, HEK293 cellswere grown in DMEM medium containing 10% calf serum, 2 m M glu-tamine, and penicillin/streptomycin (100 U ml 1 /100 g ml 1 ). Pri-mary rat cortical neurons were prepared from rat pup cortices atembryonic days (E)20 E21, cultured at 0.6 10 6 cells/ml.At 5 d in vitro(DIV) neurons were transfected withplasmids by the calciumphosphatemethod as described (Finkbeiner et al., 1997 ; Saudou et al., 1998) or by Lipofectamine 2000. For survival analysis, neurons in 24- or 96-wellplates were cotransfected with pGW1-RFP and pGW1-Venus-LRRK2 ata 1:12 molar ratio (0.51 g of DNAper well).Postnatal (P)rat midbraincultures were prepared from the ventral mesencephalon of P0P1 rats(Mena et al., 1997). For survival analysis, postnatal neurons (100,000/well) in 96-well plates were transfected using Lipofectamine (0.8 g of DNA per well). To prepare cortical cultures from mice, cortices weretaken from E21 -synuclein triple knock-out(TKO) mice( Nakamura

et al., 2011) and wild-type littermates (Ctrl). For confocal analysis, cor-tical and midbrain neurons were transfected and fixed 2448 h post-transfection.To assessthe detergent-resistance of IBs, neurons werefixed

with 1% paraformaldehyde in PBS for 10 min at 37C, rinsed twice withPBS, and treated with 12.5% Triton X-100 and 12.5% SDS for 20 minat 37C (Kazantsev et al., 1999). Neurons were then rinsed with PBS andimaged by fluorescence microscopy.

iPSC differentiation to DA neurons. Control-induced pluripotent cell(iPSC) lines derived from PD patients harboring a homozygous G2019Smutation (ND35367) were obtained from the NINDS iPSC repository atCoriell Institute. Reprogramming and characterization of the controliPSC line (normal female, 40-y-old; CRL-2524; ATCC) were as reported(Park et al., 2008; Bilican et al., 2012). Neural stem cells (NSCs) weregenerated from the control and PD iPSC lines, lifted off into suspensionand treated with dual SMAD inhibitors, SB431542 and LDN193189, for23d. NSCs were grown in suspensionas sphericalaggregates as reported(HD iPSC Consortium, 2012 ). For DA neuron differentiation, NSCswere cultured on PA6 stromal cells in serum-free insulin/transferring/selenium medium with ascorbic acid (ITS AA), as described (Park etal.,2005).ITS AA mediumwas supplementedwith bFGF andpurmor-phamine. Cells were harvested from the PA6 cells and plated ontofibronection/poly- L-ornithine (FN)-coated plates and cultured in ITSAA supplemented with bFGF, FGF8a, and purmorphamine. In the finalstage of precursor differentiation, cells were dissociated into single cellsusing Accutase and plated onto FN-coated 96-well plates (0.1 10 6

cells/well). The cells were then cultured in neurobasal media, N2, B27,ascorbic acid, BDNF, GDNF, and cAMP (adapted from Xi et al., 2012).Terminally differentiated cultures were transfected with the mApple-Map2 reporter.

Immunocytochemistry. Cortical and postnatal rat midbrain neurons andhuman differentiated DA neurons were grown on 96-well plates or 12 mmcoverslips.Neuronsweretransfectedat 4DIVor 620d into thefinal stageof precursor differentiation. At 48 h post-transfection, immunocytochemistry wasperformed asdescribed( Saudouetal.,1998 ) andlabeledwithatleastoneof the following, anti-tyrosine hydroxylase (1:1500, Pel-Freez), anti-MAP2(1:200, Millipore), anti-LRRK2 antibody (1:2000, Novus Biologicals; or1:100, Cell Signaling Technology), and anti- -synuclein (1:500, BD Biosci-ences). LRRK2 100500, LRRK2 phospho (pS935), and LRRK2 phospho(pS910) wereprovidedby Dr.DarioAlessi (MRC-PPU,DundeeUniversity,Dundee Scotland) and purchased from Epitomics. Primary antibody stain-

ing wasfollowedby secondaryantibodywitheither, anti-rabbit Cy3(1:500),anti-mouseCy5 (1:200),anti-rabbit Cy5 (1:200),or anti-mouseCy5 (1:500;Jackson Immunochemical).

Robotic microscope imaging system and image analysis. For neuronalsurvival analysis, images of neurons were taken at 1224 h intervals aftertransfection with an automated imaging platform as described ( Arrasateet al., 2004; Arrasate and Finkbeiner, 2005 ; Daub et al., 2009). Measure-ments of LRRK2 expression, IB formation, and neuron survival wereobtained from files generated with automated imaging. Digital imageswereanalyzedwith MetaMorph,ImagePro, andoriginal proprietarypro-grams that were written in MATLAB or pipeline pilot. These custom-based algorithms were then used to capture and analyze neurons in eachgroup in a high throughput and unbiased manner. Live transfected neu-rons were selected for analysis based on fluorescence intensity and mor-phology. Neurons were only selected if they hadextended processesat thestart of the experiment. The abrupt loss of cotransfected mRFP was usedto estimate the survival time of the neuron ( Arrasate et al., 2004). Datafrom each experimental group were analyzed in an identical manner sothere was little need for blinding. The expression of Venus-tagged ver-sionsof LRRK2was estimatedby measuring Venusfluorescence intensity over a region of interest that corresponded to the cell soma, using thefluorescence of cotransfected mRFP as a morphology marker (Arrasate etal.,2004). TheVenus intensity valueswere background-subtractedby usingan adjacent area of the image. The fold difference in levels of overexpressedVenus-LRRK2toendogenousLRRK2 wasdeterminedbymeasuringLRRK2(labeled with anti-rabbit Cy5 antibodies) staining intensity in Venus-LRRK2- or Venus-transfectedcells. Total neuronalprocesslength andnum-berwere calculated from tracesof neurons coexpressingmRFP and LRRK2.Tracesof neuronalprocesses were made with Fiji imaging softwareusingthe

simple neurite tracer plug-in.Statistical analysis. With longitudinal single cell data collected by

tracking individual neurons over time, survival analysis was used to ac-

Skibinski et al. LRRK2 Levels and Synuclein Mediate LRRK2 Toxicity J. Neurosci., January 8, 2014 34(2):418433 419


3/16

curately determine differences in longevity between populations of cells(Jager et al., 2008). Forlongitudinalsurvival analysis, the survival timeof a neuron was defined as the time point at which a cell was last seen alive.Statview software was used to construct KaplanMeier curves from thesurvival data. Survival functions were fitted to these curves and used to

derive cumulative hazard (or risk-of-death) curves that describe the in-stantaneous risk-of-death for individual neurons in the cohort beingtracked. Differences in the cumulative risk-of-death curveswere assessedwith the log rank test. We used Cox proportional hazards regressionanalysis (Cox analysis) to generate hazard ratios that quantified the rel-ative risk-of-death between two cohorts of neurons or the predictivevalues of variables such as the presence of a PD-associated mutation,changes in LRRK2 levels or IB formation on the risk-of-death ( Miller etal., 2010, 2011). As IB formation is a time-dependent variable, time-dependent Cox analysis (Dekker et al., 2008) was used to measure thecontribution of IB formation to neuron death. The sign and size of thecoefficient of the cellular feature describe, respectively, its predictive re-lationship to the fate of the cell and the magnitude of its effect. Hazardratios and their respective p values were generated using the coxph func-tion in the survival package for R statistical software ( Andersen and Gill,1982). All Cox models were analyzed for violations of proportional haz-ards and for outlier data points using cox.zph and dfbeta functions in R,respectively. Images of cells that were out of the focal plan of Venusfluorescence were excluded from analysis for IB formation (this was

15% of cells under analysis). Selection of IB formation was based onexpression intensity of Venus-LRRK2 and localization. For risk of IBformation, death canprecludeIB formationmakingit a competing effectof IB formation. To control for this, hazard ratios and p values for therisk-of-IB formation were determined with a multivariate regressionanalysis using the semiparametric proportional hazards model( Fine andGray, 1999), that incorporates the presence of competing risk events(competingrisk analysis).The analysis wasperformedusingthe crr pack-age for R statistical software (Scrucca et al., 2010). The cumulative risk-of-IB formation curves are the estimated cumulative subdistributionhazard for IB formation and are derived from the competing risk regres-sion model. Significance was derived from the competing risks propor-tional hazard regression model.

Coxanalysis was used to generate hazardratios for therelative risk-of-death of neurons expressing mutant LRRK2 or Venus compared withwild-type LRRK2 or Ctrl. In Table 1, time-dependent Cox analysis wasused to determinethe predictivenature of IB formationon risk-of-death.Time of IB formation was included as a time-dependent covariate,whereas Venus-LRRK2 fluorescence was included as a baseline covariatein the regression model. Violations of proportional hazards by cox.zphwere accommodated by including a time-interaction term for IB forma-tion as a covariate, which measures how the effect of IB formation onrisk-of-death varies with time. In Table 2, hazard ratios for the relativerisk-of-IB formation were determined with a multivariate regressionanalysis using the semiparametric proportional hazards model( Fine and

Gray, 1999) that incorporates the presence of competing risk events. Theanalysis was performed using the crr package for R statistical software(Scrucca et al., 2010). Covariates in the regression model included

Venus-LRRK2 fluorescence. In several models, the date of the experi-ment was included as a stratification variable. Any models that includeddateas a nominal covariate gavecomparable results. Simple linear regres-sion was usedto correlate Venus-LRRK2expression andLRRK2staining.To compare differences across two groups, the groups were statistically compared using an unpaired t test or a MannWhitney test. To comparedifferences across groups, for example when comparing average LRRK2levels or neurite length, groups were statistically compared using one-way ANOVA for repeated measures and Fishers PLSD post hoc tests.Levels of endogenous -synuclein in cohorts were determined by nor-malizing against -synuclein levels in a cohort of neurons expressingonly Venus or Ctrl. To compare non-normal data across multiplegroups, the KruskalWallis test was used.

Western blotanalysis. In experiments to confirm expression of full-lengthLRRK2, HEK293 cells were transiently transfected with pGW1-Venus,pGW1-Venus-wild-type-LRRK2, pGW1-Venus-Y1699C-LRRK2, orpGW1-Venus-G2019S LRRK2. In experiments looking at the influence of mutations within the kinase domain, HEK293 cells were transfected withpGW1-Venus-D1994A-LRRK2 constructs with or without PD-associatedmutations. In experimentsthat confirmed that Venus-LRRK2constructswere kinase active, HEK293 cells were transiently transfected with wild-type or mutant pGW1-Venus-LRRK2 or pGW1-LRRK2. At 24 h post-transfection, cells were harvested.

In experiments that confirmed that the kinase inhibitor LRRK2-IN1was active, HEK293cells weretransientlytransfected with pGW1-Venus-LRRK2 G2019S. At 24 h post-transfection, cells transfected with pGW1-Venus-G2019S were treated with DMSO or 1 M LRRK2-IN1 (kindly provided by Dr. Dario Alessi, MRC-PPU, Dundee University, DundeeScotland) and harvested 2 h later. In all experiments, lysates were pre-pared using RIPA buffer (50 m M Tris-Cl, pH 7.4, 150 m M NaCl, 0.5%deoxycholate, 1% NP-40 and 0.1% SDS) with protease inhibitors(Roche). Western analysis was carried by loading equal amounts of pro-tein per treatment condition, determined by Bradford assay (Bio-Rad).Thesamples were electrophoresedthrough a 412% gradient Bis-Tris gel(NuPage) and transferred to PVDF (Thermo Scientific). After Westernblotting, membranes were blocked with 5% nonfat milk and subse-quently incubatedin rabbit polyclonal antibody to LRRK2 (1:1000, over-

night at 4C; Cell Signaling Technology), anti-GFP (1:50,000, 30 min at37C; Abcam), -actin, (1:20000, for 1 h at RT, Sigma-Aldirch), rabbitmonoclonal anti-LRRK2 (100-500), anti-LRRK2 P-ser935 or anti-LRRK2 P-ser910 (all 1:20,000, overnight at 4C, were kindly provided by Dr. Dario Alessi and Epitomics). Immunoreactivity was measured by probing with horseradish peroxidase-conjugated anti-rabbit (1:5000,Jackson ImmunoResearch Laboratory) and anti-mouse IgG (1:5000, Cal-biochem) followed by exposure to ECL detection reagents (PerkinElmer).

ResultsMutant LRRK2 in primary neurons recapitulates featuresof PDWe developeda neuronal model that recapitulates keyfeatures of PD and the progression of LRRK2-induced degeneration in indi-

vidual neurons. To visualize cells expressing LRRK2 in real time,we generated a Venus-tagged full-length LRRK2 construct(Venus-LRRK2; Fig. 1 A,B). By immunocytochemistry (Fig.1B), Venus fluorescence levels correlated significantly withLRRK2 levels in Venus-LRRK2-expressing neurons ( Fig. 1C )but not in neurons expressing only Venus ( Fig. 1D). Primary neurons had fivefold more transfected LRRK2 than endoge-nous LRRK2 (Fig. 1E ).

Neuronal processesdegenerate early in PD (Li et al., 2009).Todetermine whether LRRK2-expressing neurons display this PDphenotype, we examined two mutations: the most common PD-associated mutation, G2019S (Kay et al., 2006) and Y1699C.G2019S, Y1699C, or wild-type Venus-LRRK2 constructs were

cotransfected into primary cortical neurons with mRFP. As re-ported in vivo (MacLeod et al., 2006), G2019S or Y1699C-LRRK2expression led to shorter total neurite lengths in neurons than

Table1. Coxproportional hazards analysis ofthe effects ofexpressionleveland IBformationof LRRK2on neuronalsurvival

Group LRRK2 related variable HR 95% CI p value

Wild-type LRRK2(n 279)

IB formation 0.73 (0.3321.62) n.s.IB formation: time 1.01 (1.0031.02) 0.0065Venus-LRRK2 fluorescence (a.u.) 2.75 (1.6374.63) 0.0001

G2019S LRRK2

(n 305)

IB formation 0.39 (0.1421.09) n.s.

IB formation: time 1.01 (1.0011.03) 0.03Venus-LRRK2 fluorescence (a.u.) 4.03 (1.54910.48) 0.0043

Y1699C LRRK2(n 116)

IB formation 0.48 (0.1631.40) n.s.IB formation: time 1.01 (0.9971.03) n.s.Venus-LRRK2 fluorescence (a.u.) 5.75 (2.43913.57) 0.0001

Venus (n 182) Venus fluorescence (a.u.) 1 (1.01.0) n.s.CI, Confidence interval; IB, inclusion body; n, number of neurons; a.u., arbitrary unit; n.s., not significant; p 0.05.

420 J. Neurosci., January 8, 2014 34(2):418433 Skibinski et al. LRRK2 Levels and Synuclein Mediate LRRK2 To


4/16

wild-type Venus-LRRK2 or Venus alone. Wild-type Venus-LRRK2 expression was associated with shorter neurite lengthsthan Venus alone ( Fig. 1F ,G).

To determine whether mutant LRRK2 recapitulates corticaland midbrain degeneration, primary neurons were transfected

with Venus-LRRK2 and mRFP. Approximately 24 h after trans-fection, automated imaging (Arrasate et al., 2004) was used tocollect consecutive images of fluorescent neurons. Using this ap-proach hundreds of neurons, expressing Venus LRRK2 andmRFP were tracked individually every 1224 h for 7 d (Fig. 1H ).The mRFP signal was used to track the cells morphology throughout its lifetime. Abrupt loss of signal corresponded to aloss of membrane integrity, and the neurons death was recordedat that time ( Fig. 1H ). Time-of-death data from hundreds of neurons were then analyzed in a KaplanMeier survival model.In cortical neurons expressingVenusor wild-typeVenus-LRRK2alone, the cumulative risk-of-death was minimal. However, neu-rons expressing G2019S or Y1699C had a significantly greaterrisk-of-death and were 50% more likely to die than controls ( Fig.1I ; hazard ratio (HR) 1.5; p 0.0001; 95% CI 1.31.8 for G2019Sand 1.31.9 for Y1699C). In postnatally derived ventral midbrainneurons cotransfected withVenus-LRRK2 constructsand mRFP,midbrain neurons expressing Y1699C or G2019S had a signifi-cantly higher risk-of-death than wild-type-LRRK2-expressingneurons ( Fig. 1J; HR 1.3; p 0.0007; 95% CI 1.11.5 for G2019Sand HR1.3; p 0.0009; 95% CI 1.31.9 for Y1699C). However,compared with cortical neurons, ventral midbrain neurons weremore susceptible to death after transfection with wild-typeVenus-LRRK2 than Venus alone (Venus ctrl compared withwild-type Venus-LRRK2; HR 1.0; p not significant (n.s.); 95%CI 0.91.2 for cortical neurons and HR 0.8; p 0.004; 95% CI0.670.93 for midbrain neurons). To determine whether TH-

positive neurons were dyingspecifically from LRRK2 expression,we compared the percentage of TH-positive neurons in LRRK2-transfected midbrain neurons at 24 and 168 h post-transfection(Fig. 1K ,L). Neurons expressing wild-type LRRK2 had similarpercentages of TH-positive neurons at both times, indicatingthatTH-positive neurons expressing LRRK2 were dying, but at a ratesimilar to TH-negative neurons ( Fig. 1L). Neurons expressingmutant G2019S LRRK2 trended toward a lower percentage of TH-positive neurons at 168 h post-transfection ( Fig. 1L), sug-gesting that TH-positive neurons might be more vulnerable tomutant G2019S LRRK2.

IB formation is not required for mutant LRRK2-induced

toxicity PD-associated mutations in LRRK2 lead to misfolding and IBformation( Smith et al., 2005 ; MacLeod et al., 2006; Higashi et al.,

2007), common features of PD. However, the importance of LRRK2 IB formation in PD pathogenesis is unknown. In primary neurons, Venus-LRRK2 is diffuselydistributed in thecytosol andexcluded from the nucleus, as reported ( West et al., 2005; Fig.2 A). Yet, in a subset of neurons, wild-type, mutant Y1699C, or

G2019S LRRK2 formed heterogeneous intracellular IBs that wereinsoluble in 1% SDS/Triton X-100. Venus expressed alone wassolubilized by SDS (Fig. 2 A,B). Immunocytochemistry failed toobserve any colocalizationof LRRK2 inclusionswith -synuclein(data not shown). We then used our longitudinal imaging plat-form to track LRRK2-Venus IB formation in real time (Fig. 2C ).As death can preclude IB formation, to correctly estimate therisk-of-IB formation, death must be treated as a competingeffect.Neurons expressingY1699CVenus-LRRK2 wereat greater riskof IB formation than neurons transfected with wild-type or G2019SVenus-LRRK2 (Fig. 2D). The relative increase in IB formation inY1699C-expressing neurons was not due to different LRRK2 lev-els, as the expression of Venus-LRRK2 before IB formation was

similar forboth G2019S andY1699C ( Fig.2E ).In other modelsof neurodegeneration, the intracellular level of a disease proteinpredicts IB formation( Arrasate et al., 2004; Barmada et al., 2010 ).To determine whether this is true for LRRK2, Venus-LRRK2 lev-els were measured in neurons 24 h after transfection and beforeIB formation. Neurons were split into cohorts based on whetherand when an IB formed. Neurons that developed IBs earlierstarted with significantly higher LRRK2 levels than those thatformed IBs later or not at all (Fig. 2F ).

To determine whether LRRK2 IB formation was necessary forLRRK2-mediated toxicity, we identified neurons that formed in-clusions at 3648 h and compared their mean time to death toneurons with similar LRRK2 levels (Fig. 2G) but that did not

form an IB at 36 48 h. If IB formation is necessary for toxicity,neurons with IBsshoulddieearlier than those without. However,we found no difference in mean time to death in neurons with orwithout IBs (Fig. 2H ), even though the levels of LRRK2 were thesame in both groups.

Time-dependent Cox proportional hazard (CPH) analysiswas used to assess whether LRRK2 IBs contribute to toxicity, actas a cellular copingresponse or are an incidental event during thedisease progression. CPH analysis is predominantly used in clin-ical studies to determine the relationshipbetween an explanatory variable(s), such as IBs, and a specified endpoint (e.g., cell death;Fleming and Lin, 2000). One advantage of CPH analysis is theability to control potentially confounding effects by includingthe

potentially confounding effect (e.g., LRRK2 level) as another ex-planatory variable simultaneously in the CPH model. A secondadvantage is that allneurons canbe included in theCPHanalysis,

Table2. Coxproportional hazards analysis ofthe effects ofkinaseactivity and-synucleinon LRRK2-mediated, toxicity,or IB formationbeforeandafter adjustment forVenus-LRRK2 levels

Group Risk HR 95% CI p value

Adjusted for Venus-LRRK2 levels

HR 95% CI p value

G2019S D1994A (n 283) vs G2019S (n 181) Death 0.752 (0.59 0.95) 0.018 0.863 (0.66 1.12) 0.2IB formation 0.521 (0.231.17) 0.11 1.35 (0.45 4.07) 0.5

Y1699C D1994A (n 260) vs Y1699C (n 354) Death 0.807 (0.66 0.99) 0.044 0.895 (0.69 1.17) 0.4IB formation 0.360 (0.2 0.753) 0.007 0.529 (0.251.11) 0.0

G2019S/0.05uM LRRK2-IN1 (n 292) vs G2019S/vehicle (n 295) Death 0.874 (0.71.09) 0.24 0.926 (0.74 1.16) 0.50G2019S/0.1uM LRRK2-IN1 (n 345) vs G2019S/vehicle (n 295) Death 0.785 (0.63 0.98) 0.04 0.856 (0.68 1.08) 0.18G2019S/0.5uM LRRK2-IN1 (n 321) vs G2019S/vehicle (n 295) Death 0.738 (0.57 0.96) 0.023 0.835 (0.64 1.09) 0.1G2019S/sh_-SYN (n 160) vs G2019S/sh_scr (n 185) Death 0.757 (0.59 0.97) 0.028 0.792 (0.6 1.02) 0.0Y1699C/sh_-SYN (n 147) vs Y1699C/sh_scr (n 93) Death 0.716 (0.51 0.98) 0.042 0.738 (0.531.02) 0.0CI, Confidence interval; n, number of neurons.



5/16

Time (h)

Y1699C (462)

Wildtype (504)G2019S (690)

0.2.4.6.8

11.21.41.6

0 20 40 60 80 100 120 140 160 180

Venus (563)

**

0

.5

1

1.5

2

0 25 50 75 100 125 150 175 200 225

Y1699C (284)

Wildtype (391)G2019S (359)

Venus (182)

2.5

C u m .

R i s k o f d e a

t h ( a

. u . )

H

******

**

***

Time (h)

Cortical

Midbrain

A

0

200

400600

800

1000

1200

1400

G 2 0 1 9 S

V e n u s

W i l d t y p e

Y 1 6 9 9 C

T o

t a l n e u r i

t e l e n g

t h ( m

)

**

*

*

G2019S

Ctrl

24 h 48 h 72 h

96 h 120 h 144 h

I

Anti-LRRK2 immunoreactivity (a.u.) V e n u s

f l u o r e s c e n c e

( a . u . )

0

20

40

60

80

100

120

0 50 100 150 200 250

A LRRK2 Ab

Venus

Venus-LRRK2Wildtype

0

1

2

3

4

5

6

F o l

d d i f f e r e n c e

t o

e n

d o g e n o u s

L R R K 2

V e n u s

L R R K 2

CB

ED

F G

TH Ab

Venus-LRRK2J

V e n u s


( a . u . )

Anti-LRRK2 immunoreactivity (a.u.)

0

500

1000

1500

2000

2500

3000

0 50 100 150 200 250

p value = 0.9854r2 =2.731e-6

Venus

W i l d

t y p e

Y 1 6 9 9 C

G 2 0 1 9 S

V e n u s

-actin

LRRK2

GFP

42kDa

285kDa

285kDa

27kDa

LRRK2p value =


6/16

regardless of their LRRK2 level or if/when they formed an IB,which increases the power and sensitivity of the analysis.

To determine the role of IBs during LRRK2-mediated tox-icity in the CPH model, we included both the LRRK2 level andif/when an IB forms in each neuron. We found that, in neu-rons expressing mutant or wild-type LRRK2, IBs did not sig-nificantly enhance or reduce mutant LRRK2-induced toxicity

(Table 1), suggesting that IB formation is an incidental eventduring neurodegeneration.

Diffuse LRRK2 has a dose-dependent toxic effect in neuronsIn other neurodegenerative diseases, intracellular toxic speciesoccur outside IBs ( Arrasate et al., 2004; Barmada et al., 2010). Todetermine whether diffuse forms of mutant LRRK2 are responsi-ble for neurodegeneration, we measured Venus-LRRK2 levels inlive neurons at 2024 h post-transfection.Cohorts that expressedlow, medium or high levels of Venus-LRRK2 were followed pro-spectively (Fig. 3 A,B). Neurons with high levels of mutantLRRK2 died at a greater rate than those with low levels (Fig.3C ,D). Wild-typeLRRK2 alsoexhibiteddose-dependent toxicity,

which is consistent with the discovery of wild-type LRRK2 as arisklocusforPD( Satakeet al., 2009; Fig.3E ). No relationshipwasnoted between control Venus expression levels and survival, in-dicating that these effects were specific to LRRK2 (Fig. 3F ,G). Tovalidate that LRRK2 levels were proportional to cell death andensureno user bias in selectingtercilesof expression, allcells wereincluded simultaneously into a Cox analysis. This independentand complementary approach, demonstrated that the level of LRRK2 expressed in the cell was proportional to the cells risk of death. No relationship was found in cells expressing Venus alone(Table 1). The average levels of Venus-LRRK2 were similar inpopulations of cells expressing mutant or wild-type LRRK2 ( Fig.2E ), suggesting elevated steady-state mutant LRRK2 levels can-not explain mutant-LRRK2-induced toxicity.

Kinase-inactivating mutations modify toxicity by reducing LRRK2 levelsLRRK2 mutations are thought to provide a toxic gain of functionto LRRK2 that includes, at least for G2019S, increased kinase

activity. To test this, neurons were transfected with versions of wild-type, G2019S, and Y1699C LRRK2 with the D1994A muta-tion, which blocks kinase activity ( Lee et al., 2010; Fig. 4 A,B).Automated imaging was used to track individually transfectedneurons, and their time of death was recorded. As reported(Greggio et al., 2006; Smith et al., 2006 ; Lee et al., 2010), blockingkinase activity modestly but significantly reduced Y1699C or

G2019S toxicity (Fig. 4C E ). However, when we measuredVenus-LRRK2 levels within individual neurons, mutationswithinthe kinasedomainalso ledto reduced steady-state levelsof LRRK2 (Fig. 4F ). If LRRK2 levels and kinase activity correlatewith mutant-LRRK2-induced toxicity, but mutations inhibitingkinaseactivityreduce LRRK2 levels, which is theprimary cause of mutant LRRK2-induced toxicity?

To dissect out the independent contributions of kinase activ-ity and levels over the course of LRRK2-mediated neurodegen-eration, cohorts of neurons expressing either kinase-active ordead constructs were matched for similar LRRK2 levels. WhenLRRK2 levels were matched, cells expressing kinase dead con-structs no longer showed a reduction in toxicity ( Fig. 4GI ). In a

complementary approach that avoids any confound of excludinglow or high-expressing LRRK2 neurons, all neurons were incor-porated into a multivariate Coxmodel. By simultaneously incor-porating LRRK2 level and kinase activity into the model, wecould determine their independent contributions to neurondeath. Before incorporating LRRK2 levels into the Cox model,kinase-dead versions reduce neuronal toxicity; however, onceLRRK2 levelswere incorporatedinto themodel,any difference intoxicity disappeared ( Table 2). Thus, the reduction in LRRK2levels fully accounted for the improvement in survival that re-sulted from the disruption of LRRK2 kinase activity.

As reported ( Greggio et al., 2006; Kett et al., 2012), blockingkinase activity significantly reduced the risk-of-IB formation inneurons expressing Y1699C-LRRK2 but not in G2019S-LRRK2-expressing neurons ( Fig. 4 J ). As LRRK2 levels influence IB for-mation (Fig. 2F ) and are modulated by kinase activity (Fig. 4F ),the reduction in IB formation for Y1699C-LRRK2 could be dueto changes in LRRK2 levelsor kinaseactivity.BeforeadjustingforLRRK2 levels, blocking kinase activity reduced therisk-of-IB for-mation in neurons expressing Y1699C-LRRK2. When we con-trolled for LRRK2 levels, blocking kinase activity had no effect(Table 2).

Pharmacologically inhibiting LRRK2 kinase activity modulates LRRK2 stability and localizationTo exclude any confounding effects of kinase-inactivating muta-tions, we tested whether LRRK2 kinase activity is required for

mutant LRRK2-induced toxicity with acute pharmacological in-hibition. LRRK2-IN1, one of the most selective and potentknown LRRK2 kinase inhibitors, dephosphorylates LRRK2 atSer910 and Ser935 by an unknown regulatory feedback loop(Deng et al., 2011 ). By Western blot analysis in HEK293cells ( Fig.5 AC ) and by immunocytochemistry in neurons ( Fig. 5D), wefound that treating G2019S Venus-LRRK2-expressing cells withLRRK2-IN1 dephosphorylated Ser910 and Ser935 in a dose-dependent manner. As LRRK2-IN1 concentrations 1 M weretoxic (data not shown), lower doses (0.05, 0.1, or 0.5 M) weretested. In cells transfected with G2019S LRRK2, LRRK2-IN1 re-duced G2019S-LRRK2-induced toxicity (Fig. 5E ; Table 2).

As reported for LRRK2 kinase inhibitors ( Herzig et al., 2011),

increasingLRRK2-IN1 concentrations decreased the steady-statelevelsof G2019S LRRK2 (Fig. 5F ). Because LRRK2 concentrationis proportional to cell death (Fig. 3C E ), Cox analysis was used to

4

(Figurelegendcontinued.) traces(purplelines)ofneuronalprocessesfromimagesof neuronscotransfected with Venus or mutant Venus-LRRK2 and mRFP. Venus-only transfected neuronswere used as controls (Ctrl). At 48 h post-transfection, mRFP images were used to trace andquantify the total number and lengths of neuronal processes. G , Average total neurite lengthwas significantly lower in neurons transfected with the PD-associated LRRK2 mutants than inneurons expressing wild-type Venus-LRRK2 or Venus alone. Neurons expressing wild-typeLRRK2also hadshorter neurites thanVenus-expressing neurons. Morethan 65 neurons in eachgroup, two experiments combined, one-way ANOVA with Fishers PLSD post hoc tests, ** p0.001,* p 0.05;F 13.2.Errorbarsare95%confidenceintervals.H ,Longitudinalimagingof two neurons expressing mRFP and wild-type Venus-LRRK2. The top neuron remains alivethroughout the imaging period. The bottom neuron underwent neurite degeneration at 72 h(arrowhead)and then died by120 h (arrow). Neurons were 5 DIVat initiation oftracking.Scalebars, 10 m. I , J , Primary cortical (I ) or postnatally derived midbrain ( J ) neurons expressingLRRK2mutantsG2019S andY1699Chave a significantly greater cumulativerisk-of-death thanneurons expressing wild-type LRRK2. Neurons per group specified in brackets on the figures,

3 experiments combined, log rank test, ** p 0.001, *** p 0.0001. K , Representativeimagesof midbrain neurons transfected withwild-typeVenus-LRRK2and stained withTH 48 hlater.Neuronsare56DIVattimeoftransfection.Scalebar,10m.L, Histogramshowingthepercentageof LRRK2-expressingneurons thatare TH-positiveat 24 or 168 h post-transfection.Neurons expressing mutant LRRK2 G2019S trended toward a lower percentage of TH-positive

neurons at7 d.Thenumberofneurons ineachgrouprangedfrom65 to184,threeindependentexperiments combined, one-way ANOVA with Fishers PLSD post hoc tests; p 0.173, F 1.805. Error bars are 95% confidence intervals.



7/16

C u m .

r i s

k o

f I B f o

r m a

t i o n

( a . u . )

ns

GE

Y1699CWildtype

G2019S

F

0

0.1

0.2

0.3

0.4

0.50.6

I B > 4 8 h

I B a

t 3 5 h

I B a

t 4 8 h

D i f f u s e

***

***

*

Y 1 6 9 9 C

W i l d t y p e

G 2 0 1 9 S0

0.1

0.2

0.3

0.4

Time (h)20 40 60 80 100 120 140 160

.1

.2

.4

.6

.3

.5

.7

***

V e n u s


( a . u . )

V e n u s


( a . u . ) Wildtype

G2019S Y1699C

0

20

40

60

80

100

120

M e a n

t i m e

t o d e a

t h ( h )

Y 1 6 9 9 C

G 2 0 1 9 S0

0.1

0.2

0.3

0.4

0.5

0.6

Y 1 6 9 9 C

G 2 0 1 9 S

V e n u s


( a . u . )

H

D

DiffuseIB

Before SDS After 1% SDS

Venus-LRRK2

Y1699C

Venus

BVenus

Venus-LRRK2G2019S

Venus-LRRK2Wildtype

mRFP

Venus-LRRK2

Y1699C

Venus-LRRK2

Y1699C

A

24 h 36 h 48 h

RFP

C

Venus- Y1699CLRRK2

62 h

ns

DiffuseIB

nsns

Merged

Figure 2. IB formation depends on the PD-associated LRRK2 mutation and LRRK2 levels, but is not required for LRRK2-dependent cell death. A, Confocal images of primary cortical neurooverexpressing wild-type Venus-LRRK2 show diffuse cytosolic LRRK2 expression. Expression of PD-associated mutants G2019S or Y1699C LRRK2 leads to the fIBs(arrowheads).Neuronswerefixedat 48h post-transfectionat 7 DIV. Scalebars,10m.B, Detergent-resistance assayshowsthatIBs wereresistantto treatmentwith 1%SDS/1% TritScalebar,10m.C ,LongitudinalimagingofaneuronexpressingmRFPandmutantY1699CVenus-LRRK2,undergoingIBformation(arrow).Neuronswere5DIVatinitiati10 m.D ,Cumulativerisk-of-IBformationisgreaterinneuronsexpressingY1699CVenus-LRRK2thanthoseexpressingwild-typeVenus-LRRK2orG2019SVenus-LRgroup, 3 experimentscombined, usinga semiparametric proportionalhazardsmodel, thatincorporates thepresenceof competingrisk events,*** p 0.0001.E , Neurons expressingwild-type,G2019S,orY1699CVenus-LRRK2hadsimilarVenus-LRRK2levels.NumberofneuronsforY1699C121,G2019S308,andwild-type283,threeexperimentscombined,FishersPLSD post hoc tests, notsignificant(ns).Errorbars are95%confidenceintervals.F , Neurons that didnot form IBshad significantly lower LRRK2levelsthanneuronsthatformed LRRK2 IBsby 35Morethan32neuronsineachgroup,threeexperimentscombined,one-wayANOVAwithFishersPLSD posthoc tests,* p 0.01,*** p 0.0001;F 36.71.Errorbarsare95%confidenceintervalsG , Venus-LRRK2 levels are matched in cohorts of neurons expressing G2019S and Y1699C Venus-LRRK2. H , In neurons expressing G2019S or Y1699C Venus-LRRK2, the average time-neurons that form anIB at36 48h issimilar tothe average time-of-deathin neurons without anIB. More than 40 neurons pergroup, threeexperimentscombined, one-wPLSD post hoc tests. Error bars are 95% confidence intervals.



8/16

test the contribution of LRRK2 steady-state levels in reducingG2019S-induced toxicity mediated by LRRK2-IN1 exposure. Be-fore adjusting for LRRK2 levels, LRRK2-IN1 reduced mutantLRRK2-induced toxicity, but not after LRRK2 levels were con-trolled ( Table 2). In the case of Y1699C, increasing LRRK2-IN1

concentrations did not reduce toxicity and only trended towardreducing LRRK2 levels (Fig. 5G,H ), again supporting the impor-tance of LRRK2 concentration in predicting neuron death.

As reported (Dzamko et al., 2010; Deng et al., 2011), increasingLRRK2-IN1 concentrations caused increased LRRK2 IB formationfor G2019S and Y1699C LRRK2 (Fig. 5I , J ). In the case of G2019SLRRK2, the redistribution of LRRK2 from a diffuse form into IBsafter LRRK2-IN1 application associated with a lowering of diffuse

LRRK2 levels and a reduction in LRRK2-mediated toxicity. Theseresults suggest that the redistribution of LRRK2 into IBs might be aviable coping response for the cell to deal with diffuse LRRK2.

0.51

1.52

2.53

3.54

0 25 50 75 100 125 150 175 200Time (h)

High levels Y1699CMedium levels Y1699CLow levels Y1699C

0.51

1.52

2.53

3.5

4

0 25 50 75 100 125 150 175 200Time (h)

High levels WildtypeMedium levels Wildtype

Low levels Wildtype

C u m .

R i s k o

f d e a

t h ( a

. u . )

***

***

E

B

0

0.1

0.2

0.3

0.4

0.5

0.6

L o w

M e

d i u m

H i g h

L o w

M e

d i u m

H i g h

L o w

M e

d i u m

H i g h V

e n u s


( a . u . )

D

Wildtype

G2019S Y1699C

0.51

1.52

2.53

3.54

0 25 50 75 100 125 150 175 200

High levels G2019SMedium levels G2019SLow levels G2019S

C

Time (h)

A

RFP

Venus-LRRK2G2019S

High LowMedium

***

0.51

1.52

2.53

3.5

4

0 20 40 60 80 100120140160Time (h)

V e n u s


( a . u . )

nsns

0

0.5

1

1.5

2

2.5

H i g h

L o w

M e d

i u m

High levels VenusMedium levels VenusLow levels Venus

F G

C u m . R

i s k o

f d e a

t h ( a . u . )

C u m . R

i s k o

f d e a t h

( a . u . )

C u m . R

i s k o

f d e a t h

( a . u . )

Figure3. Higherlevelsofvenus-LRRK2aremoretoxic. A, Representativeimages of primary neurons expressingdifferentlevelsof G2019SVenus-LRRK2.In eachneuron,a quaof LRRK2 is made using the Venus-LRRK2 fluorescence. Scale bar, 10m. B, Average levels of Venus-LRRK2 after neurons were split into separate cohorts expressing low, mediumlevels. Error bars are 95% confidence intervals. C E , Neurons expressing higher levels of Y1699C Venus-LRRK2, G2019S Venus-LRRK2, or wild-type Venus-LRRK2 harisk-of-deaththanneuronsexpressinglowerlevels.NumberofneuronsforY1699C157,G2019S345neurons,andwild-type400neurons,3experimentscombined,logranktest,*** p0.0001.F ,Cohortsofneuronsexpressinglow,medium,andhighlevelsofVenus.Errorbarsare95%confidenceintervals.G ,Inneuronsexpressinghigher,medium,orlowlevelsofVenustherdifference in the cumulative risk-of-death. Number of neurons,180, 3 experiments, log-rank test.



9/16

By genetic and pharmacological ap-proaches, LRRK2 levels, kinase activity and IB formation are interrelated vari-ables.Theability to incorporate kinaseac-tivity and LRRK2 levels simultaneously into our model enabled us to show thatchanges in toxicity and IB formation are

better explained by effects on LRRK2 levelsthan kinase activity per se.

Removing endogenous synucleinalleviates mutant LRRK2-inducedtoxicity We then turned our attention to anotherfeature of PD-associated toxicity: -synucleinaccumulation (Lin et al., 2009), a cause of PD(Singletonetal.,2003;Satakeetal., 2009;Nakamura et al., 2011 ) that correlates withneuronal deficits in the substantia nigra(SN; Kovacs et al., 2008) and is toxic in adose-dependent manner ( Nakamura et al.,2011). Indeed, expression of G2019S- orY1699C-LRRK2 significantly increased en-dogenous cytoplasmic -synuclein levels(Fig. 6 A,B) as reported ( Lin et al., 2009;Sanchez-Danes et al., 2012 ), suggesting thatmutant-LRRK2-induced toxicity may re-quire -synuclein.

Given the functional redundancy of -,-, and -synuclein ( Chandra et al., 2004 ;

Anwaret al.,2011), webeganinvestigatingaroleof synuclein in LRRK2-mediated toxic-ity by transfecting Venus-LRRK2 into pri-mary cortical neurons from -synucleinTKO andwild-typematchedCtrlmice( Na-kamuraet al.,2011 ). Incontrastto Ctrlneu-rons, toxicity induced by overexpressingG2019Sand Y1699CLRRK2wasreducedinTKO neurons ( Fig. 6C ; HR 0.634; p 0.0001, CI0.5190.774 for G2019SandHR 0.506; p 0.0001; CI 0.4070.630 forY1699C). Toxicity from wild-type Venus-LRRK2 was also reduced in TKO neurons(Fig. 6D). However, TKO or Ctrl neuronstransfected with control Venus showed nodifference in baseline survival. Further-

more, removing endogenous synuclein didnot protect neurons from toxicity inducedby mutant huntingtin expression, anothertoxic aggregation-prone protein,suggestingthatthe benefits ofsynuclein lossare specificto LRRK2 (Fig. 6E ).

Next, we investigated whether remov-ing -synuclein alone was sufficient toreduce mutant LRRK2 toxicity. ShRNA-mediated knockdown of -synuclein, con-firmedbyimmunohistochemistry( Fig.6F ),was used to reduce -synuclein levels in ratneurons transfected with LRRK2. Longitudinal imaging revealed

that reduction of -synuclein alone reduced mutant-LRRK2-mediatedtoxicity ( Fig.6G; HR0.747; p 0.001,CI 0.6270.889 forG2019S andHR 0.699; p 0.0001; CI 0.5870.833).

LRRK2 patient-derived human neurons display

-synuclein-dependent toxicity To confirm the physiological relevance of our findings, neuronsweredifferentiatedfrom G2019S-LRRK2 PD patient-derivedand

0

.2

.4

.6

.8

1

1.2

0 20 40 60 80 100120140160

Wildtype D1994AWildtype

0.2.4.6.81

1.21.4

0 20 40 60 80 100120140160

Y1699C D1994A Y1699C

0.2.4.6.81

1.21.4

0 20 40 60 80 100120140160

G2019S D1994AG2019S

C u m . R

i s k o

f d e a t

h ( a

. u . )

Time (h)Time (h)

ns**

V e n u s


( a . u . )

00.20.40.60.81.01.21.4

G 2 0 1 9 S

G 2 0 1 9 S

D 1 9 9 4 A

W i l d t y p e

W i l d t y p e

D 1 9 9 4 A

Y 1 6 9 9 C

Y 1 6 9 9 C

D 1 9 9 4 A

***

***

***

0

0.2

0.4

0.6

0.8

1.0

0

0.2

0.4

0.6

0.8

1.0

1.2

V e n u s


( a . u . )

V e n u s


( a . u . )

ns ns

0.2.4.6.81

1.21.4

0 20 40 60 80 100120140160Time (h)

G2019S D1994AG2019S

ns

G 2 0 1 9 S

G 2 0 1 9 S

D 1 9 9 4 A

Y 1 6 9 9 C

Y 1 6 9 9 C

D 1 9 9 4 A

0.2

.4

.6

.81

1.21.4

0 20 40 60 80 100120140160Time (h)

C u m .

r i s k o f

d e a

t h ( a

. u . )

ns Y1699C D1994A Y1699C

B

ED

HGF

0 . 0

0 . 1

0 . 2

0 . 3

C u m . r i s

k o

f I B f o r m a

t i o n

( a . u . )

Y1699CG2019S

Y1699C D1994AG2019S D1994A

ns

***

I

Time (h)

0 50 100 150Time (h)

pan

pS935pS910

W i l d t y p e

Y 1 6 9 9 C

G 2 0 1 9 S

W i l d t y p e

D 1 9 9 4 A

Y 1 6 9 9 C

D 1 9 9 4 A

G 2 0 1 9 S D 1 9 9 4 A

A

0

0.2

0.4

0.6

0.8

1

1.2

1.4

p 9 1 0 / p a n

L R R K 2

***

Wildtype

G2019S Y1699C

Wildtype D1994A

G2019S D1994A Y1699C D1994A

C

J

C u m . R

i s k o

f d e a t

h ( a

. u . )

C u m . R

i s k o

f d e a t

h ( a

. u . )

C u m . R

i s k o

f d e a t

h ( a

. u . )

Figure 4. Kinase-inactivating mutations reduce mutant LRRK2-induced toxicity by reducing LRRK2 levels. A, RepresentativeWesternblotsoflysatesfromHEK293cellstransfectedwithwild-type,G2019S,orY1699CVenus-LRRK2construthe D1994A mutation and immunoblotted for the detection of LRRK2 phosphorylation at Ser910 (top) and Ser9total LRRK2 (pan LRRK2, bottom). B, Quantification of phosphorylated LRRK2 at Ser910 normalized to total pan LRindependentexperimentscombined, MannWhitneytest,wild-typeversuswild-typeD1994A (S910*** p 0.0495; Z 1.964),G2019SvsG2019SD1994A(S910*** p 0.0495,Z 1.964),Y1699CversusY1699CD1994A(S910 p 0.8273; Z 0.218).ErrorbarsindicatetheSD.C ,D ,BeforecontrollingforLRRK2levels,cumulativerisk-of-deathcurvesshowthattheD1994Amto a significant reduction in toxicity in neurons expressing the mutant G2019S and Y1699C Venus-LRRK2; numG2019S 181,G2019SD1994A283,Y1699C354,andY1699CD1994A260,three experimentscombined,log ranktest,* p 0.01.E ,Graphshowingnosignificantdifferenceinthesurvivalofneuronsexpressingwild-typeVenus-LRRK2

the D1994A mutation. Number of neurons for wild-type392 and wild-type D1994A277, three experiments combined, logrank test. F , Average expression levels of LRRK2 constructs with or without the D1994A mutation in primary neuneurons as in C , D , three experiments combined, one-way ANOVA with Fishers PLSD post hoc tests, *** p 0.0001; F 22.91.Error bars are 95% confidence intervals. G , Neurons expressing mutant Y1699C or G2019S Venus-LRRK2 with or wD1994AmutationwerematchedforLRRK2levels.NumberofneuronsforG2019S170,G2019SD1994A138,Y1699C297,and Y1699C D1994A166, three experiments combined.H , I , After matching for LRRK2 levels, cumulative risk-of-deathshow neurons expressing Y1699C and G2019S with or without D1994A exhibit no significant difference in toneuronsasinG , threeexperimentscombined,log ranktest. J , Beforecontrolling forLRRK2levels,cumulative risk-of-IB focurvesshowthat theD1994A mutationsignificantly reduces therisk-of-IB formationof mutantY1699C,but notmNumberofneuronsforG2019SandG2019SD1994A(449)andforY1699CandY1699CD1994A(565),threeexperusing a semiparametric proportional hazards model that incorporates the presence of competing risk events, *** p 0.0001.



10/16

iPSCs,where LRRK2expression is under thecontrol of itsendog-enous promoter. Differentiated cells were fixed and stained formicrotubule-associated protein 2 (MAP2), TH, and DAPI ( Fig.

7 A). The percentage of neurons in the culture was calculated asthe percentage of cells with MAP2 and DAPI overlap (Fig. 7B).The percentage of dopaminergic neurons was calculated as the

percentage of MAP2-positive cells withboth MAP2 and TH staining ( Fig. 7C ).In parallel, differentiated cells transfectedwitha mApple-Map2 fluorescent reporterwere imaged and followed for at least 7 d(Fig. 7D). Tracked cells displayed neuro-nal features, including at least two pro-

jecting processes. In parallel, separatecultures were fixed, and 3040% of thesemorphologically similar cells stained pos-itive for TH (data not shown). Humancells harboring the G2019S mutation hada higher cumulative risk-of-death thancontrol cells for G2019S without the ap-plication of exogenous stressors (Fig. 7E ;HR 1.4; p 0.018; 95% CI 1.11.9). Todetermine whether the neurons that wewere tracking were TH-positive neuronsand that they were specifically dying, wecompared the percentage of transfected

neurons that were TH-positive at 24 and168 h post-transfection. In human cellsharboring the G2019S mutation and con-trol cells, there was a trend toward a lowerpercentage of TH-positive neurons at168 h post-transfection ( Fig. 7F ), indicat-ing that TH neurons were indeed dyingand potentially at a faster rate than TH-negative neurons.

In addition to recapitulating mutant-LRRK2-induced cell loss, DA neurons de-rived from G2019S iPSCs accumulatedmore -synuclein than control neurons(Fig. 7G,H ). As in primary rodent neu-rons, removal of -synucleinreducedtox-icity in iPSC-derived neurons with theG2019S LRRK2 mutation (Fig. 7I ; HR 0.676; p 0.03; 95% CI 0.50.96).

-Synuclein knockdown did not affectsurvival in an independent control line(Fig. 7I ; HR 1; p 0.83; 95% CI 0.71.5).

Removal of endogenous synucleinreduces levels of diffuse mutant LRRK2In rodent and human cells, synuclein isrequired for LRRK2-mediated toxicity.As

-synuclein accumulation disrupts pro-

teostasis and its removal enhances it(Cuervo et al., 2004; Fornai et al., 2005),synuclein might mechanistically mediatemutant LRRK2 toxicity by modulatingLRRK2 levels. In TKO neurons overex-pressing LRRK2, mutant and wild-typeLRRK2 levels were lower than in Ctrl neu-rons( Fig.8 A,B). Removingsynucleinwasspecific to LRRK2: there was no differencein Venus levels in neurons expressing only

Venus ( Fig. 8C ). Knocking down -synuclein in rat neurons over-expressing mutant LRRK2 was also associated with a reduction inmutant LRRK2 levels (Fig. 8D), indicating that the removal of

-synucleinalone wassufficienttomodulate LRRK2levels.We wereunable to detect an effect of knockdown of -synuclein caused onendogenous levels of LRRK2 (Fig. 8E ).

0

0.2

0.4

0.6

0.8

1

0 20 40 60 80 100 120 140 160

vehicle, Wildtype0.5uM LRRK2-IN1, G2019S0.1uM LRRK2-IN1, G2019S0.05uM LRRK2-IN1, G2019Svehicle, G2019S F

**nsns

0.2.4.6.81

1.21.41.6

0.05 0.1 0.50LRRK2-IN1 ( M)

***

**

V e n u s


( a . u . )

0

20

40

60

80

100

%

ibd/ibd

0.05 0.1 0.50LRRK2-IN1 ( M)

H

0

20

40

60

80

100

%

0.05 0.1 0.50LRRK2-IN1 ( M)

0

.2

.4

.6

.8

1

1.2

1.4

0.05 0.1 0.50LRRK2-IN1 ( M)

V

e n u s


( a . u . )

ns

I J

0

0.2

0.4

0.6

0.8

1

1.2

0 20 40 60 80 100 120 140 160

vehicle, Wildtype0.5uM LRRK2-IN1, Y1699C0.1uM LRRK2-IN1, Y1699C0.05uM LRRK2-IN1, Y1699Cvehicle, Y1699C

Time (h)

Time (h)

C u m .

R i s k o

f d e a

t h ( a

. u . )

*

G

ns * *G2019S Y1699C

ibd/ibd

BA C

0

0.2

0.4

0.6

0.8

1

1.2

0.05 0.1 0.50LRRK2-IN1 ( M)

p S 9 1 0 / p a n

L R R K 2

pan pS910

pS935 D M

S O

0 . 1 M

0 . 0 5 M

0 . 5 M

G2019S Venus-LRRK2*

**

0

0.2

0.4

0.6

0.8

1

1.2

0.05 0.1 0.50LRRK2-IN1 ( M)

p S 9 3 5 / p a n

L R R K 2

*****

*

D

0

0.2

0.4

0.6

0.8

1

1.2

p S 9 3 5 / p a n

L R R K 2

0.05 0.10LRRK2-IN1 ( M)

****

E

C u m . R

i s k o

f d e a

t h ( a

. u . )

Figure 5. Pharmacological inhibition of LRRK2 kinase activity reduces LRRK2 toxicity by modulating IB formation and LRRK2levels. A, Western blotof HEK293 cell lysates transfected withG2019S Venus-LRRK2, treated withDMSO or increasingconcentra-tions of LRRK2-IN1 for 2 h, and extracted for immunoblotting to detect LRRK2 phosphorylation at Ser910 (top) and Ser935(middle), and total LRRK2 (bottom). B, C , Western blot analysis was used to take measurements of LRRK2 phosphorylation atSer910 and Ser935 normalized against total LRRK2. Three experiments combined, MannWhitney test, 0M versus 0.05M(S910andS935 p 0.0495; Z 1.964), 0 Mversus 0.1M(S910andS935 p 0.0495; Z 1.964), 0 Mversus 0.5M(S910and S935 p 0.025; Z 2.236). Error bars indicate SD of the mean. D , Immunocytochemistry measurements of LRRK2 phos-phorylation at Ser935 normalized against total LRRK2, from individual rat cortical neurons overexpressing G2019S Venus-LRRK2.Morethan30neuronspergroup,one-wayANOVAwithFishersPLSD posthoc tests,* p 0.01,*** p 0.0001;F 14.579.Errorbarsindicate95%confidenceintervals.E ,ApplicationofincreasingconcentrationsofLRRK2-IN-1trendstowardsareductioninthecumulativerisk-of-deathofneuronsexpressingmutantG2019SLRRK2.Numberofneuronsforvehicle295,0.05 M 292,0.1

M 345 and 0.5 M 321, three experiments combined, log rank test, ** p 0.01. F , G , Average Venus-LRRK2 levels

measured in neurons with and without LRRK2-IN1 treatment. F , Increasing doses of LRRK2-IN1 caused a dose-dependent reduc-tion in G2019S-Venus-LRRK2 levels and G , no significant difference was found with Y1699C Venus-LRRK2. Number of neuronssame as E , H , three experiments combined, one-way ANOVA with Fishers PLSD post hoc tests, G2019S (* p 0.05, ** p 0.01;F 4.659) and Y1699C (not significant (ns); F 1.256). Error bars are 95% confidence intervals. H , Application of increasingconcentrationsof LRRK2-IN1caused no differencein riskof deathfor mutant Y1699C-expressingneurons.Number of neurons forvehicle 274,0.05 M 222,0.1 M 238,and0.5M 231,threeexperimentscombined,logranktest,* p 0.05.I , J ,Bargraphsshowingthepercentageof cellswithdiffuse mutantG2019Sor Y1699CLRRK2 expression(d),an IB(ib)or that donot havea clearly defined IB, but show signs of an IB precursor (d/ib) at 24 h post-transfection. Increasing concentrations of LRRK2-INincrease the percentage of neurons with IBs for both mutants. Number of neurons is same as above in E , H , three experimentscombined, KruskalWallis test, G2019S (* p 0.042, H 8.231) and Y1699C (* p 0.038, H 8.426).



11/16

00.25

0.50.75

11.25

1.51.75

22.25

0 20 40 60 80 100 120140 160

Ctrl, Htt 586-Q17-GFP

TKO, Htt 586-Q17-GFPTKO, Htt 586-Q136-GFP

Ctrl, Htt 586-Q136-GFP

Time (h)

ns

ns

* * * *

00.20.40.60.8

11.21.41.6

0 20 40 60 80 100 120 140 160

scr-shRNA, G2019Sa-syn-shRNA, G2019Sscr-shRNA, Y1699Ca-syn-shRNA, Y1699C

Time (h)

F

* * * * * *

Venus

Venus-

LRRK2Wildtype

Venus-LRRK2G2019S

Venus-LRRK2

Y1699C

a-SYN AbmRFP

0

0.2

0.4

0.6

0.8

1

1.2

1.4

N o r m a l

i z e d a -

S Y N l e v e

l s **G

s h_ s

c r

s h_ a

S Y N

D

0.25

.5.75

11.25

1.51.75

22.25

0 20 40 60 80 100 120140 160

TKO, WildtypeCtrl, Wildtype

TKO, VenusCtrl, Venus

C u m . R

i s k o

f d e a

t h ( a

. u . )

Time (h)

ns **

A

00.20.40.60.8

11.21.41.6

G 2 0 1 9 S

V e n u s

Y 1 6 9 9 C

N o r m a

l i z e

d e n d o

g e n o u s

s y n u c

l e i n l e v e l s

( a . u . ) *

***

W i l d t y p e

ns

B

C

TKO, Y1699CCtrl, Y1699CTKO, G2019SCtrl, G2019S

C u m .

R i s k o

f d e a

t h ( a

. u . )

0 20 40 60 80 100 120140 160Time (h)

***

E F

C u m .

R i s k o

f d e a

t h ( a

. u . )

C u m .

R i s k o

f d e a

t h ( a

. u . )

00.25

0.50.75

11.25

1.51.75

2

Figure 6. Synucleinis required for mutant LRRK2-induced toxicity. A, Representative images of primary cortical neurons transfected withVenus, wild-type, or mutant Venus-LRRlaterandstainedwith-synuclein.Neuronsare7DIV.Scalebar,10m.B,Averagelevelsofendogenous-synuclein(determinedbyanti--synucleinimmunocytochemistry) incortical neurotransfected with wild-type, Y1699C, or G2019S Venus-LRRK2. Values are normalized relative to-synuclein levels in neurons expressing a Venus control. More than 100 neurons per gexperimentscombined, one-way ANOVAwithFishersPLSD post hoc tests,* p 0.05,*** p 0.0001;F 7.707.Errorbarsare95%confidenceintervals.C , Cumulativerisk-of-deathcurves showthat TKOneuronsexpressing Y1699Cor G2019Shad significantly lowertoxicitythanCtrlneurons.More than 245neuronsper group,threeexperiments combined,log r p 0.0001.D ,Cumulative risk-of-death curves show that TKO neurons expressing wild-type Venus-LRRK2 had significantly lower toxicity than Ctrl neurons. Ctrl and TKO neurohadno difference insurvival.Morethan180 neurons pergroup, threeexperiments combined,log rank test,** p 0.001.E , Ctrl andTKO neurons transfected with mutantHtt586 -Q136-GFP showno difference in toxicity in Ctrl or TKO neurons. More than 120 neurons per group, two experiments combined, log rank test, * p 0.01, *** p 0.0001. F , Quantification of shRNA knockdown endogenous-synucleinlevels.Primaryrat cortical neurons weretransfectedwithshRNAagainst-synucleinor shRNAscrambled control.Forty-eighthoursafter transfectioncellswerstained with-synuclein. Neurons were 7 DIV when fixed. More than 34 neurons per group, two experiments combined, unpaired t test, ** p 0.001; t 3.688. Error bars are 95% confidenceintervals.G , Cumulativerisk-of-deathcurvesshowthatin rodentneuronsknockdownof -synucleinby transfection ofshRNAconstructs (a-syn-shRNA)significantlyreducemutant LRcompared with neurons transfected with scrambled shRNA (scr-shRNA). More than 150 neurons per group, three experiments combined, log rank test, *** p 0.0001.



12/16

As removing -synuclein reduces toxicity and levels of LRRK2, and given LRRK2 levels role as a powerful predictor of neuron death (Figs. 35), the benefit of removal might be ex-plained by -synucleins effect on LRRK2 levels. Before control-ling for LRRK2 levels, cells overexpressing mutant LRRK2 withknockdown of -synuclein had lower toxicity. However, onceLRRK2 levels were controlled for, there was no difference in tox-icity between cells with or without -synuclein ( Table 2). By

controlling for LRRK2 levels, we show that -synuclein is re-quired for LRRK2-induced toxicity and its effect is mediated by modulating LRRK2 levels.

DiscussionOurbasiccellularsystems modelof PD ledto several novel and significant findingsabout theroleof LRRK2-related eventsdur-ing neurodegeneration. Using an imagingplatform that longitudinally tracks primary neurons throughout their lifetime, we es-tablished rodent and human primary neu-ron models of LRRK2 that recapitulatefeatures of PD-associated LRRK2 neurode-generation, including neurite retraction,

-synuclein accumulation, and mutantLRRK2-induced neuron loss. In contrast toprevious studies ( Byers et al., 2011; Nguyenet al., 2011; Liu et al., 2012; Reinhardt et al.,2013), in our human PD model, we dem-onstrated neuronal loss without an exog-enous cell stressor. Cellular featuresassociatedwithLRRK2-mediated toxicity,including LRRK2 levels, IB formation,

and kinase activity, were tracked andlinked to cell fate.Using statistical survivaltools, we built a predictive, quantitativemodel of LRRK2 that measured the rolesand importance of these cellular featuresduring neurodegeneration, independently of each other.

The cellular concentrations of diffusemutant and wild-type LRRK2 strongly predicted cell death. Past studies relied onWestern blotting to measure LRRK2 lev-els, which were often performed in celllines that are substantially different from

neurons (LePage et al., 2005; Tsvetkov etal., 2010). Furthermore, given LRRK2subiquitous expression and linkage to asubset of diseases (Franke et al., 2010), itsfunction and metabolism may be cell-type-specific. Finally, measuring proteinlevels by population-based assays lackssingle-cell resolution. Cell-to-cell varia-tion inthe LRRK2 doseis treated as a tech-nical rather than biological variation,reducing sensitivity to detect phenotypes(e.g., celldeath) acrosscell populations. Incontrast, we measured the LRRK2 dose ineach neuron and related it to that cellsfate, harnessing cell-to-cell variability toenhance the sensitivity and power to re-solve doseresponse relationships.

Measuring LRRK2 levels to predict celldeath wascritical forinvestigatingtherole of

LRRK2kinaseactivityduringneurodegeneration.Geneticandphar-macologic blockade of kinase activity reduced mutant LRRK2 tox-icityandLRRK2levels.However, unlikepreviousstudies ( Greggioetal., 2006; Lin et al., 2009; Lee et al., 2010; Herzig et al., 2011), oursystems biology approach enabled us to measure the independentcontributions of LRRK2 levels and kinase activity during cell death,even though they are intertwined and a change in one affects the

other. Critically, we show that disrupting kinase activity reducedmutantLRRK2 toxicity by affecting LRRK2 levels rather thankinaseactivity per se.

50h 72h 96h

120h 160 h 184hC

D

IG

CtrlG2019S

C u m

R i s k o f

D e a

t h ( a

. u . )

0

.2

.4

.6

.8

1

1.2

Time (h)0 25 50 75 100125150175200225

G2019S (n=177)Ctrl (n=128)

4 experiments combined

00.20.40.60.8

11.21.41.6

Ctrl G2019S N o r m a l

i z e d a -

S Y N l e v e

l s

00.20.40.60.8

11.21.4

1.6

0 25 50 75 100 125 150 175 200 225

G2019S/sh_aSYN (82)G2019S/sh_scr (109)Ctrl/sh_aSYN (125)Ctrl/sh_scr (125)

Time (h)

2 experiments combined*

*

E

*

*

G2019S

MAP2 AbTH Ab DAPIA

0

10

20

30

40

50

60

M A P 2 p o s i

t i v e c e

l l s ( % )

B

a-SYN Ab TH Ab DAPI

0

10

20

30

40

50

60

T H p o s i

t i v e n e u r o n s

( % )

H

Ctrl G2019S Ctrl G2019S

Ctrl 24 h (37)

G2019S 24 h (46)

Ctrl 168 h (30)

G2019S 168 h (40)

20

10

30

40

0 % o

f t r a n s

f e c

t e d c e

l l s

t h a t a r e

T H - p o s

i t i v e

F

C u m

R i s k o

f D e a

t h ( a

. u . )

Figure7. LRRK2patient-derived neurons display-synuclein-dependent toxicity. A, Representativeimages fromimmunocy-tochemistryof neurons differentiatedfromiPSCs froma control individual,costained withTH, MAP2,and DAPI.Scalebar,20m.B, The percentage of neurons in the culture was calculated as the percentage of cells with MAP2 and DAPI overlap. C , ThepercentageofdopaminergicneuronswascalculatedasthepercentageofMAP2-positivecellswithcolocalizationofbothMAP2andTHstaining.D ,LongitudinaltrackingofsingleneuronsdifferentiatedfromiPSCsandtransfectedwithamApple-Map2fluorescent

reporter. The time of death is captured for each neuron that is followed. The cell indicated with a red arrowhead dies by 160 h.Degenerationof neuriteswas observedat 120h andcomplete soma loss observed at160 h.In contrasttheneuronindicatedbythegreenarrowhead,livesthroughoutthelengthoftheexperiment.Scalebar,20m.E ,NeuronsderivedfrompatientsharboringtheLRRK2 mutation G2019S had a greater risk-of-death than controls. Cells per group specified in brackets on the figure, four exper-iments combined, log rank test, * p 0.05. F , Histogram showing the percentage of mApple-transfected neurons that areTH-positive at 24 or 168 h post-transfection. TH-positive neurons expressing mutant LRRK2 G2019S trended toward a lowerpercentageofTH-positiveneuronsat6d.Approximately3040neuronsineachgroup,twoindependentexperimentscombined,one-wayANOVAwithFishersPLSD posthoc tests, p 0.246;F 1.563.Errorbarsare95%confidenceintervals.G ,Representativeimagesof TH-positiveneuronsderived fromG2019S LRRK2 andcontrol iPSCsindicateelevatedendogenous-synuclein staininginG2019SLRRK2 patient cells.Scalebar,10m.H , Average levels of endogenous-synuclein(determined by anti--synucleinimmunocytochemistry) in TH-positive neurons are significantly higher in neurons differentiated from G2019S LRRK2 iPSCs com-pared with control iPSCs. Values are normalized relative to-synuclein levels in control TH neurons. More than 100 neurons pergroup, three experiments combined, unpaired t test, * p 0.01; t 2.692. Error bars are 95% confidence intervals. I , Reducing

-synuclein levels in neurons derived from patients harboring the LRRK2 mutation G2019S caused a significant reduction inrisk-of-death compared with those with a scrambled control shRNA. Reducing-synuclein levels in neurons derived from unaf-fected individuals showed no significant difference in survival compared with those with a scrambled control shRNA. Bracketsindicate cells per group, two experiments combined, log rank test, * p 0.05.



13/16

We also found neurons with higherLRRK2 levels have a greater propensity toform IBs. However, when LRRK2-IN1was used to inhibit LRRK2 kinase activity,IB formation increased, and LRRK2 levelsdecreased. LRRK2-IN1 disrupts the bind-ing of LRRK2 to 14-3-3 protein isoforms

by dephosphorylating Ser910 and Ser935leading to increased LRRK2 aggregation(Deng et al., 2011). Reducing the 14-3-3-LRRK2 interaction destabilizes the LRRK2structure ( Nichols et al., 2010), promotingsequestration of diffuseLRRK2intoIBsandreducing diffuse LRRK2 levels elsewhere inthe cell. Unlike G2019S, LRRK2-IN1 didnot reduce Y1699C-mediated toxicity. Asthe Y1699C mutation itself impairs thebinding of LRRK2 to 14-3-3 (Nichols et al.,2010) and has a greater propensity to formIBs,LRRK2-IN1causesa smaller increasein

IBformation thanthe G2019Smutant, lead-ingto lowerreductionsinLRRK2levels.Al-though LRRK2 levels were a strongerpredictor of toxicity for both mutants, thedistinct differences of the mutants for IBformation and 14-3-3 binding indicatemechanistic differences between the PD-associated mutants that may be relevant fortherapeutic development.

In the absence of LRRK2-IN1, LRRK2IBs neither promoted nor retarded mutant LRRK2 toxicity. Thismay be due to their inefficiency in reducing toxic diffuse LRRK2levels elsewhere in the cell. In other neurodegenerative disorders,IBs are well formed and help sequester the diffuse toxic speciesinto IBs, so that the IBs act as a coping response (Arrasate et al.,2004). Therefore, in addition to therapeutics that clear LRRK2,ones that help promote more efficient sequestration of toxic dif-fuse mutant LRRK2 into more compact IBs may be a viable ther-apeutic strategy against LRRK2-mediated toxicity.

Elevated levels of mutant LRRK2 also caused the accumu-lation of the PD-associated protein -synuclein. We foundremoving endogenous synuclein significantly reduced mutantLRRK2-induced toxicity and suggested a reciprocal link between

-synuclein and LRRK2 in PD-related neurodegeneration. Todemonstrate that these key observations were due specifically tothe PD-associated mutations in LRRK2, complementary experi-ments were performedin human dopaminergicneurons with the

endogenous LRRK2 G2019S mutation and in murine neuronswith ectopic expression of mutant LRRK2. In murine neurons,thegenetic background was identical in neurons expressing wild-type or mutant LRRK2, so that any increased risk of death couldbe attributed to the PD-causing mutation in LRRK2. Similarly,the reduction of LRRK2 G2019S-induced neurodegeneration by synuclein knockdown in human dopaminergic neurons was sup-ported with an analogous murine experiment in which the geneticbackground was fully controlled. Furthermore, the knockdown of

-synuclein in human dopaminergic neurons with the LRRK2G2019S mutation only rescued survivalback to levels of thecontrolline,and hadno effect onsurvivalof the control line,consistentwiththe idea that the increased risk of death in the LRRK2 G2019Sline is

a LRRK2 mutation-dependent effect.How the removal of -synuclein facilitates LRRK2-inducedneurodegeneration is unclear. Evidence from LRRK2 and sy-

nuclein knock-out models indicates that these two proteins haveat least some normal functions that are independent of eachother, at least in some cell types. For example, -synuclein KOmice display cognitive impairments (Kokhan et al., 2012) and

subtle evidence of abnormalities in neurotransmission (Abeliov-ich et al., 2000; Cabin et al., 2002), whereas LRRK2 knock-outmice showed neither dopaminergic degeneration nor alterationsin dopamine dynamics (Andres-Mateos et al., 2009 ; Hinkle et al.,2012; Tong et al., 2012) but display age-dependent alterations inautophagic activity in the kidneys. In addition, we observed nodetectable change in endogenous LRRK2 levels with knockdownof -synuclein.

However, this does not go against the possibility that duringneurodegeneration, triggered by PD causing mutations inLRRK2, the actions of LRRK2 become critically linked to

-synuclein, at least in neurons. For example, even though-synuclein knock-out mice are resistant to MPTP-induced tox-

icity (Dauer et al., 2002) and LRRK2 KO mice show no responseto this toxin (Andres-Mateos et al., 2009 ), LRRK2 and MPTPmight act on independent pathways to cause neurodegenerationbut both require synuclein.

In the disease state, -synuclein could mediate LRRK2-mediated toxicity by modulating LRRK2levels, which are toxic ina dose-dependent manner ( Fig. 3). Without -synuclein, mutantLRRK2 levels were significantly lower. Because -synuclein dis-rupts protein homeostasis pathways ( Snyder et al., 2003; Cuervoet al., 2004; Fornai et al., 2005) and removing -synuclein bol-sters them (Fornai et al., 2005), synuclein might hinder degrada-tion and handling of mutant LRRK2, thereby leading to LRRK2saccumulation and toxicity. However, the change in LRRK2-

mediated toxicity beforeandaftercontrolling forLRRK2levels inthe absence of -synuclein, although significant, was small. Thelowering of LRRK2 levels is likely one of multiple contributing

V e n u s

f l u o

r e s c e n c e

( a . u . ) **

CB

C t r l

T K O

0

0.2

0.4

0.6

0.8

1

1.2

0

0.2

0.4

0.6

0.8

1

1.2

C t r l

T K O

V e n u s

f l u o

r e s c e n c e

( a . u . ) ns

s h_ s

c r

s h_ a

S Y N

s h_ s

c r

s h_ a

S Y N

G2019S Y1699C

*

*D

0

0.2

0.4

0.6

0.8

1

1.2

V e n u s

f l u o

r e s c e n c e

( a . u . )

C t r l

T K O

A

C t r l

T K O

** ***

0

0.2

0.4

0.6

0.8

1

1.2

V e n u s

f l u o

r e s c e n c e

( a . u . )

1.41.4 G2019S Y1699C

E n

d o g e n o u s

L R R K 2 ( a

. u . )

0.0

0.4

0.8

1.2

1

0.6

0.2

s h_ s

c r

s h_ a

S Y N

ns

E

Figure 8. Loss of -synuclein reduces LRRK2 levels. A, Venus-LRRK2 levels of the Y1699C and G2019S LRRK2 mutsignificantly lower in TKO neurons than in Ctrl mouse neurons. Number of neurons for G2019S204, G2019S TKO251,Y1699C 160,Y1699CTKO194,three experimentscombined, one-way ANOVA withFishersPLSD post hoc tests,** p 0.01,*** p 0.0001; F 8.674. Error bars are 95% confidence intervals. B, Histogram showing quantification of Venus-LRRfluorescence in Ctrl and TKO neurons cotransfected with wild-type Venus-LRRK2. TKO neurons had significVenus-LRRK2asmeasuredbyVenusfluorescence.Morethan271neuronspergroup,threeexperimentscombined,t test,

** p 0.002, t 3.179. Error bars are 95% confidence intervals. C

, In Ctrl and TKO neurons cotransfected with Venus, levVenuswere unchanged.More than188 neurons pergroup,three experimentscombined, unpairedt test, p 0.11,t 1.6.Errorbars are 95% confidence intervals.D , Venus-LRRK2 levels were also significantly lower in rat neurons with knockdowenous -synuclein(sh_aSYN)versus scrambledshRNA(sh_scr).Numberof neurons G2019S/sh_scr185, G2019S/sh_aSYN160, Y1699C/sh_scr93, Y1699C/sh_aSYN147, two experiments combined, one-way ANOVA with Fishers PLSD post hoc tests, * p 0.05;F 3.841.Error barsare 95%confidenceintervals.E , Endogenous LRRK2levelswereunchanged inrat neurwith knockdown of endogenous -synuclein (sh_aSYN) versus those with scrambled shRNA (sh_scr). Number osh_aSYN 189, sh_scr 180, three experiments combined, one-way ANOVA, p 0.153; F 2.056. Error bars are 95%confidence intervals.



14/16

mechanisms by which synuclein removal protects neurons fromLRRK2 toxicity. As -synuclein is toxic in a dose-dependentmanner ( Singleton et al., 2003; Nakamura et al., 2011 ), LRRK2-mediated toxicity may directly result from -synuclein build-up.However, the increase in -synuclein by LRRK2 is probably in-sufficient to fully explain the toxicity of mutant LRRK2.

Alternatively, mutant LRRK2 and -synuclein might lead to

neurodegeneration, via a synergistic mechanism. The presence of genetic variants within the 3 UTR of SNCA that increase-synuclein levels in the SN (Fuchs et al., 2008) cause an earlier

onset of PD in patients harboring LRRK2 mutations ( Botta-Orfila et al., 2012). A synergistic relationship between these twokey PD-associated proteins may explain the variable penetranceof LRRK2 mutations (Healy et al., 2008) and why synuclein pa-thology is common in LRRK2 patients ( Biskup and West, 2009 ).

With -synucleins role in protein homeostasis disruption(Snyderet al., 2003; Cuervoet al., 2004 ; Fornaiet al., 2005 )andasLRRK2 itself modulates -synuclein degradation (Orenstein etal., 2013) and -synuclein-mediated toxicity ( Lin et al., 2009), asynergistic relationship between LRRK2 and -synuclein may

converge on dysfunction of protein homeostasis withincells.Theaccumulation of misfolded proteins causes proteome instability (Gupta et al., 2011 ), resulting in global cellular consequences.This might explain the widespread cellular dysfunction in vesiculartrafficking, neurotransmitter release, cytoskeletal dynamics, proteindegradation ( Greggio et al., 2011), and mitochondrial dynamics(Nakamuraet al., 2011 ; Wang et al., 2012 ), observed independently in -synuclein and LRRK2-associatedneurodegeneration.

AsthecellularenvironmentmaybeimportantforLRRK2s phys-iological role (Lewis and Manzoni, 2012), cell-type specificity may contribute critically to the interplay between these proteins. In thehindbrain, LRRK2 modulated -synuclein neuropathology ( Lin etal., 2009), but it was not required for -synuclein-mediated degen-eration( Daheretal.,2012 ).Understanding howandwhere thefunc-tional pathways of LRRK2 and -synuclein converge will providegreater insight into the common mechanisms underlying neurode-generation in PD.

ReferencesAbeliovich A, Schmitz Y, Farinas I, Choi-Lundberg D, Ho WH, Castillo PE,

Shinsky N, Verdugo JM, Armanini M, Ryan A, Hynes M, Phillips H,Sulzer D, Rosenthal A (2000) Mice lacking -synuclein display func-tional deficits in the nigrostriatal dopamine system. Neuron 25:239252.CrossRef Medline

Alegre-Abarrategui J, Christian H, Lufino MM, Mutihac R, Venda LL, An-sorge O, Wade-Martins R (2009) LRRK2 regulates autophagic activity and localizes to specific membrane microdomains in a novel humangenomic reporter cellular model. Hum Mol Genet 18:40224034.CrossRef Medline

Andersen PK, Gill RD (1982) Coxs regression model for counting pro-cesses, a large sample study. Ann Statistics 10:11001120. CrossRef

Andres-Mateos E,Mejias R,SasakiM, LiX, Lin BM, Biskup S,Zhang L,BanerjeeR, Thomas B, Yang L, Liu G, Beal MF, Huso DL, Dawson TM, Dawson VL(2009) Unexpected lack of hypersensitivity in LRRK2 knock-out mice toMPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine). J Neurosci 29:1584615850. CrossRef Medline

Anwar S, Peters O, Millership S, Ninkina N, Doig N, Connor-Robson N,Threlfell S, Kooner G, Deacon RM, Bannerman DM, Bolam JP, ChandraSS, Cragg SJ, Wade-Martins R, Buchman VL (2011) Functional altera-tions to the nigrostriatal system in mice lacking all three members of thesynuclein family. J Neurosci 31:72647274. CrossRef Medline

Arrasate M, Finkbeiner S (2005) Automated microscope system for deter-mining factors that predict neuronal fate. Proc Natl Acad Sci U S A 102:38403845. CrossRef Medline

Arrasate M, Mitra S, Schweitzer ES, Segal MR, Finkbeiner S (2004) Inclu-sion body formation reduces levels of mutant huntingtin and the risk of neuronal death. Nature 431:805810. CrossRef Medline

Barmada SJ, Skibinski G, Korb E, Rao EJ, Wu JY, Finkbeiner S (2010) Cyto-plasmic mislocalization of TDP-43 is toxic to neurons and enhanced by amutation associated with familial amyotrophic lateral sclerosis. J Neuro-sci 30:639 649. CrossRef Medline

Bilican B,

skibinski g 2014 j neuroscience subrayado

Documents